ARIES^SMJTHSONIAN INSTITUTrON NOUnilXSNI ""NVINOSHillAIS S3 J HVd 8 11 ""LI B RAl i fie a?) § g ^^^^ 8 ^ 5 8 CO ARIES SMITHSONIAN INSTITUTtON NOIXniliSNI NVINOSHill/\IS SaieVHail LIBRAl 2^ ^ — 2 r» C/* 5 \ E ^ 1/5 lilXSNi_^NVmOSHimS^^S3 I 8 n^LI B R AR I ES^SMITHSONIAN INSTITUTION^ NOIIDI ARIES SMITHSONIAN INSTITUTION NOIinilXSNI NViNOSHXIWS S3IHVH8n LIBRA ^ «^ 2 ^ ^ ^ 5 * «^ 5 IXIXSNil^NVINOSHXmS S3IMVM8n'^LIBRARIES^SMITHS0NlAN"^INSTITUTI0N^N0iXnX > m g ^^^^^ m ^ ARI£$^^SMITHSONlANJNSTITUTfON^NOIXnXIXSNl""NVINOSHXIWS^S3ldVMan"L^ 2 15 .i^ 2 ilXSNI_NVIN0SHXIIf«S*^S3 1 M VM 8 n\l B RAR I ES^^SMITHSONIAN ""INSTITUTION NOIXHXI 1^5 ARIES SMITHSONIAN INSTITUTION N0IXnXIJ.SNrNVIN0SHXI!AlS^S3l HVaQII LIBRAl llllSNI^NVtNOSHllWS S3 1 UVtlS II L| BRAR I ES^SMITHSONIAN "INSTITUTION NOUni SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHill^S S3 I ^ Vd 8 H ^LI B R AR 1 15 fsiviNOSHims S3iyvyan libraries Smithsonian institution NoiiniiisNi^ c H O SMITHSONIAN INSTITUTION NOIinXUSNI NVINOSHIH/VS S3!dVMan LIBRARIES z o 15 SMITHSONIAN 2 ^ > INSTITUTION NOIiniliSNI NOIiniliSNI X o z I NVINOSHIIIAIS S3IHVHan LIBRARIES I 5 ^ .Nvi NOSHiiy^s 3 i ^ V y a n ; o 9 > J 1 'SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHi!l/\IS S3iHVyan °' ^ c/s .NViNOSHiiws S3invaan libraries Smithsonian institution NoiiniiisNi^ z. . tn #/» //I 2j H SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHlll/VS S3 1 ^VM BIT LIBRARIES NviNosHiiws S3iyvyan libraries Smithsonian institution NoiiniiisNi : w z to 2" v.. (/> z VOLUME 32, NUMBER 1 A WEST AMERICAN JOURNAL OF BOTANY Contents A Reconsideration of the Nomenclature and Taxonomy of the Festuca altaica Complex (Poaceae) in North America Vernon L. Harms Full-glacial Vegetation of Death Valley, California: Juniper Woodland Opening to Yucca Semidesert Philip V. Wells and Deborah Woodcock A Cytotaxonomic Contribution to the Western North American Rosaceous Flora E. Durant Mc Arthur and Stewart C. Sanderson Serotiny and Cone-habit Variation in Populations of Pinus coulteri (Pinaceae) IN THE Southern Coast Ranges of California Mark Borchert A New Species of Erythronium (Liliaceae) from the Coast Range of Oregon Paul C. Hammond and Kenton L. Chambers 11 24 29 49 NOTES AND NEWS Yellow Jackets Disperse Vancouveria Seeds (Berberidaceae) Olle Pellmyr 56 NOTEWORTHY COLLECTIONS California Colorado 57 57 REVIEW ANNOUNCEMENTS 58 23, 28, 48, 59 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Madrono (ISSN 0024-9637) is published quarterly by the California Botanical So- ciety, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to Barbara Keller, California Academy of Sciences, Golden Gate Park, San Francisco, CA 941 18. £'JzYo/'— Christopher Davidson Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1985- Sterling C. Keeley, Whittier College, Whittier, CA Arthur C. Gibson, University of California, Los Angeles 1986— Amy Jean Gilmartin, Washington State University, Pullman Robert A. Schlising, California State University, Chico 1987— J. RzEDOwsKi, Instituto Politecnico Nacional, Mexico Dorothy Douglas, Boise State University, Boise, ID 1988 - Susan G. Conard, USDA Forest Service, Riverside, CA William B. Critchfield, USDA Forest Service, Berkeley, CA 1989 — Frank Vasek, University of California, Riverside Barbara Ertter, University of Texas, Austin CALIFORNIA BOTANICAL SOCIETY, INC. Officers for 1984-85 President: Rolf W. Benseler, Department of Biological Sciences, California State University, Hayward, CA 94542 First Vice President: Dale W. McNeal, Department of Biological Sciences, Uni- versity of the Pacific, Stockton, CA 95204 Second Vice President: Gregory S. Lyon, 1 360 White Oak Drive, Santa Rosa, CA 95405 Recording Secretary: Robert W. Patterson, Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: Barbara Keller, Department of Botany, California Academy of Sciences, San Francisco, CA 941 18. Treasurer: Cherie L. R. Wetzel, Department of Biology, City College of San Fran- cisco, San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, Isabelle Tavares, Department of Botany, Uni- versity of California, Berkeley 94720; the ^'iy/Zcr of Madrono; three elected Council Members: Lyman Benson, Box 801 1, The Sequoias, 501 Portola Rd., Portola Valley, CA 94025; James C. Hickman, Department of Botany, University of California, Berkeley 94720; Thomas Fuller, 171 West Cott Way, Sacramento, CA 95825; and a Graduate Student Representative, Joseph M. DiTomaso, Department of Botany, University of California, Davis, CA 95616. A RECONSIDERATION OF THE NOMENCLATURE AND TAXONOMY OF THE FESTUCA ALTAICA COMPLEX (POACEAE) IN NORTH AMERICA Vernon L. Harms The W. P. Fraser Herbarium and Biology Department, University of Saskatchewan, Saskatoon, Saskatchewan S7N OWO, Canada Abstract The recent taxonomic treatment of the North American members of the Festuca altaica complex by Looman (1979) is briefly reviewed and some nomenclatural prob- lems are discussed. The nomenclatural histories of the included taxa are briefly summarized, and a revised taxonomic treatment is presented, recognizing three sub- species within the complex in North America, and including F. altaica subsp. hallii. In Budd's Flora of the Prairie Provinces, Looman (1 979) presented a revised taxonomic treatment, including some major taxonomic and nomenclatural changes, for the ecologically important Festuca altaica complex (rough-fescues) in North America. This was offered with minimal explanation. Although further elaboration may still be forthcoming from the author, four years have passed with only an unexplained nomenclatural correction (Looman 1981). Mean- while, several comments with regard to this recent "revision" seem pertinent lest Looman's changes be too uncritically accepted. Looman (1979) recognized three separate North American species in the Festuca altaica complex, as follows: {\) F. altaica Trinius of central to northeastern Asia and northwestern North America, oc- curring on this continent in the Alaskan and northern Rocky Moun- tains {=F. altaica sensu stricto of other treatments); (2) "F. doreana Looman" (a name that has never been validly published and that was replaced by F. campestris Rydb. without explanation in the "reprint" of Budd's Flora [198 1]), applying to a species of the some- what more southern Rocky and Cascade Mountains and foothills {F. scabrella var. major, F. altaica var. major, F. scabrella sensu lat., or F. altaica subsp. or var. scabrella sensu lat. of other modem treatments); and (3) F. hallii (Vasey) Piper, a species of the Northern Great Plains grasslands, extending southward to Colorado along the eastern foothills of the Rocky Mountains (=F. scabrella s. str., or F. altaica subsp. or var. scabrella of most other modem treatments). Looman's (1 979) treatment contrasts with that of most other modem authors, who have recognized the latter two taxa as conspecific, either varietally distinguished or not, under F. scabrella Torrey, or Madrono, Vol. 32, No. 1, pp. 1-10, 15 February 1985 2 MADRONO [Vol. 32 else as one subspecies or two varieties of a more comprehensive F. altaica. The western mountain variant with larger, 3-5-floreted spikelets, unequal glumes, and taller culms (within the usually ac- cepted North American species, F. scabrella sensu lat.), has often been distinguished as var. major Vasey (="F. doreana'^ and F. cam- pestris sensu Looman 1977 and 1981, respectively), leaving the typ- ical varietal epithet, scabrella, to refer to the northern Great Plains taxon {=F. hallii sensu Looman), with smaller, 2-3-floreted spike- lets, sub-equal glumes, and shorter culms. Treated under an all- inclusive F. altaica, the two variants have either been merged under F. altaica subsp. scabrella (Torrey) Hulten or distinguished as F. altaica var. major (Vasey) Gleason and var. scabrella (Torrey) Brei- tung, respectively. Looman' s (1979) quite significant taxonomic revision of the North American members of the Festuca altaica complex represents an innovative contribution towards a better understanding of the mor- phological variations and taxonomic distinctions within this group. Presented therein for the first time are various clarifying taxonomic differences that are potentially useful in separating the included taxa of this complex. Unfortunately, because of its concise and unelab- orated presentation in manual format, submerged within this re- gional flora, Looman's treatment of this group may easily be over- looked by interested agrostologists, plant taxonomists and ecologists. Nevertheless, despite its value as a useful taxonomic contribution, there are some nomenclatural problems in Looman's (1979 and 198 1) taxonomic treatment of the rough-fescue complex. In the first place, the name '^Festuca doreana Looman" was not validly pub- lished in the original (1979) printing of Budd's Flora, nor has it been validated since. Latin diagnosis (as required under Article 36.1 //it was intended as a new species description) and bibliographic data for the cited synonym "F. scabrella var. major Vasey" (as required for a basionym under Articles 32.1 and 33.2 of the International Code of Botanical Nomenclature) are lacking. Furthermore, the name "F. doreana" would have been nomenclaturally superfluous even with the latter data, because it would have been based on the same basionym (viz. F. scabrella var. major Vasey, Contr. U.S. Natl. Herb. 1:278-279, 1897) as was the earlier species name, F. cam- pestris Rydberg [Mem. N.Y. Bot. Gard. 1:57, 1900]. If the even older specific name, Festuca scabrella Torrey, should be ruled out as not applicable to this taxon of the Rocky and Cascade Mountains, as Looman (1979) did, then the next in priority is F. campestris Rydberg. There appears no valid reason to exclude the latter name, since Rydberg (1900) seems clearly to have based it on F. scabrella var. major Vasey, simultaneously citing as a usage synonym, "F. scabrella Coulter, Man. R.M., 424, not Torr." Looman must have recognized the unacceptability of the name. 1985] HARMS: FESTUCA ALTAICA COMPLEX 3 "F. doreana," as shown by his substitution of the name F. campestris in the second printing (1981) of Budd's Flora and the use of the latter name in Looman (1983), although no nomenclatural expla- nations were given in either case. I believe that the still older specific name, F. scabrella, has priority over F. campestris for this western mountain taxon because Drummond's type material of the former appears best referred here. Looman (1979) evidently considered it necessary to find a new name for this taxon because of his view that the type of F. scabrella Torrey was referable instead to F. altaica s. str. He gave no explanation for this dispensation, which was remi- niscent of the much earlier, similar referral by Piper (1906). The taxonomic placement of Thomas Drummond's type material of F. scabrella Torrey is critical to the nomenclature of this group. An examination of the "Ex Herb. Torrey" holotype (now at GH), and the original "Gray Herb.!" isotype (also at GH) appeared readily enough to preclude their identification as the northern Great Plains- Eastern Foothills taxon, F. hallii sensu Looman because of the dis- tinctly unequal glumes and the spikelets mostly with 4 florets and exceeding 8 mm in length. Furthermore, the leaves of the "Ex Herb. Torrey" holotype specimens varied from 1.5 to 2.5 mm wide and were only loosely involute, although the leaves on the "Gray Herb.!" isotype were more strongly involute. Unfortunately the quality of the type materials was too inadequate and the spikelets too immature to assure their unequivocal identification as either F. altaica sensu str. or the more southern mountain taxon that Looman called "F. doreana" (in 1979) and F. campestris Rydb. (in 1981). Most of the individual plants appeared depauperate for either taxon, and had rather scanty panicles. Coloration of herbage and spikelets was ob- scure. Although measurements of the lengths of spikelets, glumes and lemmas fell more into the expected range for F. altaica sensu str. than for the more southern mountain taxon, these are question- ably reliable for the particular maturation stages. The rather con- tracted panicles with mostly ascending-erect upper branches, on the other hand, suggested the latter, as did the relatively inconspicuous glume-borders. Thus, identification of the available type material seemed inconclusive, although I was most inclined to place it with the more southern mountain taxon F. campestris sensu Looman 1981. Neither does the original diagnosis in Hooker (1840) give conclu- sive clues to the proper taxonomic placement of the F. scabrella type material. The unequal glumes, 3-4-floreted spikelets, nearly glabrous and only loosely involute leaves, and (lower?) panicle branches spreading, would seem to exclude the northern Great Plains taxon, F. hallii, and rather imply either F. altaica s. str. or the more southern mountain taxon that Looman called "F. doreana" and F. campestris (in 1979 and 1981). 4 MADRONO [Vol. 32 The spikelet size described as 0.75 in. (= 1 .9 cm), and the "upright" and "erect" panicles, seem best referable to the more southern moun- tain taxon, but the "purplish-green" spikelet color suggests F. altaica sensu str. Neither panicle shape nor spikelet color, however, seem very reliable diagnostic characters. The considerable emphasis given to the scabrous lemmas, leaf-sheaths, and culms, in the original species description, now seems unwarranted because such more or less pubescent forms occur in each of the three presently recognized taxa. Festuca hallii has the most consistently pubescent leaf-sheaths. The taxonomic placement of Thomas Drummond's type collec- tion of F. scabrella Torrey can be clarified by reference to its geo- graphical source. Drummond's type was collected in 1825 or 1826, on the John Franklin Second Expedition, and is listed in Hooker (1840) as from "alpine districts of the Rocky Mountains." There has been frequent difficulty in confidently determining the locality, elevation, and habitat of many Drummond collections. But with the specimen label data and notations in Hooker (1840), supplemented by information from Drummond's own travel accounts (Drummond 1827, 1830; and in Franklin 1928, pp. 308-313), especially as clar- ified by the chronological and geographical tracing of his expedition by Bird (1967) and Ewan and Ewan (1981, pp. 63-64), it can be concluded with some confidence that Drummond's "Rocky Moun- tain" collections labelled as above came from the present-day Jasper National Park region. Drummond's collections from this region ap- parently ranged from "Jasper House" (53°20'N, 117°51'W), Rock Lake (called "Lac-la-Pierre" (53°27'N, 1 18°16'W), and the length of the Snake Indian (called "Assinaboyne") River (ca. 53°10-22'N; 1 18°00-50'W), southward along the upper Athabasca River (called "Red-Deer River, one of the branches of the Athapescow") to its headwaters near the Columbia Ice Fields (53°1 3-25'N, 1 1 7°1 5-20'W) and up the Whirlpool River toward Athabasca Pass (at 52°23'N, 1 18°1 1'W), on the Columbia Portage. The collections made on Drummond's side-expedition north- westward to the headwaters of the Peace River in September 1826, seemingly were labelled from "north of Smoking (=Smoky) River" and/or "lat. 55°" or "lat. 56°." His collections from Athabasca Pass itself, in October 1926, appear labelled, "height of land," "summit" or "near summit of Rocky Mountains." His British Columbian col- lections, made during his brief October 1926 expedition through and southwest of Athabasca Pass along the Wood (called "Portage") River to the Columbia River, apparently bore the labels "Grande Cote," "Portage (River)," "sources of the Columbia," "the Colum- bia (River)" or "west side of the Rocky Mountains." Thus, at least tentatively discounting those Rocky Mountain collections by Drum- mond from the Peace River, Athabasca Pass and British Columbia, all presumably bearing the special label notations indicated above, 1985] HARMS: FESTUCA ALTAIC A COMPLEX 5 the type locality of F. scabrella can be narrowed down with consid- erable assurance to the east (i.e., Alberta) side of the Continental Divide and to within the coordinates: 52°15'-53°30'N and 1 17°15 - 118°50'W. Interestingly, the presumed type locality of F. scabrella falls within the overlapping known ranges of the taxa that Looman (1 98 1) called Festuca hallii and F. campestris, and somewhat less than 100 miles south of the known range of F. altaica sensu str. (as I interpret these three taxa). In this general area of east-central British Columbia and west-central Alberta, the three taxa are characteristic of lower, mid- dle and high mountain elevations, respectively, so the altitudinal placement of the type collection of F. scabrella assumes importance. But, as his botanical colleague, John Richardson (1851, Pt. 3, p. 521), pointed out, "It is unfortunate that the vertical limits of the species gathered by Drummond in the mountains were not better noted ... (as that) would have conveyed much information with respect to the distribution of plants." The habitat notation in Hooker (1840), "alpine districts," is not as unequivocal as it at first might seem. The word "alpine" was used frequently for Drummond col- lections in Hooker (1 880) in references to "alpine woods," "marshes," etc., habitats, sometimes in apparent contrast to "open elevated places," "summits," or "barren places," and at other times in ap- parent contrast to "mountain woods." So the notation of "alpine districts" for Drummond's collections does not necessarily, if at all, imply an alpine tundra habitat at high elevations above tree-line, but quite possibly a sub-alpine zone. The habitat indication, "alpine districts," given in Hooker (1840) may also be queried because such a notation was not included on Drummond's exsiccatae labels. The cumulative evidence from the characteristics of the type ma- terials available of Festuca scabrella (viz., the holotype and an iso- type, both in GH), the original species diagnosis, and the presumed geographical location and possible habitat of the type collection, allows its most likely, although still somewhat tentative, referral to the larger more southern mountain taxon that Looman called F. campestris Rydb. (in 198 1). Such a taxonomic placement of the type material, would give the name F. scabrella Torrey priority over F. campestris. Acceptance of these conclusions concerning the type of F. scabrella raises a nomenclatural problem with respect to the most frequent traditional taxonomic treatment (although not Looman's) that has recognized the two varieties, scabrella and major, within F. sca- brella. A problem results because the names, F. scabrella Torrey (in Hooker, Fl. Bor.-Am. 2, 252, 1840; type from alpine districts of the Rocky Mountains, Canada, Thomas Drummond s.n.) and F. sca- brella var. majorVasey {Conthh. U.S. Natl. Herb. 1:278-279, 1897; type from Spokane Co., Washington, Suksdorf 118), are both based 6 MADRONO [Vol. 32 on type specimens that are interpreted as belonging to the more southern Rocky and Cascade Mountain taxon rather than to the northern Great Plains-Eastern Foothills taxon. If the latter is rec- ognized as distinct from the former, the epithet scabrella is un- available for it. For this reason, rather than because of Looman's (1979) referral of the F. scabrella type to synonymy under F. altaica sensu str., I concur with Looman's substitution of the epithet hallii, which is based on Melica hallii Vasey (Bot. Gaz. 6:296, 1881; lec- totype [indicated by Piper, 1 906] from northern Colorado, Hall and Harbour 621 [US]). Upon examination of the lectotype of F. hallii and duplicates of it, as well as several later collections by W. A. Weber et al. from Larimer and Huerfano Counties, Colorado, it seems apparent that all of these do indeed belong to the same taxon as does the rough-fescue of the northern Great Plains and Eastern Foothills grasslands. This identification seems definite despite their rather surprising occurrence at high alpine-meadow elevations. Some short rhizomes are even evident on the type materials, a taxonomic characteristic also noted by Weber (1961) on his Colorado collec- tions. Aside from the nomenclatural problems, Looman's recognition of the three taxa within the North American F. altaica complex seems basically well conceived and acceptable, although not nec- essarily with these variants treated as species. If they were to be treated as distinct species, the appropriate names for the (1) Be- ringian, (2) more southern Rocky and Cascade Mountain, and (3) northern Great Plains taxa of the rough-fescue complex as distin- guished by Looman, would be, respectively: {\) F. altaica Trinius, (2) F. scabrella Torrey and (3) F. hallii (Vasey) Piper. If distinguished at the varietal level, the appropriate names would be, respectively: (1) F. altaica Trinius var. altaica, (2) F. altaica var. major (Vasey) Gleason, and (3) F. altaica var. hallii (Vasey), a combination that has never been made. Recognition of these taxa at the subspecific level seems most pref- erable, however, for the following reasons. Although I agree that three North American taxa are at least broadly distinguishable within the F. altaica complex, personal experience in both field and her- barium strongly suggests to me that these are rather less discrete than is implied by Looman (1979). In many specimens, especially those from the mid-latitude (circa 49°-55°) in the Rocky Mountains and eastern foothills, the characters that distinguish the taxa may often appear overly subtle or seemingly intergradient. Thus it would seem preferable to accept these taxa at an infraspecific rather than a specific level. On the other hand, despite evidences of apparent intergradation (i.e., the presence of morphological intermediates), they occupy broadly separate geographical ranges for the most part. 1985] HARMS: FESTUCA ALTAICA COMPLEX 7 The recognition of these taxa at a subspecies rank seems most ap- propriate. The following taxonomic treatment (viz., key and nomenclatural summaries) is presented for Festuca altaica sens. lat. in North Amer- ica, based largely on the diagnostic criteria given by Looman (1 979), to distinguish and circumscribe the three recognized taxa, but at a subspecific level. The generalized distributional information ap- pended for each taxon, and forming a partial basis for the ranges given in Fig. 1, has been extracted from Johnson (1958) and Hitch- cock (1950), and a variety of regional floras including Gleason (1952), Harrington (1954), Scoggan (1957, 1978), Hulten (1968), Hitchcock et al. (1969), Voss (1972), Taylor and MacBryde (1977), Cronquist etal. (1977), McGregor et al. (1977), Porsildand Cody (1980), Boivin (1981), and Packer (1983), and from the numerous herbarium spec- imens personally reviewed. Key to Subspecies of Festuca altaica in North America 1 . Plants tufted, not at all rhizomatous; culms often over 6 dm high; spikelets over 1 0 mm long; fertile florets 3-6 per spikelet; glumes distinctly unequal, the first glume distinctly shorter than the first lemma; leaf-blades either flat or involute, 7-nerved; at least the 8 MADRONO [Vol. 32 lower panicle branches spreading to somewhat reflexed or ± ascending. 2. Culms (3-)4-6(-8) dm high; herbage yellowish to dark-green; spikelets often ± reddish, (8-)9-13 mm long, with 3-5 florets; glumes with conspicuous translucent borders; lemmas 5-8 mm long; leaf-blades 1-2.5 mm wide; panicles open, lax, ± ovoid, with longer, spreading, and weaker branches, often ± secund subsp. altaica 2. Culms (3-)5-10 dm high; herbage more grayish to bluish- green; spikelets mostly green to stramineous, (8-) 10- 16 mm long, with 4-6 florets; glumes with less conspicuous translu- cent borders; lemmas 6-9 mm long; leaf-blades (1-) 1.5-3 (-4) mm wide; panicles tending to be more erect, contracted, stiffer and narrower, not at all secund subsp. scabrella 1 . Plants less strongly tufted, somewhat rhizomatous and mat-form- ing; culms 2-6 dm high; spikelets mostly green to stramineous, 7-8 mm long; florets 2-3 per spikelet, the 3rd often sterile; glumes subequal, about equal to the first lemma; leaf-blades involute, less than 1.5 mm wide, obscurely 5 -nerved; the lower panicle branches more strongly ascending to contracted-appressed .... subsp. hallii 1 . Festuca altaica Trinius subsp. altaica. "Northern Rough Fes- cue."—/^, altaica Trinius in Ledebour, Fl. Altaica, 1:109-110. 1829.— Type: Central Asia, Altai Mts.: "In summa alpe ad fon- tem fl. Acjulac rarissima," (tr.: "Very rare on mountain summit at source of Acjulac River"), C. B. Trinius (Holotype: LE). Distribution, e.c.-n.e. Asia; Alaska, Yukon, w. Mack. Distr., s. to n. B.C. and in Rocky Mts. to e.c. B.C. and w.c. (and n.w.?) Alta. 2. Festuca altaica Trinius subsp. scabrella (Torrey) Hulten. "Mountain Rough Fescue." — F. scabrella Torrey, in Hooker, Flora Boreali-Amer. 2:252. 1840.— i^. altaica subsp. scabrella (Torrey) Hulten, Flora Alaska and Yukon, v. 2, p. 241. 1942.— F. altaica var. scabrella (Torrey) Breitung, Amer. Midi. Natu- ralist 58:12. 1957 (as to basionym, not concept). — "i^. altaica forma scabrella (Torrey) Looman," Budd's Flora Can. Pr. Prov., p. 128. 1979 (as to basionym, not concept; non rite publ.).— Type: Canada: "Alpine districts of the Rocky Mountains," 1827, T. Drummond s.n. (Holotype: the "Ex Herb. Torrey" specimen now at GH!; isotypes: the "Gray Herb.!" specimen at GH!; BM). F. scabrella var. major Vsisey, Contr. U.S. Natl. Herb. 1:278. 1893.- F. altaica var. major (Vasey) Gleason, Phytologia 4:21. 1952.— F. campestris Rydberg, Mem. N.Y. Bot. Gard. 1:57. 1900 (basi- onym: F. scabrella var. ma/or Vasey). — "F. doreana Looman," 1985] HARMS: FESTUCA ALTAICA COMPLEX 9 Budd's Flora Can. Pr. Prov., pp. 128-129. 1979 (basionym indicated as F. scabrella var. major Vasey; non rite publ.).— Type: USA: Washington: Spokane Co.: "on prairies," 1884, SuksdorfllS (Holotype: US!). Distribution. Rocky Mts. of w.c.-s.w. Alta., e.c. and s. B.C., s. in Rocky and Cascade Mts. to e. Ore., s. Ida., and w. Mont.; disjunct eastern isolates in Great Lakes region (n. Mich, and s. Ont.), the Gaspe Peninsula region (e. Que.), e.c. Que. (and w.c. Lab.?), Ungave Bay (n. Que.), and w. Newfoundland. The Michigan and e. Canadian disjunct populations of this complex seem best referred, at least tentatively, to subsp. major, but this conclusion needs verification, especially since, for reasons of geographical proximity, subsp. hallii might seem the more likely taxon to occur there. Yet the phytogeo- graphical pattern of Cordilleran taxa with such eastern isolates is well known for various other groups. 3. Festuca ALTAICA Trinius subsp. hallii (Vasey) Harms comb, no v. "Plains Rough Fescue."— M^//c<3 hallii Vasey, Bot. Gaz. 6:296. 1881. -Festuca hallii (Vasey) Piper, Contr. U.S. Natl. Herb. 1 0: 31. 1906.— i^. altaica sub var. hallii (Vasey) St. Yves, Candollea 2:271. 1925.— i^. scabrella subsp. hallii (Vasey) W. A. Weber, Univ. Colo. Stud., Ser. Biol. 7:8. 1961. -Type: USA, Rocky Mountains, northern Colorado, Lat. 39°-40°, 1862, Hall and Harbour 621 (Lectotype [as indicated by Piper, Contr. U.S. Natl. Herb. 10:31. 1906]: US!; isolectotypes: US-3!, F-photo- graph!). F. scabrella auct. pro parte, non Torrey. Distribution. This is the rough fescue variant of the northern Great Plains grasslands and parklands (w. Alta., c. Sask., to s.e. Man. and n.w. Dak.), apparently disjunct near Thunder Bay, Ont., extending s. along the eastern foothills of the Rocky Mts. in Mont, and Wyo., to s.c. Colo, (in Colorado, at high elevations, rare, and perhaps disjunct, as suggested by Weber, 1961). Literature Cited Bird, C. D. 1 967. The mosses collected by Thomas Drummond in Western Canada, 1825-1827. Bryologist 70:262-266. BoiviN, B. 1981. Flora of the prairie provinces. Part 5— Gramineae. Mem. de I'Herbier Louis-Marie, Universite Laval, Quebec. Cronquist, a., a. H. Holmgren, N. H. Holmgren, J. L. Reveal, and P. K. Holmgren. 1977. Intermountain flora. Vol. 6. New York Bot. Garden. Drummond, T. 1827. Letters (from Drummond). Linnaea 2:519-525. . 1 830. Sketch of journey to the Rocky Mountains and to the Columbia River in North America. Hooker's Bot. Misc. 1:178-219. EwAN, J., and N. D. Ewan. 1981. Biographical dictionary of Rocky Mountain naturalists. Bohn, Scheltema, and Holkema, Utrecht/ Antwerpen. 10 MADRONO [Vol. 32 Franklin, John. 1828. Narrative of a second expedition to the shores of the Polar Sea in the years 1825, 1826, and 1827. Charles E. Tuttle Co. Rutland, VT. Gleason, H. a. 1952. The New Britton and Brown Illustrated Flora of the North- eastern United States and Adjacent Canada. New York Botanical Garden, NY. Harrington, H. D. 1954. Manual of the plants of Colorado. Sage Books. Hitchcock, A. S. 1950. Manual of the grasses of the United States (2nd ed., rev. by A. Chase). USDA Misc. Publ. 200. Hitchcock, C. L., A. Cronquist, M. Ownbey, and J. W. Thompson. 1 969. Vascular plants of the Pacific Northwest, Part 1 . Univ. Wash. Press. Hooker, W. J. 1840. Flora Boreali-Americana, Vol. 2. H. G. Bohn, London. HuLTEN, Eric. 1968. Flora of Alaska and neighboring territories. Stanford Univ. Press, Stanford, CA. Johnston, Alexander. 1968. Note on the distribution of Rough Fescue {Festuca scabrella Torr.). Ecology 39:536. LooMAN, J. 1979. Festuca, fescue. In J. Looman and K. F. Best, Budd's Flora of the Canadian prairie provinces, pp. 1 27-1 32. Research Branch, Agriculture Can- ada, Publ. No. 1662. . 1981. Festuca, fescue. In J. Looman and K. F. Best, Budd's Flora of the Canadian prairie provinces, pp. 127-132. Research Branch, Agriculture Canada, Publ. No. 1662. (Reprint with corrections of 1979 publication.) . 1983. Ill range and forage plants of the Canadian prairies, Publ. No. 1751, Agriculture Canada, Research Branch, Ottawa. Marie- ViCTORiN, Frere. 1964. Flore Laurentienne. Universite de Montreal, Mon- treal, PQ. McGregor, R. L., T. M. Barkley, et al. 1977. Atlas of the flora of the Great Plains. Iowa State Univ. Press, Ames. Packer, J. G. 1983. Flora of Alberta, 2nd ed. Univ. Toronto Press, Toronto. Piper, C. V. 1906. North American species of Festuca. Contr. U.S. Natl. Herb. 10: 1-48. PoRsiLD, A. E. and W. J. Cody. 1 980. Vascular plants of the continental Northwest Territories, Canada. Natl. Mus. Nat. Sci., Ottawa. Richardson, John. 1851. Arctic searching expedition: a journal of a boat- voyage through Ruperts Land and the Arctic Sea in search of the Discovery ships under command of Sir John Franklin. Longman, Brown, Green and Longmans, Lon- don. Rydberg, p. a. 1 900. Catalogue of the flora of Montana and the Yellowstone National Park. Mem. N.Y. Bot. Gard. 1: (No. 3835). ScoGGAN, H. J. 1957. Flora of Manitoba. Natl. Mus. Canada, Ottawa. . 1978. The Flora of Canada, Part 2— Pteridophyta, Gymnospermae, Mono- cotyledonae. Natl. Mus. Canada, Ottawa. Taylor, R. L. and B. MacBryde. 1977. Vascular plants of British Columbia, — A descriptive resource inventory. Techn. Bull. No. 4, The Botanical Garden, Univ. British Columbia. Vasey, George. 1881. Some new grasses. Bot. Gaz. 6:296-298. . 1893. Descriptions of new or noteworthy grasses from the United States. Contr. U.S. Natl. Herb. 1:267-280. Voss, E. G. 1972. Michigan flora, Part 1 — gymnosperms and monocots. Cranbrook Inst. Sci., Bloomfield Hills, Michigan. Weber, W. A. 1961. Additions to the flora of Colorado. Univ. Colorado Stud., Series in Biology, No. 7:1-26. (Received 10 August 1982; accepted 8 May 1984.) (See note added in proof, p. 59.) FULL-GLACIAL VEGETATION OF DEATH VALLEY, CALIFORNIA: JUNIPER WOODLAND OPENING TO YUCCA SEMIDESERT Philip V. Wells and Deborah Woodcock Department of Botany, University of Kansas, Lawrence 66045 Abstract Full-glacial (13,000-19,000 yr bp) wood rat (Neotoma) deposits from Death Valley establish a 1 200-1 500-m displacement of juniper woodland below modem, moun- taintop relicts of Juniperus osteosperma. At an elevation of 425 m, however, there was full-glacial (19,550 yr bp) semidesert dominated by chaparral yucca (Fwcca whip- plei) with minor Joshua tree ( Y. brevifolia). The late Pleistocene climate must have been much less arid and more equable with cooler summers than at present. Modem, hot-desert vegetation appeared at the 425-m site between 1 1,000 and 10,000 yr bp. The shift from pluvial woodland to hyperarid desert at 775 m was time-transgressive during 13,000-9000 yr bp, as is documented by three dated transitional stages of semidesert from this site. Death Valley, California, a northern extremity of the Mohave Desert, is now one of the hottest and driest places on earth, but there is evidence that the extreme aridity of the modem climate is less than 1 1,000 years old. Wood rat {Neotoma) middens (Wells 1976) from very low elevations in Death Valley (Fig. 1) provide a detailed macrofossil record (thousands of leaves, twigs, seeds, etc., in many separate deposits) of late Pleistocene and Holocene vegetation over the past 20,000 years, whence the magnitude and nature of climatic change may be inferred. A more generalized indication of vegeta- tional change during this time span is apparent from the pollen records at Tule Springs to the east in Nevada (Mehringer 1967) and at nearby Searles Lake, California (Roosma 1958). Neotoma macrofossil records from the Amargosa Range, which flanks Death Valley on the east, indicate that late-Pleistocene Juniper Woodland (dominated by Juniperus osteosperma) grew in the ele- vational range 1130-1280 m (Fig. la, b), on slopes now occupied by creosote bush scrub. The highest site (a), on the northeast slope of Pyramid Peak at 1280 m, yielded a late-glacial (11,800 yr bp) record of relatively mesophytic juniper woodland with montane gooseberry, mountain mahogany, and shrubby ash (Wells and Berger 1967). A new and significant (though rare) component of this as- semblage in the context of the other records reported here (Table 1) is Yucca whipplei. At much lower elevations on the west flank of Death Valley (Fig. Madrono, Vol. 32, No. 1, pp. 11-23, 15 February 1985 12 MADRONO [Vol. 32 Fig. 1 . Death Valley and surrounding region of eastern California (see inset). Solid line delimiting Death Valley and adjacent basins is the 610-m contour, which esti- mates the pleniglacial extent of juniper woodland. Stippled area in Death Valley approximates the high stands of Pleistocene Lake Manly (see text). Present-day wood- lands and montane zones (black areas) are restricted to elevations above 1950 m on high mountains: G, Grapevine Pk., 2679 m (Grapevine Mts.); H, Hunter Mtn., 2272 m and T, Telescope Pk., 3368 m (Panamint Range); M, Maturango Pk., 2694 m (Argus Range). Locations of Quaternary Neotoma macrofossil records: (a,b) Amargosa Range: a, 1280 m (Wells and Berger 1967); b, 1 130 m (M. D. Kelly, in Van Devender 1977); (c) Panamint Range, 775 m, 425/414 m, and 260 m (this report). 1985] WELLS AND WOODCOCK: PLEISTOCENE VEGETATION • - S ■S + > I ^ o O 52 -2 u. C t« ■I o o (J cd 5 T3 oo O TO o '7 (/I Q -35 o O ^ c« (-! _ o.S g -H ^ o S < «/5 - o 43 °° W CO t '3; Z CO ^ w S S _j O CO oa ^ < CO ^ ^ M + O _ t2 00 - -go .sp io c/i CO < S hJ oo TO O CO CO S . ^ o o c X -I" a o 6^ o o as ^ On ^ +1 ^ O OS 11 o +1 ^ >/-> : +1 OS +1 OS O O ^ OS +1 OS +1 o o 5o O OS +1 if +1 + + + + + + + + + : + + + + + + + + P ? ^ ^ S ■-2 ^ ^ Q § O ^ i s ^ ^ s: -2 "§ O ?S ^ ^ C ^ C >-i >^ a I 5^ 5^ « ^ ^ ^ r<« >0 -ff* r- V a, O ^ 14 MADRONO [Vol. 32 Ic) new Neotoma records extend the late-glacial (ca. 13,000 yr bp) lower limit of juniper woodland to below 775 m, a level that now supports an extremely sparse, hot-desert scrub of Larrea and Am- brosia dumosa. Pleniglacial (= full-glacial) deposits from a south- facing site at 425 m contain an unusual abundance of yucca leaves (chiefly Y. whipplei; some Y. brevifolia), a number of semidesert or cool-desert shrubs, and only occasional twigs of juniper in a large volume of matrix; two of these assemblages yielded dates of 19,550 and 17,130 yr BP (Table 1). The classical Wisconsinan glacial max- imum is dated at 18,000 yr bp. Juniperus osteosperma (cf Fig. 2a) is now dominant only at ele- vations above 1950 m in the higher parts of the Panamint Range, some 15-20 km distant from its documented Pleistocene occur- rences in the foothills bordering the western floor of Death Valley (Fig. 1). Junipers are no longer extant in the lower and drier Funeral Range (to 2040 m) on the east side of the valley. The late-glacial abundance of juniper (thousands of leafy twigs) at the 775-m, north- facing site, together with the recovery of a few traces of juniper twigs from the pleniglacial, Fwcc^^-dominated deposits at 425 m (south- facing), indicates a Pleistocene displacement of 1200-1500 m for juniper. This is one of the most extreme and well-documented Qua- ternary shifts thus far established for any species or zone in western North America; the exceptional magnitude is especially significant because of the modem aridity of Death Valley. The geographic extent of the vegetational change near Death Val- ley is indicated in Fig. 1 , where the drastically shrunken mountaintop areas of the modem woodland and higher zones are shown in black; the minimal late-Pleistocene (Wisconsinan) extent of juniper wood- land is estimated by the 610-m contour. Thus, most of the mapped area above the valley floor was probably wooded during the last glacial. The extraordinary vegetational displacement documented in Death Valley reflects both the magnitude of relatively recent climatic changes and the unusually large range of elevation locally available for their expression. A Zone of Chaparral Yucca in Death Valley In contrast to the juniper-dominated woodland assemblages, the semidesert of yuccas indicated by the deposits at 425 m has no modem analog at higher elevations on nearby mountains, so far as is now known. The principal pleniglacial species at the lowest (425- m) site was Yucca whipplei (Fig. 2b), a mild-climate rosette-shrub with very distinctive, grooved leaf-surfaces (lacking a smooth cu- ticular sheath). This Yucca is now apparently absent in the Death Valley region, despite the extensive elevational/climatic gradients available on the higher mountain ranges. On the other hand, the 1985] WELLS AND WOODCOCK: PLEISTOCENE VEGETATION 15 0.5 cm 0.5 cm Fig. 2. Drawings of representative specimens among the many thousands of plant macrofossils in late Pleistocene Neotoma middens from near the floor of Death Valley: (a) Juniperus osteosperma: leafy twig and seeds, (b) Yucca whipplei: leaf base and tip, seed, (c) Purshia glandulosa: upper and lower surface of leaf (d) Ambrosia dumosa: leaf fragments and fruit. arborescent Yucca brevifolia, a more xerophytic and cold-tolerant species, was also present near the floor of Death Valley at 19,550 yr BP but now occurs only above about 1700 m on the slopes of some of the higher mountains. Of course, Joshua tree woodlands are the most characteristic modem feature of the higher or less arid sectors of the Mohave Desert, of which Death Valley is a northern extremity. The pleniglacial combination of Yucca whipplei and Y. brevifolia, however, is decidedly unusual today, occurring on the extreme west- em margins of the Mohave Desert (e.g., foothills of Tehachapi Mountains) in a zone of transition to Mediterranean climate. Minor components of the pleniglacial Yucca community in Death Valley (Table 1) included Chrysothamnus teretifolius (a cool-desert species), Atriplex confertifolia, and Opuntia basilaris. The latter two species are favorite food plants of the desert wood rat, Neotoma lepida; therefore their scarcity in the pleniglacial assemblages at the 42 5 -m site was undoubtedly real. The full-glacial Neotoma middens clearly 16 MADRONO [Vol. 32 indicate a relatively cool semidesert near the bottom of Death Valley, comparable to the "high desert" community of the western Mohave Desert that borders the Transverse Ranges of California. The Yucca-dominatQd semidesert may have existed only locally, however. There are much more numerous macrofossil records of low-elevation coniferous woodland occurring throughout most of the Mohave Desert, even as late as 8000-10,000 yr bp (Wells and Berger 1967, King 1976, Van Devender 1977, Wells 1983). Under full-glacial conditions, juniper- Joshua tree woodland descended as low as 258 m, even in the northern Sonoran Desert, just to the southeast of the Mohave (Wells 1974, 1983). These low-elevation Mohavean woodlands were accompanied by semidesert shrubs, but there has been no firm pleniglacial evidence of Larrea and Ambrosia, the modem codominant shrubs of the hyperarid hot desert now prevalent at most of the low-elevation Neotoma sites in the Mohave Desert (Wells and Hunziker 1976). In fact, increasingly numerous late-Pleistocene Neotoma records establish a monotonous pattern of low-elevation woodland in the Southwest, from southern Cali- fornia to western Texas and adjacent areas of Mexico (Wells 1966, 1969, 1974, 1976, 1979, 1983; Van Devender and Spaulding 1979). The geography of the Chihuahuan Desert (relatively high base levels) permitted species-rich pluvial woodlands of pinyon, juniper and oaks to dominate the lowlands of most, if not all, of this modem desert region (Wells 1 966, 1974). Only the more subtropical Sonoran Desert, with elevations descending to sea level at lower latitudes, provided a pleniglacial refugium for the more extreme xerophytes (Wells and Hunziker 1976). Thus, the pleniglacial record of Yucca semidesert on the lower slopes of Death Valley is the first direct evidence that the ubiquitous Pleistocene woodlands of the arid Southwest did not descend to local base level everywhere during the glacial maximum. Restriction of the extent of any desert vegetation at that time is indicated by several limiting factors: 1) increasing basal elevations in all directions from the uniquely low and rain-shadowed trough of Death Valley; 2) greater lowering of the more mesophytic pinyon pine and evergreen oak (as well as juniper) components of the woodland zone toward the southeast at present (associated with the increasing proportion of summer rain in that direction); a very similar pattem is docu- mented in the pleniglacial macrofossil record (Wells 1979), with juniper descending to local base level in the eastem Mohave, north- em Sonoran, and probably all of the Chihuahuan Desert; and 3) occupation of the bottoms of enclosed basins by extensive glacio- pluvial lakes (Hubbs and Miller 1948). A potentially extensive zone of semidesert or desert vegetation on the floor of Death Valley was preempted by Pleistocene Lake Manly (cf Fig. 1); the highest stand of this lake is not well known, but it may have risen to within 250 1985] WELLS AND WOODCOCK: PLEISTOCENE VEGETATION 17 m of the lowest Neotoma record of Yucca semidesert at 425 m above sea level (Hunt and Mabey 1966). Presence of Atriplex confertifolia in the full-glacial deposits at 425 m suggests the possibility of zones of Atriplex and other halophytes around the fluctuating margins of Lake Manly, comparable to the existing salt-desert vegetation bor- dering the shrunken remnants of Pleistocene lakes in the cooler, northern Great Basin (cf. Wells 1983). Time-Transgressive Vegetational Shifts A major upward shift of the Yucca semidesert in Death Valley occurred during 13,000-11,000 yr bp, signifying a large climatic change. A deposit dated to 1 1,210 yr bp at the 775-m site is com- posed, not of juniper as at 13,060 yr bp, but of Yucca whipplei (previously unrecorded at this site) and of shrubs now characteristic of a cooler zone transitional between woodland and desert (Table 1). Absence of juniper from this still relatively early (1 1,210 yr bp) record at the 775-m site is doubly significant because this site is a relatively mesic, north-facing wall of a deep canyon, and Juniperus osteosperma is a very xerophytic woodland species. This juniper persisted at other localities in the Mohave Desert at comparably low or somewhat higher elevations as late as 10,000 yr bp, or even as late as 8000 yr bp at a few higher sites (Wells and Jorgensen 1964, Wells and Berger 1967, Van Devender 1977, King 1976, Wells 1983). The explanation for the much earlier upward shrinkage of juniper woodland in Death Valley probably lies in the great vertical relief of the graben and the extreme intensity of rain-shadow induced by the mountain barriers around it. No other valley within the Mohave Desert has so great an elevational span between existing lower limit of juniper woodland (at ca. 1900 m) and local base level (—86 m at Badwater) as now obtains in Death Valley, a span of nearly 2000 m. In most other sectors of the Mohave, the span is considerably less than 1000 m, well within the robust tolerance limits of Juniperus osteosperma under the glaciopluvial climate. In Death Valley, the uniquely wide environmental gradient from mountain slope to valley bottom apparently exceeded the tolerance limits of Juniperus os- teosperma even under the cooler climate of the glacial maximum. Populations of the juniper growing as much as 1200-1500 m below their present lower limits were probably near the end of their tether; and they could have been trimmed back by climatic oscillations too small to affect junipers growing at slightly higher and less marginal levels elsewhere in the Mohave. An outstanding feature of the 1 1,210-year-old deposit at the 775- m (north-facing) site in Death Valley is the similarity in composition to the full-glacial (19,550 yr bp) semidesert assemblage at the lowest (and south-facing) site at 425 m (Table 1). Shared dominant species 18 MADRONO [Vol. 32 include Yucca whipplei and Chrysothamnus teretifolius; presence of Purshia glandulosa (Fig. 2c) at the 775-m site suggests that woodland was not far above at 1 1 ,2 1 0 yr bp. The time-transgressive consistency of this ywcc<2-semidesert community seems to be an example of a simple cliseral shift of 350 m along this segment of the elevational gradient. A nearly contemporaneous (1 1,600 yr bp) record from 500 m higher (at 1280 m) shows Yucca whipplei coexisting with juniper woodland (Table 1). The further shift to modem vegetation at the 775-m site involved at least two other stages (Table 1). About 1500 years later, at 9455 yr BP, Yucca whipplei had dropped out locally, but Purshia glan- dulosa, Chrysothamnus teretifolius, and other shrubs of the high desert/woodland transition still persisted. By 9090 bp, however, only two transitional species, Haplopappus cuneatus and Opuntia basi- laris, remained; of the Pleistocene species, only O. basilaris survives in the modem creosote bush scmb at the 775-m site today. The unfolding in ordered time sequence at this location of three transitional phases of semidesert vegetation, almost entirely distinct from both late-glacial juniper woodland and modem desert scmb, is strong evidence of a slow, secular upward shift of the woodland/ desert boundary past this site in the interval 13,000-9000 yr bp. A gradual and protracted warming and desiccation of climate during this late-glacial/Holocene phase of transition is clearly indicated here. Thus, the suggestion (Van Devender 1977) of an abmpt shift from woodland to desert throughout the Southwest at about 8000 yr BP (on the basis of a few coincidental dates at scattered sites) is refuted by the present evidence and by a more detailed review of the whole Mohavean data set (Wells 1983). The earliest appearance of vegetation similar to the existing, hot- desert scmb in Death Valley was prior to 10,000 yr bp. At the lowest and most xeric (south-facing) 425-m site, which supported Yucca semidesert during the full-glacial (17,000-19,500 yr bp), there is a hiatus of seven millennia, followed by a Neotoma record of Ambrosia dumosa, dated at 10,230 yr bp (Table 1). This deposit is composed entirely of remains of this Ambrosia (Fig. 2d), but lacks Larrea tridentata. Larrea and Ambrosia dominate the existing desert scmb at the 425-m site and throughout Death Valley from below sea level (except on saline deposits) upward to 1 500 m or more. Only the absence of Larrea denies characterization of the 10,000- yr BP desert scmb community as really modem; the absence of Larrea at 10,230 yr BP, however, may not be of climatic significance at the latitude of Death Valley. This is merely another instance, already indicated by the weight of macrofossil evidence throughout the Southwest (Wells and Hunziker 1976), of the late arrival of Larrea over much of its present range in North America. Within the Mohave Desert region, the earliest firmly established Neotoma record of Lar- 1985] WELLS AND WOODCOCK: PLEISTOCENE VEGETATION 19 rea tridentata is from the Mohave River valley at 670 m (west- facing), just north of Ord Mountain and east of Daggett; the radio- carbon date of 7400 ± 100 yr bp (UCLA-759) was on a branch and leaves of Larrea itself, which dominated the deposit (Wells and Berger 1967). Woodland conifers were lacking in this record, but Juniperus osteosperma still dominated as late as 7800 yr bp at higher elevations (1006 m, south-facing site) on the south side of Ord Mountain, as indicated by the Neotoma records of King (1 976). This juniper is now absent on Ord Mountain, which lacks woodland despite its substantially high elevation (to 1920 m), and Larrea ascends the mountain from all sides. The oldest records of Larrea are from the much lower elevations and latitudes of the more subtropical Sonoran Desert. The Larrea- dominated Neotoma deposits from the low Wellton Hills (at 162 m), in southwestern Yuma County, Arizona, have been dated to at least 10,580 yr bp, but this does not establish the earliest possible occurrence of Larrea at elevations approaching sea level (see review in Wells and Hunziker 1976). It should be emphasized that these early, low-elevation records of Larrea are about 4° south of the substantially higher Death Valley Neotoma sites reported here. A time-transgressive development of the Larrea scrub zone in the Sonoran and Mohave Deserts is apparent. In Death Valley, two very late Holocene Neotoma middens dated to 900 yr bp (very close to the 425-m site) and 1990 yr bp (at 260 m) serve as a control on the unusual composition of pure Ambrosia at 10,230 yr bp. Both of the very recent deposits contain Larrea and Ambrosia in subequal amounts (Table 1). Thus, there is no indication of Neotoma dis- crimination against Larrea. Although the timing of arrival for Larrea is yet to be established in Death Valley, it seems clear that it was preceded by its ubiquitous modem associate in the Mohave, Am- brosia dumosa. Late-Glacial/Holocene Climatic Change in Death Valley With regard to the magnitude and nature of late Quaternary cli- matic change in Death Valley, the evidence presented here sheds some light on the perennial temperature vs. precipitation contro- versy (Brakenridge 1978, Wells 1979). A pluvial increase in precip- itation is apparent from the present moisture requirements of Ju- niperus osteosperma and its 1 200-1 500-m displacement as the dominant species of the late-Pleistocene woodlands in Death Valley. This juniper is a relatively xerophytic species, descending at its lower limits to semidesert of Artemisia tridentata, Atriplex confertifolia, or Coleogyne ramosissima; but rarely does /. osteosperma descend into the hot desert of Larrea and Ambrosia. Juniperus osteosperma forms woodlands (even in the cooler, northern Great Basin) only 20 MADRONO [Vol. 32 Fig. 3. Distribution of yuccas of the Yucca whipplei group (sect. Hesperoyucca). Modem records of Y. whipplei are in southern Cahfomia and Baja California; the closely related (if not conspecific) Y. newberryi is restricted to northwestern Arizona; documented occurrences indicated by dots (after McKelvey 1938, Hastings et al. 1972). Late Pleistocene macrofossil records of yuccas with the highly distinctive leaf morphology of the Yucca whipplei group are indicated by "A"; occurrences at very low elevations indicate continuity of range then. under mean annual precipitation of 200 mm or more (Beeson 1974). Today, much of Death Valley below 775 m receives 75-100 mm or less of precipitation per year; and there are frequent years with 25 mm or less. The late Pleistocene (13,060 yr bp) record of juniper woodland at 775 m in Death Valley suggests that the precipitation then was about twofold greater than today and probably less variable. 1 985] WELLS AND WOODCOCK: PLEISTOCENE VEGETATION 2 1 The late-Pleistocene climate was undoubtedly cooler than today; but limits to cooling are set by the pleniglacial (19,500 yr bp) dom- inance of Yucca whipplei at the lowest site (425 m) and late-glacial (1 1,600 yr bp) occurrences of the same yucca as high as 1280 m. At present, Y. whipplei appears to be absent in the Death Valley region, despite a large available elevational gradient of over 3000 m on the adjacent Panamint Range (Telescope Peak: 3368 m). It is now widely distributed in chaparral vegetation under the mild, Mediterranean isoclime (winter rain, summer drought) of cis-montane California. But Y. whipplei (the polycarpic subsp. caespitosa) extends to the extreme western margins of the Mohave Desert, where there is a transition from woodland/chaparral to high desert under a similar climatic rhythm (Fig. 3). It (chiefly as the polycarpic subsp. eremica) occurs also in the relatively cool, foggy desert of northern Baja Cal- ifornia. There is one outstanding exception to this restriction of Y. whipplei to the Mediterranean isoclime of the Pacific slope: an aberrant, widely disjunct population in the lower, western end of the Grand Canyon, Arizona under a hot-desert climate (Fig. 3). It differs in having reduced placental wings and has been regarded as a distinct species, Y. newberryi McKelvey. Furthermore, the Grand Canyon population appears to be entirely monocarpic (McKelvey 1938). The low-elevation Neotoma records from Death Valley and the lower Colorado River valley (Wells and Hunziker 1976) document a much wider and more continuous distribution of yuccas of Y. whipplei affinity (Hesperoyucca) as recently as the last (Wisconsinan) glacial of the Pleistocene (Fig. 3). The climatic characteristics of regions with modem populations of Yucca whipplei include very mild winter temperatures and substantial winter rain; and for most populations, summer temperatures are only mildly hot (relative to Death Valley). Thus, the abundance of Y. whipplei under a full-glacial (19,500 yr bp) climate in Death Valley is an indication of cool but relatively mild winters then, coupled with greater precipitation; summers were almost certainly cooler then. A cold, dry pleniglacial climate (Bra- kenridge 1978) in Death Valley, cold enough to account for juniper woodland 1200-1500 m below its present lower limits, would have been too cold for Yucca whipplei. A cool and equable, pleniglacial climate (Wells 1979) with greater precipitation better explains the present evidence. Yucca whipplei may have disappeared in the Death Valley region because montane elevations moist enough for its sur- vival are too cold in winter; and the area of mild winter climate near the valley ffoor now becomes too hot and dry in summer. At the close of the last (Wisconsinan) glacial, the climate of Death Valley suffered a drastic but gradual decline in equability and pre- cipitation during the interval 1 1 ,000-9000 yr bp, giving rise to the extremely hot summers and limited precipitation of the modern 22 MADRONO [Vol. 32 desert. The macrofossil evidence indicates that the desertification process began on the floor of Death Valley (already a Yucca semi- desert during the last glacial) and spread gradually up the slopes of adjacent mountains in the wake of the shrinking juniper and Joshua tree woodlands. Acknowledgments Research support from the National Science Foundation (DEB-78-1 1 87 to P.V.W.) is gratefully acknowledged. Literature Cited Beeson, C. D. 1 974. The distribution and synecology of Great Basin pinyon-juniper. M.S. thesis, Univ. Nevada, Reno. Brakenridge, G. R. 1978. Evidence for a cold, dry, full-glacial climate in the American Southwest. Quaternary Res. 9:22-40. Hastings, J. R., R. M. Turner, and D. K. Warren. 1972. An atlas of some plant distributions in the Sonoran Desert. Technical Reports on the Meteorology and Climatology of Arid Regions No. 2 1 , Univ. Arizona, Tucson. HuBBS, C. L. and R. R. Miller. 1948. The Great Basin, with emphasis on glacial and postglacial times. The zoological evidence. Univ. Utah Bull. 38:18-144. Hunt, C. B. and D. R. Mabey. 1966. Stratigraphy and structure. Death Valley, CaHfomia. U.S. Geol. Surv. Prof. Paper 494-A. King, T. J. 1976. Late Pleistocene-early Holocene history of coniferous woodlands in the Lucerne Valley region, Mohave Desert, California. Great Basin Naturalist 36:227-238. McKelvey, S. D. 1 938. Yuccas of the Southwestern United States: Part One. Arnold Arboretum, Jamaica Plain, Mass. Mehringer, P. J. 1967. Pollen analysis of the Tule Springs area, Nevada. Nevada State Museum, Anthropological Papers 13:129-200. RoosMA, A. 1958. A climatic record from Searles Lake, California. Science 128: 716. Van Devender, T. R. 1977. Holocene woodlands in the southwestern deserts. Science 198:189-192. and W. G. Spaulding. 1979. Development of vegetation and climate in the southwestern United States. Science 204:701-710. Wells, P. V. 1 966. Late Pleistocene vegetation and degree of pluvial climatic change in the Chihuahuan Desert. Science 153:970-975. . 1969. Preuves paleontologiques d'une vegetation tardi-Pleistocene (datee par le "'C) dans les regions aujourd'hui desertiques d'Amerique du Nord. Rev. Geogr. Phys. Geol. Dynam. 11:335-340. . 1974. Postglacial origin of the present Chihuahuan Desert less than 11,500 years ago. In R. H. Wauer and D. H. Riskind, eds.. Transactions of Symposium on the Biological Resources of the Chihuahuan Desert Region, pp. 67-83. U.S. Govt. Print. Office, Wash., D.C. . 1976. Macrofossil analysis of wood rat (Neotoma) middens as a key to the Quaternary vegetational history of arid America. Quaternary Res. 6:223-248. . 1979. An equable glaciopluvial in the West: pleniglacial evidence of in- creased precipitation on a gradient from the Great Basin to the Sonoran and Chihuahuan Deserts. Quaternary Res. 12:311-325. . 1983. Paleobiogeography of montane islands in the Great Basin since the last glaciopluvial. Ecol. Monogr. 53:341-382. and R. Berger. 1967. Late Pleistocene history of coniferous woodland in the Mohave Desert. Science 155:1640-1647. 1985] WELLS AND WOODCOCK: PLEISTOCENE VEGETATION 23 and J. H. Hunziker. 1976. Origin of the creosote bush (Larrea) deserts of southwestern North America. Ann. Missouri Bot. Gard. 63:843-861. and C. D. Jorgensen. 1964. Pleistocene wood rat middens and climatic change in Mohave Desert: a record of juniper woodlands. Science 143:1171- 1174. (Received 9 June 1983; accepted 25 July 1984.) ANNOUNCEMENT To Madrono authors: The California Botanical Society Council has adopted a new policy of assessing page allotments. Co-authors may now split the total page number, rather than each author's being charged with the full page total. Thus, co-authors of a 20-page paper will each be debited for 10 pages. In addition. Madrono will now give Society members 10 free pages each year or 20 pages in a two-year period. This policy will begin with volume 32, January 1985, and the slate will be cleaned. Thus, beginning on 1 January 1985, all members of the Society will be eligible to publish a 10-page paper without an editorial fee. On 1 January 1986, those who have not used any of this allowance will be able to publish a 20-page paper without a fee. Pages published in 1985 will be deducted from this free allotment of 20. An author publishing a 20-page paper in 1986 would be eligible to publish a 10-page paper in 1987, but would not be able to publish a 20-page paper free until 1988. In other words, page allotments are granted on an annual basis and may be accumulated for two years only. The cycle begins in the first full year of membership. Sale of Surplus Copies of Madrono A certain number of surplus issues have been set aside for sale at half price, first come, first served. Available are full sets of volume 1 and some numbers in volumes 2, 1 1-21, and 23 (including vol. 1 1, no. 2 and vol. 20, no. 3). Write to the corre- sponding secretary for a complete list of available issues and prices. Botanical Systematics Symposium The Rancho Santa Ana Botanic Garden will sponsor and host a symposium on Trends in Systematic and Evolutionary Botany on 25-26 May 1985, at the Garden in Claremont, California. The purpose of this symposium is to examine current trends and, if possible, suggest and identify promising trends in systematic and evolutionary botany in the coming decade. Invited papers will be presented on pollination biology (H. Baker), chemical systematics (T. Swain), morphology (J. Skvarla), cladistics (M. Donoghue), physiological ecology (P. Rundel), aspects of modem floristics and tra- ditional systematics (G. Prance), and research in botanical gardens (P. H. Raven). Low-cost housing will be available at nearby Pomona College. Attendance will be limited. For further information write: Systematics Symposium, Rancho Santa Ana Botanic Garden, 1 500 North College Avenue, Claremont, California 91711. This will be the first of an annual series of symposia planned at the Rancho Santa Ana Botanic Garden and is intended to provide a forum primarily for botanists in the southwestern United States and adjacent regions of Mexico. A CYTOTAXONOMIC CONTRIBUTION TO THE WESTERN NORTH AMERICAN ROSACEOUS FLORA E. DuRANT McArthur and Stewart C. Sanderson USDA Forest Service, Intermountain Forest and Range Experiment Station, Shrub Sciences Laboratory, Provo, UT 84601 Abstract New chromosome counts are reported for Cercocarpus montanus Raf. var. mon- tanus and C. ledifolius Nutt. (2 varieties), x = n = 9\ Chamaebatia australis (Brandg.) Abrams and C.foliolosa Benth., x = « = 9; Chamaebatiaria millifolium (Torr.) Max- im., x = n = 9; Coleogyne ramosissima Torr., x = n = Holodiscus dumosus (Nutt.) Heller, 2x = n= 18; Kelseya unijlora (Wats.) Rydb., x = « = 18; Peraphyllum ra- mosissimum Nutt., x= n = 17; and Petrophytum caespitosum (Nutt.) Rydb., x = n = 18. The counts are related to rosaceous subfamilial taxonomy. All taxa are relatively narrowly (geographically or ecotypically) distributed endemics. Cercocarpus and Hol- odiscus are the most wide ranging taxa. Rosaceae, consisting of four subfamilies (Rosoideae, Prunoideae, Spiraeoideae, and Pomoideae), is pandemic but most common in the northern temperate region— especially western North America and eastern Asia (Raven and Axelrod 1974, Robertson 1974, Gold- blatt 1976, Jones and Luchsinger 1979). In this paper we report chromosome counts for 10 species in 8 genera for three of the subfamilies. These counts are first reports for three varieties, nine species, and seven genera. Base chromosome numbers of the four subfamilies of Rosaceae are summarized in Table 1 . Materials and Methods Chromosome counts were made from root tips of germinated seedlings or from plants brought from the field to a mist bench and stimulated to produce additional roots; or from pollen mother cells (PMCs). Fixation was in 5% acetic acid or 1 N HCl. Roots were hydrolyzed in 1 N HCl for four hours at room temperature. Staining was by iron acetocarmine, air-evaporated on the slide to increase its concentration. Preparations were mounted in Hoyer's medium and examined microscopically under light field (meiosis) or phase contrast (mitosis) (Sanderson et al., in review). Counts were made from several plants (2-5) for each sampled population. Voucher herbarium specimens of representative samples were placed in the herbarium at the Intermountain Station's Shrub Sciences Laboratory (SSLP). Madrono, Vol. 32, No. 1, pp. 24-28, 15 February 1985 1985] McARTHUR AND SANDERSON: CYTOTAX., ROSACEAE 25 Table 1. Subfamilies of Rosaceae with Characteristic Chromosome Numbers AND Generic Examples. References: Omcluff(1967), Federov (1969), Moore (1973, 1974,1 977), Robertson ( 1 974), Goldblatt (1976,1981), McArthur et al. ( 1 983), Baker et al. (1984), and Table 2. *New generic counts. Table 2. Subfamily Com- mon x's Other x's Representative genera (x) Rosoideae 7,9 8, 14 Fragaria, Geum, Potentilla, Rosa, Rubus, Sanguisorba (7); Alchemilla, Coleogyne* (8); Adenostoma, Cercocarpus, Chamaeba- tia*, Cowania, Dryas, Purshia (9); Fallugia (14). Prunoideae 8 Exochorda, Oemleria, Prunus (8). Spiraeoideae 9 7 Physocarpus, Spiraea, Aruncus (7, 9); Cha- maebatiaria*, Holodiscus*, Kelseya*, Petro- phytum*, Luetkea (9). Pomoideae 17 14, 15 Quillaja (14); Vauquelinia (15); Amelanchier, Cotoneaster, Crataegus, Mains, Peraphyl- lum*, Pyracantha, Pyrus, Sorbus (17). Results and Discussion The counts reported in Table 2 reveal no major surprises. Each corresponds to a base number consistent with previous records in its subfamily, unlike the new base numbers recently reported for Fallugia {x= 14) (Rosoideae) (McArthur et al. 1983, Baker et al., 1984) and for Quillaja {x = 14) and Vauquelinia {x = 15) (Po- moideae) (Goldblatt 1976). Each generic count is new except for Cercocarpus, which had been reported x = « = 9 for one population of C betuloides Nutt. (Morley 1949)— best referred to as C. montanus Raf. var. glaber (Wats.) F. L. Martin (Martin 1950). We report original counts for the typical variety of C. montanus Raf. and two varieties of C ledifolius Nutt. Both species of Chamaebatia (Rosoideae) and its namesake genus Chamaebatiaria (Spiraeoideae) are x = n = 9. Perhaps our most interesting new count is that of Coleogyne ra- mosissima Torr. (Rosoideae), a relictual endemic from the Mohave- Cold Desert ecotone (Bowns and West 1976). This species had x = n = S chromosomes. This is only the second unequivocal x = S count for Rosoideae (Table 2; Robertson 1974). Coleogyne is quite different from most other Rosoideae (and Rosaceae) in having 4- rather than 5-merous flowers. Alchemilla, the others = 8 Rosoideae genus, is also 4-merous. Holodiscus dumosus (Nutt.) Heller, 2x = n = 18, was the lone new count we uncovered above the diploid basic number. Higher base numbers are not unexpected among the Rosaceae, where polyploidy is quite common in some groups (references cited for Table 1 ). 26 MADRONO [Vol. 32 Table 2. Chromosome Counts of Rosaceous Shrubs. *First report for taxon. Taxa Location, collection number Chromo- some count Cercocarpus ledifoHus Nutt. var. intercedens C. K. Schneider Cercocarpus ledifolius Nutt. var. ledifolius Cercocarpus montanus Raf. var. montanus Chamaebatia australis (Brandg.) Abrams Chamaebatia foliolosa Benth. Chamaebatiaria millifolium (Torr.) Maxim. Coleogyne ramosissima Holodiscus dumosus (Nutt.) Heller Kelseya uniflora (Wats.) Rydb. Peraphyllum ramosissimum Nutt. Big Horn Mts., near Dayton, Sher- idan Co., WY, 1230 m, S. B. Monsen s.n., Jun 1982. Near Adin, Lassen Co., CA, 1370 m, J. A. Young s.n. Near Soda Springs, Caribou Co., ID, 2120 m,J. N. Davis AB 849. Near Verdi, Washoe Co., NV, 1 520 m, J. A. Young s.n. Mineral Mountain Pass, Beaver Co., UT, 2050 m, /. A^. Davis S- 12. Near Cedar City, Iron Co., UT, 2100 m, W. R. Stewart s.n. (U 23). Weber Canyon, Morgan Co., UT, 2000 m, J. N. Davis AB 681. Salt Creek Canyon, Juab Co., UT, 1800 m, M. Black s.n., 6 Jun 1961 (U 8). Tecate Peak, San Diego Co., CA, 760 m, M. Kottman s.n., 20 Dec 1982. Near Auburn, Placer Co., CA, 500 m, McArthur 1361. Marysvale Canyon, Sevier Co., UT, 1800 m, McArthur and Sanderson 1339 (V 1). Ophir Canyon, Oquirrh Mtns., Tooele Co., UT, 2000 m, Mc- Arthur 1370. Tobin Wash, Washington Co., UT, 1250 m, T. B. Moore s.n., 29 Jun 1978 (U 4). Motoqua, Washington Co., UT, 1250 m, J. E. Bowns s.n. Chalk Cr., Pavant Range, Millard Co., UT, 1860 m, S. Goodrich 16897. Pass Canyon, Lost River Range, Custer Co., ID, 1840 m, San- derson 1367. Near Monticello, San Juan Co., UT, 2130 m, W. R. Stewart s.n., 18 Aug 1965 (U 9). 2n = 2n = 2n = 2n = 2n = 2n = 2n = 2n = n = 9* 2n = 18* n = 9* n = 9 2n= 16* 2n= 16 n= 18* 2n= 18* 2n = 34* 1985] McARTHUR AND SANDERSON: CYTOTAX., ROSACEAE 27 Table 2. Continued. Chromo- some Taxa Location, collection number count Petrophytum caespitosum Rock Canyon, Wasatch Range, n = 9* (Nutt.) Rydb. Utah Co., UT, 1700 m, Mc- Arthur and Sanderson 1365. Pass Canyon, Lost River Range, n = 9 Custer Co., ID, 1840 m, San- derson 1366. Valley Mts., near Gunnison, San- n = 9 pete Co., UT, 1920 m, 5. Good- rich s. n. Kelseya uniflora (Wats.) Rydb. and Petrophytum caespitosum (Nutt.) Rydb. (Spiraeoideae) are restricted western North American endemics; both have x = n = 9 chromosomes. Peraphyllum ramo- sissimum Nutt. (Pomoideae), in common with the mainstream members of its subfamily, has x = n = 17 chromosomes. We sug- gested earlier (McArthur et al. 1983) that the shrubby x = 9 Ro- soideae {Adenostoma, Cercocarpus, Chamaebatia, Cowania, Dryas, Purshia) of western North America might be closely allied with the shrubby x = 9 Spiraeoideae {Chamaebatiaria, Holodiscus, Kelseya, Luetkea, Petrophytum, Physocarpus, Spiraea) of the same area. Both groups are characterized by high endemism and monotypicism and by sclerophyllous or microphyllous leaf habit. These six Rosoideae may have more in common with the seven Spiraeoideae than with mesic, x = 1 Rosoideae, such as Fragaria, Potentilla, Rosa, Rubus, and Sanguisorba. For example, Chamaebatia of Rosoideae and Chamaebatiaria of Spiraeoideae resemble one another closely in their unusual leaf form, although they differ in fruit type. However, one unifying factor for the x = 9 Rosoideae is presence of actino- mycete root nodulation (Klemmedson 1979, Nelson 1983). Spi- raeoideae are not nodulated. We also point out that Physocarpus and Spiraea have x = 1 members in Asia in addition to their more common x = 9 North American and Asian taxa (McArthur et al. 1983). Acknowledgments This work was facilitated by funds from the Pittman Robertson Wildlife Restoration Project W-82-R. We thank J. E. Bowns, J. N. Davis, S. Goodrich, M. Kottman, S. B. Monsen, and H. C. Stutz for plant materials and stimulating discussion. Literature Cited Baker, M. A., D. J. PiNKAVA, B. Parrtt, and T. J. RiGHETTi. 1984. On Cowania and its intergeneric hybrids in Arizona. Great Basin Naturalist 44:484-486. 28 MADRONO [Vol. 32 BowNS, J. E. and N. E. West. 1976. Blackbrush {Coleogyne ramosissima Torr.) on southwestern Utah rangelands. Utah Agric. Exp. Sta. Res. Report 27, Logan, UT. Federov, a. a. 1969. Chromosome numbers of flowering plants. Izdateyur Nauk, Leningrad. GoLDBLATT, P. 1976. Cytotaxonomic studies in the tribe Quillajeae (Rosaceae). Ann. Missouri Bot. Card. 63:200-206. , ed. 1981. Index to plant chromosome number 1975-1978. Monogr. Syst. Bot. Missouri Bot. Gard. No. 5. St. Louis. Jones, S. B., Jr. and A. E. Luchsinger. 1979. Plant systematics. McGraw-Hill Book Co., NY. Klemmedson, J. O. 1979. Ecological importance of actinomycete-nodulated plants in the western United States. Bot. Gaz. (Suppl.) S91-S96. Martin, F. L. 1950. A revision of Cercocarpus. Brittonia 7:91-1 11. McArthur, E. D., H. C. Stutz, and S. C. Sanderson. 1983. Taxonomy, distri- bution and cytogenetics of Purshia, Cowania, and Fallugia (Rosoideae, Rosa- ceae). In A. R. Tiedemann and K. L. Johnson, compilers. Proceedings— research and management of bitterbrush and cliffrose in western North America, pp. 4- 24. USDA Forest Service, Gen. Techn. Report INT- 152, Ogden, UT. Moore, R. J., ed. 1973. Index to plant chromosome numbers 1967-71. Regnum Vegetabile, vol. 90. Utrecht. . 1974. Index to plant chromosome numbers 1972. Regnum Vegetabile, vol. 9 1 . Utrecht. . 1977. Index to plant chromosome numbers 1973-1974. Regnum Vege- tabile, vol. 96. Utrecht. MoRLEY, T. 1949. In Documented chromosome numbers of plants. Madrono 10: 95. Nelson, D. L. 1983. Occurrence and nature of actinorhizae on Cowania stansbu- riana and other Rosaceae. In A. R. Tiedemann and K. L. Johnson, techn. com- pilers. Proceedings— research and management of bitterbrush and cliffrose in western North America, pp. 225-239. USDA Forest Service Gen. Techn. Report INT- 152, Ogden, UT. Ornduff, R., ed. 1967. Index to plant chromosome numbers 1956. Regnum Ve- getabile, Vol. 50. Utrecht. Raven, P. H. and D. I. Axelrod. 1974. Angiosperm biogeography and past con- tinental movements. Ann. Missouri Bot. Gard. 61:539-673. Robertson, K. R. 1 974. The genera of Rosaceae in the southeastern United States. J. Arnold Arbor. 55:303-332, 344-401, 611-662. (Received 12 March 1984; accepted 26 July 1984.) ANNOUNCEMENT All-expense-paid Trip to Berkeley Yes, people desiring one or even two free trips each year to Berkeley, California may apply for the position of editor, MADRONO, A West American Journal of Botany. The president of the California Botanical Society is accepting applications at this time. Qualifications include 3-5 hrs per day of free time during Dec, Mar., June, and Sept. (work load varies during other months); proficiency in English; and a very durable red pencil. Office expenses incurred while editing MADRONO will be reimbursed by the Society. I have found the experience of working on MADRONO educational and highly rewarding, and I recommend the job to anyone who is interested.— C. Davidson SEROTINY AND CONE-HABIT VARIATION IN POPULATIONS OF PINUS COULTERI (PIN ACE AE) IN THE SOUTHERN COAST RANGES OF CALIFORNIA Mark Borchert Los Padres National Forest, Goleta, CA 931 17 Abstract The cone habit of Pinus coulteri exhibits considerable variation among plant com- munities in the southern Coast Ranges of California. Serotiny is prevalent in P. coulteri/ chaparral, P. coulteri/ Quercus chrysolepis, and P. coulteri/ Cupressus sargentii communities that are periodically burned by wildfire. The bulk of pine regeneration in these types occurs in the first year after a fire, after which it rapidly declines, and ceases within 20 years. By contrast, in the P. coulteri/ Q. agrifolia community nearly all cones open at maturity or soon thereafter. Pines in this habitat are generally less subjected to fire-caused mortality; regeneration, although sporadic, is continuous. The quantity of stored viable seed is reduced in all community types by animal depredations and varying degrees of spontaneous cone opening. Despite these losses, the amount of stored, viable seed retained in serotinous stands is 50 times greater than quantities stored in nonserotinous stands. Three closely related species of the genus Pinus, P. torreyana, P. sabiniana, and P. coulteri, constituting the subsection Sabinianae, show tendencies toward seed retention (McMaster and Zedler 198 1; Critchfield, pers. comm.). Pinus torreyana retains seed for up to 15 years in cones that open gradually over time (McMaster and Zedler 1981). Minnich (1980) suggested that the often abundant regener- ation of P. coulteri following wildfire originates from seed stored in partially open cones and maturing cones persistent on fire-killed trees. Despite these tendencies, however, serotiny, defined as the retention of mature seeds in closed cones, is not known to occur in this group. In this study I describe serotiny, cone-habit variation in relation to plant community types, and factors that influence seed retention in P. coulteri growing in the southern Santa Lucia and La Panza Ranges, part of the southern Coast Ranges of California. I also discuss the possible adaptive value of canopy-stored seed in fire- prone habitats and management implications of seed storage and other life history traits of P. coulteri. The Study Area The study area includes a highly fragmented P. coulteri distri- bution in part of the Los Padres National Forest (Fig. 1). The tree ranges in elevation between 610 m and 1525 m, and average annual Madrono, Vol. 32, No. 1, pp. 29-48, 15 February 1985 30 MADRONO [Vol. 32 precipitation generally exceeds 450 mm. Stands grow on a variety of soil types (Table 1), but gravelly loams derived from marine bedrock parent material are the most common. The climate of the region is Mediterranean. At lowland localities like Santa Barbara (36 m) and San Luis Obispo (60 m), most pre- cipitation falls from November-April. Mean monthly precipitation and temperature range from 30-100 mm and 1 1-1 5°C, respectively. May-October is dry and warm: mean monthly precipitation and temperature for this period vary from 0.5-10 mm and 16-20°C, respectively. Data from Santa Ynez Peak (1310 m) and Figueroa Mountain (FM; 960 m) (Fig. 1) indicate that precipitation at higher elevations extends into May and ranges from 30-186 mm for the six month period. Wildfires are very frequent in the region. Since 1912, major fires (>4000 ha) have burned on the average of once every three years, 1985] BORCHERT: SEROTINY IN PINUS COULTERI 31 2 ^ o ® o In a 6 9 c« y 9i C/3 q rr ON O — «' 6 ^ go 00 *:5 — (u t« C N c N 11 ll o # o in IT) o i i I r*^ rn r~ CO o c o cd r ■3 I ■§ s OS O) C/2 e o ^ O PQ 60%) north-facing aspects. Toxicodendron diver- silobum is the only common understory species, but its cover is low (Campbell 1980). The P. coulteri/ C. sargentii forest grows at a single locality in the area (CRBA) (Fig. 1). Study Methods Twelve stands were sampled from most parts of the P. coulteri distribution in the study area (Fig. 1). The stands were selected subjectively to represent a variety of community types of differing ages. Each stand had to be native (i.e., unplanted), relatively acces- sible, undisturbed, and homogeneous in vegetation. It is estimated that at least 50% of the major P. coulteri stands are represented. The CRBA stand is anomalous because it straddled a sheltered fuel- break. However, there were no signs of pine cutting in this stand, although C sargentii was thinned. Elevation, slope angle, and aspect were recorded at each site. Soil classification was taken from a third order soils map (U.S. Forest Service data on file on file at the Supervisor's Office, Goleta). De- pending on the size and density of the stand, tree density was esti- mated with circular plots or by the plotless point-center-quarter method (Cottam and Curtis 1956). Small (<0.75 ha) or relatively dense stands were sampled with plots varying in size from 0.05-0.2 ha. Each plot was made large enough to include 30-60 pines >2.5 1985] BORCHERT: SEROTINY IN PINUS COULTERI 33 cm in diameter measured at 1.4 m (dbh). Larger sample sizes were necessary to characterize multi-aged stands. In low-density stands larger than one hectare (SCRE, ZL), the point-center-quarter method was employed. In each stand a total of 12-15 points was sampled along 2-3 transects oriented parallel to the contours. Tree density estimates using this method had standard errors less than 10% of their mean values. All P. coulteri seedlings (<2.5 cm dbh), saplings (2.5 cm- 10 cm dbh) and trees (> 10 cm dbh) were counted in plots and at sampling points on the transects. Fewer than 5 seedlings were encountered in each of the WH, ZL, and LPM stands. Stems >5 cm were cored at 60 cm to determine age. In addition, two to four stems were cut at ground level in several stands to estimate the number of years re- quired to reach 60 cm; as a result, four years were added to the core- height age of trees in all stands. No attempt was made to estimate the mean and variance in the age required to reach 60 cm in each stand. All other tree species >2.5 cm dbh were counted and mea- sured. Shrub cover was visually estimated in 2.3 m^ circular subplots taken at each sampling point along transects or at 8-15 locations within the plots. Total herbaceous cover also was recorded for each subplot and the dominant species noted. Open and closed cones were censused and aged with binoculars. Since over 95% of cones in 15- to 50-year trees are borne on the bole, cone age can be determined indirectly by counting the number of branch whorls. Generally, trees in this age class produce whorls annually. Nevertheless, in order to verify the accuracy of whorl counts, one to two branches were cut from a sample of 5-15 trees per site, and the branch ages compared to their corresponding whorl counts. The two were highly correlated (r = 0.948, p < 0.01, n = 14, SCRW; r = 0.974, p < 0.01, n = 31, SCRE; r = 0.947, p < 0.01, n = 19, CRBA). Branch cones in trees older than 50 years could not be aged ac- curately using binoculars because branch growth had slowed enough to prevent distinction between recent cones, which remain closed during the normal maturation period, and those that remained closed longer. Accurate aging could only have been accomplished by count- ing bud-scale scars, which would have required tree felling or branch cutting. Thus, only the number of open and closed cones was re- corded in these trees. All stand samples had at least 10 trees 15-50 years old. To simplify analysis of cone age data, cones were grouped into one-year age class intervals. Thus, cones 0-12 months are in the first age class, 13-24 months in the second, and so on. Cone pro- duction is assumed to have commenced on 1 June, when pollination first was observed in the SCR and PM stands. 34 MADRONO [Vol. 32 A total of 25 1 closed cones, varying in age from 16 months to 24 years, was gathered from trees in PM and SCRE stands: 156 cones from 26 trees in the SCRE stand, and 95 cones from 25 trees in the PM stand. Cones were selected from trees that appeared visually to have an adequate number of cones in certain size classes. Cones remaining closed for more than two years after pollination are termed serotinous in this study. Each cone was first measured and examined for external signs of animal damage, and then immersed in boiling water for 60-75 seconds to melt the resinous bonding material that seals the scales together. Once the scales had separated, the seeds were hand-extracted. In addition, residual seeds were removed from 28 open cones (5-17 years) gathered from the SCRE stand. Empty seeds were removed by floating in water. Germination trials were conducted on a random sample of 20 filled seeds from each cone. Because of the small number of filled seeds in each open cone, they were combined into one lot from which three samples were tested. After soaking in an aerated water bath for 24 hours, each sample was dipped in a captan solution and sown on a moist layer of cotton in a petri dish. Following a strati- fication period of 60 days at 10°C in continuous light, the light- temperature regime of the samples was altered to 8 hours of light at 25°C and 16 hours of darkness at 15°C. A seed was considered germinated if it produced a radicle at least 3 mm long after 14 days. The small sample size of cones in some age classes required the grouping of germination results for analysis. A description of cone maturation was constructed from small samples of cones (4-10 at each sample age) collected 12, 14, 16, 18, and 22 months after pollination (1 June) in the LPM and SCRE stands, and 16-month cones in the PM stand. Results Age structure and establishment. Stand age data indicate two gen- eral types of age structures: unimodal, bell-shaped distributions in- dicative of even-aged stands, and pulsed distributions suggesting irregular episodes of successful regeneration (Figs. 2-3). Most P. coulteri stands growing in association with chaparral (SCRE, SCRW, PM), Q. chrysolepis (MC, TPK), or C. sargentii (CRBA) are characterized by an even-aged structure that results from a stand-killing wildfire. Evidence corroborating this explanation is provided by Minnich (1978, 1980), who noted high levels of fire- caused pine mortality in both P. coulteri/ chaparral (90%) and P. coulteri/ Q. chrysolepis (81%) types in the eastern Transverse Ranges. Comparable pine losses probably occur in C sargentii thickets, which are prone to crown fires (Vogl et al. 1977). The PMS stand is anomalous. It is an old-aged pine stand in 1985] BORCHERT: SEROTINY IN PINUS COULTERI 35 SCR 0.3- 0.2- 0.1 - TPK PM 20 20 40 0.4 0.3- 0.2 0.1 40 20 20 40 60 MC ' 1 — ' n ^ 40 60 CRBA 20 0.4 0.3 0.2 0.1- 40 BPBR 20 40 60 PMS n — I — 120 80 100 140 AGE Fig. 2. Age-frequency histograms for the P. coulteri/ chaparral (BPBR, PM, PMS, and SCR); P. coulteri/Q. chrysolepis (MC, TPK); and P. coulteri/C. sargentii (CRBA) stands. The SCRE and SCW are nearly identical and have been combined into SCR. 36 MADRONO [Vol. 32 chaparral with several episodes of pine establishment. The oldest cohort likely originated after a fire in the 1 830s. Another documented fire swept the area in 1921, but evidently spared most of the 90- year-old trees. Fire scar analyses of 140-year-old trees indicate that a fire that burned an adjacent area in 1939 also spread to the PMS site. The isolated 80- and 100-year-old trees suggest that chaparral stands relatively well-protected from stand-converting fires may ex- hibit pulsed regeneration at an advanced age. In a discussion of 40- year-old, south-slope, montane chaparral of the type often associated with P. coulteri, Hanes (1971) noted the formation of openings in the canopy resulting from shrub mortality. These openings were colonized by Salvia mellifera seedlings. Similarly, P. coulteri seed- lings may establish in these sites. My observations of canopy open- ings in comparatively young stands (<60 years), however, suggest that such establishment is probably rare and most likely occurs on the periphery of the stand. Complete or nearly complete stand destruction is followed within a year by a burst of pine regeneration (Minnich 1977, Griffin 1982). Seeds fall from heat-opened, maturing cones, residual seed in open cones, and older closed cones stored in the canopy. Age-frequency histograms (Fig. 2) show the appearance of new trees consistently later than the first year after the fire. To a large extent, this discrep- ancy can be attributed to the variable growth period necessary for seedlings to reach the core height. In dense stands, for example, numerous seedlings do not reach 60 cm even after 6 years (Griffin, pers. comm.). Inadequate precipitation and poor site conditions may also retard seedling growth. As vegetative cover increases with stand age, tree establishment declines and eventually ceases. Despite the variety of species asso- ciated with P. coulteri among the even-aged stands, the length of establishment, as determined by the age difference between the oldest and youngest trees in each stand, was remarkably constant, averaging 23 ± 1.86 years (n = 7). The actual estabHshment period is probably around 20 years, since very late appearing trees would likely require more time to reach the core height. Vale (1979) found that successful regeneration of P. coulteri in a chaparral stand on Mt. Diablo spanned 19 years. Although first-year seedlings originate from fire-released, cone- stored seed, later pine regeneration probably developed from other sources. These include: (a) occasional trees that survived fires. Some trees, especially those on ridgetops, are undamaged by fire and con- tinue to supply seed to burned areas for decades. Such survivors were present in both the SCR and TPK stands; (b) singed, but un- opened cones on fire-killed trees. After a recent fire (1981) in a P. coulteri/Q. agrifolia stand, some large, heavily scorched pines had unbumed or lightly singed cones on the tips of high branches. Con- 1985] BORCHERT: SEROTINY IN PINUS COULTERI 37 Table 2. Cone and Seed Characteristics of the PM and SCRE Stands. All attributes are significantly different (p < 0.05, one-way ANOVA) between the two stands, except full seeds/undamaged cone. Cone size is length x maximum width. Means are presented ± one standard error. Cone and seed characteristics Study site PM n SCRE n Cone size 209.3 ± 4.2 133 159.0 ± 5.3 44 Seed length (cm) 1.43 ± 0.02 75 1.26 ± 0.03 75 Seed weight (gm) 0.35 ± 0.01 2420 0.29 ± 0.01 1780 Full seeds/undamaged closed cone 150.5 ± 5.6 111 151.3 ± 5.9 88 Percent empty seeds/ closed cone 15.7 ± 1.3 134 9.8 ± 1.1 82 Full seeds/open cone 2.2 ± 0.4 28 no data ceivably, these cones might open slowly enough to furnish viable seed to the area for several years after the fire; (c) early reproducing trees of the immediate post-bum cohort. Reproduction can occur in trees as young as 10 years (Minnich 1980). Seed released from precociously reproducing trees appeared to be the likely source of trees where no obvious survivors were observed (PM, CRBA, MC); and (d) residual seed trapped in the basal scales of open cones (Table 2). Irregular pulses of pine regeneration characterize the P. coulteri/ Q. agrifolia stands (Fig. 3). Some even-aged cohorts date back to fires, whereas others probably coincide with years of good precipi- tation and seed production. Typically, small groups of pine seedlings appear in grassy openings or at the edge of the oak canopy. Saplings are often spindly, and some heavily shaded individuals die. Never- theless, many manage to reach a fire-resistant size. Because ground fuels in this habitat consist of grasses and discontinuous patches of shrubs and subshrubs, wildfires are generally lower in intensity, and crown fires are probably rare. For example, fire-caused pine mor- tality averaged only 30% in a similar P. coulteri/ Q. kelloggii type in the eastern Transverse Ranges (Minnich 1977). Both forests are open, free of brush and Q. chrysolepis. Age of first reproduction. Based on the age of the oldest serotinous cones in the SCRE and PM stands, as well as observations of saplings at other locations, reproduction in most P. coulteri begins 12-15 years after germination. PubUshed figures range from 8-20 years (Krugman and Jenkinson 1974) to 1 5-20 years (Minnich 1980). High tree density may significantly delay reproduction. Thus, although the two SCR stands are the same age, reproduction in the extremely dense SCRW stand began five years after the adjacent low-density SCRE stand. Late-establishing trees often grow poorly because of 38 MADRONO [Vol. 32 0.3 CO o o I- o o GC 0. 0.2 LPM ^ 0.1 Q > Q 02 HL I I 20 40 80 100 WH 0.1 0.2 0.1 0.2 n 20 40 60 80 100 FM 20 40 80 20 40 60 80 100 120 AGE Fig. 3. Age-frequency histograms for the P. coulteri/Q. agrifolia stands. brush competition or suppression by overstory trees. Thus, cones are infrequently encountered in trees of reproductive age. Trunk and branch cone production. Young trees typically bear whorls of up to four cones (rarely five) on the bole. Cones do not appear at the ends of branches until limbs are strong enough to support the heavy cones (the heaviest in the Pinaceae). The age of first branch-cone production differed markedly among stands. Trees in the low-density, fast-growing SCRE stand produced branch cones 20-25 years after germination, whereas cones occurred in small numbers on 36- to 40-year-old trees in the considerably denser CRBA, PM, TPK and MC stands. Limb cones were not produced in the extremely dense SCRW stand. As the stand ages, cone pro- duction shifts almost entirely to tips of the major branches and apex of the crown (Minnich 1980). It is not uncommon, however, to find 1985] BORCHERT: SEROTINY IN PINUS COULTERI 39 tiers of serotinous cones dating back 20 years on the trunks of trees as old as 50 years. Cone maturation. Twelve-month (June) cones are similar in size to mature cones. However, 12-month cones are green, firm, and pulpy compared to mature cones, which are hard, brittle, and light caramel in color. Seeds at this stage are full-sized, but white and soft, lacking any trace of a hard seed coat. By 1 4 months (August), the umbos of outward-facing scales begin to turn dark brown. Cones are typically moist, but the scales are fibrous and separate individually when heated. The seed coat is hard and light brown instead of black. Only 1 1% of the seeds were solid; of these, 50% were viable. Cones burned at this stage of ripening probably contribute only marginally to pine regeneration. Most cones turned to mature-cone color after 1 6 months, although some retain a reddish hue. Seeds are fully formed and, despite the somewhat milky texture of the endosperm, have a high germination rate (Table 3). By 1 8 months (December), cones have completely matured, and judging by the degree of cone development at 16 months, most probably reach full maturity between 1 5 (September) and 1 6 months (October) after pollination. Cone weathering. Serotinous cones retain a light-caramel color for 2-3 years. Almost invariably by the fifth year, however, outward- facing scales show signs of weathering as the apophyses begin to gray. In succeeding years the signs of weathering spread, and the abundant resinous covering of newly-ripened cones gradually wears away. Fifteen-year cones are almost entirely gray and usually devoid of external resin. Cones reaching 24-26 years are heavily weathered, usually to the point of disintegrating. This weathering sequence con- forms closely to that described for Pinus banksiana cones (Roe 1 963). Cone and seed characteristics. Table 2 summarizes cone and seed characteristics of the PM and SCRE stands. Cone size, seed length and weight, and the percent empty seeds per cone differ significantly between the two populations. Despite these differences, however, the mean number of full seeds per undamaged cone is similar in the two stands. Seed viability. The viability of seeds from closed cones of all ages from the SCRE and PM stands is high, 83-100%, and shows no decline with cone age. This trend contrasts with similar studies of other serotinous pines, P. clausa (Cooper et al. 1959), P. banksiana (Schantz-Hansen 1 94 1 , Beaufait 1 960) and P. /7w«^^«5 (Barden 1979), in which seed viability decreases with cone age. High viability ap- pears to be related to protection afforded by the tightly sealed scales that cover the seed with a hard, woody layer 1-1.5 cm thick. High 40 MADRONO [Vol. 32 1^ O O -o >■ S r- C50 W-^ 00 OO r- On ir^ OO t-- 00 ON +1 IT) >sO (N 00 vooos-^or--ir^t-~- =^ S O 73 ^ 1 a: 2^ la w o UJ X5 < 5 Q OS r> vo (N 00 iTi rsi 00 Tt -H in O O c<-) m — Tj- d c5 d d oo-^-Hoo>A)r-r-m d r-' od 00 ro (N (N +1 vo^r--moooo(NON r-orj-'^osooo Tt-^t~-ONOOOOTt r~- rn N b 0^ ^ w ^ 2 oauSo^Puonv^H 1985] BORCHERT: SEROTINY IN PINUS COULTERI 41 seed viability (97%) from open cones suggests that the seed coat may also provide additional protection, although it was not possible to determine how long seeds in open cones were exposed at the time of collection. Insect damage to cones. In early stages of development (<16 months), cones are highly susceptible to insect attack. For example, 71% of the SCRE cones in the 2-4-year age classes were infested by Dioryctria auranticella (ponderosa pine coneworm). Damage was complete in 25% of the collected cones. However, in the 5-7 -year age classes partially damaged cones decreased to 36%, 1 1%, and 9%. Insect damage was absent in cones older than 7 years. Directly or indirectly, insect infestations caused an average loss of 67 full seeds per attacked cone, or a 45% reduction. Seed losses may have been elevated further by the mining of Camponotes anthrax (carpenter ants), a frequent inhabitant of damaged cones. Complete cone losses to insects in the PM stand were far less frequent than in the SCRE stand (<1%). Insect-damaged cones in 2-9 -year age classes varied from 25-40%, whereas cones older than 9 years showed no signs of recent insect attack. Altogether, there was a reduction of 57 full seeds per damaged cone, or a 38% seed loss. Nearly all the cones in the PM collection were infested by larvae of Chrysophana canocola (flatheaded cone borer). Although this species is not known to attack pine seeds (Essig 1958), its extensive tunneling may weaken the cone's resistance to weathering. The diminishing percentage of insect-attacked cones in increas- ingly older age classes indicates that damaged cones probably weath- er, disintegrate and fall before unattacked cones. However, there was no evidence to suggest that insect tunneling caused serotinous cones to open or encouraged squirrel consumption. Variation in serotiny. The degree of serotiny differed markedly among the P. coulteri/ chaparral, P. coulteri/Q. chrysolepis, P. coul- teri/C. sargentii stands (Fig. 4). The highest degree of serotiny was exhibited by the PM stand; fully 52% of the cones 6 years or older were closed. By contrast, none of the MC cones was closed after 4 years. Between these extremes the other stands form a gradient, but exhibit no discernible pattern of variation. Attempts to assess serotiny in the P. coulteri/Q. agrifolia stands were only partially successful. Trees younger than 50 years were present in all the stands and were particularly abundant in the WH and FM stands. These stands were notable for their scarcity of cones in comparison to the even-aged stands regardless of tree age (Table 3). The small cone populations suggest that either (a) they opened at maturity and were blown from the trees, or (b) they remained closed at maturity but were cut by Sciurus griseus anthonyi (western gray squirrels) before they could age further. 42 MADRONO [Vol. 32 100 n 90- 80- Z LU O > 60- 20- 10- 1 5 10 15 20 25 CONE AGE Fig. 4. Cumulative cone-age distributions for the serotinous BPBR (A), CRBA (O), MC (■), PM (A), SCR (•), and TPK (□) stands. The SCRE and SCRW stands have identical distributions displayed as SCR. Several lines of evidence point to cone opening at maturity. Few trees in the age class 1 5-50 years with an adequate sample of attached cones had closed cones older than 2 years. On the periphery of the LPM and FM stands in chaparral, pines appeared less prone to squirrel depredations than those in the central portion of the stand, judging by the numerous attached cones in the older age classes (4- 6). Significantly, nearly all the cones in these outlying trees were open. And finally, the percentage of open cones/trees in these stands was nearly 8 times (p < 0.01; one-way ANOVA, arcsine transfor- mation) that of the even-aged stands (Table 3). There are at least three factors that clearly influence the length of time that cones stay closed and attached to the tree: (a) insect damage, (b) cone cutting by squirrels, and (c) spontaneous cone opening. As discussed previously, insect-attacked cones diappear from the cone population before undamaged cones. Because coneworm infestations are generally widespread, insect damage probably has a significant but highly variable effect on cone persistence. Squirrel cone cutting and consumption was observed in all stands. 1985] BORCHERT: SEROTINY IN PINUS COULTERI 43 Both green and mature cones were cut and consumed on the ground, or less frequently, partially consumed while attached to the tree bole or limbs. Scales of some cones were partially or entirely stripped from the main axis while others appeared to be excavated with only the half shell remaining. Western gray squirrels are known to depend heavily on acorns and pine seeds for winter food supplies (Stienecker and Browning 1970, Stienecker 1977). Evidence for this dependence was readily observed in the P. coulteri/Q. agrifolia habitat. For example, a ran- dom sample of 125 ground cones in the ZL stand showed that 44% had signs of squirrel damage, including removal of basal scales, partial excavation, or complete scale removal. Seventy-five percent of a sample of 1 08 cones on the ground in the FM stand were squirrel- damaged. In most the scales had been stripped completely from the cone axis. Observations suggest that P. coulteri cones are a relatively dependable year-round food source for squirrels. Moreover, squirrel population density and stability may be closely tied to cone pro- duction by this pine in southern California, a relationship perhaps comparable to that of serotinous P. contorta and squirrels of the genus Tamiasciurus in the Cascade Mountains of Washington (Smith 1970). Spontaneous cone opening occurred in varying degrees in nearly all stands (Table 3). Perry and Lotan (1977) cited three factors ex- plaining differential cone opening: differences in scale tension, en- vironmental effects on bonding strength, and genetic differences in the bonding oleoresin. In a study of P. contorta, they found opening differences between new and old serotinous cones from the same tree, suggesting that melting characteristics of the resin seal may be influenced by the environment. A complex interaction of genetic and environmental factors probably determines cone opening in P. coulteri, but differentiating between the two will require further study. Stored viable seed. Table 3 gives estimates of stored viable seed for sampled stands. The number and viability of seeds in open and closed cones in the Q. agrifolia stands are assumed to be the same as the chaparral stands. A comparison of seed populations between the two community groups reveals a marked difference. Although there is considerable within-group, interstand variability, the P. coulteri/ chaparral, P. coulteri/Q. chrysolepis, P. coulteri/ C. sargentii stands taken together average 50 times the number of stored viable seeds in the P. coulteri/Q. agrifolia stands (p < 0.01; one-way AN- OVA, logarithmic transformation). I attribute most of this between- group difference to the higher incidence of both squirrel cone cutting and spontaneous cone opening (i.e., less serotiny) in the P. coulteri/ Q. agrifolia stands. 44 MADRONO [Vol. 32 Discussion There is a discernible pattern of cone-habit variation among sam- ple stands in the study area. The open-cone habit predominates in the P. coulteri/Q. agrifolia forest, where fire-caused mortality is low and pine regeneration relatively unrestricted by competing vegeta- tion. Serotiny tends to be well developed in communities subject to killing crown fires. Previous studies (Minnich 1977, Griffin 1982) indicate that pine establishment in fire-prone communities is rela- tively synchronous, peaking in the first post-fire year. Thereafter, establishment declines and eventually ceases within 20 years as com- peting vegetation gradually dominates the site. Habitat-related cone-habit variation similar to that described above has been recorded for a number of pine species (Little and Dorman 1952; Lotan 1967, 1968; Givnish 1981). Indeed, several investi- gators have proposed that serotiny has evolved in direct response to fire (Perry and Lotan 1979, Givnish 1981, McMaster and Zedler 1981). Extensive evidence linking fire frequency and serotiny was presented for P. rigida in the Pine Barrens of New Jersey (Givnish 1981). Similarly, results of this study suggest that cone-habit vari- ation in P. coulteri is strongly influenced by fire. Indeed, the pattern of variation corresponds favorably to that predicted by a model for the evolution of serotiny in Mediterranean-climate conifers pro- posed by McMaster and Zedler (1981). They argue that serotiny is selected for when: (a) stand-killing fires bum over extensive areas, (b) interfire intervals are too short for reproduction by second-gen- eration trees, and (c) fire size is too large for seed dispersal from adjacent unbumed areas to be a significant factor for recolonization of burned areas. As discussed previously, stand-immolating fires are common in the P. coulteri/ chsiparral, P. coulteri/Q. chrysolepis and P. coulteri/ C. sargentii communities, and average fire size is usually large relative to the seed dispersal abilities of P. coulteri. The average fire-free interval, however, is often long enough (Byrne 1979) to permit some reproduction by second generation trees. Thus, limited seed release is favored, as was observed in even the most serotinous stands. When any of the above conditions are sufficiently relaxed, open- cone behavior is favored. Hence, the generally low fire intensity in P. coulteri/Q. agrifolia forests assures that open-cone trees can leave offspring that survive and reproduce. Closed-cone behavior, on the other hand, is not favored in this fire regime because seed-releasing crown fires are infrequent. The retention of substantial quantities of viable seed has obvious adaptive value to fire-killed pine stands that regenerate in the face of heavy post-bum tree and chaparral competition. The accumu- lation of seed in serotinous cones insures that the maximum number 1985] BORCHERT: SEROTINY IN PIN US COULTERI 45 of seed is available to take advantage of a transient post-bum en- vironment conducive to seedling growth. In addition, cone-stored seed acts to buffer year-to-year fluctuations in seed production (McMaster and Zedler 1981). If stand regeneration were dependent entirely on seed in ripening cones, a crown fire coincident with a poor cone year could severely curtail stand replacement. Serotinous cones guarantee that at least some seed will fall after the fire. An abundant seed rain may also be crucial to pine reestablishment in habitats with a limited supply of microsites sufficient for seed germination and seedling growth. In this regard, Wilson and Vogl (1965) noted locales in the Santa Ana Mountains where successful P. coulteri establishment was confined to rivulets, channels, and eroded areas. The dissemination of large numbers of propagules may enhance the chances of seed encounter with these "safe sites" (Harper etal. 1965). Finally, serotiny may confer superior competitive status to P. coulteri in mixed conifer stands burned by recurrent fire. In at least one stand in the Santa Lucia Mountains, Griffin (1982) recorded a very high post-bum P. coulteri/ P. lambertiana seedling-to-tree ratio (6.65) compared to the prefire ratio (0.15). He attributed this in- version of relative species abundances in part to the release of stored seed by P. coulteri. Lotan (1976) cites the serotinous cone habit as a major reason for the aggressive reinvasion of disturbed sites by P. contorta at the expense of other conifers. Understanding the variation in cone habit as well as other life history traits of P. coulteri is vital to the management of this species. In southem Califomia fire suppression activities have created ex- tensive tracts of old, highly flammable vegetation subject to cata- strophic fires (Minnich 1983). In recent years, land management agencies have introduced prescribed buming as a means of creating a less flammable vegetation mosaic of younger age classes. This means that P. coulteri in its various habitats will be managed in- creasingly under controlled bum conditions (Dougherty and Riggan 1981). Bum objectives, of course, vary depending on the desired outcome whether it be complete or partial stand regeneration. When the goal is complete stand tumover, knowing whether there is adequate stored seed for natural regeneration is essential (Lotan 1976). Another con- sideration is timing the burn so that it coincides with maximum viability of seeds in maturing cones. The limited data presented here suggest that ripening seed would not contribute measurably to pine regeneration until after mid-September. Obviously, the quantity of stored viable seed is but one factor affecting the post-bum abundance of P. coulteri. Others include fire intensity, precipitation, and post- fire competition. One aspect of seed retention not adequately investigated in this 46 MADRONO [Vol. 32 Study is the relationship between canopy storage of seed and stand age. In many serotinous species stored viable seed increases with stand age (Roe 1963, Vogl 1973, Zedler 1981). In this study, the sample size of stands of differing age growing in similar environ- mental settings is too small to observe clear trends. My general impression from field observations, however, is that stored viable seed decreases rather than increases with stand age. In older trees, most cones are bunched at the crown apex or grow singly at the ends of branches. Branch-cone production tends to be spotty, and closed cones seldom accumulate well back on the limbs in the same way they do on the trunks of young trees. Additionally, there appears to be more active cone cutting in older stands, perhaps because large old trees furnish an abundance of denning sites resulting in higher squirrel densities. If net seed accumulation diminishes with stand age, then it may become increasingly difficult to secure natural pine regeneration by prescribed burning. Further, trees in excess of 100 years appeared especially susceptible to insect and disease attack as well as to wind breakage. Consequently, excessive periods of protection from fire may be detrimental to stand persistence. Acknowledgments I thank Drs. Richard Minnich, Robert Haller, W. B. Critchfield, and James Griffin for reviewing this manuscript. My special thanks to James Griffin for suggesting seed retention as a research topic and Valdo Calvert for his diligent field assistance. Literature Cited Anonymous. 1972. Decennial census of the United States climate. Climatic sum- mary of the United States. California. Supplement for 195 1 through 1960. NOAA/ EDIS, Washington, D.C. Barden, L. S. 1979. Serotiny and viability of Pinus pungens in the southern Ap- palachians. Castanea 44:44-47. Beaufait, W. R. 1960. Some effects of high temperatures on the cones and seeds of jack pine. Forest Sci. 6:194-199. Byrne, R. 1 979. Fossil charcoal from varved sediments in the Santa Barbara Chan- nel: an index of wildfire frequencies in the Los Padres National Forest (735 ad to 1 520 ad). USDA For. Serv., Pacific Southwest Range Exp. Sta., Unpubl. Rep., Res. Agreement PSW-47. Campbell, B. 1980. Some mixed hardwood communities of the coastal ranges of southern California. Phytocoenologia 8:297-320. Cooper, R. W., C. S. Schopmeyer, and W. H. Davis. 1959. Sand pine regeneration on the Ocala National Forest. USDA For. Serv., Southeastern For. Exp. Sta., Res. Rep. 30. CoTTAM, G. and J. T. Curtis. 1956. The use of distance measures in phytosocio- logical sampling. Ecology 37:451-460. Dougherty, R. and P. J. Riggan. 1981. Operational use of prescribed fire in southern California. In E. C. Conrad and W. C. Oechel, eds.. Proceedings of the symposium on dynamics and management of Mediterranean-type ecosystems, pp. 502-510. USDA For. Serv., Pacific Southwest Range Exp. Sta., Gen. Techn. Rep. PSW-58. 1985] BORCHERT: SEROTINY IN PINUS COULTERI 47 EssiG, E. O. 1958. Insects and mites of western North America. McMillan, NY. GivNiSH, T. J. 1981. Serotiny, geography and fire in the Pine Barrens of New Jersey. Evolution 35:101-123. Griffin, J. R. 1982. Pine seedlings, native ground cover, and Lolium multiflorum on the Marble-Cone bum, Santa Lucia Range, California. Madrono 29:177-188. Hanes, T. L. 1971. Succession after fire in chaparral in southern California. Ecol. Monogr. 41:27-52. Harper, J. L., J. T. Williams, and G. R. Sagar. 1965. Behavior of seeds in soil. I. The heterogeneity of soil surfaces and its role in determining the establishment of plants from seed. J. Ecol. 53:273-286. Krugman, S. L. and J. L. Jenkinson. 1974. Pinus L., Pine. In C. S. Schopmeyer, techn. coordinator. Seeds of woody plants in the United States, pp. 598-638. USDA For. Serv. Agric. Handb. 450. Little, E. L., Jr. and K. W. Dorman. 1952. Geographic differences in cone opening in sand pine. J. Forest. 50:204-205. Lot AN, J. E. 1967. Cone serotiny of lodgepole pine near West Yellowstone, Mon- tana. J. Forest. (Washington) 13:55-59. . 1968. Cone serotiny of lodgepole pine near Island Park, Idaho. USDA For. Serv., Intermountain For. Range Expt. Sta. Res. Pap. INT-52. . 1976. Cone serotiny-fire relationships in lodgepole pine. Proc. Tall Timbers Fire Ecology Conf 14:267-278. McMaster, G. S. and P. Zedler. 1981. Delayed seed dispersal in Pinus torreyana (Torrey pine). Oecologia 51:62-66. MiNNiCH, R. A. 1977. The geography of fire and bigcone Douglas-fir, Coulter pine and western conifer forests in the eastern Transverse Ranges. In H. A. Mooney and E. C. Conrad, techn. coordinators. Proceedings of the symposium on the environmental consequences of fire and fuel management in Mediterranean eco- systems, pp. 443-450. USDA For. Serv. Gen. Techn. Rep. WO-3. . 1978. Fire and the biogeography of conifer forest in the eastern Transverse Ranges, California. Ph.D. dissertation, Univ. California, Los Angeles. . 1980. Wildfire and the geographic relationships between canyon live oak. Coulter pine, and bigcone Douglas-fir forests. In T. R, Plumb, techn. coordinator, Proceedings of the symposium on the ecology, management and utilization of California oaks, pp. 55-61. USDA For. Serv., Pacific Southw. For. Range Exp. Sta., Gen. Techn. Rep. PSW-44. . 1983. Fire mosaics in southern Cahfomia and northern Baja California. Science 219:1287-1294. Perry, D. A. and J. E. Lot an. 1977. Opening temperatures in serotinous cones of lodgepole pine. USDA For. Serv., Intermountain For. Range Exp. Sta., Res. Note INT-228. and . 1979. A model of fire selection for serotiny in lodgepole pine. Evolution 33:958-968. Roe, E. I. 1963. Seed stored in some jack pine stands, northern Minnesota. USDA For. Serv., Lake States For. Exp. Sta., Res. Pap. LS-1. Schantz-Hansen, T. 1941. A study of jack pine seed. J. Forest. 39:980-990. Smith, C. C. 1970. The coevolution of pine squirrels {Tamiasciurus) and conifers. Ecol. Monogr. 40:349-371. Stienecker, W. 1977. Supplemental food habits of the western gray squirrel. Cal- ifornia Fish and Game 63:1 1-21. and B. M. Browning. 1970. Food habits of the western gray squirrel. California Fish and Game 56:36-48. SuDwoRTH, G. B. 1908. Forest trees of the Pacific Slope. Dover, NY. Vale, T. R. 1 979. Pinus coulteri and wildfire on Mount Diablo, California. Madrono 26:135-140. VoGL, R. J. 1 973. Ecology of knobcone pine in the Santa Ana Mountains, California. Ecol. Monogr. 43:125-143. , W. P. Armstrong, K. L. White, and K. L. Cole. 1977. The closed-cone 48 MADRONO [Vol. 32 pines and cypresses. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, pp. 295-358. Wiley-Interscience, NY. Wilson, R. C. and R. J. Vogl. 1965. Manzanita chaparral in the Santa Ana Moun- tains, California. Madrono 18:47-61. Zedler, p. H. 1981. Vegetation change in the chaparral and desert communities in San Diego County, California. In D. C. West, H. H. Shugart, and D. B. Botkin, eds.. Forest succession: concepts and applications, pp. 406-430. Springer Verlag, NY. (Received 1 June 1984; accepted 27 July 1984.) ANNOUNCEMENT ASPT Herbarium Travel Awards The American Society of Plant Taxonomists is pleased to announce the availability of competitive awards for travel by graduate students to the nation's herbaria. The awards will not exceed $500 and will be used to help pay expenses to and from any herbarium (or herbaria) in the United States and per diem expenses during the visit. Competitions for awards will be held twice a year: The first competition deadline is 1 January 1985, with the second deadline 1 July 1985. The grants program will last a minimum of three years (six competitions). Interested Master's or Ph.D. graduate students should send a curriculum vitae, two letters of recommendation (including one from the major professor), a two or three page outline of the proposed research emphasizing the role that the visit to the herbarium will play, and a letter from the Head Curator, Chairman or Director of the institution(s) to be visited indicating willingness to receive the visitor. Awards will be announced by 1 March from the January competition and during the annual banquet of the ASPT from the July competition. Students are encouraged to obtain additional funds from their home institutions (or elsewhere) to extend their research visits even further. This compe- tition is open to students of both cryptogamic and phanerogamic groups. Completed appHcations and additional questions should be directed to Tod F. Stuessy, Chair- man, ASPT Committee for Systematics Collections, Department of Botany, Ohio State University, 1735 Neil Avenue, Columbus, OH 43210. (Phone: (614) 422-5200 or (614) 422-8952.) Berkeley Herbarium Severe space limitations require that the Herbarium of the University of California, Berkeley (UC) box and store all vascular plant specimens originating from Europe, Africa, Asia, and the Pacific Basin. Specimens from these areas will be unavailable for loan or routine consultation. However, special arrangement can be made to see them. Please write to the Director of the Herbarium to arrange to use these collections. Collections from North and South America remain unaffected; exchange programs will also remain unaffected. We regret the inconvenience that the inaccessibility of these collections may cause researchers and will try to resolve our space problems so that Eastern Hemisphere specimens will be available again as soon as possible. UC has acquired additional space to house these collections and hopes to obtain storage cases within the next year. For the next decade, the collection at UC will be housed in two locations within close proximity. As the University of California completes the renovation that began this year of all biological science facilities, the collections of the University Herbarium will be brought together in a single expanded and modernized facility.— Thomas Duncan, Director— UC and JEPS. A NEW SPECIES OF ERYTHRONIUM (lAllACEAE) FROM THE COAST RANGE OF OREGON Paul C. Hammond 2435 E. Applegate St., Philomath, OR 97370 Kenton L. Chambers Department of Botany and Plant Pathology, Oregon State University, Corvallis 97331 Abstract A new species, Erythronium elegans, is described from three sites in the northern Coast Range of Oregon. On morphological grounds the species is closely allied with E. montanum of the Olympic Mountains and northern Cascade Range, and with E. klamathense of the southern Cascades. Its variability in certain traits such as flower color, stamen width, and leaf mottling is suggestive of past hybridization with the geographically associated species, E. revolutum. During 1982, the U.S. Forest Service initiated a project to study the distribution and ecology of Erythronium revolutum Smith along the Oregon Coast in order to assess conservation requirements for the species (Bierly and Stockhouse 1982). In the course of this study, large populations of a distinctly different Erythronium were discov- ered on the top of Mt. Hebo (945 m) in Tillamook County just 10 miles inland from the Pacific Ocean. Two additional populations of this taxon have since been found, one on Saddleback Mountain in northern Lincoln County 1 5 miles south of Mt. Hebo, and another on Fanno Ridge north of Valsetz in Polk County. The species ex- hibits a combination of characteristics seen in no other described taxon, and it forms a link between sections Concolorae and Par- dalinae as defined by Applegate (1935). In this paper we describe this Erythronium as new and examine evidence concerning its origin and relationships. Erythronium elegans Hammond & Chambers, sp. nov. Herba perennis, foliis duo concoloris raro maculosis petiolo sub- terraneo lamina lanceolata prostrata, scapo 16-30 cm alto, floribus l-2(-4) cemuis tepalis reflexis lanceolatis 2-4(-5) cm longis albis basi luteis dorsaliter plerumque roseis, filamentis 0.5-2.0 mm latis, antheris luteis, stylo 1-3 cm longo, stigmate profunde trifurcato, capsula clavata 2.5-3.5 cm longa. Corm 2.0-5.5 cm long, 8-1 5 mm wide, enclosed in papery sheaths and producing new cormlets laterally; subterranean stem between Madrono, Vol. 32, No. 1, pp. 49-56, 15 February 1985 50 MADRONO [Vol. 32 Fig. 1 . Erythronium elegans. A. Habit of flowering and non-flowering individuals. B. Inner tepal and pair of stamens. C. Androecium and exserted style. D. Gynoecium. E. Capsule. Scale equals 10 cm in A, 1 cm in other Figs. leaves and corm 3.5-10 cm long; leaves usually uniformly deep green or mottled with a few pale lines, rarely with well-developed brown mottling; leaf of non-flowering plants single, 6-8 cm long, 4-5 cm 1 985] HAMMOND AND CHAMBERS: NEW ER YTHRONIUM 5 1 wide, the blade broad, ovate-lanceolate, usually abruptly narrowed to a slender, nearly wingless petiole; leaves of flowering plants two, more or less prostrate, 7-13(-15) cm long, 2-4(-8) cm wide, the blade narrowly lanceolate, usually with strongly undulate margins, gradually narrowed to a short, evidently winged petiole; scape 1 6- 30 cm tall; flowers l-2(-4), nodding, with perianth strongly reflexed in bright sunshine to only slightly spreading under shady or cloudy conditions; perianth segments lanceolate, 2-4(-5) cm long, (3-)5- 10(-15) mm wide, with well-developed basal appendages, white or pale pink with bright yellow stripes at the base, often reddish on abaxial surface; filaments narrow to somewhat dilated, 0.5-2.0 mm wide; anthers golden yellow, not connivent around style; style fili- form, 1-3 cm long; stigma deeply divided with recurved lobes; cap- sule broadly clavate, blunt, 2.5-3.5 cm long, 6-8 mm wide. Type: USA, OR, Tillamook Co.: Mt. Hebo, T4SR9WS13, ca. 945 m, open sites on rocky slopes and cliffs, P. Hammond s.n., 28 May 1982 (Holotype: OSC; isotype: UC). Paratypes: OR, Tillamook Co.: Mt. Hebo, open slopes with Gaultheria and Vaccinium, Pseudotsuga belt, 945 m, Constance and Beetle 2656 (ORE, WILLU); Mt. Hebo, 914 m, plentiful under Pseudotsuga, Thuja, Picea and in open meadows with Gaultheria, Vaccinium, grasses, Lupinus, Fragaria, mosses, GreenleaflSOl (OSC, ORE, NY); Mt. Hebo, A. M. Phillips s.n., 24 May 1954 (OSC). Lincoln Co.: Lost Prairie, Saddleback Mtn. at edges of bog, T 7 S R 9 W S 2, 5'. Lofton s.n.. May 1982 (MOC). Polk Co.: Fanno Ridge, bog at headwaters of Little Luckiamute River, T8SR8WS 14, L. Scofields.n., 25 May 1983 (MOC). Populations of this species extend over the entire length of the Mt. Hebo escarpment, an area of some three square miles, and occupy meadows, rocky cliffs, brushland, and open coniferous forest. In many areas thousands of plants carpet the ground in a manner like that of E. montanum in subalpine meadows of the northern Cascade Range. The Saddleback Mountain and Fanno Ridge pop- ulations of E. elegans are much smaller and are largely restricted to the edges of sphagnum bogs. Additional populations of the species may be expected in mountain bogs, meadows, or rocky balds of northern Lincoln County and in Polk and Tillamook Counties. Populations of E. elegans show considerable variation in leaf mot- tling, flower color, and stamen filament width. Most plants have completely unmottled leaves that are uniformly deep green or exhibit a few pale lines along the veins. Well-developed brown mottling is found in occasional plants, however, and may be so extensive that the leaves are almost entirely brown. The frequency of such extreme mottling is roughly estimated at about one in 10,000 plants on Mt. Hebo, but some local colonies consist largely of mottled plants. The majority of individuals have a pinkish- white perianth, often with 52 MADRONO [Vol. 32 deeper pink on the abaxial side of the tepals. However, flower colors vary from deep rose-pink at one extreme to pure white at the other. The dilation of the stamen filaments is also variable. In a sample of 174 flowers on Mt. Hebo, 62 (36%) had thin filaments about 0.5 mm wide, and 1 12 (64%) had dilated filaments of 1.0-2.0 mm. In most plants the stamens spread away from the style, but in a few plants they are constricted around the style, as in E. revolutum. Sympatric with E. elegans on the rocky cliffs of Mt. Hebo is E. grandiflorum Pursh var. pallidum St. John. Its leaves are similar in shape to those of E. elegans but differ in their blue-green color. The two species exhibit floral differences as well, and there is no evidence of hybridization between them. In western Oregon there are four additional species of Erythronium with which E. elegans can be compared. These are E. revolutum, E. montanum Watson, E. ore- gonum Applegate, and E. klamat hense Applegate. Applegate (1935) divided the genus Erythronium into two sections: the Pardalinae, found in lowland habitats and having mottled leaves and white to lavender flowers; and the Concolorae, occurring in mountain hab- itats and having unmottled leaves and white to yellow flowers. Er- ythronium montanum and E. klamathense belong to sect. Conco- lorae, while E. oregonum and E. revolutum belong to sect. Pardalinae. However, E. elegans, as described here, is intermediate between the two sections in certain of its characteristics. In order to understand this relationship, it is useful to review the biogeography and ecology of the five allied species. Both E. montanum and E. klamathense have white flowers, nar- row filaments, and unmottled leaves. They occupy habitats at high elevations in the Cascade Range, including subalpine meadows and wet forests. Erythronium montanum is distributed from Vancouver Island and the Cascades of southern British Columbia southward to the Olympic Mountains and through the Cascades to Monument Peak in northern Linn County, Oregon. It differs from E. klamath- ense by the larger size of the plants and flowers (tepals 25-40 mm long), deeply divided stigma lobes, erect flowers, long and slender aerial petioles that raise the leaves above the ground, and abruptly expanded leaf blades. In contrast, E. klamathense is restricted to the southern Cascade Range extending from Black Rock Lookout in Douglas County, Oregon, south to the Mt. Shasta region of Sis- kiyou County, California. The species is smaller than E. montanum (tepals 15-25 mm long) and has entire stigmas, pendulous flowers, short subterranean petioles, prostrate leaves, and gradually expanded leaf blades. Erythronium revolutum and E. oregonum, the closest geographical associates of E. elegans, both have dilated stamen filaments and distinctly mottled leaves. The flowers of E. oregonum are creamy- white, with the stamens spread away from the style (contrary to the 1985] HAMMOND AND CHAMBERS: NEW ERYTHRONIUM 53 illustrations in Hitchcock et al. 1969, p. 792), whereas those of E. revolutum are deep pink and the filaments are appressed to the style (as illustrated by Hitchcock et al. 1969, and Gilkey 1945). These two taxa occupy mostly lowland habitats west of the Cascades from southern British Columbia to southern Oregon; E. revolutum con- tinues south to Mendocino County, California. Except in the north- em part of this range, the two species are rarely sympatric because E. revolutum occupies the coastal rain-forest zone and E. oregonum occurs in drier forests and woodlands of the interior. As observed by us in Oregon, E. revolutum is best developed in low-lying riparian habitats in the portion of its range south of the Siuslaw River, Lane County. These habitats include streamside benches above the high- water line, rotting log-jams, and rocky ledges. North of the Yaquina River, Lincoln County, the species shifts to moist coniferous forests on mountain slopes, reaching 900 m elevation on Onion Peak, Clat- sop County. In the vicinity of Mt. Hebo, E. revolutum occurs on mountainsides above Powder Creek and Limestone Creek on the north and east sides of the peak below the main escarpment, where it is separated by 600 m elevation from the main population of E. elegans. It should be noted that E. oregonum is sometimes present in open meadows or "balds" on the tops of high Coast Range peaks, including Roman Nose Mountain in Douglas County, Marys Peak in Benton County (elevation 1200 m), and Rickreall Ridge in Polk County. These meadow habitats are superficially similar to those occupied by E. montanum in the Cascade Range, and by E. elegans on Mt. Hebo. The population of E. oregonum on Rickreall Ridge is ecologically isolated from E. elegans on nearby Fanno Ridge, however, since the former is on an exposed rocky ridge and the latter is in a cold bog. In summary, the habitats of E. elegans are more similar overall to those of E. revolutum in this part of Oregon than to the habitats of E. oregonum. There may have been sympatry between E. revo- lutum and E. elegans at some time in the past, with an opportunity for introgression into the latter species of such traits as pink flower color, dilated filaments, and mottled leaves. These traits are found in all three disjunct populations of E. elegans, suggesting that in- trogression is not a recent event. There is presently no evidence that E. elegans and E. oregonum have had genetic contact through hy- bridization, however. The morphological traits that suggest introgression by E. revo- lutum affect only a small minority of the plants of E. elegans, and we do not believe that the new species is itself of hybrid origin. The basic affinities of E. elegans appear instead to be with E. montanum and E. klamathense. The habitats of the three taxa are similar, although at present they are widely separated geographically. Eryth- ronium elegans is like E. montanum in its flower size and its deeply 54 MADRONO [Vol. 32 Species and Locations E. montanum Squaw Mt,, Clackamas Co, OR Mt, Rainier, Pierce Co. WA E. elegans Mt. Hebo. Tillamook Co. OR E. oregonum Philomath, Benton Co, OR Salem. Marion Co, OR E. revo/utum Beaver,Tillamook Co, OR Pioneer Mt,, Lincoln Co, OR Fig. 2. Width of stamen filaments in four species of Erythronium. Range of variation is given for each population, with the frequency of the most common class shown as a labeled rectangle. divided stigma lobes; it resembles E. klamathense in having pen- dulous flowers, prostrate leaves with short subterranean petioles, and blades tapering gradually at the base. Figures 2 and 3 illustrate the variation in two critical morphological features, the width of the stamen filaments and the angle formed by the tapering base of the leaf blade, in selected populations of four species under discussion. All measurements were taken from living plants in the field. The filament width in E. elegans is commonly in the range of E. mon- tanum but varies toward the broad filaments of E. revolutum and E. oregonum. The basal angle of the leaf blades of E. montanum is variable, but the most frequent condition is to have a larger angle (that is, a more abruptly tapering blade) than in E. elegans. Figure 2 also illustrates that filament width cannot be used to separate E. oregonum from E. revolutum, despite the assertion by Applegate (1935, p. 100). In the above discussion, E. elegans has been compared with its congeners for all the characteristics commonly used by taxonomists to distinguish species of Erythronium. We hypothesize that some of the present variability of this new taxon derives from past hybrid- ization with E. revolutum, but that the origin of the species lies in an ancestral complex from which both E. montanum and E. klam- athense are also derived. Putatively primitive traits recognizable in E. elegans are its separate stigma lobes, its pendulous flowers, and Sample Width of Filaments Size (mm) ,0,5 1,0 1.5 2.0 2.5 3.0 3.5. I 1 1 I I I 1 1 lOQo/o 1985] HAMMOND AND CHAMBERS: NEW ERYTHRONIUM 55 Species and Locations Sample Angle of Leaf Base Size (degrees) I 15 20 25 30 35 40 45 50, I — I — I — r-T — I — I — \ — I 62 62 E. montanum Squaw Mt,, Clackamas Co, OR 80 Mt, Rainier, Pierce Co, WA E. elegans Mt, Hebo, Tillamook Co, OR E. orego num Philomath, Benton Co, OR Corvallis, Benton Co, OR E. revolutu m Beaver, Tillamook Co, OR Pioneer Mt, Lincoln Co, OR Fig. 3. Angle of leaf base in four species of Erythronium. Measurements are the angle, in increments of five degrees, from the petiole midline along the diverging margin of the blade at its base. Range of variation is given for each population, with the frequency of the two most common classes shown as a labeled rectangle. its tapering, short-petioled leaves; advanced traits in the complex may be the united stigma lobes of E. klamathense, and the erect flowers and long-petioled leaves of E. montanum. It is reasonable to find a species with Cascadian affinities growing at high elevations in the northern Oregon Coast Range. In Clatsop County, Saddle Mountain (Detling 1954), Onion Peak (Chambers 1973) and Su- garloaf Mountain (Chambers 1974) harbor plant species that are disjunct from the Olympic Mountains and the Cascade Range. Erythronium elegans may once have had a wider range in the coastal mountains of the Pacific Northwest, and have been reduced to a few relict populations by postglacial climatic changes. Although modi- fied by genetic contact with E. revolutum, it is recognizable as a distinct and phylogenetically interesting species. Acknowledgments We thank the following persons for advice and assistance with this study: K. Bierly, H. Chrostowski, M. Johnson, S. Lofton, and L. Scofield. The illustration was prepared by Julie Kierstead. Literature Cited Applegate, E. L 1935. The genus Erythronium: a taxonomic and distributional study of the western North American species. Madroiio 3:58-1 13. Bierly, K. F. and R. E. Stockhouse. 1 982. Coast fawn lily {Erythronium revolutum) 56 MADRONO [Vol. 32 sensitive species conservation status report. USD A For. Serv., Pacific Northwest Region, Siuslaw Natl. For., Corvallis, OR. 82 pp. Chambers, K. L. 1973. Floristic relationships of Onion Peak with Saddle Mountain, Clatsop County, Oregon. Madrono 22:105-114. . 1974. Notes on the flora of Clatsop County, Oregon. Madrono 22:278-279. Detling, L. R. 1954. Significant features of the flora of Saddle Mountain, Clatsop County, Oregon. Northwest Sci. 28:52-60. GiLKEY, H. M. 1945. Northwestern American plants. Oregon State Coll. Monogr., Studies in Botany, No. 9. Hitchcock, C. L., A. Cronquist, M. Ownbey, and J. W. Thompson. 1969. Vascular plants of the Pacific Northwest. Part 1 . Univ. Washington Press, Seattle. (Received 9 April 1984; accepted 30 July 1984.) NOTES AND NEWS Yellow Jackets Disperse Vancouveria Seeds (Berberidaceae).— During a field study in Lewis and Clark State Park (WA: Lewis Co.) in early July 1983, I had the opportunity to study seed dispersal in Vancouveria hexandra (Hook.) Morr. & Dene, (inside-out-flower). Observations were made on 3 and 4 July, at which time plants were in all stages, from open flowers to ripe seeds. The common yellow jacket Vespula vulgaris (L.) was fairly abundant in the area, and on eight occasions individuals were observed collecting seeds of V. hexandra. When an insect searching through an area covered by the plant found a dehisced pod, it alighted on the pod stalk next to it. It then bit loose the seed and its large, white appendage and flew a few meters before alighting on a low branch. It then bit the appendage off within a few seconds and dropped the seed to the ground. The yellow jacket then disappeared carrying the appendage. Apart from the eight successful searches I observed, there were also four occasions when yellow jackets searched in vain for seeds in sterile patches or in patches with unripe pods. Attempts to bite open unripe pods were never observed to be successful. Searches lasted for up to four minutes before the yellow jacket left the study area. The chemical nature of the appendage appears to be unknown, but may possibly consist of polysaccharides. Berg (Amer. J. Bot. 59: 109-1 22. 1 972), in an extensive experimental study, showed that ants are effective seed vectors of Vancouveria. In repeated experiments, where seeds were put out within a few meters from nests of ant species in three genera, all seeds were dragged by'^he appendages to the nests within 0.5-2.5 h. In contrast to other members of the genus, the pods of V. hexandra dehisce while the seeds are still green; this leaves some time before they fall spontaneously to the ground. My observations suggest that this can be an adaptation to allow seed dispersal by vespids. It should result in a more random distribution of the seeds than that provided by ants, which drag them to their nests along their trails. Berg observed ant-mediated transport up to 2.5 m from the starting point. Dispersal by the winged yellow jacket also increases the possibility of dispersal over a somewhat longer range. To my knowledge, this is the first reported case of seed dispersal by yellow jackets. Vance Tartar (pers. comm.) has observed vespids collecting seeds of Trillium, another myrmecochorous genus. Further studies of vespids as seed vectors in ant-dispersed species may prove this phenomenon to be of wider occurrence. The support of B. J. D. Meeuse during the work in the Pacific Northwest is gratefully acknowledged. This study was made while on a joint travel grant from the Swedish Natural Science Research Council (NFR) and the Royal Swedish Academy of Sci- ences.—Olle Pellmyr, Dept. of Entomology, Uppsala Univ., Box 561, S-751 22 Uppsala, Sweden. (Received 6 April 1984; accepted 24 May 1984.) NOTEWORTHY COLLECTIONS California DicENTRA PAUciFLORA Wats. (Papaveraceae). — Ventura Co., 2 km ne. of Reyes Peak (34°39'30"N, 1 19°16'15"W). Few small scattered colonies observed on steep n.- facing slope immediately above Beartrap Cr. Plants growing in deep leaf mold within dense forest of Calocedrus decurrens, Pinus ponderosa, Abies concolor, Quercus chrysolepis, 1500 m, 12 Jun 1983, Odion 68 (UCSB), 67 (RSA). Verified by K. R. Stem, Calif State Univ., Chico. (Voucher) Significance. Previously known from northern California and southern Oregon. Small populations also occur in the Greenhorn Mts. (Twisselman, Fl. Kern Co. 1967) and San Emigdio Range (Emmel 562, RSA), both in Kern Co. This is the first record in Ventura Co. and within the areas covered by floras of Munz 1974 (Southern Calif) and Smith '76 (the Santa Barbara Region). —Dennis C. Odion, Dept. Biol. Sci., UCSB. Spirodela punctata (G. F. W. Meyer) Thompson (Lemnaceae).— USA, CA, Humboldt Co., Clam Beach, 15 km n. of Areata, e. side Hwy 101 (T7N RIE S17 nw. 1/4), 8 m, 25 May 1980, Richards 101 (HSC). Dense population at north end of pond, associated with Spirodela polyrrhiza and Azolla filiculoides. Previous knowledge. Known from Asia, s. Pacific, se. US, MO, IL. Reported in CA from San Diego, Fresno, Yolo, Madera and Santa Clara Counties and from the UC Berkeley campus (Alameda Co.). This species is listed as S. oligorrhiza (Kurz) Hegelm. by Daubs (111. Biol. Monogr. 34. 1965; Rhodora 64. 1962) and Mason (A Fl. Marshes Calif 1957). It is distinguished from S. polyrrhiza by its smaller size and the presence of only 2-3 roots, and from Lemna spp. by purple lower surface of frond and multiple roots. Without close examination, however, these features can easily be overlooked and the plants mistaken for Lemna. Significance. Northernmost occurrence in CA and range extension of ca. 480 km n. of Berkeley. This species was also collected on the UC Davis campus in 1967 and mislabeled S. polyrrhiza. I looked for this population in June 1983 and am doubtful it still exists. I found another population of S. punctata in 1 980 ca. 5 km w. of Blue Lake, CA (ca. 25 km se. of Clam Beach), Richards 102 (HSC). Winter storms in 1982 destroyed this population and it has not recovered. Spirodela punctata from both Humboldt Co. locations have been observed in flower between June and early No- vember. WoLFFiA BOREALis (Engclm.) Landolt (Lemnaceae).— USA, CA, Mendocino Co., 11.2 km s. of Willits, w. side of Hwy 101 (T17N R13W S16 se.'A), 442 m, 25 Oct 1981, Richards 104 (HSC). Dense population at surface of small pond, mixed with Lemna minor and Azolla filiculoides. Verified by W. P. Armstrong. Previous knowledge. Known from WA, OR, s. Canada, c. and ne. US. In CA reported from Shasta and San Diego Counties. This species has been listed incorrectly as W. punctata Griseb. by most Amer. authors. The original W. punctata described by Grisebach was synonymous with W. brasiliensis Weddell, a species native to s. US and in C. and S. America. Significance. First record of this sp. in Mendocino Co., a sw. range extension of ca. 250 km from Fall River Mills, Shasta Co.— Daniel V. Richards, Humboldt State Univ., Dept. of Biol. Sci., Areata, CA 95521. Colorado Phacelia const ancei Atwood (Hydrophylaceae). — San Miguel Co., ne. -facing slopes in gypsum soils, 2.5 km se. of Gypsum Gap (T43N R16W S6, 38°02'N, 108°39'W), 1830 m, 16 Aug 1983, Kelley 83-147 (Mesa College Herb., CS). Madrono, Vol. 32, No. 1, pp. 57-58, 15 February 1985 58 MADRONO [Vol. 32 Previous knowledge. Common on soils derived from the Moenkopi formation in se. UT and ne. AZ. Significance. This is the first record of the species in CO.— Walt Kelley, Dept. of Biology, Mesa College, Grand Junction, CO and Dieter H. Wilken, Dept. of Botany, Colorado State Univ., Ft. Collins 80523. REVIEW Intermountain Flora: Vascular Plants of the Intermountain West, U.S.A. Volume 4. By Arthur Cronquist, Arthur H. Holmgren, Noel H. Holmgren, James L. Reveal, and Patricia K. Holmgren. The New York Botanical Garden, New York. 1984. $77.50 (U.S.), $79.00 (non U.S.). ISBN 0-89327-248-5. The publication of volume 4 of the Intermountain Flora represents another land- mark of botanical excellence for western United States. The Flora, this time published by the New York Botanical Garden, is the third volume of their proposed six volume series, covers Cronquist's Subclass Asteridae and includes 28 families, leaving the Asteraceae for volume 5. This volume was written largely by Cronquist and Noel Holmgren. Cronquist authors 1 3 families and co-authors four others, two with Jim Reveal (Lamiaceae, Cuscutaceae), and one each with Arthur Holmgren (Caprifoli- aceae), and C. L. Hitchcock (Solanaceae). Noel Holmgren authors nine families in- cluding the Scrophulariaceae, the largest in the book (215 species, including 104 species of Penstemonl). Pat Holmgren contributes the Campanulaceae and, with Noel, the Asclepiadaceae. The format used is the same as in the preceding volumes, complete with excellent illustrations of each species contributed largely by Jeanne R. Janish, Bobbi Angell and Robin Jess. The book provides complete citation of synonymy, very readable, well-written descriptions, ample discussions of distribution, and perhaps most im- portantly, comments on past treatments of groups and the rationale for the present treatments. These commentaries, to me, are the best part of the book because they provide historical perspectives and often in-depth discussions of nomenclatural prob- lems; and when coupled with the references, allow for direct access to literature of a group. Even if a person has no direct interest in the Intermountain area, the book is an important reference bringing together much information on the plants of western North America. While the book is invaluable to practicing botanists, it will also be very useful to beginning botanists, non-botanical professionals and amateurs who can use the illustrations to help them through the keys (or one may rely entirely on the pictures to identify plants in question). In this regard the Intermountain Flora is potentially useful to a large segment of the Intermountain population. The book is also, to some degree, user friendly in that both technical and artificial keys are provided in some large genera and species that are known outside the area, but that may eventually be found in the region. Regarding taxonomy, Neogaerrhinum is separated from Antirrhinum, Fraseria from Swertia, Gentianella and Gentianopsis from Gentiana, and Leucophysalis from Chae- maesaracha. Tiquilia replaces our Coldenia. However, Glandularia is not distin- guished from Verbena. The careful work of Dempster and Erhendorfer in the Galium multiflora complex has been critiqued and reduced to a more "practical" taxonomy (3 varieties) in a lengthy discussion by Cronquist. Likewise, Grant's Gilia inconspicua complex has been reduced and Ipomopsis is not distinguished from Gilia. While Grant admits some elements in the Gilia inconspicua complex may be disregarded in favor of more easily discemable morphotaxa, he (pers. comm. 1 984) feels that the reduction has gone too far and that taxa such as Gilia opthalmoides and G. clokeyi MADRoto, Vol. 32, No. 1, pp. 58-59, 15 February 1985 1985] REVIEW 59 are recognizable, morphologically and ecologically. Ipomopsis has been accepted as distinct from Gilia by all recent workers in Polemoniaceae, and in all but one recent western flora, and its recognition leaves a much "cleaner" Gilia. But we are reminded that taxonomy still consists mostly of opinion. At least their preferences have been openly discussed and not proposed by fiat. As in past works from the New York Botanical Garden, infraspecific taxa are recognized primarily at the rank of variety and 38 new combinations are made at that level, only three of which represent simple subsp.-to-var. changes. The volume is very well edited and has a list of new combinations and an index in the back. It represents a fine contribution to systematic botany that reflects well on the authors, their institution and on the field of systematic botany as a whole. While some may feel the price is high, we must remember what happened to the nickle candy bar. One should consider it an investment. The Flora is sure to have a strong influence on plant systematics in the years to come.— James Henrickson, Biology Department, California State University, Los Angeles 90032. PUBLICATION ANNOUNCEMENT Las Gramineas de Mexico Alan Ackerman Beetle and collaborators. Secretaria de Agricultura y Recursos Hidraulicos under the auspices of COTECOCA [Comision Tecnico Consultiva par el Determinacion Regional de los Coeficientes de Agostadero]. 260 pp., 1983. This first of four volumes includes a key to the subfamilies, tribes and genera, and treats genera and their species through the letter A. The following three volumes, intended for publication in 1984, will complete the taxonomic treatment and in addition contain bibliography, synonymy, glossary and general index. Species treatments in- clude full page illustrations and maps giving distributions within states. According to a notice in Macpalxochitl [newsletter of the Sociedad Botanica de Mexico], institutions or investigators interested in obtaining a copy of this book may address their request to Ing. Horacio Garcia Aguilar, Secretario de Agricultura y Recursos Hidraulicos with a copy for Victor Jaramillo Villalobos [Director General de COTECOCA]. Such a letter should state the reasons for wishing to have a copy of the work.— Annetta Carter Note added in proof (See Harms, pp. 1-10, this issue): The following article was published too late to be incorporated into the text of Harms, Madrono 32(1): 1-10: Pavlick, L. E. and J. Looman. 1984. Taxonomy and nomenclature of rough fescues, Festuca altaica, F. campestris (F. scabrella var. major), and F. hallii, in Canada and the adjacent part of the United States. Canad. J. Bot. 62(8): 1739- 1749. Although not taken into account, it is not believed to alter the premises and conclu- sions of the paper in Madrono. The reader is referred to the Pavlick and Looman paper in conjunction with the present paper for amplified morphological (specifically anatomical) and cytological data, and for differing taxonomic interpretations. 60 MADRONO [Vol. 32 Postal Sarvic* STATEMENT OF OWNERSHIP, MANAGEMENT AND CIRCULATION Requirtd by 39 U.S.C. 36851 lA TITLE OF PUBLICATION MADRONO A WEST AMERICAN JOURNAL OF BOTANY JB. PUBLICATION NO. 2. DATE OF FILING 10/1/84 0 0 2 4 9 6 3 7 3. FREQUENCY OflSSUE^ Quarterly 3A. NO. ANN FOU □ F ISSUES PUBLISHED ^ALLY 3B. ANNUAL SUBSCRIPTION PRICE $25.00 County of Alameda 5. COMPLETE MAILING ADDRESS OF THE HEADQUARTERS OF GENERAL BUSINESS OFFICES OF THE PUBLISHER (Not prinierj California Botanical Society, Dept. of Botany, Life Science Bldg., University of California Berkeley, CA 94720 6. FULL NAMES AND COMPLETE MAILING ADDRESS OF PUBLISHER, EDITOR, AND MANAGING EDITOR (Thii iirm MUST NOT be blank) PUBUSHER (Name and Complete Mailing Addresi) California Botanical Society, Dept. of. Botany, Life Science Bldg., University of California Berkeley, CA 94720 EDITOR (Name and Complete Mailing Address) I Christopher Davidson, Idaho Botanical Garden, P.O. Box 2140, Boise, Idaho 83701 MANAGING EDITOR (Name and Complete Mailing Address) Same 7. OWNER (If owned by a corporation. Its name and address must be slated and also Immediately thereunder the names and addresses of stockholders oviTiinjr or holding 1 percent or more of total amount of stock. If not o^vned by a corporation, the names and addresses of the individual owners must be given, tf owned by a partnership or other unincorporated firm, its name and address, as well as that of each Individual must he given. If the publica- tion is published by a nonprofit organization. Its name and address must be stated.) (Item must be completed.) FULL NAME COMPLETE MAILING ADDRESS Caiitornia iiotanicai bociety (a non-profit Ucpc. Of BuLauy, ^ife Science Building corporation) University of California, Berkeley, CA y^/iiU a KNOWN BONDHOLDERS. MOBTGAOEES. AND OTHER SECURITY HOLDERS OWNING OR'ifOLOING 1 PERCENT OR MORE OF TOTAL AMOUNT OF BONDS, MORTGAGES OR OTHER SECURITIES (If there are none, so state) FULL NAME COMPLETE MAILING ADDRESS NONE 9. FOR COMPLETION BY NONPROFIT ORGANIZATIONS AUTHORIZED TO MAIL AT SPECIAL RATES (Section 4 2}.t 2 DMM only) The purpose, lunction, and nonprofit «taiu» of thit organiiation and the exempt statut for Federal income lax purpoiei (Check one) (II (21 HAS NOT CHANGED DURING i 1 HAS CHANGED DURING (If changed, publisher must submit explanation of PRECEDING 12 MONTHS | | PRECEDING 12 MONTHS change with this statement.) 10. EXTENT AND NATURE OF CIRCULATION AVERAGE NO. COPIES EACH ISSUE DURING PRECEDING 12 MONTHS ACTUAL NO. COPIES OF SINGLE ISSUE PUBLISHED NEAREST TO FILING DATE A. TOTAL NO. COPIES (Net Press Run) 1250 1250 B. PAID CIRCULATION 1 . Salei through dealerj and carrier!, street vertdori and counter tales r-O- -0- 3. Mail Subscription 1051 1064 C. TOTAL PAID CIRCULATION (Sum of lOBI and I0D2) 1051 1064 0. FREE DISTRIBUTION BY MAIL, CARRIER OR OTHER MEANS SAMPLES, COMPLIMENTARY, AND OTHER FREE COPIES 7 7 E. TOTAL DISTRIBUTION ^Sum o/C«rtrf D; 1058 1071 F. COPIES NOT DISTRIBUTED 1. Olfice use, left over, unaccounted, spoiled after printing 192 179 2. Return from Ne«vi Agents -0- -0- G. TOTAL (Sum of E. Fl and 2-should equal net press run shown In A) 1750 1250 11. 1 certify that the statements made by me above are correct and complete > SIGNATURE AND TITLE OF EDITOR, PUBLISHER, BUSINESS MANAGER, OR OWNER {See instruction on reverse} 62 For his brilliant contributions to plant physiology, ecology, and our understanding of life on Earth, in a career that has spanned more than half a century, we take great pleasure in dedicating volume XXXI of Madrono to FRITS WARMOLT WENT. Bom in 1903 in Utrecht, the son of a professor of botany in the University there. Frits Went first found employment at the Botanical Gardens in Bogor (then Buitenzorg) from 1927 to 1933. While still a graduate student, he initiated the first quantitative studies of plant hormones, one of the fields of investigation in which his work has been most successful. He spent the next quarter century at the California Institute of Tech- nology, where he developed the first Phytotron and continued to make distinguished scientific contributions, especially in the areas of plant growth and desert plant ecology. As Director of the Missouri Botanical Garden from 1958 to 1963, he was responsible for the first geodesic-dome greenhouse, the Climatron, still a splendid dis- play of tropical plants and an important element in the rejuvenation of that 125-year-old institution. For the past two decades, Frits Went has been pursuing questions of desert ecology at the Desert Research Institute of the University of Nevada. We salute him in his 82nd year for the genius that he has brought to bear on the science of botany during the course of a long and distinguished career. 63 Subscriptions— Membership Membership in the California Botanical Society is open to individuals ($18 per year; students $ 1 0 per year for a maximum of seven years). Members of the Society receive Madrono free. Family memberships ($20) include one ten-page publishing allot- ment and one journal. Emeritus rates are available from the Corresponding Secretary. Institutional subscriptions to Madrono are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Madrono should be sent to the Corresponding Secretary. Information for Contributors Manuscripts submitted for publication in Madrono should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For ar- ticles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed ending with figures, and number each page. Do not use a separate cover page, "erasable" paper, or footnotes. Manuscripts prepared on dot matrix printers may not be considered. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including cap- tions) to the Madrono page, and designed for reduction to 'A original size. Scales should be included in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff card- board and sent flat. No illustrations larger than 22 x 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in Madrono 22:288. 1974. In- stitutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Herbariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Periodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted ten pages in Madrono each year or 20 pages over a two-year period. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $60.00 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing Madrono on a reasonable schedule, with equity among all members for access to its pages. Printer's fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author's changes after typesetting @ $3.00 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. CALIFORNIA BOTANICAL SOCIETY VOLUME 32, NUMBER 2 APRIL 1985 llADRONO A WEST AMERICAN JOURNAL OF BOTANY Contents Southern California White Fir (Pinaceae) Frank C. Vasek Some Floral Nectar-Sugar Compositions of Species from Southeastern Arizona and Southwestern New Mexico C. Edward Freeman and Richard D. Worthington Paronychia ahartii (Caryophyllaceae), a New Species from California Barbara Enter Chromosome Counts in Section Simiolus of the Genus Mimulus (Scrophulariaceae). X. The M. glabratus Complex Robert K. Vickery, Jr., Steven A. Werner, Dennis R. Phillips, and Steven R. Pack The Identity of Cracca Bentham (Fabaceae, Robinieae) in the United States Matt Lavin A New Subspecies of Perennial Linanthus (Polemoniaceae) from the Klamath Mountains, California T. W. Nelson and R. Patterson The Missing Fremont Cannon— An Ecological Solution? Jack L. Reveal and James L. Reveal NOTES AND NEWS Inhibition of Abies concolor Radicle Growth by Extracts of Ceanothus velutinus Susan G. Conard 118 Pitfalls in Identifying Ventenata dubia (Poaceae) Kenton L. Chambers 1 20 NOTEWORTHY COLLECTIONS Arizona 1 2 1 California 1 22 Nevada 1 23 Oregon 1 24 Utah 125 Wyoming 125 ANNOUNCEMENTS 129, 130 ERRATUM 130 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY 65 78 87 91 95 102 106 Madrono (ISSN 0024-9637) is published quarterly by the California Botanical So- ciety, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to James R. Shevock, Botany Dept., California Academy of Science, San Francisco, CA 94 11 8. Christopher Davidson Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 198 5 — Sterling C. Keeley, Whittier College, Whittier, CA Arthur C. Gibson, University of California, Los Angeles 1986— Amy Jean Gilmartin, Washington State University, Pullman Robert A. Schlising, California State University, Chico 1987— J. RzEDOwsKi, Institute Politecnico Nacional, Mexico Dorothy Douglas, Boise State University, Boise, ID 1988— Susan G. Conard, USDA Forest Service, Riverside, CA William B. Critchfield, USDA Forest Service, Berkeley, CA 1989— Frank Vasek, University of California, Riverside Barbara Ertter, University of Texas, Austin CALIFORNIA BOTANICAL SOCIETY, INC. Officers for 1985-86 President: Charles F. Quibell, Department of Biology, Sonoma State University, Rohnert Park, CA 94928 First Vice President: Rodney G. Myatt, Department of Biological Sciences, San Jose State University, San Jose, CA 951 14 Second Vice President: David H. Wagner, Herbarium, Department of Biology, University of Oregon, Eugene, OR 97403 Recording Secretary: Virgil Thomas Parker, Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: James R. Shevock, Department of Botany, California Academy of Sciences, San Francisco, CA 94 1 1 8 Treasurer: Wayne Savage, Department of Biological Sciences, San Jose State Uni- versity, San Jose, CA 951 14 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, Rolf W. Benseler, Department of Biological Sciences, California State University, Hayward, CA 94542; the Editor of Madrono; three elected Council Members: James C. Hickman, Department of Botany, Uni- versity of Cahfomia, Berkeley, CA 94720; Thomas Fuller, 171 West Cott Way, Sacramento, CA 95825; Annetta Carter, Department of Botany, University of California, Berkeley, CA 94720; and a Graduate Student Representative, Joseph M. DiToMASO, Department of Botany, University of California, Davis, CA 95616. SOUTHERN CALIFORNIA WHITE FIR (PINACEAE) Frank C. Vasek Department of Botany and Plant Sciences, University of California, Riverside 92521 Abstract Three lines of evidence are evaluated to determine the varietal relationships of southern California white fir. Based on nine seedling traits studied by Hamrick and Libby, southern California populations first cluster with those from Arizona and Utah, then later cluster with those from northern California. Based on relative amounts of camphene and 3-carene, three equivalent chemical races were found by Smedman et al. (1969) and Zavarin et al. (1975). A greater similarity of the southern California race to the Rocky Mountain race is rather tenuous and depends on the significance of rather small differences in four sesquiterpenes. Based on leaf morphology of adult trees (this study), two broad groups were defined. Northern California populations have long leaves with an emarginate to rounded tip, few stomatal rows on the upper surface, and a definite twist at the leaf base. The other populations have more stomatal rows on the upper leaf surface, acute tips, and very little to no twist at the base. They are distributed from Colorado-New Mexico with long, thin, narrow leaves, to Baja California with short, thick, wide leaves. The southern California populations largely fall within the latter group, are clearly more similar to the Rocky Mountain popu- lations, and hence are referable to Abies concolor var. concolor. White fir, Abies concolor (Gord. and Glend.) Lindl., is a common and widespread montane forest tree in the western United States (Fowells 1965, Hamrick and Libby 1972, Markstrom and Jones 1975, Sudworth 1908, 1916). Regionally-correlated variation has been formalized by the recognition of two varieties: A. concolor var. lowiana in California, and A. concolor var. concolor in the Rocky Mountain region. The pertinent nomenclature has been reviewed by Lamb (1914), Fowells ( 1 965), Liu ( 1 97 1 ), and Hamrick and Libby (1972). On a broader view, the white fir complex is interpreted by Hamrick (1966) and by Hamrick and Libby (1972), on the basis of seedling morphology and growth characteristics in experimental nursery plots, to consist of four population groups: I, A. grandis in Oregon and northwest California; II, A. concolor var. lowiana in the Sierra Ne- vada and northeastern California; III, A. concolor var. concolor in southern California and Arizona; and IV, another form of A. con- color var. concolor in Utah and southern Nevada. All the California white fir have been included in var. lowiana by Griffin and Critchfield (1972) except those in the high mountains of the eastern Mojave Desert. Griffin and Critchfield do not assign varietal names, but they clearly imply a "California" form and a Madrono, Vol. 32, No. 2, pp. 65-77, 26 April 1985 66 MADRONO [Vol. 32 "Rocky Mountain" form on map 3 and on page 10. The Rocky Mountain white fir extends intermittently from the Wasatch Plateau region (Markstrom and Jones 1975, Vasek and Thome 1977) to the Kingston and Clark Mountains of eastern San Bernardino County, California (Henrickson and Prigge 1975, Miller 1940, Mehringer and Ferguson 1969, Munz 1959). The influence of the Rocky Moun- tain white fir extends into southern California and is evident to some extent in populations as far north as the Tehachapi Mountains (Grif- fin and Critchfield 1972). Southern California white fir has been interpreted as var. lowiana by Griffin and Critchfield (1972) and as var. concolor by Hamrick and Libby (1 972). The taxonomic and phytogeographic relationships of southern California white fir therefore remain unsettled. This paper addresses the question of varietal classification of southern California white fir and briefly speculates on its paleohistory. Methods and Materials Two approaches were used. First, information about morphology and growth of seedlings (Table 1 ) and terpenoid chemistry of mature trees (Table 2) was gathered from the literature for evaluation. Second, new information about the morphological variation of mature trees was gathered via small population samples from rep- resentative localities throughout the range (Table 3). For each sam- pled tree, an exposed sun branch was selected and cut off^ with a pole pruner. Ten consecutive leaves were removed from the center of a stem growth segment at least 1 year old (i.e., current season leaves were not sampled). The branch was pressed as a voucher and later scored for degree of pubescence. The leaves were transported in labeled coin envelopes, air dried, and then measured or scored in the laboratory for the first six traits listed in Table 4. The average of 1 0 leaves was used to represent the plant. The first three characteristics— leaf width, leaf thickness, and leaf length— were measured to the nearest 0. 1 mm with the aid of vernier calipers. The shape of the leaf tip was rated on a scale of 1 to 4 from acute to obtuse to rounded to emarginate after Hamrick (1966). The twist at the base of the leaf was rated on a scale of 1 to 5 progressively from no twist to twists of 90°, 180°, 270°, and 360°. The number of rows, or partial rows, of stomata on the upper leaf surface was counted several times under a dissecting microscope until a consis- tent maximum count was obtained. The seventh trait, degree of stem pubescence, was rated on a scale of 0 to 4 respectively for no pubescence, a few scattered hairs, small intermittent patches of hairs, uniformly distributed but not dense (hairs shorter than the spaces between), uniform and dense (hairs 1985] VASEK: WHITE FIR 67 longer than the spaces between). A stepwise discriminant analysis was used to analyze the data (BMDP7M). Results Seedling morphology. A study of variation in 43 test populations was conducted (Hamrick and Libby 1972) by growing seedlings in a nursery. An interpopulation difference index was compiled based on the sum of significant differences for nine characteristics. A matrix of such difference indexes sorts the test populations into four groups. Group I {Abies grandis from Oregon and northwest California) is not of direct concern to the present question. Groups II {A. concolor var. lowiana from northern California), III (A. concolor var. concolor from southern California and Arizona), and IV (A. concolor var. concolor from Utah and eastern Nevada) are of present concern. These groups are each less homogeneous (by inspection of Hamrick and Libby's 1972 data) than Group I. One or two, or very few, populations contribute toward high interpop- ulation difference indexes within each group (Table 1, lower left). Nevertheless, Groups II and III are about equally homogeneous as judged by the average wi thin-group interpopulation difference index (Table 1 , upper right). But Group IV is more heterogeneous and can readily by separated into two components: Utah (BF, BG) and Ne- vada (AW, AX, AY). Furthermore, Arizona population BB is some- what anomalous and could be separated out. Nevertheless, both Utah and Arizona cluster with southern California as one group using the available interpopulation difference index as a clustering tool. Arizona joins first if BB is omitted, and Utah joins first if BB is included. The combined group, an expanded Group III, then clusters next with northern California, leaving the Nevada popula- tions unclustered. On this analysis, the southern California, Arizona, and Utah populations form one group assignable to A. concolor var. concolor. The northern California populations form one group as- signable to A. concolor var. lowiana. The Nevada populations, the reduced Group IV, are the most different but could be considered another form of A. concolor var. concolor because of reasonable similarity to the Utah populations. Terpenoid chemistry. The terpenes in several species of Abies, specifically including the three groups of A. concolor under concern here, were studied by Smedman et al. (1969). Chemical differences support the interpretation that three groups of populations are in- cluded in A. concolor. However, the method employed used bulk resin samples from up to 50 trees, in unspecified sample areas, which permits "good averaging" but precludes any test of significance be- tween averages. Consequently, the results (Table 2) are difficult to interpret. 68 MADRONO [Vol. 32 fi ^ X3 a til ■;4 ^ a o O < < < < < > 70 MADRONO [Vol. 32 Table 2. Terpenoids of Abies concolor. A, Monoterpenes. Data from Smedman et al. (1969) and Zavarin et al. (1975). B, Sesquiterpenes. Data from Smedman et al. (1969). Southern California populations are similar to: C, Northern California; D, Rocky Mountains; E, both; or F, neither. Key: x = trace; — = none; ( ) = tenuous similarity. 1 Q^Q 1 Q7S ly 1 J Similar- ity ivy NC SC RM NC SC RM A a-Pinene 9 10 23 8.9 10.0 16.4 /3-Pinene 53 42 28 66.3 41.2 22.7 (a + /3-Pinene) 62 52 51 75.2 51.2 39.1 D Camphene X X 12 1.3 X 9.4 C 3-Carene 0.5 20 11 0.7 13.1 0.6 D Myrcene 0.5 0.5 8.0 0.4 0.6 1.3 (C) Limonene 2 2 2 2 2 2.6 E |3-Phellandrene 35 26 16 17 25.4 6.8 C Terpineole 0.5 X 0.1 0.2 E B Famesene 3 4 (D) /3-Elemene 5 14 6 F a-Selinene 14 10 10 (D) /3-Selinene 15 6 (F) i8-Bergamotene 8 2 3 D 7-Cadinene 1 8 3 F a-Muurolene 1 3 9 C a-Cubabene 2 21 9 F a-Copaene 1 3 8 C Longifolene 5 2 1 (D) Longicyclene 3 (D) The monoterpenes were studied by Zavarin et al. (1975) in resin from individual trees in widely distributed and amply described sample populations of all three groups (Table 2A). The monoterpene components varied extensively (Zavarin et al. 1975), but medial differences confirmed three population groups. Trees from northern California populations (var. lowiana) produce little camphene and little 3-carene. Trees from Rocky Mountain populations (var. concolor) produce fairly large amounts of cam- phene and 3-carene. Trees from southern California populations produce large amounts of 3-carene but no camphene. The three chemical races were compared by means of a non- weighted similarity index. In terms of monoterpenes, the southern California populations were more similar to var. lowiana (Zavarin et al. 1975) with a similarity index of 0.64. However, inspection of Table 2A indicates significant similarity only in amounts of cam- phene, /5-phellandrene, and perhaps myrcene. The southern Cali- fornia populations were more similar to var. concolor in amounts 1985] VASEK: WHITE FIR 71 of 3-carene and the total of a- and /3-pinene. (Since a- and /5-pinene are stereo-isomers, one increases as the other decreases. Hence, they are better combined as one chemical trait.) The greater similarity to var. lowiana is therefore reduced to a tenuous similarity in one component (myrcene). In terms of sesquiterpenes, the southern California populations approach var. concolor (Rocky Mts.) "being overall closer to the latter" (Zavarin et al. 1975). However, the sesquiterpene similarity index was calculated from the Smedman et al. (1969) data for which no test of significance is available. Given the extreme variability of terpenoid data (see Zavarin et al. 1970, Zavarin 1975), differences in sesquiterpenes (Table 2B) of less than 5-10% are probably not significant. Therefore, similarity of southern California samples to northern California samples seems fairly certain for two compounds, a-muurolene and «-copaene. Sim- ilarity of southern California populations to Rocky Mountain pop- ulations seems fairly certain for only one compound, |5-bergamotene, and somewhat tenuous for four other compounds (Table 2B). The overall greater similarity of southern California to Rocky Mountain populations, as claimed by Zavarin et al. (1975), is demonstrated only if the rather small differences in famesene, a-selinene, longi- folene, and longicyclene are significant. Morphology of mature trees. Two groups of samples are evident (Table 4). The northern California populations stand out in having rather long leaves with a definite twist at the base (sun branches), a rounded to emarginate leaf tip, and a low number of stomatal rows on the upper surface. The other populations are distributed along a line from Colorado-New Mexico to Baja California. The Colorado- New Mexico populations have the narrowest, thinnest, and longest leaves with the most acute tips but with little twist. The Baja Cal- ifornia population has the widest, thickest, and shortest leaves with the very least twist. All the extreme values for the six leaf traits occur in the geograph- ically extreme populations: northern California, or Colorado-New Mexico, or Baja California. The other populations fall in between the above extremes except for the contrasting pattern in stem pu- bescence for which the southern California populations score the lowest and the desert mountain populations score the highest. The variation and distribution patterns for the seven character- istics were analyzed by a stepwise discriminant analysis that ranked the average values for each character for each region. The program selected the most discriminating characters in the sequence: leaf twist, leaf thickness, stomatal rows, leaf length, pubescence, leaf width, and leaf tip shape. The cumulative proportion of the total dispersion was respectively 0.721, 0.898, 0.949, and 0.978 for the 72 MADRONO [Vol. 32 Table 3. Collection Sites. Sam- ple County Locality Date size Northern California (including Kern Co.): Alpine W. Woodfords Sep 1979 8 Alpine Ebbetts Pass Rd.; Silver Crk. C. G. Sep 1979 5 Alpine Ebbetts Pass Rd.; Alpine Lake Sep 1979 5 Calaveras Ebbetts Pass Rd.; Hell's Kitchen oep 1 V / V c J Vista Kern Greenhorn Mts. May 1977 17 Lassen 5W Susanville Sep 1976 5 Mariposa oiN wawona Oct 1981 6 Mariposa 15S Wawona Oct 1981 12 Tulare Giant Forest Sep 1979 6 Tulare General's Highway Sep 1979 7 Tuolumne Sonora Pass Rd.; Strawberry CaTN 1 Q7Q oep ly /y Q O Tuolumne Sonora Pass Rd.; 5E Strawberry ep ly /o 1 r\ San Diego Palomar Mts. Sep 1977 18 Desert Mts. (Nevada-California): Clark (Nev.) Charleston Mts. Alio 1 Q7A z Clark (Nev.) Charleston Mts. Mar 1 Q77 ividi ly / / 1 0 1 u Clark (Nev.) Charleston Mts. Cpn 1 Q7Q ocp ly 1 y u San Bernardino Clark Mts. \/fci\/ 1 078 J (Cal.) San Bernardino New York Mts. Jul 1979 2 (Cal.) Baja California (Mexico): Norte Sierra San Pedro Martir Oct 1 976 16 Arizona: Coconino Mormon Lake Oct 1977 7 Coconino Happy Jack Oct 1977 8 Coconino Miller Ridge Ort 1 077 WCl 17// Coconino IW Baker Butte Oct 1977 5 Coconino Woods Canyon Lake Jet. Oct 1977 14 Gila Colcom Summit (Young) Oct 1977 6 Utah: Iron E. Cedar City Jul 1976 10 Kane Strawberry Peak Jul 1976 4 Salt Lake Big Cottonwood Canyon: Storm Mt. Dec 1977 6 Salt Lake Little Cottonwood Cyn: Snow Bird Dec 1977 6 Salt Lake Big Cottonwood Cyn: 6200 ft Dec 1977 5 Salt Lake Little Cottonwood Cyn: Lisa Falls Dec 1977 6 1985] VASEK: WHITE FIR 73 Table 3. Continued. County Locality Sam- ple Date size Colorado-New Mexico: Rio Grande (Colo.) Rio Grande (Colo.) Costilla (Colo.) Costilla (Colo.) Conejos (Colo.) Rio Ariba (N.M.) Rio Ariba (N.M.) S. Girard S. Del Norte W. LaVeta Pass E. LaVeta Pass N.E. La Manga Pass 9N Chama 3N Chama Sep 1978 Sep 1978 Sep 1978 Sep 1978 Sep 1978 Sep 1978 Sep 1978 Table 4. Mean and Standard Deviation for Seven Morphological Traits OF Abies concolor in Seven Geographical Regions. 1 , northern California; 2, south- em California; 3, desert mountains (Nevada-California); 4, Baja California; 5, Ari- zona; 6, Utah; 7, Colorado-New Mexico, a = scale of 1 to 4 from acute to obtuse to rounded to emarginate; b = scale of 1 to 5 from no twist to twists of 90°, 180°, 270°, and 360°. c = scale of 0 to 4 from no hairs to dense pubescence. Maximum and minimum average values for each trait are underlined. Leaf Stem Geo- Leaf thick- Leaf Tip Leaf Rows pubes- graphic width ness length shape twist stomata cence area n (mm) (mm) (mm) (a) (b) (no.) (c) Means 1 101 2.06 0.90 34.24 2^ 16.3 2.0 2 55 2.05 0.88 26.23 2.1 1.3 17.1 _L2 3 25 2.07 0.82 23.39 2.2 1.0 17.5 23_ 4 16 2.46 1.16 22.48 2.0 LO 21.4 2.3 5 55 2.12 0.79 32.67 2.1 1.2 20.1 1.5 6 37 2.07 0.79 32.79 2.0 1.1 18.3 2.4 7 54 1.95 0.76 34.59 L9 1.1 19.9 2.8 Total 343 2.07 0.85 31.26 2.2 1.5 18.1 2.0 Standard deviations 1 0.16 0.10 7.31 0.41 0.52 3.06 0.92 2 0.19 0.08 6.32 0.33 0.33 2.84 0.99 3 0.22 0.12 3.75 0.44 0.06 2.54 1.04 4 0.22 0.14 3.67 0.05 0.04 2.52 1.13 5 0.22 0.09 5.52 0.29 0.18 2.47 1.42 6 0.17 0.09 5.88 0.12 0.14 2.35 1.75 7 0.16 0.09 5.20 0.19 0.13 2.34 0.84 Total 0.18 0.10 6.07 0.32 0.33 2.70 1.14 74 MADRONO [Vol. 32 7.50- 6 25 - 5.00 - -5.00 - i -750 -600 -4.50 -3.00 -1.50 0.00 1.50 300 4.50 6j00 750 9.00 CANONICAL VARIABLE I Fig. 1 . Distribution of white fir samples relative to two canonical variables. Num- bers represent group means; symbols represent individual trees: northern California, 1, 0; southern California, 2, •; desert mountains, 3, A; Baja California, 4, A; Arizona, 5, □; Utah, 6, ■; Colorado-New Mexico, 7, 4. Asterisks indicate points with oc- currence of two or more plants. Distribution of southern California is enclosed by a four-sided figure, northern California by a six-sided figure. first four characters entered. The remaining characters effected neg- Ugible dispersion, although all seven had significant F values. These characters were combined into two canonical variables against which all the individual samples were plotted (Fig. 1) as a formalized de- scription of character distribution. On this scheme, the northern California populations fell out to the right and are completely distinct from all other populations except for a very slight overlap with southern California. The remaining populations fall in a line between the Baja California populations and the Colorado-New Mexico pop- ulations. The southern California populations are slightly off line toward the right but clearly overlap strongly with populations from Arizona, Utah, and especially the desert mountains. Discussion Variation in white fir has now been studied from several view- points. On the basis of seedling morphology, Hamrick and Libby (1972) proposed several groups of white fir: Abies concolor var. 1985] VASEK: WHITE FIR 75 lowiana in northern California, A. concolor var. concolor in the Rocky Mountains, and another form of A. concolor var. concolor in southern CaUfomia-Arizona. Examination of their data indicates that some groups, especially Group IV, Utah-Nevada, are markedly heterogeneous. Simple clustering, based on interpopulation differ- ence indexes, places the Utah populations in an expanded Group III with those from Arizona and southern California. The expanded Group III is more similar to Group II, northern California, than to the Nevada populations residual in Group IV. The reorganized Group III indicates a stronger relationship among populations from south- em California, Arizona, and Utah than Hamrick and Libby (1972) had proposed. On the basis of terpenoid chemistry, Zavarin et al. (1975) dem- onstrate three chemical races. The extreme variation in chemical composition makes difficult an unambiguous interpretation of the relationship among the three chemical races. Nevertheless, Zavarin et al. (1975) suggest a somewhat greater overall similarity of the southern California populations to those in Arizona and the Rocky Mountains than to those in northern California. On the basis of mature tree leaf morphology, the present study shows two groups of populations, the northern California (mostly Sierra Nevada) populations, corresponding to A. concolor var. low- iana, and all the other populations being definitely referred to A. concolor var. concolor. Populations of the latter variety are distrib- uted along a line (Fig. 1) from Colorado-New Mexico to Baja Cal- ifornia, a pattern that approximates a latitudinal gradient. (Inversion and reversal of Fig. 1 emphasizes the geographical correlation.) The southern California populations fall slightly offline toward the north- em Califomia populations. A slightly higher incidence of leaves with a basal twist (Table 4) may account for the difference. Furthermore, the individual trees within the overlap area with northem Califomia all have a leaf twist and all occur on Palomar Mountain. Their occurrence there may suggest some ancient introgression at a time when Sierran forests extended southward (early Quatemary?) and met Rocky Mountain forests that had extended southwest across the southem Great Basin (Miocene to late Pliocene?) (Axelrod 1976). The rather close relationship between southem Califomia and Arizona populations, as proposed by Hamrick (1966) and Hamrick and Libby (1972), and confirmed in the present study, is rather surprising in view of the long separation of the two areas by the intervening Sonoran Desert and the current climatic differences (Ax- elrod 1976). Nevertheless, several plants evidence phytogeographic relationships between the two areas. For example, Rhus ovata occurs in chaparral, and Arctostaphylos pringlei occurs in montane forests or woodlands, of the two areas (Keamey and Peebles 1960, Munz 1959). In addition, Acer grandidentatum brachypterum, known from 76 MADRONO [Vol. 32 early Pleistocene fossils in the San Jacinto Mountains (Axelrod 1 966), presently occurs in the mountains near Prescott in west-central Ar- izona. Among herbaceous plants, Clarkia rhomboidea occurs in montane forests of both the Rocky Mountain region and the Cali- fomian region (Mosquin 1964). Analysis of translocation hetero- zygosity in interpopulational hybrids of C. rhomdoidea indicates disjunct distribution of a "southern type" in Arizona and southern California and a skewed horseshoe-shaped distribution of a "north- em type" in mountains around the Great Basin. Mosquin (1964) interprets that late Wisconsin distributions in broad, shallow horse- shoe-shaped bands across the Great Basin region, and subsequent retreat from the Great Basin upon increase in aridity, would account for the present distribution and disjunctions. Since the montane forests of central Arizona and southern Cali- fornia have long been separate (Axelrod 1976), the similarity of white fir trees of the two regions may derive from a distribution similar to that proposed by Mosquin (1964) for Clarkia in Late Pleistocene. In the case of white fir, links across the present desert may have occurred prior to the Late Pleistocene, as implied by Wells and Berger (1967) and Mehringer and Ferguson (1969), and may extend clear back to the Miocene, when conifer forests were widespread in the Great Basin (Axelrod 1976). A forested region is also indicated by late Pliocene-early Pleistocene pollen floras from the Coso Moun- tains, Panamint Valley, Little Lake, and other stations in the north- em Mojave Desert (Axelrod and Ting, 1 960, 1961). Rocky Mountain forest vegetation most probably ranged to southem Califomia well before the Pleistocene and left populations of white fir that clearly are referable to Abies concolor var. concolor. Acknowledgments I wish to thank the following for their aid and encouragement: C. Huszar for statistical aid; S. James, J. Lippert, and N. Smith-Huerta for technical assistance; A. Wyckoff, A. Sanders, M. Vasek, L. LaPre, P. Rowlands, and O. Clarke for collecting some of the samples; J. R. Haller, E. Zavarin, and R. Scora for helpful discussion; and B. Bickler for typing the manuscript. Literature Cited Axelrod, D. I. 1966. The Pleistocene Soboba flora of southem Califomia. Univ. Calif Publ. Geol. Sci. 60:1-79. . 1 976. History of the coniferous forests, Califomia and Nevada. Univ. Calif Publ. Dot. 70:1-62. Powells, H. A. 1965. Silvics of forest trees of the United States. USDA For. Serv. Agric. Handb. 271, Washington, D.C. Griffin, J. R. and W. B. Critchfield. 1972. The distribution of forest trees in Califomia. U.S. For. Serv. Res. Paper PSW-82. Hamrick, J. L. 1966. Geographic variation in white fir. M.S. thesis, Univ. Cahf., Berkeley. 1985] VASEK: WHITE FIR 77 and W. J. LiBBY. 1972. Variation and selection in western U.S. montane species. I. White fir. Silvae Genet. 21:29-35. Henrickson, J. and B. Prigge. 1975. White fir in the mountains of Eastern Mojave Desert of California. Madrono 23:164-168. Kearney, T. H. and R. H. Peebles. 1 960. Arizona flora. Univ. Calif Press, Berkeley. Lamb, W. M. 1914. A conspectus of North American firs (exclusive of Mexico). Soc. Amer. Foresters Proc. 9:528-538. Liu, T. 1971. A monograph of the genus Abies. Dept. of Forestry, College of Ag- riculture, National Taiwan Univ., Taipai. Markstrom, D. C. and J. R. Jones. 1975. White fir— an American wood. USDA For. Serv. Bull. 237. Mehringer, p. J. and C. W. Ferguson. 1969. Pluvial occurrence of bristlecone pine {Pinus aristata) in a Mohave Desert mountain range. J. Arizona Acad. Sci. 5:284-292. Miller, A. H. 1940. A transition island in the Mohave Desert. Condor 42:161- 163. MosQUiN, T. 1964. Chromosomal repatteming in Clarkia rhomboidea as evidence for post-Pleistocene changes in distribution. Evolution 18:12-25. MuNZ, P. A. 1959. A California flora. Univ. Calif Press, Berkeley. Smedman, L. a., K. Snajberk, E. Zavarin, and T. R. Mon. 1969. Oxygenated monoterpenoids and sesquiterpenoid hydrocarbons of the cortical turpentine from different Abies species. Phytochemistry 8:1471-1479. Sudworth, G. B. 1908. Forest trees of the Pacific slope. USDA For. Serv. U.S. Govt. Printing Office, Washington, D.C. . 1916. The spruce and balsam fir trees of the Rocky Mountain region. USDA Bull. 327. Vasek, F. C. and R. F. Thorne. 1977. Transmontane coniferous vegetation. In J. Major and M. G. Barbour, eds., Terrestrial vegetation of California, Chap. 23. Wiley-Interscience, NY. Wells, P. V. and R. Berger. 1 967. Late Pleistocene history of coniferous woodland in the Mohave Desert. Science 155:1640-1647. Zavarin, E. 1975. The nature, variability and biological significance of volatile secondary metabolites from Pinaceae. Phytochem. Bull. 8:6-15. , K. Snajberk, and J. Fisher. 1975. Geographic variability of monoterpenes from Abies concolor. Biochem. Syst. Ecol. 3:191-203. , K. Snajberk, T. Reichert, and E. Tsien. 1970. On the geographic variabihty of the monoterpenes from the cortical blister oleoresin of Abies lasiocarpa. Phy- tochemistry 9:377-395. (Received 15 March 1984; accepted 29 July 1984.) SOME FLORAL NECTAR-SUGAR COMPOSITIONS OF SPECIES FROM SOUTHEASTERN ARIZONA AND SOUTHWESTERN NEW MEXICO C. Edward Freeman and Richard D. Worthington Department of Biological Sciences, University of Texas, El Paso 79968-0519 Abstract The floral nectar-sugar compositions of 34 species from southeastern Arizona and southwestern New Mexico were determined by high-performance Uquid chromatog- raphy (HPLC). Of these, 26 species have not been reported previously. Among the species surveyed were hummingbird flowers (18 species), hawkmoth flowers (seven species), bee flowers of various kinds (five species), butterfly flowers (two species), and those whose pollinators were not known (two species). Both the hummingbird and hawkmoth nectars were high in sucrose, averaging 71% and 81% respectively. The nectars of the purportedly butterfly flowers were very different at 76% and 46% sucrose. Bee nectars from large-flowered species with large corolla tube openings were high in sucrose (average = 76%) whereas small-flowered bee species with small corolla tube openings were lower in sucrose (average = 35%). It is clear from the work of Baker (1978) and Baker and Baker (1975, 1979, 1983) that the sugar composition of floral nectars is worthy of careful examination in regard to its differential attrac- tiveness to various groups of potential pollinators. It has been found that the most common sugars in nectars are the hexoses, glucose and fructose, and the disaccharide sucrose. These are the so-called "big three" sugars of nectars (Baker and Baker 1983). Other sugars are occasionally present in very small amounts. We present here the floral nectar-sugar compositions of a series of species from southeastern Arizona and southwestern New Mexico. The pollinators of many of these species have been studied but their sugar compositions have not been reported. Methods and Materials Sugar compositions by mass were quantified using HPLC. The methods employed are described in earlier publications (Freeman et al. 1983, Freeman et al. 1984). HPLC was used because the anal- ysis is direct, eliminating the derivatization steps of other techniques that add to the error term. This makes HPLC much more accurate. In addition, HPLC is much more rapid. Sugar compositions were calculated to the nearest 0.1%. Voucher specimens are deposited at UTEP. Nomenclature follows Lehr (1978) and Lehr and Pinkava (1980). Madrono, Vol. 32, No. 2, pp. 78-86, 26 April 1985 1985] FREEMAN AND WORTHINGTON: FLORAL NECTAR 79 Results and Discussion The nectar-sugar compositions of 34 species are presented in Table 1. Freeman et al. (1984) have previously reported on seven species {Bouvardia glaberrima, Epilobium canum subsp. latifolia [as Zauschneria latifolia], Fouquieria splendens, Mimulus cardinalis, Penstemon barbatus, P. pseudospectabilis, and Silene laciniata). Sherbrooke and Schereens (1979) have reported on Erythrina Jla- belliformis. About half of the species surveyed in this study are known or suspected to be hummingbird-pollinated. These include Anisacanthus thurberi, Aquiliegia triternata, Bouvardia glaberrima, Castilleja patriotica, Erythrina flabelliformis, Fouquieria splendens, Lonicera arizonica. Lobelia cardinalis, Mimulus cardinalis, Penste- mon barbatus, P. pinifolius, P. pseudospectabilis, Ribes pinetorum. Salvia lemmoni, Silene laciniata, and Epilobium canum subsp. la- tifolium. These nectars ranges in sucrose composition from 55% (Mimulus cardinalis) to 93% (Ribes pinetorum). The mean value of 71% sucrose is very similar to the means of other groups of hummingbird flowers (Freeman et al. 1984, Free- man, unpubl. data). In addition, many of these species also have a hexose imbalance; i.e., fructose is present in much larger quantities than glucose. A reversed imbalance is found in Aquilegia triternata, which has about twice as much glucose as fructose. In this species hexoses are present in very low quantities. The sugar composition of the floral nectar of Erythrina flabelliformis collected in this study from the Dragoon Mountains of Arizona is virtually identical with a sample collected and analyzed earlier from the same mountain range (Sherbrooke and Schereens 1979) also using HPLC as the analysis technique. Several species previously reported by Freeman et al. (1984) from other localities had very similar sugar composi- tions in the study area, including Bouvardia glaberrima, Penstemon barbatus, and P. pseudospectabilis. Others varied somewhat. The sucrose composition of Silene laciniata in the Chiricahua Mountains averaged about 1 0% higher than a sample of the same species from the White Mountains, also in eastern Arizona. Mimulus cardinalis and Epilobium canum subsp. latifolium nectars, however, were low- er in sucrose in southeastern Arizona (Freeman et al. 1984). Lonicera arizonica was very similar in sugar composition to a sample of L. involucrata, also hummingbird-pollinated, reported previously (Freeman et al. 1984). Brown and Kodrick-Brown (1979) report a population of Lobelia cardinalis in the White Mountains of eastern Arizona that did not produce nectar. They suggested that this population was mimetic to several common hummingbird-pollinated species in the area. Other populations of that species in the Chiricahua Mountains in south- eastern Arizona and near Montezuma Wells National Monument MADRONO [Vol. 32 0 as od oo ^ in in (N* d V3 8^ vo ON oo q (N r-! cn in mi ob +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 q — 00 O 00 in ON in >n in od d in od as O NO en NO (N NO d r«S (N ro d in +1 +1 +1 +1 +1 +1 +1 +1 + 1 r- (N r~-" (N in in rn en ON ON d 1 " O § § g o 6 Q fS iS 43 Q r a O ■55) U C3 O E .1 6 C3 O d o XI o S o S Q I 5 1985] FREEMAN AND WORTHINGTON: FLORAL NECTAR 8 1 S -i-i U+l 2 o o 6 -7= ^ o ^ o + 1 +1 00 od o o $5 ^ .52 a -c o s la § I S O (/5 O |x I i •c s: o cd 82 MADRONO [Vol. 32 (U I/) « +1 2 o o o O od +1 +1 — ; (N > ^ ^ U •2 la Co oo O oo Os q q 3 . S -1 cd 1985] FREEMAN AND WORTHINGTON: FLORAL NECTAR 83 2-0 2 o ,6 ± 0.6 .3 ± 11.4 .6 ± 7.7 ,7 ± 1.1 60.9 30.1 ,4 ± 16.6 ,0 ± 8.4 ,4 ± 8.4 76.4 (N* o 00 o >n vo r- vo m vo O r- d -f-l vd +1 +1 ^. d vq vd Ov ON CO ^ m vd CN Ov OO m o d +1 +1 +1 q d +1 q rn od 00 +1 vq - Ui 1 1 a < I- ? 2 > 3 S d i J ^ o 84 MADRONO [Vol. 32 in central Arizona do produce nectar. We found a population of L. cardinalis in the Guadalupe Mountains of trans-Pecos Texas that also did not produce nectar, even when the plants were removed from the field and grown in a greenhouse. We found ample pro- duction of nectar in a population sampled in the Huachuca Moun- tains in this study, although the nectar was rather low in sucrose for a hummingbird flower (57%). The hawkmoth-pollinated taxa sampled in this study include Ipo- mopsis longijlora, I. thurberi (Grant and Grant 1968), Castilleja sessiliflora (Cruden et al. 1983), Datura meteloides (Grant 1983, Grant and Grant 1983), Aquilegia chrysantha (Miller 1982), Oeno- thera caespitosa (Grant 1983), and presumably Calylophus hartwegii because of its great similarity to the other hawkmoth-pollinated Oenothera species (Cruden et al. 1983). All have the high-sucrose nectars described by Baker and Baker (1983) for this pollinator syndrome, averaging 81%. Ipomopsis thurberi has the lowest sucrose composition among this group and its overall composition is very similar to that of the closely related hummingbird taxon /. aggregata (Freeman et al. 1984, Freeman, unpubl. data), from which it differs only in color. Castilleja sessiliflora has a nectar-sugar composition like the other hummingbird-pollinated Castilleja species studied to date (Freeman et al., 1984, Freeman, unpubl. data). Two of the three populations for which floral morphology suggests large-bee pollination had high-sucrose nectars. Penstemon steno- phyllus and one population of a similar species, P. dasyphyllus, had 71% and 61% sucrose, respectively. In contrast, another population of P. dasyphyllus averaged only 38% sucrose. The reason for this difference in populations separated by only about 130 km and at the same latitude is not known. The smaller flowered P. linarioides, however, produces a nectar much lower in sucrose, and is perhaps pollinated by short-tongued bees, as its size and nectar-sugar com- position suggest (Baker and Baker 1983). Schaffer and Schaffer (1 977) studied nectar secretion and pollinators of four species of Agave from Arizona, of which three {A. schottii, A. toumeyana, and A. parvijlord) are in subgenus Littaea and are probably closely related. While the three species vary in time of daily nectar secretion and sugar concentrations, all are pollinated by large bees (Bombus so- norus and Xylocopa arizonensis). Agave parviflora, reported here, has a nectar-sugar composition like those of A. schottii and A. tou- meyana, which are atypical of the agaves surveyed to date (Freeman et al. 1983). The pollinators of Swertia radiata (or Frasera speciosd) have been studied in detail by Beattie et al. (1973). A wide variety of insects, primarily hymenopterans, dipterans, and lepidopterans, visited flowers of this species in Colorado. The nectar-sugar com- position is very similar to the majority of the bee-pollinated taxa in this study. 1985] FREEMAN AND WORTHINGTON: FLORAL NECTAR 85 Two purportedly butterfly-pollinated species, Nyctaginea capitata (Cruden et al. 1983) and Ipomopsis macombii (Grant and Grant 1965), were sampled. Baker and Baker (1983) have described but- terfly nectars as being predominately either sucrose-rich or -domi- nated. Nyctaginea capitata, at 76% sucrose, fits that description. However, /. macombii, at an average of 46% sucrose, is considerably more hexose-rich. A study of butterfly and skipper nectars, utilizing more sensitive analytical techniques, is needed in order to more adequately define them statistically. Only then will it be possible to determine if either of these nectars is anomalous. Hedeoma hyssopifolium, and Mertensia franciscana are conifer- ous forest species with unknown pollinators. Hedeoma hyssopifo- lium has the typical high-sucrose nectar of the family Lamiaceae (Baker and Baker 1983). The nectar of Mertensia is low in sucrose, which, along with open flowers, suggests pollination by short-tongued bees or flies. Acknowledgments We wish to thank the Research Ranch Foundation of Elgin, Arizona, for a summer fellowship grant to CEF during 1983 when the nectar samples were collected. We also thank M. Murray for collecting nectar samples from the Dragoon Mountains of Arizona. Literature Cited Baker, H. G. 1978. Chemical aspects of the pollination biology of woody plants in the tropics. In P. B. Tomlinson and M. H. Zimmerman, eds., Tropical trees as living systems, p. 57-82. Cambridge Univ. Press, Cambridge. and I. Baker. 1975. Nectar constitution and pollinator-plant coevolution. In L. E. Gilbert and P. H. Raven, eds.. Animal and plant coevolution, p. 100- 140. Univ. of Texas Press, Austin. and . 1979. Sugar ratios in nectars. Phytochem. Bull. 12:43-45. and . 1982. Chemical constituents of nectar in relation to pollinator mechanisms and phylogeny. In M. N. Nitecki, ed., Biochemical aspects of evo- lutionary biology, p. 131-171. Univ. of Chicago Press, Chicago. and . 1983. Floral nectar sugar constituents in relation to pollinators. In C. E. Jones and R. J. Little, eds.. Handbook of experimental pollination biology, p. 117-141. Van Nostrand-Reinhold Co., NY. Beattie, a. J., D. E. Breedlove, and P. H. Raven. 1973. The ecology of the pollinators and predators of Frasera speciosa. Ecology 54:81-91. Brown, J. H. and A. Kodrick-Brown. 1979. Convergence, competition, and mim- icry in a temperate community of hummingbird-pollinated flowers. Ecology 60: 1022-1035. , , T. G. Whitham, and H. W. Bond. 1981. Competition between hummingbirds and insects for the nectar of two species of shrubs. Southw. Nat- uralist 26:133-145. Cruden, R. W., S. M. Hermann, and S. Peterson. 1983. Patterns of nectar pro- duction and plant-pollinator coevolution. In B. Bentley and T. Elias, eds.. The biology of nectaries, p. 80-125. Columbia Univ. Press, NY. Freeman, C. E., W. H. Reid, and J. E. Becvar. 1983. Nectar sugar composition in some species of Agave (Agavaceae). Madroiio 30:153-158. , , , and R. Scogin. 1984. Similarity and apparent convergence 86 MADRONO [Vol. 32 in the nectar-sugar composition of some hummingbird-pollinated flowers. Bot. Gaz. 145:132-135. Grant, K. A. and V. Grant. 1965. Flower pollination in the Phlox family. Co- lumbia Univ. Press, NY. and . 1 968. Hummingbirds and their flowers. Columbia Univ. Press, NY. Grant, V . 1983. The systematic and geographical distribution of hawkmoth flowers in the temperate North American flora. Bot. Gaz. 144:439-449. and K. A. Grant. 1983. Behavior of hawkmoths on flowers of Datura meteloides. Bot. Gaz. 144:439-449. Lehr, J. H. 1978. A catalog of the flora of Arizona. Northland Press, Flagstaff", AZ. and D. J. Pinkava. 1980. A catalog of the flora of Arizona, Supplement I. J. Arizona-Nevada Acad. Sci. 15:19-32. Miller, R. B. 1982. Hawkmoth pollination of Aquilegia chrysantha in southern Arizona. Bot. Soc. Amer. Misc. Publ. No. 162, p. 39. ScHAFFER, W. M. and M. V. Schaffer. 1977. The reproductive biology of Aga- vaceae: I. Pollen and nectar production in four Arizona agaves. Southw. Natu- ralist 22:157-168. Sherbrooke, W. C. and J. C. Schereens. 1979. Ant- visited extrafloral (calyx and foliar) nectaries and nectar sugars of Erythrina Jlabelliformis Kearney in Arizona. Ann. Missouri Bot. Gard. 66:472-481. (Received 6 April 1984; accepted 21 August 1984.) PARONYCHIA AHARTII (CARYOPHYLLACEAE), A NEW SPECIES FROM CALIFORNIA Barbara Ertter Plant Resources Center, University of Texas, Austin 78712 Abstract A tiny annual Paronychia that has been known from the Sacramento Valley of California for 46 years is described as Paronychia ahartii. Its affinities appear to be with the Mediterranean P. ambica, from which it is nevertheless clearly differentiated by the erect, scarious, bilobed apices of its sepals. In the course of studying a new variety of dwarf rush from the Peter Ahart ranch in Butte County, California, the following incon- spicuous Paronychia was brought to my attention with the request that I provide a name and description for it. The request is herewith fulfilled. Paronychia ahartii Ertter, sp. nov. Herbae annuae minutae glomerulum argenteum ca. 1 cm diam. efformantes. Sepala elliptico-lanceolata basi pilis uncinatis induta, costa in aristum ca. 1 mm longa exeunti, marginibus latis hyalinis ultra costam productis et inter se in limbum erectum bilobum ca. 1 mm longum coadunatis (Fig. 1). Inconspicuous annual plants 0.5-1.2 cm tall, 0.5-1.8 cm across, consisting of a tight silvery glomerule dominated by stipules, bracts, and sepals, arising from a slender taproot; leaves linear to oblan- ceolate, drying reddish-stramineous, 2.5-7.5 mm long, 0.5-1.2 mm wide, often inconspicuously ciliate, tipped with a colorless awn to 0.8 mm long; stipules and bracts similar, conspicuous, concealing flowers, scarious, broadly ovate, 3-6 mm long, 2-4 mm wide, acute to acuminate; flowers few, sessile, 4.2-5 mm long; hypanthium 0.5- 1 mm long, often dark red-brown resinous-papillate below the free portion of the sepals; sepals 5, lanceolate to elliptic, 3.5-4.5 mm long, 1.5-2.5 mm wide, the herbaceous midrib linear, 2.5-3 mm long, 0.2-0.5 mm wide, green to stramineous, sometimes red-flecked proximally, terminated by a spreading colorless awn 1-1.5 mm long, the edges of the midrib thickened, covered with upwardly spreading hairs with tightly coiled tips, the conspicuous scarious margins 0.5- 1 mm wide on each side of the midrib, united beyond the midrib Madrono, Vol. 32, No. 2, pp. 87-90, 26 April 1985 88 MADRONO [Vol. 32 Fig. 1 . Paronychia ahartii. A. Habit. B. Flower, with enlargement of crozier-tipped hair. C. Inside of sepal, showing stamen and staminodia. D. Utricle with seed. to form an erect scarious tip 1-1.5 mm long, the apical 0.5 mm bilobed; staminodia (petals?) filiform, ca. 1 mm long, equaling or exceeding the stamens; filaments flattened, 0.5-1 mm long; anthers ovoid, 0.2 mm long, orange; styles (including stigmatic portion) ca. 0.5 mm long, bilobed, persistent; fruit a. thin-walled utricle with a compressed ovoid body 1.3 mm long and beak ca. 0.5 mm long, the apex papillose; seed lenticular, ca. 1 mm long, brown, borne on a flattened curved funiculus ca. 1.5 mm long. Type: CALIFORNIA. Tehama Co., 8.5 mi s. of Corning, rolling plains, 12 Jun 1955, /. T. Howell 30307. (Holotype: CAS; isotypes: B, GH, K, NY, RM, TEX, U, US.) Paratypes: CALIFORNIA. Butte Co., Ahart Ranch, Honcut, 5 1985] ERTTER: NEW PARONYCHIA 89 May 1974, Ahart s.n. (CAS); same, 8 May 1980, Enter et al. 3326 (NY); 3 mi n. of Chico, 24 Apr 1938, Hoover 3242 (CAS); Hwy 99 between Williams and Oroville, 19 Apr 1958, Langenheim 4480 (JEPS). Shasta Co., Hwy 44 ca. 1 1 mi se. of Millville, 23 Apr 1958, Bacigalupi et al. 6318 (JEPS, TEX). Tehama Co., ca. 7 mi s. of Coming, 22-23 Apr 1958, Bacigalupi et al. 6290 (JEPS, TEX); Jel- ley's Ferry, 20 May 1942, Hoover 5879 (CAS). Very rare on poor clay of swales and higher ground around vernal pools in the northern Sacramento Valley of Butte, Shasta, and Te- hama Counties, California, from 30 to 500 m elevation. Flowering from April to June. Although this diminutive species has been collected several times since R. F. Hoover first discovered it in 1938, these collections remained unidentified beyond tentative placement in Paronychia. Rimo Bacigalupi and Alice Howard worked with the specimens, but no publications resulted from their studies. Although it was sus- pected of being another example of a Mediterranean annual estab- lished in California, with the appearance of Chaudhri's (1968) world- wide monograph of the genus it became evident that the collections did indeed represent a distinctive new species. In Chaudhri's monograph, Paronychia ahartii would be associated with P. arabica (L.) Del. subsp. annua (Del.) Maire & Weiller var. annua of the Middle East and north Africa. According to Chaudhri, P. arabica "is the most variable species of this genus, and, for that matter, of the entire subtribe, and exhibits marked variability in duration, leaf form, length of stipules, size and form of the glomer- ules, bracts, flower-size, and the form of the tepals as well as the structure of their awns and the anthers." Nevertheless, P. ahartii is easily distinguished from P. arabica and appears to be unique in the genus by virtue of the prominent, erect, bilobed apices of the sepals formed by the prolongation of the broad scarious margins beyond the awned herbaceous midrib. Its extremely reduced size is also unusual. It might at first seem curious that a rare Califomian endemic could have a Mediterranean progenitor. Nevertheless a comparable situ- ation involving North American derivatives of a basically Mediter- ranean genus can be found in Loeflingia, as summarized by Bameby and Twisselmann (1971). Paronychia franciscana Eastwood, the only other species of Paro- nychia in California, is considered to be introduced from Chile (Munz 1959). This species is a coastal, mat-forming perennial to 4 dm across, with herbaceous sepals lacking scarious margins. Confusion with P. ahartii is therefore not likely. If Paronychia ahartii is indeed as rare as it appears to be, its continued existence is precarious. Not only is it restricted to the highly developed Sacramento Valley, but a reproduction rate of less 90 MADRONO [Vol. 32 than ten seeds per individual does not bode well under any circum- stances. It is in response to Lowell Ahart's plea for a name to place in his checklist of the flora of his ranch that this species is finally being described. The epithet honors Ahart (b. 1938) not only for his per- sistence and interest in this inconspicuous plant, but also in recog- nition of his careful collections of the flora of the Sacramento Valley and Sierran foothills. Acknowledgments The cooperation, encouragement, and advice of John Thomas Howell, Rupert Bameby (who helped with the Latin), James Hickman, and Lowell Ahart are all greatly appreciated. I am especially appreciative of the drawings done by Julia Larke. Literature Cited Barneby, R. C. and E. C. Twisselmann. 1971. Notes on Loeflingia (Caryophyl- laceae). Madrofio 20:398-408. Chaudhri, M. N. 1968. A revision of the Paronychiinae. Meded. Bot. Mus. Herb. Rijks Univ. Utrecht 285:1-440. MuNZ, P. A. 1959. A California flora. Univ. California Press, Berkeley, CA. (Received 5 March 1984; accepted 19 July 1984.) CHROMOSOME COUNTS IN SECTION SIMIOLUS OF THE GENUS MIMULUS (SCROPHULARIACEAE). X. THE M. GLABRATUS COMPLEX Robert K. Vickery, Jr., Steven A. Werner, Dennis R. Phillips, and Steven R. Pack Department of Biology, University of Utah, Salt Lake City 84112 Abstract Chromosome numbers were ascertained from aceto-carmine squash preparations for members of the Mimulus glabratus complex that had been little studied previously. Representative populations of M. glabratus vaT.fremontiifrom Chihuahua, Durango, and Baja California Sur were found to have « = 15 chromosomes. Populations from Colombia, the disjunct South American range of M. glabratus var. glabratus, have « = 3 1 chromosomes. Populations from Peru of M. andicolus and M. pilosiusculus have « = 46. An intergrading population between M. glabratus var. glabratus {n = 3 1 ) and M. andicolus (n = 46) was found at Pasto on the southern border of Colombia. This cytological study is an integral part of our long-range, ex- perimental studies on the evolution of species in Mimulus (Vickery 1950, 1964, 1978). The chromosome counts here reported are for populations of the M. glabratus complex of related species— M glabratus H.B.K. (and its varieties), M. andicolus H.B.K., and M. pilosiusculus H.B.K. Not only do these counts help provide baseline data for the larger project, but they are of intrinsic interest for a better understanding of this highly polymorphic and plastic complex. Materials and Methods The study populations sampled areas of the Western Hemisphere range of the complex that had been little studied previously (McArthur et al. 1 972, Vickery 1978), although they represent some of the main taxa comprising the complex (Table 1). Cultures of 20 to 30 plants of each population were grown in the University of Utah greenhouse. The chromosome counts were obtained from aceto-carmine squash preparations of pollen mother cells as before (Mia et al. 1964, McArthur et al. 1972). Twenty or more cells were studied from five or more plants of the culture of each population. Representative cells were recorded with sketches, camera lucida drawings or pho- tographs (Fig. 1). Madrono, Vol. 32, No. 2, pp. 91-94, 26 April 1985 92 MADRONO [Vol. 32 Table 1 . Chromosome Counts in the Mimulus glabratus Complex of Related Species and Varieties. All populations, except as noted, were collected by R. K. Vickery, Jr. and grown under his culture numbers. Vouchers are in the Garrett Her- barium of the University of Utah (UT). Mimulus glabratus var. fremontii (Bentham) Grant. « = 1 5 Cuauhetemoc, Chihuahua, Mexico, 2060 m, culture no. 12183; Aldama, Chihua- hua, Mexico, 1 1 50 m, culture 12185; Durango, Durango, Mexico, 1677 m, culture 12215; San Bertola Oasis, Baja Cahfomia Sur, Mexico, 75 m, culture 12223. Mimulus glabratus H.B.K. var. glabratus. « = 31 Sierra de Toluca, Toluca, Mexico, 2830 m, culture 7306; Duitama, Dept. Boyaca, Colombia, 2760 m, culture 13021; Aquitania, Dept. Boyaca, Colombia, 2975 m, culture 13026; Lago Tota, Dept. Boyaca, Colombia, 3010 m, culture 13029. Mimulus andicolus H.B.K. n = A6 Rio Grande, Dept. Ancash, Peru, 3000 m, culture 13066 {Emma Cerrate de Ferreya #6547); Anta, Dept. Cuzco, Peru, 3468 m, culture 13096 (Leonardo Florez 3/7/ 81). Mimulus andicolus H.B.K. x M. glabratus H.B.K. var. glabratus. n = 40-48 Pasto, Dept. Nariiio, Colombia, 2750 m, culture 13033 has n = 40-48 typically, but ranges from n = 3\ to/7 = 55 chromosomes (the median is between « = 44 and n = 45). Mimulus pilosiusculus H.B.K. n = 46 Thermas Baiios de Yura, Dept. Arequipa, Peru, 2475 m, culture 13069; Bolneario Tingo, Dept. Arequipa, Peru, 2250 m, culture 13070; Chilina, Dept. Arequipa, Peru, 2350 m, culture 13071. Results and Discussion Mimulus glabratus var. fremontii (Benth.) Grant has the diploid 71=15 chromosome number (see Table 1 and Vickery 1978) throughout its range from eastern Canada to western Mexico, except for a single questionable 2n = 2S count from Manitoba (Love and Fig. 1 . Anaphase I configurations of pollen mother cells from plants of the in- tergrading population from Pasto, Colombia (culture 13033). 1985] VICKERY ET AL.: CHROMOSOMES IN MIMULUS 93 Table 2. Chromosome Counts Observed in Pollen Mother Cells of the Pasto, Colombia, Population (13033) of M. andicolus H.B.K. x M. glabratus H.B.K. var. glabratus. Chromosome number, n = Number of cells observed 31 1 35 1 36 2 37 1 39 1 40 8 41 4 42 3 44 7 45 5 46 10 47 6 48 4 51 1 53 1 55 J_ 56 median Love 1982) and except for populations in the Rio Grande drainage, where M. glabratus var. fremontii has the tetraploid « = 30 chro- mosome number (McArthur et al. 1972). The present study shows that the pervasive n = \5 chromosome number occurs also in the populations of the geographic races of the Chihuahuan desert of northern Mexico (Table 1, e.g., culture numbers 12183, 12185) and of the Mexican Mesa Central (e.g., culture number 12215) as well as in the distinctive erect, delicate but wiry form from a palm oasis (San Bertola) of southern Baja California (culture number 12223). The last mentioned is suggestive of the typically erect and branched form of the tetraploid, « = 30 populations of M. glabratus var. fremontii from Texas. Except for the erect, more or less wiry forms that probably represent separate taxa, the rest of the M. glabratus var. fremontii group constitutes a diploid, polymorphic complex of geographic races and sibling species separated by an intricate net- work of partial to complete barriers to gene exchange (Vickery 1978). Mimulus glabratus var. glabratus has the aneuploid tetraploid chromosome number, = 3 1 , both in its Meso-American range in Mexico and Guatemala (e.g., culture 7306 and see Vickery 1978) and in its South American range in Colombia (e.g., culture numbers 13021, 13026, 13029). Mimulus glabratus var. glabratus appears to intergrade morphologically and chromosomally with the Ecuadorian and Peruvian M. andicolus (n = 46) in the southern Colombia pop- ulation (culture 13033) near Pasto. The chromosome numbers we observed in microsporocytes of this population ranged from = 3 1 94 MADRONO [Vol. 32 to « = 55 (Table 2). The median number of chromosomes fell be- tween n = 44 and n = 45. This suggests to us that the chromosome number is truly lower and variable, as well as showing aberrant segregations, such as 45/47, 44/48, etc., from the normal hexaploid n = 46 chromosome number of M. andicolus. The Pasto population appears to be closer to M. andicolus than to M. glabratus var. gla- bratus both chromosomally and morphologically. We found, as ex- pected from earlier work (Vickery 1978), that our two populations of M. andicolus (cultures 13066 and 13096) from central Peru were n = 46. Lastly, we found three populations (cultures 1 3069, 1 3070, 1 307 1) of M. pilosiusculus from southern Peru to have n = 46 chromosomes. These counts agree with our earlier reports (Mc Arthur et al. 1972, Vickery 1978) for related forms from farther south in Argentina and Chile. Thus, this study fills in several important geographic lacunae— Chihuahua, Baja California, Colombia, Peru— in the north-to-south series of polyploid and aneuploid adaptive radiations of the Mimulus glabratus complex (Vickery 1978). Acknowledgments We thank the Biomedical Research Support Program, grant S07-RR07092, and the National Science Foundation, grant DEB 79-11554, for their support of this project. Literature Cited Love, A. and D. Love. 1982. In A. Love, ed., lOPB chromosome number reports LXXV. Taxon 31:324-368. McArthur, E. D., M. T. Alam, F. A. Eldredge II, W. Tai, and R. K. Vickery, Jr. 1972. Chromosome counts in section Simiolus of the genus Mimulus (Scroph- ulariaceae). IX. Polyploid and aneuploid patterns of evolution. Madrono 2 1 : 417-420. Mia, M. M., B. B. Mukherjee, and R. K. Vickery, Jr. 1964. Chromosome counts in the section Simiolus of the genus Mimulus (Scrophulariaceae). VI. New num- bers in M. guttatus, M. tigrinus and M. glabratus. Madrono 17:156-160. Vickery, R. K., Jr. 1950. An experimental study of the races of the Mimulus guttatus complex. Proc. 7th Int. Bot. Cong. Stockholm, p. 272. . 1 964. Barriers to gene exchange between members of the Mimulus guttatus complex (Scrophulariaceae). Evolution 18:52-69. . 1 978. Case studies in the evolution of species complexes in Mimulus. Evol. Biol. 11:405-507. (Received 20 July 1984; accepted 30 October 1984.) THE IDENTITY OF CRACCA BENTHAM (FABACEAE, ROBINIEAE) IN THE UNITED STATES Matt Lavin Department of Botany, University of Texas, Austin 78712 Abstract The identity and circumscription of species of Cracca Bentham in the United States are confused in the literature. As clarified herein, two species of Cracca, C. glabella (Gray), comb, et stat. nov. and C. sericea (Gray) Gray, occur in the United States, where they are restricted to oak woodlands and associated grasslands of southeastern Arizona. Cracca Bentham is composed of approximately 27 species found from sea level to high elevations along the Andean cordillera of northern Argentina to Colombia, in Central America, and through Mexico to the United States in southeastern Arizona. The genus includes large shrubs, subshrubs and herbaceous forms, and is closely allied to Coursetia DC. The genus Cracca Bentham has been beset by much taxonomic and nomenclatural confusion since its erection by Bentham in 1854 (see Wood 1949 and White 1980). Unknown to Bentham, Linnaeus (1753) had also proposed a genus Cracca, comprising six species now included in Tephrosia. The work of Alefeld (1862) and sub- sequent work of Kuntze (1891) established Cracca Bentham as a later homonym. Benthamantha Alefeld was erected to include those species of Cracca Bentham, and Cracca L. was used to incorporate the species of Tephrosia Persoon. This latter arrangement was fol- lowed by botanists such as Rydberg (1924) during the early 1900s. Eventually, Cracca Bentham was conserved over Cracca L. (Taxon 8:293, 1959), and Tephrosia was reinstated to include Linnaeus's species of Cracca. Compounding the nomenclatural problems is the fact that the species of Tephrosia Persoon and Cracca Bentham are superficially very similar, belonging to two closely related tribes, the Tephrosieae and Robinieae, respectively. Many species of Cracca Bentham are filed in herbaria under "barbistyled" Tephrosia, and various species of Tephrosia are filed under Cracca. Important morphological fea- tures that distinguish the two genera are summarized as follows: Inflorescences of axillary racemes with 1 flower per node, never Madrono, Vol. 32, No. 2, pp. 95-101, 26 April 1985 96 MADRONO [Vol. 32 terminal; wing petals free from the keel; mature pod parti- tioned; leaflets stipellate, venation reticulate . . Cracca Bentham Inflorescences of terminal and/or axillary pseudoracemes with 2-3 flowers per node; wing petals lightly adherent to the keel; mature pod not partitioned; leaflets not stipellate, venation penni- parallel Tephrosia Persoon The two most important features distinguishing the two genera are the position of the inflorescence and the number of flowers per node. These are two important systematic characters that distinguish the tribe Tephrosieae from Robinieae. The circumscription and nomenclature of the two species in the United States have also been unsettled since Gray (1853) described Cracca edwardsii from a plant collected "Near Monterey [sic], Mex- ico, Dr. Edwards, in herb. Torr." and from plants collected by Charles Wright in Arizona. The one species of Cracca that occurs naturally in the Monterrey area is identical to the lectotype (NY!, designated by Rydberg 1924) of C edwardsii and belongs to a species complex confined to northeastern Mexico (Coahuila, San Luis Potosi, Ta- maulipas, Veracruz and Hidalgo). It is quite distinct from the two species that occur in southern Arizona and northwestern Mexico. Gray, however, reported C edwardsii from southern Arizona and northern Sonora, having confused Charles Wright's collection 963 (number assigned by Gray) with C. edwardsii. A further complica- tion arose when Gray distributed Wright's coUections because Wright 963 actually consisted of two taxa of Cracca, neither of them C. edwardsii. Later, Gray (1882) rectified this oversight by describing the vars. sericea and glabella of C. edwardsii, but he lacked sufficient material to distinguish between the species of Arizona and Mon- terrey. As a result, the majority of the specimens of the genus Cracca from Arizona have the wrong name applied to them. Rydberg (1924) treated the genus Cracca Bentham as Bentha- mantha Alefeld, a later synonym, and listed Benthamantha wrightii, B. edwardsii, and B. glabella from the United States. He treated Cracca sericea as a synonym of C. edwardsii. Kearney and Peebles (1960) omitted Cracca sericea from their treatment. They listed Cracca edwardsii from Arizona and treated C glabella as a variety of C edwardsii. They also state that the two Arizona taxa intergrade with each other. Based on personal prelim- inary field and herbarium studies, this condition is not known to occur, and I am unaware of any definite references to intergradation between these taxa other than the vague one in Kearney and Peebles. Wiggins (in Shreve and Wiggins 1 964) also treated the genus Crac- ca Bentham as Benthamantha Alefeld. He did not mention Cracca sericea in the text but listed B. edwardsii and treated B. glabella as a variety of B. edwardsii. More recently, Kartesz and Kartesz (1980) listed only Cracca caribaea from the United States; C. glabella and C sericea were 1985] LAVIN: CRACCA 97 omitted from their work altogether. Cracca caribaea, a widespread weed, is not known from the United States but gets as close as southern Tamaulipas and southern Baja California. Actually, only Cracca sericea and C. glabella occur in the United States, where they are restricted to the pine-oak woodlands and associated grasslands of southeastern Arizona (Santa Rita Mts. to the Chiricahua Mts.). They are distinct species that do not intergrade and belong to two separate phyletic lines within the genus. Cracca sericea is closely allied to the widespread, lowland C. caribaea and belongs to a group of Cracca that has an erect habit, tap roots, and relatively few, large, elliptic leaflets. In fact, if it were not for the tannin deposits on the leaflets and racemes congested above the leafy portions of the plant, C. sericea would be almost indistinguishable from C. caribaea. Cracca glabella, on the other hand, belongs to the C pumila complex, which is centered at high elevations in the Transverse Volcanic Axis and the Sierra Madre Occidental of Mexico. This species complex is characterized by plants with a prostrate habit, fusiform tuberous roots, and many, small, oval leaflets. Cracca gla- bella is distinguished from C pumila only by its larger, yellowish flower, sericeous ovary and different pattern of tannin deposits on the lower surface of the leaflets. Cracca glabella is remarkable in that it is one of the few herbaceous species of the genus in Mexico that has developed showy corollas, a feature more characteristic of South American craccas. The species has large, yellow flowers, and a deep reddish, relatively long-tubed calyx. It has thickenings or protuberances at the base of the banner that appear to act as nectar-guides. The filament of the vexillary stamen is thickened and clasped by the well developed auricles of the basal portion of the staminal tube. Cracca glabella is thus quite distinctive compared to most other Mexican species of Cracca, which are predominantly self-fertile and have flowers that are inconspic- uous and/or predominantly cleistogamous. In essence, to regard these two taxa as synomymous, or as mere varieties, obscures or ignores phylogenetically important characters of the genus. Cracca edwardsii, in contrast, is a small, erect subshrub with fu- siform tuberous roots and leaves lacking tanning deposits. The in- florescences are reduced and do not overtop the leafy portion of the plant. Additionally, the flowers are very small, inconspicuous and cleistogamous. It inhabits dry, limestone foothills of the Sierra Madre Oriental. Key to species of Cracca Bentham in the United States Stems prostrate, decumbent; roots fusiform tuberous; leaflets ovate- elliptic, rounded at both ends, (9-)l 1-21 per leaf, largest leaf- 98 MADRONO [Vol. 32 lets 7-15 X 4-7 mm, adaxial surface glabrate, rarely lightly sericeous, abaxial surface sericeous but with hairs restricted to leaf margins and main veins; abaxial calyx lobes at least 4 mm long; all petals yellow, banner commonly with a reddish mid- vein; ovary sericeous; leaflet tannin deposits (evident upon aging or drying) restricted to veins of abaxial surface; stipels minute, rarely to 1 mm C glabella Stems erect, from a taproot, roots not tuberous; leaflets elliptic, acute at both ends, 7-13 per leaf, largest leaflet 14-28 x 6-14 mm, adaxial surface evenly sericeous, occasionally glabrate in older leaves, abaxial surface evenly to densely sericeous; abax- ial calyx lobe to 3.5 mm long; banner reddish-pink, whitish at base, wings and keel whitish; ovary glabrous, granuliferous; leaflet tannin deposits (evident upon aging or drying) restricted to the center of the leaflet on the abaxial surface; stipels 0.5- 1.5 mm long, rarely minute C sericea Cracca glabella (Gray) Lavin, comb, et stat. nov. — Cracca edwardsii Gray var. glabella Gray, Proc. Am. Acad. Arts 17:201, 1882.— Benthamatha grayi Alefeld var. glabella (Gray) Britton & Baker, J. Bot. 38:19, 1900. -Benthamantha glabella (Gray) Rydb., North American Flora 24:247, 1924.— Benthamantha edward- sii (Gray) Rose var. glabella (Gray) Wiggins, Veg. and Flora Sonoran Desert. Vol. 1:690, 1964.— Type: hills between the Barbocomori and Santa Cruz [the Huachuca Mts., Arizona], 23 Sep, 1 85 1 . C Wright 963 p.p. Lectotype (designated by Rydberg 1924): US!; isotypes: MO! NY! UC! US!. Cracca glabella is restricted to pine and pine-oak forests at high elevations (2000-2200 m) in the Huachuca (Lanner and Garden Canyons), Chiricahua (Rucker Valley) and Patagonia Mountains of southeastern Arizona and the Sierra Madre of west-central Chihua- hua. It flowers from July through September. Only five collections of Cracca glabella are known from Arizona; the most recent collec- tion is from 1928. It may have been eliminated from the Arizona flora by overgrazing. Based on preliminary investigations, the genus is very sensitive to the pressures of grazing. Gray (1 853), in his original description of Cracca edwardsii, stated that the collections made by Wright were from "Valleys in the moun- tains between the San Pedro and the Sonoita, Sonora (in flower and with young fruit); and on hills between the Barbocomori and Santa Cruz (with ripe pods): Sept. (963)." The collection "in flower and with young fruit" is of Cracca sericea, whereas the collection "with ripe pods" is of C glabella. Wright's collection numbers, for what were obviously two separate collections, were discarded by Gray, who distributed both collections under a new number, 963. This practice has been documented by McKelvey (1955) and Johnston (1940): 1985] LAVIN: CRACCA 99 Not only did Gray ignore Wright's field-numbers but he also frequently united and distributed under a single distribution number, two or even more collections which Wright had col- lected under different field-numbers, frequently at distant sta- tions and at different seasons. If Gray thought two or more of Wright's collections represented the same species and if there was any advantage in uniting them, he did so regularly without scruples. Cracca glabella was first collected in September, 1851 by Charles Wright (no. 963) in southeastern Arizona. Gray (1882), in his orig- inal description, cited one additional collection by Lemmon in 1881. Although Wright's collection consists of both C glabella and C sericea, Rydberg (1924) indicated the collections of Wright as type for C. glabella when he stated the type locality to be "hills between the Barbocomori [Arizona] and Santa Cruz, Sonora." This is the same locality that Gray (1853) indicated the plants to have ripe pods, and these plants correspond precisely with C. glabella in the mixed Wright collection 963. The exact locality at which Wright first collected C glabella can be deduced from the protologue combined with Wright's field notes. It was most likely collected on 23 September 1851, as Wright went around the north end of the Huachuca Mountains, leaving the Bar- bocomori River drainage and starting down into that of the Santa Cruz River. Johnston (1940) says that Wright's collections dated September 23 were probably all collected before he reached Santa Cruz and probably all from within present day Santa Cruz County, Arizona. This is supported by the herbarium record, which docu- ments C. glabella as occurring only at relatively high elevations in southeast Arizona and west-central Chihuahua. Suitable habitats are missing in Sonora near Santa Cruz. Cracca sericea (Gray) Gray, Proc. Am. Acad. Arts 19:74 (1883) 1884. —Cracca edwardsii Gray var. sericea Gray, Proc. Am. Acad. Arts 1 7:20 1 , ISS2. —Brittonamra sericea (Gray) Kearney, Trans. N.Y. Acad. Sci. 14:32, IS94. —Benthamantha sericea (Gray) Britton & Baker, J. Bot. 38:19, 1900. -Type: Santa Rita Mts., Arizona, 6 May, 1881. C. G. Pringle 292. Lectotype (here designated): GH!; isotypes: F! NY!. Benthamantha wrightii Rydberg, North American Flora 24:246, 1924. Type: between San Pedro River and the Sonoita [River, east side of the Huachuca Mts., Arizona]. C. Wright 963 p.p. Holotype NY!; isotypes: UC! US!. Cracca sericea inhabits moderate elevations (1300-1900 m) in southeastern Arizona, northeastern Sonora, central and western Chi- huahua, and northern Sinaloa and Durango. It flowers from July to September and occasionally in the spring (March through May) dur- 100 MADRONO [Vol. 32 ing wet years. One of the more abundant species of Cracca, C. sericea is a conspicuous understory member of oak woodlands and asso- ciated grasslands of this area and, in typical form, has conspicuous reddish-flowered racemes that overtop the leaves. Tannins, which are evident on the older leaflets and dried herbarium specimens, are deposited centrally in each leaflet, a pattern unique to this species. Cracca sericea was first collected, along with C. glabella, by Charles Wright (no. 963) in September, 1 85 1 , in southeastern Arizona. Gray included these combined collections as syntypes of C edwardsii in his original description of that species. Thirty years later. Gray (1882) recognized that some of the Arizona collections were different from the type of C. edwardsii and named these var. sericea and var. glabella. Later, Gray (1884) raised var. sericea to the species rank as more collections became available for comparison. Gray (1 882) based var. sericea on Pringle 292 from the Santa Rita Mts. and Lemmon "136 & 588" from Spring Creek Canyon of the Santa Catalina Mts. The latter collection consists of a single sheet and although Rydberg designated (on the herbarium sheet) Lemmon "136 & 588" as type of C. edwardsii var. sericea, an account was never so published. The sheets of Pringle 292 are a bit more nu- merous and the specimen at GH is annotated by Gray. This collec- tion is, therefore, designated as the type. Rydberg (1924) also based Benthamantha {=Cracca Bentham) wrightii on Wright 963. The collection, as mentioned earlier, con- tains two distinct species. Rydberg expressly distinguished between the two on the pertinent sheets and his intention, therefore, is ob- vious. His designation of the type locality is that given by Gray (1853), but one qualification is needed here. Gray understood this locality to be between the "San Pedro" and the "Sonoita" in Sonora. The "Sonoita" Wright referred to is a river that drains the east slope of the Huachuca Mts. (Johnson 1940). If Wright's path is retraced, the locale between the San Pedro River and the Sonoita River lies roughly near the northeast end of the Huachuca Mts. in Arizona. It is possible to narrow the collections down to three of Wright's col- lection numbers, but as with Cracca glabella, the real collection number may never be known. Acknowledgments I thank Dr. R. C. Bameby and my colleagues at TEX for helpful suggestions and the curators of the following herbaria for loans: ARIZ, ASU, CAS, CHAPA, COLO, F, ISC, GH, LL, MEXU, MICH, MO, NMC, NY, TEX, UCB, UNM, and US. Literature Cited Alefeld, F. 1862. Namensanderung zweier Leguminosen-Gattung. Bonplandia 10: 264. 1985] LAVIN: CRACCA 101 Bentham, G. 1854. In G. Bentham and A. Oersted. Leguminosae Centroameri- canae. Vidensk. Meddel. Dansk Naturhist. Foren. Kjobenhavn. 1853 (Nr. 1-2): 1-19. Gray, A. 1853. Plantae Wrightianae. Texano-Neo-Mexicanae. Part II. Smithsonian Contr. Knowl. V, Art. 6. Oct. 1952. . 1882. Contributions to North American botany. Proc. Am. Acad. Arts 17: 199-230. . 1884. Contributions to North American botany. Proc. Am. Acad. Arts 19: 1-96. Johnston, I. M. 1940. Field notes of Charles Wright for 1849 and 1851-1852, relating to collections from Texas, New Mexico, Arizona and adjacent Sonora and Chihuahua. A copy with commentary by I. M. Johnston, Feb., 1 940 (Unpubl. manuscript at TEX). Kartesz, J. and R. Kartesz. 1980. A synonymized checklist of the vascular flora of the United States, Canada, and Greenland. Vol. 2. Univ. North Carolina Press, Chapel Hill. Kearney, T. H. and R. H. Peebles. 1960. Arizona flora (2nd ed.) with supplement. Univ. California Press, Berkeley. KuNTZE, O. 1891. Rev. Gen. 1:164-165. Linnaeus, C. 1753. Cracca. In Species Plantarum 2:752. McKelvey, S. D. 1955. Botanical exploration of the Trans-Mississippi West, 1 790- 1850. The Arnold Arboretum of Harvard University, MA. Rydberg, p. a. 1924. Robinianae. North American Flora 24:220-249. Shreve, F. and I. L. Wiggins. 1964. Vegetation and flora of the Sonoran Desert. Vol. 1 . Stanford Univ. Press, Stanford, CA. White, P. S. 1980. Flora of Panama— Fabaceae: Cracca. Ann. Missouri Bot. Gard. 67:596-599. Wood, C. E. 1949. The American barbistyled species of Tephrosia (Leguminosae). Contr. Gray Herb. 170:193-384. (Received 27 July 1984; accepted 30 November 1984.) A NEW SUBSPECIES OF PERENNIAL LINANTHUS (POLEMONIACEAE) FROM THE KLAMATH MOUNTAINS, CALIFORNIA T. W. Nelson P.O. Box 6435, Eureka, CA 95501 R. Patterson Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Abstract Linanthus nuttallii subsp. howellii Nelson & Patterson is described from the ser- pentine soil of the southern Klamath Mountains of California. It is near L. nuttallii subsp. nuttallii and subsp. pubescens morphologically, differing by its smaller leaf lobes as well as dense pubescence. During a recent expedition by the first author to the southernmost portion of the Klamath Mountains in Tehama County, California, an unusual population of perennial Linanthus was noticed. Further field investigation and examination of herbarium material support the taxonomic distinctness of this population, and it is herein pro- posed as a new subspecies. Linanthus nuttallii Milliken subsp. howellii Nelson & Patterson, subsp. nov. Caules, folia et calyces dense brevisetosi; folia divergens ad levitor arcuata, 5-7 partita, segmenta 3.5-7 mm longa; pollinis grana 34- 36 /Lim; chromosomatum numerus 2« = 18 (Fig. 1). Herbaceous perennial from woody root crown; stems reclining to slightly ascending at tips, forming a compact mat 7-19 cm in di- ameter, 4-8 cm high, densely gray pubescent, the trichomes short- bristly; intemodes 3-9(-12) mm long; leaves opposite, divergent to slightly arcuate, palmately partite into 5-7 narrowly oblanceolate divisions, 3.5-7 mm long, the trichomes short-bristly, mostly spreading at right angles to the leaf surface; flowers sessile to sub- sessile in dense subcapitate clusters; calyx narrow-campanulate, tri- chomes short-bristly, the lobes lanceolate-subulate, pungent, 7-1 1 mm long; corolla funnelform, white with yellow throat, 8-1 1 mm long, densely white villous on exterior, the tube included in the calyx; stamens inserted at base of the throat, included to barely exserted; MadroSo, Vol. 32, No. 2, pp. 102-105, 26 April 1985 1985] NELSON AND PATTERSON: NEW LINANTHUS 103 Fig. 1 . Linanthus nuttallii Milliken subsp. howellii Nelson & Patterson. A. Habit. B. Perianth with corolla opened. C. Node showing leaf detail. From Nelson and Nelson 5847. pollen grains 34-36 ^im in diameter, yellow; chromosome number 2n= 18. Flowering period is from June to early July. Type: USA. California. Tehama Co., w. side of Mt. Tedoc along NF road 45, ca. 1 .5 miles n. of Tedoc Gap, Yolla Bolly Quad. T28N, R9W, Sec. 28, 1500 m, 22 Jun 1980, Nelson and Nelson 5847 (Holotype: CAS; isotypes: GH, HSC, MO, NY, OSC, SFSU, WTU). Paratypes: USA. California. Tehama Co., nw. side of Tedoc, 16 104 MADRONO [Vol. 32 50 km Fig. 2. Map showing distribution of Linanthus nuttallii subsp. howellii and L. nuttallii subsp. nuttallii in northwestern California. Jun 1972, Stebbins s.n. (JEPS); slopes of Tedoc Mtn., Tedoc Gap road, 30 Jun 1974, Lester 338 (HSC); along Tedoc road ca. 10 mi from junction with State route 36, 21 Jul 1978, Nelson and Nelson 4441 (HSC); along forest road 45 at milepost 1 1, 25 Jun 1915, Nelson and Nelson 4930 (JEPS); along forest road 27N1 3 2.5 mi s. of Tedoc Gap, 1 Jul 1983, Nelson and Nelson 7436 (CAS); along forest road 27N13 0.7 mi s. of junction with forest road 27N12, 1 Jul 1983, Nelson and Nelson 7437 (BYU, JEPS, NY, MO); along forest road 27N12 0.3 mi from junction with forest road 27N13, 1 Jul 1983, Nelson and Nelson 7438 (DAV, RSA). Distribution. Linanthus nuttallii subsp. howellii is apparently very rare. It is limited in distribution to Mt. Tedoc and immediate vicinity in the southernmost Klamath Mountains of western Tehama Coun- ty, California, at elevations of 1500-1800 m (Fig. 2). It represents the southernmost extension of perennial Linanthus in the North Coast Ranges. It is disjunct from the nearest population of subsp. nuttallii by 90 km. It is found on serpentine soil in association with Jeffrey pine woodlands where it is often a common understory species. Linanthus nuttallii subsp. howellii, although similar in most fea- tures to the other perennial taxa in the genus, is easily distinguished by its low-growing habit, short leaf lobes and dense short-bristly indumentum on stems, leaves and calyces. Typical subsp. nuttallii 1985] NELSON AND PATTERSON: NEW LINANTHUS 105 is usually less hairy overall and possesses longer ((6-) 10-3 2 mm) leaf lobes; subsp. pubescens, which is considerably disjunct, generally possesses longer leaf lobes. The newly described species, as are most narrow endemics, is noteworthy for being much less variable mor- phologically than the other subspecies of L. nuttallii. Pollen grain diameter in subsp. howellii presents an additional intriguing problem. The grains of diploid perennial taxa have di- ameters ranging from 23-28 jum, whereas those of related tetraploid taxa measure 33-38 ixm (Patterson 1977). Pollen grains of subsp. howellii measure 34-36 jum, well within the range of tetraploid pe- rennials; however, subsp. howellii is diploid. This seeming discrep- ancy cannot be explained readily, and it points out the difficulty in relying on pollen grain diameter measurements in determining poly- ploid level. The discovery of the disjunct Tedoc Mountain populations of subsp. howellii and the apparent restriction to serpentine soil may be important in interpreting the nature of the entire perennial species complex. Throughout its range, this complex (Sect. Siphonella) oc- curs as a series of disjunct populations, sometimes separated from the nearest population by only a few kilometers. This pattern is suggestive of a formerly more continuous range that has subse- quently fragmented. There is no evidence supporting great vagility in perennial Linanthus species; hence it is unlikely that the Tedoc populations result from "long" distance dispersal. A possible ex- planation for the distribution seen is that populations of perennial Linanthus represent relict stands that persist due to special features of a given region. Perhaps near Tedoc the ability to survive on serpentine soil allows Linanthus nuttallii to persist in this part of its range, some distance south of its nearest population. While this hypothesis is highly speculative it may be worth future study in this group. The new subsp. of Linanthus nuttallii is named for John Thomas Howell, long a student of the California flora and worker with plants of the Tedoc region Acknowledgments Appreciation is extended to Jane Nelson, Chris Bern, and Barbara Williams for their assistance in the field, and to the curators of CAS, DS, JEPS, and UC for facilitating access to herbarium material. Literature Cited Patterson, R. 1977. A revision oi Linanthus sect. Siphonella (Polemoniaceae). Madrofio 24:36-48. (Received 3 October 1984; accepted 3 December 1984) THE MISSING FREMONT CANNON-AN ECOLOGICAL SOLUTION? Jack L. Reveal RBR & Associates, Suite 904, Centre City Bldg., 233 A St., San Diego, CA 92107 James L. Reveal Department of Botany, University of Maryland, College Park 20742 Abstract Donald Jackson and Mary Lee Spence (1970) proposed that the cannon John Charles Fremont abandoned on 29 January 1844 is to be sought in the Mill Creek Canyon area of the Sierra Nevada, while Ernest Allen Lewis (1981) suggests the site was on an unnamed peak ("Mt. 8422") above West Walker River in the Sweetwater Mountains of northern Mono Co., California. Both suggestions fail to account fully for the descriptive details given in Fremont's 1845 report, especially as they relate to geological, ecological, topographic and vegetational features. We propose that Fremont's men abandoned his twelve-pound howtizer in a cache near the base of a steep hill a short distance north of Cottonwood Meadow along Cottonwood Creek in northwest quarter of Sec. 23, T.7N., R.23E. This site is on the western edge of the Sweetwater Mountains in the Toiyabe National Forest. On 29 January 1844, John Charles Fremont and his men aban- doned a small cannon on the western edge of the Sweetwater Moun- tains in northern Mono County, California. The fate of this twelve- pound mountain howitzer' has been the subject of several local legends, sought after for years, had geographical place-names as- signed to areas where the cannon was believed to have been left, and has been discussed in numerous articles (e.g., Jackson 1967, Russell 1957) and two recent books (Jackson and Spence 1 970, Lewis 1981). Lewis tells an appealing story of the Fremont cannon and of certain other cannons that were not Fremont's but proclaimed by museums and assorted charlatans to be the very one that he trans- ported to California. In our opinion, both books fail to present a realistic view of the probable route of the Fremont (1 845) expedition on the critical days of 26-29 January, when the howitzer was aban- doned. We propose an alternative route and suggest a probable site where it was left. On 26 January, Fremont and Kit Carson made a reconnaissance of the country that lay ahead of the party. The expedition was camped just downstream from the junction of the East Walker River and Swauger Creek, about four miles north of the present-day town of Madrono, Vol. 32, No. 2, pp. 106-117, 26 April 1985 1985] REVEAL AND REVEAL: MISSING FREMONT CANNON 107 XI * FREMONT'S ROUTE JANUARY 27, 28, 29, 1844 Fremont and Fitzpatrick on Reconnaissance January 27 Party's Route: Lewis Reveal Campsites of: January 27-28 - Lewis @ Reveal (§) January 28-29 - Lewis © Reveal (B) January 29-30 - Fremont @ Preuss @ Cannon Abandoned - Lewis (e) Reveal © Fig. 1. Fremont's route, 27-29 January 1844. 108 MADRONO [Vol. 32 Bridgeport.^ Fremont wrote that "one of its branches [East Walker River]" was "coming directly from the south," and the other branch [Swauger Creek], "issued from a nearer pass, in a direction S. 75° W., forking [into Robinson and Buckeye Creeks] at the foot of the mountain, and receiving part of its water from a little lake."^ Fre- mont and Carson went up Swauger Creek and entered Huntoon Valley probably on the north side of the creek just northwest of the former Bridgeport Ranger Station. Fremont states that they went in a "northwesterly direction up the valley [Huntoon], which here [at Huntoon Campground] bent to the right .... The little stream grew rapidly smaller, and in about twelve miles we had reached its head,"^ the last water coming immediately out of the mountain on the right [south flank of the Sweetwater Mountains]; and this spot we selected for our next encampment [Fig. 1]." Later, Fremont wrote: "To the left, the open valley [Pimentel Meadows] continued . . . forming a beautiful pass [Devil's Gate] . . . which we deferred until the next day . . . ." On 27 January, Fremont and Thomas Fitzpatrick went quickly ahead leaving Carson to follow with the camp. The two men traveled rapidly up Huntoon Valley and "Arriving at the head of the stream, we began to enter the pass— passing occasionally through open groves of large pine trees [Pinus Jeffrey i], on the warm side of the defile [north side of Pimentel Meadow, or the south-facing slope] .... Continuing along a narrow meadow [Pimentel], we reached in a few [two] miles the gate of the pass [Devil's Gate], where there was a narrow strip of prairie, about fifty yards wide." Fremont and Fitzpatrick passed through Devil's Gate and onto the headwaters of Hot Creek in the West Walker River drainage. "On either side [of Devil's Gate] rose the mountains, forming on the left [Bush Mountain] a rugged mass, or nucleus, wholly covered with deep snow, presenting a glittering and icy surface. At this time, we supposed this [Devil's Gate, the Sweetwater Mountains to the north and crest of the Sierra Nevada extending southward] to be the point into which they [the mountains] were gathered between the two great rivers [San Joaquin and Sacramento], and from which the waters flowed off to the [San Francisco] bay. This was the icy and cold side of the pass [they were on the west side] .... On the left, the mountains [Mahogany Ridge] rose into peaks; but they were lower and secondary, and the country had a somewhat more open and lighter character. On the right were several hot springs [Fales Hot Springs]." As Fremont and Fitzpatrick moved into the area where the hot springs would have been seen on the right, they were almost a mile west of Devil's Gate and three-tenths of a mile east of Fales Hot Springs. In going through Devil's Gate, which is a massive grano- diorite outcrop, Fremont was impressed "by the majesty of the 1985] REVEAL AND REVEAL: MISSING FREMONT CANNON 109 mountain, along the huge wall of which we were riding." Fremont next states: "Here there was no snow [we believe "here" means the hot springs area]; but immediately beyond was a deep bank [into the meadow below the north edge of Wheeler Flat], through which we dragged our horses with considerable effort. We then immediately struck upon a stream [Hot Creek], which gathered itself rapidly, and descended quick; and the valley [of Hot Creek] did not preserve the open character of the other side [e.g., Huntoon Valley on the east or "other" side of Devil's Gate], appearing below to form a caiion." Fremont then writes that they "climbed one of the peaks on the right, leaving our horses below; but we were so much shut up, that we did not obtain an extensive view." We believe the men climbed the steep ridge north of the present U.S. Highway 395 on the north- western edge of Devil's Gate (Fig. 1). Here they would not have had an "extensive view." Fremont wrote that the "valley of the stream [Hot Creek] pursued a northwesterly direction, appearing below to turn sharply to the right, beyond which further view was cut off." From their likely vantage, the rim of Burcham Flat would have cut off the Sonora Junction area and the West Walker River as it flows from Pickle Meadows to the west. Fremont would have seen Hot Creek flowing westerly, then making a sharp turn to the right and going north into West Walker Canyon. Jackson and Spence suggest the men climbed a peak called Mt. 8422 some three miles north of Burcham Flat (Fig. 1), and some five airline miles north-northeast of Sonora Junction. This is ex- ceedingly unlikely since Mt. 8422 provides an excellent vista of much of the region. Lewis is of the opinion the men climbed "the steep escarpment up to Burcham Flat." Looking at Lewis's map, we be- lieve he places their climb at a point above the gauging station some 1.2 mi north-northeast of Sonora Junction. Had this been the case, Fremont would have seen the West Walker River coming down from the Sierra Nevada. As no mention is made of this river (nor is it shown on Preuss's maps of the region made for Fremont's report), it is unlikely they were ever aware of this important branch of the Walker River. Fremont states that after viewing the countryside from his vantage point he "resolved to continue our road the next day down this valley, which we trusted still would prove that of the middle stream between the two great rivers [San Joaquin and Sacramento]." It was clear to any observer that the crest of the Sierra Nevada to the west would not have allowed a river to flow through to the Pacific. Fre- mont was looking for the fabled Buenaventura River, and he hoped the small stream he was on was one of its headwaters and that by following this stream to the north he would find his way through the Sierra Nevada. From Fremont's view of the territory beyond Devil's Gate, the valley of Hot Creek and its "right" turn. West 110 MADRONO [Vol. 32 Walker Canyon, would certainly have been an attractive route. By cutting across the eastern edge of Burcham Flat, as he would the following day, Fremont could avoid deep snow and the impassability of Walker River Canyon. Fremont completes his report of the day by saying that toward "the summit of this peak, the fields of snow were four or five feet deep on the northern side . . . At that time of the year, these conditions would surely apply to the summits above Devil's Gate, and equally be true of Burcham Flat and Mt. 8422. The camp that evening was established on the north side of Pi- mentel Meadow at the point where Swauger Creek enters Huntoon Valley (Fig. 1). The slopes on both sides of Swauger Creek are south- facing and essentially devoid of conifers. The sagebrush (Artemisia tridentata) covered slopes were nearly snow-free and likely harbored some grass for the animals. Lewis is of the opinion Fremont estab- hshed his camp "one or two miles up the canyon [of Swauger Creek] from U.S. 395." The aspen (Populus tremuloides) dominated canyon floor with conifers interfingering onto the bottom-lands would have been heavily choked with snow. Passage up and into such a canyon would have been difficult. The north-south trending exposures would not have afforded pasture for the party's horses and mules. The events of 28 January are only brieffy described by Fremont. The camp went through Devil's Gate and traveled twelve miles, making camp "on a high point where the snow had been blown off', and the exposed grass afforded a scanty pasture for the animals." During the day, the "snow and broken country together made our travelling difficult: we were often compelled to make large circuits, and ascend the highest and most exposed ridges, in order to avoid snow, which in other places was banked up to a great depth." Jackson and Spence are of the opinion that Fremont crossed the West Walker River and climbed up onto an 8600-foot peak in the Sierra Nevada south of Grouse Meadow on the west side of West Walker Canyon and south of Mill Creek. Fremont does not report crossing any river, nor is such a crossing shown on Preuss's maps. In graphic terms, Lewis (p. 99) describes the party's journey of 28 January. He tells how Fremont and his men nearly exhausted them- selves climbing steep embankments and criss-crossing exposed ridges to reach the south end of Burcham Flat by early afternoon. The actual topography along this route, however, belies his description. Burcham Flat is only about 1 50 feet higher in elevation than Hot Creek at Wheeler Guard Station, and just east of this Forest Service post there is a shallow draw where the party could easily have as- cended onto the flat. Above that point, the elevation between the stream's edge and that of the flat is even less. Lewis's map shows the Fremont route as turning northwestwardly at Fales Hot Springs (where access to the flat is not difficult and thus contradictory of his 1 985] REVEAL AND REVEAL: MISSING FREMONT CANNON 1 1 1 text), crossing the middle of Burcham Rat, and climbing the south slope of Mt. 8422, where a camp was established "in the saddle near the top of the mountain."^ We agree that Fremont, to avoid deep snow, had to ascend the highest and most exposed ridges. Our proposed route (Fig. 1) accepts Fremont at his word. Our experience with Burcham Flat in the winter indicates that the snow on the flat would have been deep and the area difficult to traverse. The terrain over which Fremont would have passed as suggested by Lewis is shown in two photographs on page 92 of his book. They depict a fairly moderate slope over a distance of two miles. The average gradient is only ten percent be- tween Burcham Creek (7400 ft) and the saddle of Mt. 8422. The topography to the east and north of the flat, where we believe the expedition traveled, is one of "exposed ridges" and "steep ascents" compelling "large circuits" which fits Fremont's characterization of his travel that day. Lewis places Fremont's camp in a small saddle on his Mt. 8422 above West Walker River. We believe Fremont came around toward "Mt. 8422" from the southeast, following the ridges, and camped in the saddle east of the mountain at Cottonwood Pass. The Cottonwood Pass camp for the night of 28-29 January might have been in the pass itself, where the prevailing winds would have exposed some grass for the animals, or in a somewhat more protected site less than half a mile to the east. At the latter place, the snow might have been deeper, but the camp would have had some pro- tection from the wind. There was ample firewood in both sites con- sisting mainly of aspen and Rocky Mountain juniper [Juniperus scopulorum]. Fremont speaks of the camp having "scanty pasture" for his animals, and this leads us to suggest the camp was in the pass.^ The campsite suggested by Lewis is exposed and without tree cover, an unlikely camp for Fremont to have selected given the protection and cover provided at Cottonwood Pass. The rocky sad- dle proposed by Lewis does harbor some grass, but the swale, in our opinion, would have been covered with deep snow and thus what grass might have been present would have been buried. The pass could have been seen by Fremont even if he had traveled northward across Burcham Flat as suggested by Lewis. Looking at "Mt. 8422" from the south, Cottonwood Pass is clearly observable. The small saddle on the western slope can be seen as well, but Cottonwood Pass, at 8066 ft elevation, is certainly a more inviting campsite than the western saddle on "Mt. 8422," nearly 250 ft higher. Fremont reports that they "did not succeed in getting the howitzer into camp" the evening of the 28th, and thus a party was sent back on the 29th to get it. Some time was required to "bring up the gun . . . ." While this was being done, Fremont, Fitzpatrick "and a few 112 MADRONO [Vol. 32 men" left the camp and "followed a trail down a hollow where the Indians had descended, the snow being so deep we never reached the ground . . . when we reached a little affluent to the river at the bottom, we suddenly found ourselves in presence of eight or ten Indians." If Fremont was on the Sierra Nevada above Grouse Meadows as proposed by Jackson and Spence, he would have had to cross the meadows before descending Mill Creek and encountering the In- dians. Fremont makes no mention of this large expanse of several hundred acres. The campsite suggested by Jackson and Spence is some twelve miles from the Swauger Creek campsite on a straight line path down Hot Creek, across West Walker River and up to the saddle above Grouse Meadows. Because Fremont states his route was circuitous, the Jackson and Spence route seems unlikely. The route we propose from the Swauger Creek camp to the Cottonwood Pass camp measures closer to ten miles than twelve, which suggests Fremont's actual route was even more circuitous than we have in- dicated. Lewis states that Fremont and his men "rode down the north face of the mountain along a gradual trail to the [West Walker] river." No such route is described by Fremont. It is unlikely that the local Indians would have established a trail along the north slope and ridge of "Mt. 8422" as proposed by Lewis. It is more likely the trail was through the pass and down the hollow to Cottonwood Creek, the approximate route of today's Cottonwood Road, and that this was the trail followed by Fremont in the belief that it would lead to Indian villages in the valley [Antelope Valley] he had seen from the vicinity of his encampment.^ While it is true one can see more of Antelope Valley to the north from the upper reaches of "Mt. 8422," the western slopes of the valley can still be seen from Cottonwood Pass. Likewise, a walk from the pass toward the summit of "Mt. 8422" would not have been too difficult. From there Fremont would have had an excellent view of Antelope Valley, as suggested by Lewis. We believe the "little affluent" Fremont mentions is Cottonwood Creek. Lewis believes the "little affluent" is an unnamed spring and runoff channel found on the northwest side of Mt. 8422. Lewis's map shows Fremont's route as along the western edge of "Mt. 8422" and the ridge (not on the crest of the ridge, as given in the text) going toward an unnamed mountain peak illustrated as similar in size to "Mt. 8422." Because no such mountain is to be found, we suspect Lewis's unnamed peak is Mt. 6926, a point south of Deep Creek on the ridge north of Mt. 8422. On the western slope between elevations 8422 and 6926 is an intermittent stream that reaches the West Walker River at Toll House Campground. This is on the route 1985] REVEAL AND REVEAL: MISSING FREMONT CANNON 1 1 3 suggested by Lewis and we believe is what he illustrates as the "little affluent." Fremont presented gifts to the several Indians he found "on the hill side above our heads." Such a locality description is possible for our proposed Cottonwood Creek route, and would place this incident where the canyon narrows a short distance down stream from Cottonwood Meadows at an elevation of about 7 1 00 ft. Lewis states correctly that the upper end of his spring-fed draw is "twenty foot wide and ten foot deep." It is hardly likely that Fremont would have elected to enter such a narrow draw, especially with snow on the ground. Furthermore, it is not a place where the Indians could wait "above our heads" as stated by Fremont. The Cottonwood Creek route is much wider, with elevated slopes along both sides of the stream. If Fremont met the Indians in the canyon below Cot- tonwood Meadow as we suggest, where the timbered slopes are high and steep, and where the distance between the two slopes is more than one hundred feet, the Indians would have been "out of reach" and would have "thought themselves safe" as he describes. At this point Fremont makes a critical statement regarding the fate of his howitzer: "The principal stream [Cottonwood and Deep Creeks] still running through an impracticable canon, we ascended a very steep hill, which proved afterwards the last and fatal obstacle to our little howitzer, which was finally abandoned at this place." Jackson and Spence are of the opinion the cannon was left on Mill Creek. An examination of the Mill Creek area provides places where Fremont could have had Indians above him and out of reach, but no steep hills that would have to be ascended on 29 January. Also, Fremont would not have found in Mill Creek an "impracticable canon" which would have forced him to leave the drainage. Like- wise, Jackson and Spence (p. 623) state that Fremont's expedition departed the mountains in Antelope Valley "which it entered from the mouth of West Walker Canyon." To accomplish this, the party would have had to climb over the high ridge separating Mill Creek and the West Walker River, an unnecessary passage and much more difficult than simply continuing down Mill Creek. Lewis says that Fremont followed his "little affluent" down to the West Walker River. Fremont clearly does not say this. Upon seeing or perhaps learning from the Indians of the difficulties of taking animals farther down Cottonwood Creek and into Deep Creek, he "ascended a very steep hill" to get out of the "little affluent." If our route is correct then Fremont and his men turned westward where the canyon narrows below Cottonwood Meadows, and ascended the southeast-facing slope. That steep slope is covered with single-leaf pinyon (Pinus monophylla). There are no signs of fire and the trees are of an age estimated to be in excess of 250 years. The sideling 114 MADRONO [Vol. 32 ground and undulations in the snow would have made pulling the wheeled cannon difficult. Likewise, the descent into the Walker River Canyon along a timbered slope with a mixed stand of pinyon and Jeffi"ey pine (which also show signs of considerable age and no fire), would have made getting the cannon down to the river difficult. By climbing out of Cottonwood Canyon, Fremont would avoid Deep Creek Canyon, and by crossing over the ridge of his "steep hill" he could get onto a southwest-facing slope and thus avoid deep snow. Fremont never saw the howitzer's last resting place, nor does he state in what condition it was left. Lewis speculates the cannon was left unattended at the campsite on "Mt. 8422," and that it "was a jubilant group that rode away from the little howitzer, sitting atop its carriage in a small meadow . . . ." Fremont's own words clearly refute this statement. Fremont says that the howitzer was "left by Mr. Preuss in obe- dience to my orders." Lewis suggests that Carson was sent "back to the camp" with instructions to leave the cannon, but most likely Carson was put in charge of the camp party and was moving ahead of Preuss and the cannon party who were, as Fremont reports, bring- ing the cannon up from where it had been left the day before. We speculate that Fremont's orders, if any were actually given at the time, were to attempt to get the cannon beyond "this place," but if impossible, his men were to abandon it.^ Perhaps a written message was left on the trail, a common practice among wilderness travelers. Lewis recounts Preuss' s dislike for the howitzer, but Fremont, if not Preuss, had a military man's attachment to the cannon (see Note 1). It is unlikely they would have abandoned such a weapon to an unknown fate. With the party were mountain-men skilled at con- cealing bales of fur, traps and other equipment. Because Fremont states "I reluctantly determined to leave it [the howitzer] there for the time [our emphasis]," it is reasonable to speculate that the how- itzer was disassembled and, along with its ammunition, safely cached. We believe the site of the cache for the cannon and its ammunition was situated near the bottom of the steep side-slopes below Cotton- wood Meadow at the point Fremont abandoned Cottonwood Creek. The "very steep hill" is the one in the northwest quarter of Sec. 23, T.7N., R.23E. (Mt. Diablo Base and Meridian). It has a gradient of 30 to 35 percent, which could be considered "very steep," especially in the winter and with laden horses and mules. The soil in this area is of decomposed granite and, on the south-facing slope, it would have been easy to dig rapidly a suitable cache for the disassembled howitzer. If the cannon and ammunition had been left exposed, certainly some remains would have been discovered. Cattlemen, sheepherders and hunters could have traveled in the area unaware of a hidden cache, but if it were unburied some evidence of the cannon would have been found. We believe that downslope slough- 1985] REVEAL AND REVEAL: MISSING FREMONT CANNON 1 15 ing of weathering rock and soil have continued to cover the cache and that this accounts for the lack of physical evidence of the cannon at this site. After describing his ascent from Cottonwood Creek, Fremont states that "We passed through a small meadow a few miles below, crossing the river . . . and, after a few more miles of very difficult trail, issued into a larger prairie bottom, at the farther end of which we encamped, in a position rendered strong by rocks and trees. The lower parts of the mountain were covered with nut pine [Pinus monophylla].'' He indicates that camp was established in the afternoon and in his table of distances states the party traveled seven miles this day. The presence of single-leaf pinyon, as nut pine is now called, and the description of a "larger prairie bottom," leads us to believe that Fremont and his small party camped at China Garden, today a sagebrush flat with scattered conifers. Given the heavy grazing and other disturbances that have happened in this area since the 1890s, the grasses Fremont saw have been replaced with sagebrush and other shrubs, thus changing the vegetational characteristics of the area. We suspect that Fremont came down to the West Walker River and crossed it in the vicinity of Shingle Mill Flat.^ This we take to be Fremont's "small meadow." Preuss and his group did not advance as far as China Garden, "but encamped in the upper meadow," which we believe to be Shingle Mill Flat. If our route is correct, the distance from Cotton- wood Pass down Cottonwood Creek, across the West Walker River, and then down the river to China Garden is seven miles as noted by Fremont. The Lewis route is only five miles, whereas that sug- gested by Jackson and Spence is a minimum of ten airline miles. The events of the 29th were trying for Fremont and his men. The cannon was left not in the Sierra Nevada but on the western edge of the Sweetwater Mountains, geologically and floristically a part of the Great Basin (Cronquist et al. 1972, Harper and Reveal 1978, Reveal 1980). Lewis believes the cannon is yet to be discovered. He is probably correct. We suggest that the route proposed here (Fig. 1) best fits Fremont's own description and that the routes indicated by Jackson and Spence and by Lewis fail to account for all the factors described by him. We also suggest Fremont and his men cached the cannon and its ammunition. Perhaps the howitzer was found by Indians who moved it elsewhere or even destroyed it as local legend tells. However, some of the five hundred pounds of ammunition ought to be in the area of the original cache and likely might be discovered by a careful, and we must add, legal search. '° Acknowledgments The research reported here has been made possible in part by funds from contracts with the Toiyabe Natl. Forest for studies by the senior author on range-trend plots 116 MADRONO [Vol. 32 on the western slope of the Sweetwater Mountains in the 1980s, and by National Science Foundation grants GB-22645 and BMS75- 13063 to the junior author for ongoing studies on the flora of the Intermountain West. Our initial investigations began in 1 958 (the subject of a "term paper" at Sonora Union High School, California) and continued in 1 960 when working for the U.S. Forest Service at Lee Vining (pater) and Wheeler Guard Station {filius). We wish to thank Joan Bonin of RECON, Inc., San Diego, California, for providing valuable assistance in preparing the map, and the Bridgeport Ranger District for use of aerial photographs. This is Scientific Article A3840, Contribution No. 6820 of the Maryland Agricultural Experiment Station. Notes ' Fremont obtained the field piece "from the United States arsenal at St. Louis, agreeably to the orders of Col. S.W. Kearney." This acquisition was deemed unnec- essary by Col. J. J. Abert, head of the Bureau of Topographic Engineers, and a letter was sent to Fremont asking him to justify his actions (see Jackson and Spence 1970, for its text). According to Fremont's wife, Jessie, the daughter of Missouri Senator Thomas Hart Benton, she intercepted the letter and in a dramatic note urged her husband to depart immediately. Jackson (1967) questions that this event ever hap- pened. Fremont clearly had no authority to request or take a field piece into Spanish California, an action that could have been considered an act of war. The howitzer was placed "under the charge of Louis Zindel, a native of Germany, who had been 19 years a non-commissioned officer of artillery in the Prussian army . . . ." ^ The junction of the two streams (and others) is now covered by the waters of Bridgeport Reservoir. ^ Although Fremont used the singular, we believe he is referring to Twin Lakes. Robinson Creek arises from these lakes. We are not certain how he learned of any lake, but it is possible Indians told him. The maps of the expedition do not show such a lake. Meadow irrigation in Bridgeport Valley has changed greatly the config- uration of the four streams. See USGS 1911 Bridgeport Quadrangle. " Lewis's map shows that Fremont and Carson rode to the little spring tributary to Swauger Creek in Sec. 8, T.6N., R.24E. This spring is at 9200 ft, 1800 ft in elevation above Pimentel Meadow. Had the men made such a difficult climb, certainly it would have been reported. ^ Mt. 8422 as defined by Lewis (and maintained here for purposes of consistency) often refers to the entire elevated area north of Burcham Flat and west of Cottonwood Pass. USGS point 8422 on the Fales Hot Springs Quadrangle is actually a small nob above West Walker River and west of an unnamed, higher, north-south trending ridge that is approximately 8560 ft in elevation. Evidently, this higher ridge is Lewis's Mt. 8422, thereby causing some confusion in his book. When we refer only to the higher ridge and not the entire mountain top, we use "Mt. 8422." Lewis's camp was in a saddle between 8422 and 8560. ^ An examination of our proposed campsites for this night indicates there was a much better stand of aspen/juniper in Cottonwood Pass when Fremont was there. The reduction of tree cover in the area probably has been due to livestock use and sheep trailing through the pass. ^ Fremont describes "yellow spots" in Antelope Valley to the north. Such spots are still visible today. They are the grassy western slopes above the valley floor and west of U.S. Highway 395. ^ Lewis suggests Fremont sent Carson back to the morning camp with his orders and that the "courier probably arrived in camp at about the same time as the cannon party returned with the cannon." There is a small basin or flat on top between Mt. 8422 and the higher ridge to the east. This was both Lewis's camp and where he suggests the howitzer was left. ^ Preuss's maps have been criticized for their lack of details concerning the critical days of 26-30 January 1 844. We disagree. Preuss's maps (see maps 3 and 5 in Jackson 1 985] REVEAL AND REVEAL: MISSING FREMONT CANNON 1 1 7 and Spence) clearly show Bridgeport Valley and the route up Huntoon Valley, where the party turned westward toward Devil's Gate. Swauger Creek is shown to continue northward onto the southern flank of the Sweetwater Mountains. The mapped route shows the expedition going through the pass down to the Little Walker River, which arises from the south. West Walker River, which comes from the Sierra Nevada to the west, is not shown. The route then follows the edge of the mountains east of West Walker River, not in the mountains west of the river as suggested by Jackson and Spence. The two maps differ slightly in this respect. The 1845 map shows the route along the immediate bank of the river, making three crossings, and with two campsites shown in the canyon. On a later map released in 1848, no camps are shown, and the route is east of the river, seemingly— in part— to be in the mountains. Both maps indicate Fremont crossed the river at the mouth of the canyon, but we believe he crossed only once and then in the vicinity of Shingle Mill Flat. Lewis's map also shows a route down the east side of the river, and although his location can be supported by the Preuss maps, Fremont states he crossed the river upon reaching it. The east bank of the river defies passage today and it was probably an impossible route also in Fremont's day. It should be noted that on 1 February, Fremont, then on the East Carson River in Carson Valley, gave his latitude as 38°37'38". This is an error or a misprint: it should be 47', as indicated by Preuss's maps. No doubt the confusion reported by Jackson and Spence regarding the placement of Fremont's camp this night is due to this error. '° Any search activity for the Fremont cannon that would involve surface distur- bance, excavation or disturbance of artifacts must be authorized by an archaeological permit issued by the U.S. Forest Service, the land managing agency, via U.S. De- partment of Interior under the provisions of the National Antiquities Act of 1 906 and the Archaeological Resources Protection Act of 1979 [16 U.S.C., Sec. 470aa, et seq.]. Literature Cited Cronquist, a., a. H. Holmgren, N. H. Holmgren, and J. L. Reveal. 1972. Intermountain flora. Vascular plants of the Intermountain West, U.S.A., Volume one. Hafner Publ. Co., NY. Fremont, J. C. 1845. Report of the exploring expedition to the Rocky Mountains in the year 1842, and to Oregon and North CaHfomia in the year 1843-1844. Gales and Seaton, Washington, D.C. 693 pp. Harper, K. T. and James L. Reveal, eds. 1978. Intermountain biogeography: a symposium. Great Basin Naturalist Mem. 2:1-268. Jackson, D. 1967. The myth of the Fremont howitzer. Bull. Missouri Hist. Soc. 23:205-214. and M. L. Spence. 1 970. The expeditions of John Charles Fremont. Volume 1. Travels from 1838 to 1844. Univ. Illinois Press, Urbana. Lewis, E. A. 1981. The Fremont cannon high up and far back. Arthur H. Clark Co., Glendale, CA. Reveal, J. L. 1980. Intermountain biogeography— a speculative appraisal. Ment- zelia 4:1-92. Russell, C. P. 1957. Fremont's cannon. California Hist. Quart. 36:359-363. NOTES AND NEWS Inhibition of Abies concolor Radicle Growth by Extracts of Ceanothus ve- lutinus. — Ceanothus velutinus Dougl. ex Hook is a major shrub species that invades recently burned areas and clearcuts at middle elevations in the Sierra Nevada, Cal- ifornia, where it competes strongly with natural and planted conifer seedlings (Conard and Radosevich, Forest Sci. 28:243-304, 1982; Zavitkovski et al., J. For. 67:242- 245, 1969). Establishment of natural Abies concolor {Gord. & Glend.) Lindl. seedhngs in brush fields with a high cover of Ceanothus velutinus is frequently spotty, and many established seedlings appear somewhat unhealthy for several years. The earlier study by Conard and Radosevich was concerned primarily with competitive inter- actions. Chemical (allelopathic) effects, however, are also a possible partial expla- nation for reduced growth of conifers in the presence of Ceanothus velutinus. To investigate this possibility, I conducted a series of laboratory and growth chamber experiments. Methods. The first two experiments compared germination and growth of Abies concolor seedlings under Ceanothus velutinus canopies and without Ceanothus can- opies. Seeds were germinated in 1.4-liter pots containing either a 20- to 25-cm tall Ceanothus seedling or no seedling (control). Pots were filled with potting mix (1:1:: peat:sand plus balanced fertilizer). Twenty Abies seeds were placed in each pot, with two pots per treatment. Seeds were watered regularly with distilled water. Seeds in the Ceanothus treatment were watered through the shrub canopy. To minimize po- tential moisture competition, the soil was kept near field capacity at all times in both treatments. Both experiments were conducted in a controlled environment chamber (23''C day, 10°C night, 16-h photoperiod). Average daytime light intensity was 0.6 mmol m~2 sec"'. In the first experiment, germination was recorded after 1 month, and a random sample of eight seedlings selected from each treatment was measured for root and shoot length. In the second experiment, the treatments were the same as those de- scribed, except that the seedlings were grown for 7 weeks, at the end of which ger- mination and survival were recorded. This experiment was repeated three times, with two pots per treatment each time. On the basis of the results of these first two experiments, and of a series of extremely inconclusive experiments that had been designed to look at possible effects of Cea- nothus root exudates on A. concolor seedlings (S. Conard, unpubl. data), I hypothe- sized that the Ceanothus foliage was the most likely source of allelopathic compounds. To test this hypothesis, I conducted two additional experiments to examine more closely the effects of Ceanothus velutinus foliage extracts on radicle growth of ger- minating Abies concolor seeds. Leaf extracts were prepared by soaking crushed, fresh- frozen Ceanothus foliage for 1 h in enough distilled water to cover the leaves. Leaves and water were placed in a Waring blender (trade names are used for information only; no endorsement by the U.S. Department of Agriculture is implied) for 3 min. The resulting mixture was strained through cheesecloth and vacuum-filtered through Whatman No. 1 filter paper to remove suspended solids. The osmolality of these extracts was measured with a vapor pressure osmometer (Wescor, Model 5130) to determine if observed effects could be attributed to osmotic potential of the extracts. Extracts were stored in the refrigerator for up to 2 weeks until use. In one experiment, standard glass-chromatography plates were covered with chro- matography filter paper that had been moistened and rolled to remove air bubbles. Plates were set on an angle in covered plastic trays with their bottom edges in 250 ml of either Ceanothus extract or distilled water. A row of five Abies concolor seeds was placed 2.5 cm from the top edge of each plate. On half of the plates, seeds were Madrono, Vol. 32, No. 2, pp. 118-121, 26 April 1985 1985] NOTES AND NEWS 119 covered with a 2.5-cm wide strip of chromatography paper moistened with distilled water. Each of the four treatment combinations was replicated twice. Radicle growth and germination were recorded 10 and 22 days after the beginning of the experiment, which was conducted at room temperature (20° to 25°C) in indirect sunlight in the laboratory. In a second similar experiment, seeds were germinated in 1-pint plastic freezer containers. A 1 -cm thick pad of open-cell foam placed in the bottom of each container was covered with a square of chromatography paper cut so it could be folded over the edges of the foam to contact the bottom of the container and act as a wick for treatment solutions. Nine seeds of Abies concolor were placed in a square grid in each container. The same two treatment solutions were used in this experiment as in the previous one. To minimize fungal attack, 10 mg captan was added per 450 cc of each treatment solution. Twenty ml of the appropriate treatment solution was added along the edges of each freezer container after the foam pad and seeds were in place. Containers were covered with plastic wrap and placed on a windowsill in the labo- ratory. Each treatment was replicated three times. Germination, radicle growth, emer- gence of cotyledons, and evidence of fungal attack or root dieback were recorded 2 1 days after the start of the experiment. Results of all experiments were analyzed by analysis of variance. In experiments with more than two treatments, the least significant difference (LSD) was used to compare treatments. Arcsine square-root transformations were applied to percentage data where appropriate. Germination. Germination of seeds grown with Ceanothus and watered through their canopies averaged 47%. This germination was not significantly different (paired t = 0.1727, df = 3) from germination in control pots (45%). I also observed no sig- nificant differences in germination among treatments in either the chromatography plate bioassay or the foam pad bioassay, where germination averaged across treat- ments was the same (78% ± 4 SD) for both experiments. Germination percentages observed in the two bioassay experiments ranged from 60 to 89%. Because treatments appeared not to affect germination, seeds that did not germinate were omitted from calculations of growth parameters. Survival. Although treatments did not affect germination, they substantially affected survival after 7 weeks. The survival rate of Abies seedlings grown in pots with Ce- anothus (0.20 ± 0.02 SE) was significantly lower (p < 0.005) than survival of seedlings grown in the same soil without Ceanothus (0.81 ± 0.06 SE). Radicle growth. The germinated Abies seedlings that were harvested in the first experiment showed dramatically different patterns of root growth, despite nearly identical shoot growth (Table 1 ). I also observed that the roots of seedlings that had been germinated in pots with Ceanothus velutinus were withered and broke off readily, whereas roots of the controls were healthy. With chromatography plate and foam pad bioassay s I more closely evaluated the possible effects of Ceanothus foliage extracts on root growth of germinating Abies seedlings in the absence of potentially confounding variables such as moisture competition, shading, and root exudates. In the chromatography plate bioassay, radicle growth of germinated Abies seeds averaged 126.8 ± 3.6 mm for the control and 71.2 ± 1.2 mm for the control with filter paper strips. Values for the comparable treatments with Ceanothus extract were 45.4 ± 0. 1 mm for the control and 55.6 ± 1 1.8 mm for the control with filter paper strips. All differences, except between the two extract treatments, were statistically significant at the 0. 1 level or greater. The foam pad bioassay produced similar results, with average radicle growth of 5.3 ± 0.6 mm for seeds exposed to the Ceanothus treatment and 41.9 ± 3.8 mm for seeds exposed to the distilled water treatment. The extract treatments showed highly significant decreases (p < 0.00 1 ) in radicle growth compared with those of the control. No differences were observed among treatments in the degree of fungus attack (range = 3-16%) or the number of radicles with tip dieback (16-25%). Significantly more of 120 MADRONO [Vol. 32 Table 1 . Stem and Root Growth of Abies concolor Seedlings Grown for 1 Month in Pots with or without Ceanothus velutinus (±One Standard Error). Stem length Root length Treatment (cm) (cm) Ceanothus 3.3 ± 0.3 2.4 ± 0.2 Control 3.5 ± 0.3 11.4 ± 0.4 the seedlings in control treatments, however, had cotyledons emerged at the end of the experiment (44%) than those in the Ceanothus extract treatment (0%). The os- molality of Ceanothus extract used in these experiments was 77 ± 1 mOS/kg (about -1.8 X 10"^ MPa). It is unlikely, therefore, that osmotic potential of the solutions affected the results to any great degree. Discussion. The results of these experiments suggest the possibility of an allelopathic inhibition of radicle growth of Abies concolor seedlings by Ceanothus velutinus. At least one compound (cinnamic acid), found by Craig et al. (Phytochemistry 10:908, 1971) in the leaves of C. velutinus has been implicated as inhibiting seedling growth in some species (Rice, E. L., 1974, Allelopathy, Acad. Press, NY, p. 256). Further experiments are required to isolate the causes of the responses described here and to determine whether extracts of foliage and litter produced under field situations are sufficiently concentrated and persistent to have measurable effects. If responses similar to those described here are observed in the field, Ceanothus velutinus may be expected to affect adversely the natural regeneration of Abies concolor— especially in dry years or in other situations where rapid root growth could be critical to seedling survival.— Susan G. Conard, Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S.D.A., 4955 Canyon Crest Dr., Riverside, CA 92507. (Received 16 Feb 1984; accepted 30 Oct 1984.) Pitfalls in Identifying Ventenata dubia (Poaceae). — The annual Eurasian grass species Ventenata dubia (Leers) Coss. & Dur. appears to be expanding its range in the Pacific Northwest, and botanists who may encounter it need to be aware of some pitfalls in making a correct identification of this potential weed. Its occurrence as an adventive species in Idaho and Washington was first reported by Baker (Leafl. Western Bot. 10:108-109. 1964), and it was subsequently described and illustrated by Hitch- cock et al. (Vase. PI. Pac. N.W. 1:724. 1969). A recent collection from Polk County {Halse 2857, 21 Jun 1984, OSC) documents its invasion of the Willamette Valley in western Oregon. The spread of V. dubia by human agency may accelerate if it becomes a contaminant of the various crop grasses grown for seed in the Pacific Northwest. Ventenata is generally considered to be taxonomically allied with Trisetum and Avena (Hackel, The true grasses, Henry Holt and Co., 1 890, p. 121; Bews, The world's grasses, Longmans, Green and Co., 1929, p. 174). The two upper florets of its 3- flowered spikelets are fertile; as commonly occurs in members of tribe Aveneae, these lemmas bear a conspicuous dorsal, geniculate awn. The lowest floret, however, is usually staminate, and its lemma has a straight terminal awn, as is found in various members of tribes Poeae, Brachyeltreae, Stipeae, etc. This floret is incorrectly de- scribed as awnless by Gould and Shaw (Grass systematics, Texas A&M Univ. Press, 1983, p. 179) and Hackel (op. cit.). Both glumes are shorter than the first lemma, and disarticulation occurs in the rachilla above the lowest floret. Because the dorsally-awned florets are shed at maturity and the terminally-awned one is retained within the glumes, this grass is deceptive when presented for identi- 1985] NOTES AND NEWS 121 fication in post-mature condition. It convincingly mimics a 1 -flowered member of tribes like Stipeae or Eragrosteae, whose genera may have terminally awned lemmas and relatively short glumes. The deception is increased when, as sometimes happens, a caryopsis is formed in the lowest floret. Distinctive features of the species are its conspicuously 7-nerved glumes and smooth, obconical pedicels. Probably the best protection against being fooled by this grass, when it has shed its dorsally-awned fertile florets, is simply to be aware of the problems described here. I hope that this note will save others the several hours I wasted trying to identify this species before serendipitously discovering the cause of the difficulty. — Kenton L. Chambers, De- partment of Botany and Plant Pathology, Oregon State University, Corvallis 97331. (Received 31 October 1984; accepted 3 December 1984.) NOTEWORTHY COLLECTIONS Arizona Antennaria microphylla Rydb. (Asteraceae).— Coconino Co., San Francisco Mt., ridge between Agassiz and Humphreys Pks., fellfield, 3658 m, 28 Jul 1983, J. D. Morefield 1563 and C. G. Schaack (DHA); Schaack 1182 and J. D. Morefield (DHA, NY). (Verified by A. Cronquist.) Significance. New addition to the Arizona alpine. A depauperate specimen of A. microphylla may have been the source of McDougall's A. rosulata Rydb. report for the Arizona tundra {Seed plants of Northern Arizona, Museum of Northern Arizona, Flagstaff, 1973). The presence of A. rosulata in the alpine could not be confirmed in the field or herbarium (MNA) and this species should be deleted from Schaack's alpine list (Madroiio 30:Suppl., p. 79-88, 1983). Aquilegia chrysantha Gray (Ranunculaceae). — Coconino Co., San Francisco Mt., nne. -facing slope below the knob designated 3685 m (12,089 ft) on the ridge between Agassiz and Humphreys Pks., melt-water channel and alpine meadow, 3612- 3627 m, 27 Aug 1983, C. G. and B. J. Schaack 1167 and J. D. Morefield {X^WK). Significance. New addition to the Arizona alpine. Though identified here as A. chrysantha, "yellow" columbine at highest elevation, ca. 3048 m or above, often has spurs and sepals tinged blue or purple and an overall habit that differs from more typical specimens below. Carex haydeniana Olney (Cyperaceae). — Same site as the A. chrysantha collec- tion (above), melt-water channel, ca. 3642 m, 16 Jul 1983, D. Morefield 1580 and C. G. Schaack; C. G. and B. J. Schaack 1173 and J. D. Morefield (DHA). Significance. New addition to the Arizona alpine. PiNus FLExiLis Jamcs (Pinaceae). — Coconino Co. , San Francisco Mt. , ridge between Agassiz and Humphreys Pks., sw.-facing slope, ca. 3627 m, 16 Jul 1983, C. G. & B. J. Schaack 1175 and J. D. Morefield (DHA). Significance. New addition to the Arizona alpine. Only a single depauperate (ca. 30 cm), individual is presently known from this area, growing with rock protection at the upper end of an open extended krummholz. Though protected, the condition of the plant suggests that prolonged survival in this rigorous environment is unlikely. Clark's Nutcracker is known for the area and periodic introduction of Limber Pine seed into the alpine is expected. Madrono, Vol. 32, No. 2, pp. 121-128, 26 April 1985 122 MADRONO [Vol. 32 PoTENTiLLA NivEA L. (Rosaceae).— Coconino Co., San Francisco Mt., nw. pro- jecting ridge of Agassiz Pk., fellfield, 3688 m, 12 Jul 1983, C G. Schaack 1153 and J. D. Morefield (DHA); 0.4 mi n. of Agassiz Pk. summit on an exposed andesitic ridgetop, 3667 m, 16 Jul 1983, D. Morefield 1568 and C. G. Schaack 1163 (DHA); ridge leading to the summit of Humphreys Pk., fellfield, 3780-3795 m, 27 Aug 1983, C. G. Schaack 11 94 and J. D. Morefield (ASV , DHA, NY). (Verified by N. Holmgren.) Significance. First AZ record. Cannon and Lloyd s.n. (Aug 1904) (NY) were ap- parently the first to collect and recognize P. nivea from the San Francisco Mountain (N. Holmgren, pers. comm.). This specimen was unknown to Kearney and Peebles (Arizona flora, Univ. of CaUf. Press, Berkeley, 1960) and subsequent authors, and collectors using these keys identified the trifoliolate P. nivea with the digitate P. concinna Richards. Potentilla nivea is primarily restricted to the ridgetops within the Arizona tundra. This habitat is subject to increasing foot traffic, particularly on Humphreys Peak, where the greatest number of plants were observed. Since this species is poorly represented in the alpine, and apparently finds its southernmost North American station and only Arizona occurrence on the San Francisco Mountain, protection seems warranted.— C. G. Schaack and J. D. Morefield, Deaver Her- barium, Dept. of Biol. Sciences, Northern Arizona University, Flagstaff" 8601 1. CowANiA suBiNTEGRA Kcamcy (Rosaceae). — Yavapai Co., n. end of Dead Horse Ranch St. Park near Cottonwood, low calcareous mesa top on Verde Formation with Krameria, Guttierrezia, Canotia and Parthenium, ca. 1050 m: ffowering and pro- ducing abundant fruit, 10 May 1984, A^. B. Herkenham s.n. (DHA); 16 Mar 1984, /. Anderson 84-2 (ASU). Yavapai Co., Coconino Natl. Forest, 0.5 km ne. of Bridge- port on U.S. Hwy 89A on calcareous Verde Formation, 1 020 m, 23 Sep 1984, Schaack and Schaack 1365 (DHA); 1 km n. of Bridgeport along Rocking Chair Rd., 1005 m, 23 Sep 1984, Schaack and Schaack 1363 (DHA). Significance. First reports for Yavapai Co., extensions of 130 km e. from the type locality in Mohave Co., and 200 km nw. from nw. Graham Co. These are the only known localities for this species, and Federal Endangered status has recently been granted. Many of the state park plants appeared to have been heavily grazed. A complex series of putative hybrids involving C. mexicana D. Don var. stansburiana (Torr.) Jeps. occurs in the Bridgeport area, and is currently under investigation by Schaack and Morefield.— Clark G. Schaack, James D. Morefield, James M. RoMiNGER, Deaver Herbarium, Dept. Biol. Sci., Northern Arizona Univ., Flagstaff" 860 1 1 ; and John Anderson, Dept. Botany and Microbiol., Arizona St. Univ., Tempe 85287. California Dedeckera eurekensis Reveal & Howell (Polygonaceae). — Mono Co., Inyo Natl. For., w. ffank of White Mts. ca. 5.5 km up Coldwater Canyon on s.-sloping talus above stream with Atriplex, Artemisia and Mirabilis, 2070 m (observed to 2200 m), 19 Jun 1984, Morefield 2126 (DHA, MICH, Morefield, NY, RSA, US); BLM land ca. 0.8 km up Coldwater Canyon on n. -sloping talus with Petalonyx, Hecastocleis, Eriogonum and Pericome, 1570 m, 19 Jun 1984, Morefield 2132 (DHA, MICH, Morefield, NY, RENO, RSA); Inyo Co. and Inyo Natl. For., 1 .9 km S45°E of Southern Belle Mine in Gunter Cr. Canyon, crevices in n. -sloping calcareous rock with He- castocleis, Brickellia and Psorothamnus, 1 570 m, 18 Jun 1 984, Morefield 2124 (DHA, KANU, MICH, MNA, Morefield, NY, RENO, RSA, UNLV, VDB), all collections in bloom. Specimens to be distributed. Previous knowledge. Ca. 10 generally small, disjunct stands known from the Last Chance, Inyo, Panamint and White Mts., all in Inyo Co., mostly on steep, unstable carbonate cliffs and talus from 1220-2070 m (Mary DeDecker, pers. comm., 1984; Novak and Strohm, Madrono 28:86-87, 1981; Reveal and Howell, Brittonia 28:245- 251, 1976). 1985] NOTEWORTHY COLLECTIONS 123 Significance. The Coldwater Canyon specimens represent the largest continuous stand known for the species (ca. 2.5 km^), the first records for Mono Co., and an extension 1 2 km n. from the one small stand previously reported for the White Mts. Along with the Gunter Cr. stand (ca. 0.3 km^), the known species population is approximately tripled. Dedeckera remains rare throughout its range, but does not appear endangered at this time. Opuntia pulchella Engelm. (Cactaceae). — Inyo Co., White Mts., BLM land 5.3 km N70°E of Antelope Spgs. along trail between Beer and Birch Cks. in Deep Springs Valley, plants in bloom on disturbed sandy flats dominated by Hilaria jamesii, 1560 m, 30 May 1984, Morefield 1982 (DHA, Morefield, NY, RSA); Deep Springs Valley, 30 May 1982 (flowering), A. C. Sanders 2464 (RSA); also observed in CA just n. of Deep Springs College (Inyo Co.) and at the mouth of McAfee Cr., e. side of the White Mts. (Mono Co.), both on disturbed sandy flats. Significance. First records for CA, confirming Munz's (1959, p. 313) prediction, and an extension 53 km s. from the nearest known locality in NV. The species should be considered rare and endangered in CA, because all the sites found are subject to trampling during heavy grazing.— James D. Morefield, Deaver Herbarium, Dept. Biol. Sci., Northern Arizona Univ., Flagstaff" 8601 1. Dalea ornata (Dougl.) Eat. & Wright (Fab aceae). — Lassen Co., flats on nw. side of Shaffer Mt., 12 km nnw. of Litchfield, growing in areas of churning vertisol clay soil (T30N R14E S5), 1670 m, 7 Jun 1984, Tiehm, Evans and Young 8592 (CAS, NY, RSA, UTC). Significance. First record for CA and a nnw. range extension of over 1 1 0 km from the Reno, NV area. Previously known from e. WA, e. OR, sw. ID and w. NV. IvEsiA baileyi S. Wats. (Rosaceae).— Lassen Co., Skedaddle Mts., Wendel Canyon, ca. 13 km e. of Wendel, growing in rock crevices along canyon walls (T29N R16E S22), 1600 m, 21 Jun 1979, Schoolcrafi 87 (NY); #2 canyon of Amedee Mt., growing in moist shade of rock crevices, 28 Jun 1983, Schoolcraft 1032 (NY). Significance. First record for CA and a range extension of about 50 km nw. from the Virginia Mts., Washoe Co., NV. Previously known from se. OR, sw. ID, and n. NV. We here consider var. setosa S. Wats, to be worthy of specific rank as /. setosa (S. Wats.) Rydb.— Schoolcraft and Tiehm, see note below. Nevada Cryptantha celosioides (Eastw.) Pays. (Boraginaceae).— Washoe Co., s. side of Mahogany Mt., 1 . 1 km ese. of Denio Camp on n. side of Little High Rock Cr., growing on whitish ash deposits (T39N R22E S23), 1610 m, 30 Jun 1983, Tiehm 8041 (CAS, NY, RSA, UTC). Significance. First record for NV and a se. range extension of over 150 km from southern OR. Previously known from ND to ID and se. OR. The corollas are unusually small, being at most 6 mm broad. Eriogonum CROSBY ae Rcvcal (Polygon aceae).— All from whitish ash deposits in Washoe Co., and all determined by J. L. Reveal. East of Butcher Flat (T42N R23E S36), 1700 m, 28 Jun 1983 Lisa Ganio 95 (MARY); e. of Grassy Ranch (T41N R22E S18), 1790 m, 22 Jun 1983 Lisa Ganio 96 (MARY); s. side of Mahogany Mt., 1.1 km ese. of Denio Camp, n. of Little High Rock Creek (T39N R22E S23), 1610 m, 30 Jun 1983, Tiehm 8040 (CAS, MARY, NY, RSA, UTC); Granite Range, e. side of Grass Valley Cr., 3.2 km nnw. of Grass Valley Ranch (T38N R22E S32), 30 Jun 1983, Tiehm 8043 (CAS, MARY, NY, RSA, UTC). Significance. First records for Nevada and a sse. range extension of about 90 km. Previously known only from the type area in Lake Co., OR (Brittonia 33:442, 198 1). 124 MADRONO [Vol. 32 Eriogonum prociduum Reveal (Polygonaceae).— Washoe Co., Hays Canyon Range, near upper end of Hays Canyon, ca. 18 km e. of Eagleville, CA, growing with Artemisia on a small gravelly, rather barren knob (T39N, R19E, S2), 2180 m, 14 Jun 1979, Schoolcraft 62 (MARY) (verified by J. L. Reveal); 13 Jul 1982, Schoolcraft 781 (NY); 5 Jul 1983, Tiehm 8056 (CAS, MARY, NY, RSA, UTC). Significance. First records for NV and a range extension of about 30 km ese. from Modoc Co., CA. Previously known from Lassen and Modoc Cos., CA and Lake Co., OR. IvEsiA RHYPARA Erttcr & Rcvcal (Rosaceae).— Washoe Co., Yellow Rock Canyon area, ca. 8 km n. of Mahogany Mt., growing on steep yellowish ash deposits along canyon walls, in two separate populations (T40N R22E SI), 1640 m, 14 Jun 1983, Schoolcraft 995 (TEX) (verified by B. Ertter); 7 Jul 1983, Tiehm 8100 (NY). Significance. Previously known from two areas, one in ne. Malheur Co., OR, and one in w. Elko Co., NV. This is a w. range extension of about 290 km from the Elko Co. site. Mentzelia packardiae Glad (Loasaceae). — Elko Co., near the IL Ranch, growing on barren tuffaceous clay hills (T42N R50E S6), 1600 m, 15 Jun 1982, Williams and Tiehm 82-133-1 (CAS, RENO) (verified by H. J. Thompson). Significance. First record for Nevada and a se. range extension of about 190 km from the type area in Leslie Gulch, Malheur Co., OR. Phacelia thermalis Greene (Hydrophyllaceae). — Washoe Co., 9.8 km w. of Buffalo Meadows road on Buckhom Rd. to Ravendale, growing along a small rocky wash (T35N R19E), 1800 m, 25 Jun 1984, Tiehm, Nachlinger and Schoolcraft 8818 (CAS, NY, RSA, UTC). Signiftcance. First record for NV and a short range extension w. from adjacent Lassen Co., CA. Previously known from ne. CA, se. OR, sw. ID and MT. Scutellaria holmgreniorum Cronq. (Lamiaceae).— Washoe Co., nw. of Norton Place, growing on a rocky vertisol with Chrysothamnus (T34N R19E S29), 1740 m, 16 May 1984, Schoolcraft 1167 (NY); Washoe Co., 3 km w. of Buffalo Meadows road on Buckhom Rd. to Ravendale, plants growing in clay soil with vertisols (T35N R19E), 1550 m, 25 Jun 1984, Tiehm, Nachlinger and Schoolcraft 8808 (CAS, NY, RSA, UTC). Significance. First record for NV and a ne. range extension of about 24 km. Pre- viously known only from e. Lassen Co., CA.— Gary Schoolcraft, Bureau of Land Management, 2545 Riverside Dr., Susanville, CA 96130 and Arnold Tiehm, New York Botanical Garden, Bronx, NY 10458. Oregon Amsinckia carinata Nelson & Macbride (Boraginaceae).— Malheur Co., ca. 7 km sw. of Harper via Hwy. 20, n. of Malheur River on low, rocky hills with outcrops of welded tufr(T20S R41E S22/23), 850-925 m, 15 Jun 1984, Joyal 496 (OSC, NY, ORE, US); ca. 8 km sw. of Harper (T20S R41E S 22/27), 900-925 m, 16 Jun 1984, Joyal 510 (OSC, NY, UTC, DS). Plants numerous in rocky soil of upper slopes, with Atriplex spinosa and Hordeum jubatum; Amsinckia tessellata occurs on lower slopes and in disturbed sites on hilltops. Previous knowledge. Known only from the type collection, J. B. Leiberg 2234 (isotype, OSC!), cited by Nelson and Macbride as from "Oregon: rocky soil, Malheur Valley, near Harper Ranch, alt. 1 100 ft., June 10, 1896." Leiberg's specimen label, as well as his field notes (consulted courtesy of the Smithsonian Institution archives), give the elevation as "1100 m." Mentioned by Suksdorf (Werdenda 1:112, 1931), Ray and Chisaki (Amer. J. Bot. 44:530, 1957), and Cronquist (Vascular plants of the 1985] NOTEWORTHY COLLECTIONS 125 Pacific Northwest 4:183, 1959; Intermountain flora 4:478, 1984) without citation of additional collections. Significance. This represents the rediscovery of a species that was thought to be extinct. The species was omitted from the standard floras for the Pacific states prior to the mention by Cronquist; however, Ray and Chisaki suggested that it "appears to be a distinct and interesting species . . . not known to exist in nature today." The U.S. Fish and Wildlife Service (Federal Register 48:53643, 1983) listed Amsinckia carinata as a possibly extinct taxon for which more information was needed to support a proposal for endangered or threatened status. The rediscovery of populations near the type locality should make it possible to clarify the species' status under the Endangered Species Act. — Elaine Joyal, Dept. Botany PI. Pathol., Oregon St. Univ., Corvallis 97331. Utah Aster sibiricus L. (Asteraceae).— Summit Co., Uinta Mts., 19 km nw. of Kings Pk. (T2N R13E S18 sw.'A), 3140 m, with scattered Engelmann spruce, 31 Aug 1981, Goodrich 16211 (BRY, NY). (NY collection verified by A. Cronquist.) Significance. First record of this species for UT and a range extension of ca. 240 km s. of the nearest locale in nw. WY. Gnaphalium viscoscum H.B.K. (Asteraceae).— Uintah Co. (TIS R23ES10 se.'A), e. side of Cow Cr., 2350 m, 17 Aug 1983, Neese 14964 (BRY). Significance. First record of this species for UT and a range extension of ca. 280 km w. of the nearest locale in nc. CO. Draba douglasii Gray (Brassicaceae). — Box Elder Co., Grouse Creek Mts., 29 km 263 degrees sw. of Rosette (T12N R17W S2 s.Va), windswept rocky ridge, 2438 m, 23 Jun 1982, Goodrich and Atwood 17127 (BRY, GH, SSLP). (GH specimen verified by R. C. Rollins.) Significance. First record of this species for UT and a range extension of ca. 50 km s. of the nearest locale in Twin Falls Co., ID. Astragalus robbinsii (Oakes) Gray (Fabaceae). — Summit Co., Wasatch Natl. Forest, Whitney Guard Station, 38 km 52 degrees ne. of Kamas (TIN R9E S3 nw.'A), 2730 m, 15 Aug 1983, Goodrich 19670 (BRY, NY). Significance. First record of this species for UT and a range extension s. and w. from the Rocky Mountains of WY and CO. Ribes laxiflorum Pursh (Grossulariaceae).— Juab Co., Deep Creek Mts., Thomas Cr., waterfall area, 2895 m, 20 Jun 1959, D. W. Lindsay 279 (UT); Toms Cr., 15 km 260 degrees w. of Callao (TllS R18W S16 NW 'A), 2615 m, wet area with boulders and downed timber, Engelmann spruce-aspen community, 1 3 Jul 1 983, Goodrich 19052 (BRY, ID, NY, SSLP). Significance. These specimens represent a disjunction of ca. 800 km e. from the Pacific Coast. The UT specimens seem closer to R. I. var. laxiflorum than to R. I. var. coloradense (Cov.) Jancz.— Sherel Goodrich, USDA For. Serv. Intermt. For. and Range Exp. Sta., Ogden, UT, stationed at Shrub Sciences Laboratory, Provo, UT 84601, DuANE Atwood, USDA For. Serv. Uinta Natl. For., Provo, UT 84601, and Elizabeth Neese, Monte L. Bean Life Science Museum, Brigham Young Univ., Provo, UT 84602. Wyoming Aristida oligantha Michx. (Poaceae). — Weston Co., 5.6 km se. of Upton, abun- dant in gray shale soil in semibarren pine-prairie (T47N R64W), 1280 m, 5 Aug 1973, Stephens 70098 (KANU). 126 MADRONO [Vol. 32 Significance. Reported for Weston Co., WY (Barkley, T. M. 1977. Atlas of the flora of the Great Plains. Iowa State Univ. Press), but without location. A range extension of 200 km nw. from Sheridan Co., NE. BoiSDUvALiA GLABELLA (Nutt.) Walp. (Onagraceae). —Campbell Co., along trib- utary of East Fork of S Bar Creek, ca. 13.7 air km w. of Gillette on Montgomery Rd., common in loamy clay in vernal pool (T50N R73W S29 nw.V4), 1433 m, 18 Sep 1978, Nelson 1994 (RM); Crook Co., ca. 4.8 air km n. of Oshoto, muddy pond margin (T54N R67W S32), 1250 m, 26 Jul 1978, Dueholm 4897 (RM). Significance. First records for WY, a range extension of ca. 80 km w. from Butte Co., SD. Celtis occidentalis L. var. occidentalis (Ulmaceae). — Sheridan Co., Big Horn Mts., Tongue River Canyon, ca. 8 air km wsw. of Dayton, woodland with Acer and Populus (T56N R87W SIO), 1371 m, 6 Sep 1933, Stell S-1 (USES); woodland with Prunus and Rhus (T56N R87W S8 and 9), 1463 m, 23 Sep 1935, Dickson 401 (USES); Tongue River Canyon Trail, population of 1-2 ha on calcareous substrate, from canyon wall to river, containing all age classes (T56N R87W S4 se.'A), 1341 m, 30 Oct 1983, Hamann 456 (RM). Significance. Second report of an apparently native population in WY; the other stand is in Goshen Co., ca. 410 km to the sw. Claytonia lanceolata Pursh var. flava (A. Nels.) C. L. Hitchc. (Portulaca- ceae). — Fremont Co., Waynes Cr., occasional at edge of meadow (T43N R105W Sll), ca. 2834 m, 21 Jun 1956, Gierisch 1877 (USES). Significance. First record for WY, a range extension of 185 km se. from Henry's Lake, Fremont Co., ID, the type locality. The only other known location is Anaconda, Deer Lodge Co., MT (Blankinship 768, RM; Henderson, D. M. 1981. In Vase. Pit. Spp. of Concern in Idaho. Univ. Idaho, For. Wldlf and Range Expt. Sta. Bull. 34). Eragrostis trichodes (Nutt.) Wood var. trichodes (Poaceae).— Goshen Co., ca. 1 1.2 air km nne. of Lingle, common in sandy soil on roadside (T26N R62W SIO se.V4), 1341 m, 27 Sep 1978, Nelson 2435 (RM). Significance. Listed as "reported" for Goshen and Laramie Cos., WY (Beetle, A. A. 1977. Grasses of Wyoming. Univ. Wyoming Agric. Expt. Sta. Res. J. 39R), but without locations. A range extension of ca. 100 km w. from Box Butte and Morrill Cos., NE. Euphorbia serpens H.B.K. (Euphorbiaceae). — Crook Co., ca. 9.7 air km se. of Moorcroft, wet ditch (T49N R67W S34 and 35), 1310 m, 24 Jul 1978, Dueholm 4696 (RM). Significance. First record for WY, a range extension of ca. 100 km sw. from Butte Co., SD. Larix occidentalis Nutt. (Pin aceae).— Teton Co., Teton Range, Darby Creek Canyon, ca. 9.7 air km sse. of Alta, in forest of Picea and Pinus (T43N Rl 18W S22), ca. 2286 m, 4 Jul 1983, Hartman 15720 (RM). Significance. First record for WY, a range extension of ca. 370 km ese. from Valley Co., ID, and se. from Granite Co., MT. An apparently native stand of 14 or more trees (14-16 m in height with a DBH of 25-30 cm), extending along creek bottom for 2-3 km. Leptodactylon watsonii (A. Gray) Rydb. (Polemoni aceae). — Fremont Co., Wind River Indian Reservation, vicinity of Boysen Dam, limestone rock crevices with Lesquerella and Petrophytum (T5N R6E S8), 1767 m, 23 Jun 1980, Dorn 3458 (COLO, RM). (Determined by W. A. Weber, COLO; verified by D. Wilken, CS.) 1985] NOTEWORTHY COLLECTIONS 127 Significance. First record for WY and a range extension ne. of ca. 275 km from Bear Lake Co., UT. LoEFLiNGiA SQUARROSA Nutt. subsp. TEXANA (Hook.) Bameby & Twisselmann (Caryophyllaceae).— Weston Co., Cambria Canyon, n. of Newcastle (T45, 46N R61W), 1300-1525 m, 29 Jul 1896, A. Nelson 2534 (COLO, RM). (Distributed as Gilia pungens caespitosa; the annotation as Leptodactylon pungens [Torr.] Nutt. by C. L. Porter, 1968, questioned and brought to our attention by W. A. Weber, COLO.) Significance. First record for WY, a range extension of 200 km nw. from Dawes Co., NE. Subspecies artemisiarum Bameby & Twisselmann is known from n.-cent. Sweetwater Co. (Bameby and Twisselmann. 1970. Madrono 20:398-408). Linaria canadensis (L.) Dum. Cours. var. texana (Scheele) Penn. (Scrophulari- aceae).— Campbell Co., ca. 54.4 km n. of Gillette, sandy prairie pasture, 30 Jun 1968, Stephens and Brooks 23858 (KANU); ca. 30.6 air km nne. of Gillette, eroded slopes and drainage (T53N R71W S25), 1190 m, 21 Jun 1978, Hartman, Dueholm and Sanguinetti 6830 (RM); ca. 18.5 air km nne. of Weston, open slopes and drainage (T55N R70W S14), 1 190 m, 22 Jun 1978, Hartman, Dueholm and Sanguinetti 6925 (RM); Cow Cr., ca. 10.5 air km s. of Weston, plains with Artemisia (T53N R71W S26), 1 190 m, 21 Jun 1978, Dueholm, Hartman and Sanguinetti 2394 (RM); Horse Cr. Ranch, ca. 14.5 air km nw. of Weston, plains with Artemisia (T55N R72W S26), 1220 m, 8 Jul 1978, Dueholm and Sanguinetti 3544 (RM); Crook Co., ca. 27 km nw. of Belle Fourche, SD, few plants in alkaline soil in moist prairie (T56N R60W), 975 m, 8 Jul 1973, Stephens 66672 (KANU); ca. 12.9 air km nne. of New Haven, shaley ridge with Quercus and Pinus (T56N R66W S7 and 8), 1 160 m, 26 Jul 1978, Dueholm 4991 (RM). Significance. Reported from Campbell and Crook Cos., WY (Barkley 1977), but without location. A range extension of ca. 60 km sw. from Harding Co., SD. MoNARDELLA ODORATissiMA Bcnth. subsp. GLAUCA (Greene) Epling (Lamiaceae).— Lincoln Co., Salt River Range, head of Strawberry Cr., on w. side of pass, abundant on grassy s. exposure (T33N Rl 17W), 2895 m, 20 Aug 1927, McDonald 779 (USES); ca. 9.6 air km e. of Etna, rocky se. exposure with Balsamorhiza, Apocvnum, Cirsium, and Pseudotsuga, 2500 m, 27 Jul 1979, Shultz 621 (USES). Significance. First records for WY, a short range extension e. from adjacent ID. Opuntia macrorhiza Engelm. var. macrorhiza (Cactaceae).— Goshen Co., 1 1.2 km w. of Ft. Laramie, common in valley in prairie hills (T26N R65W), 1310 m, 11 Aug 1973, Stephens 70884 (KANU); ca. 6.4 km n. of Torrington, roadside with Kochia, Salsola, and Artemisia (T25N R61W), 1280 m, 20 Aug 1977, Dorn 3010 (RM); along the North Platte River, ca. 1.6-2.4 air km sw. of Torrington, sandy soil in disturbed bottomland with Tribulus (T24N R61W S17 ne.'A), 1250 m, 26 Sep 1978, Nelson 2379 (RM); along the North Platte Ditch, ca. 3.7 air km nw. of Tor- rington, common in sandy soil in upland grassland with Andropogon and Panicum (T24N R61 W S5 ne.'A), 1250 m, 29 Sep 1978, Nelson 2570 (RM); ca. 6.4 air mi nw. of Ft. Laramie, frequent on sandy gravel roadcut (T26N R64W S7 ne.'A), 1325 m, 29 Sep 1978, Nelson 2571 (RM). Significance. Reported from Goshen Co., WY (Barkley 1977), but without location. A short range extension w. from adjacent NE. Pectis angustifolia Torr. var. angustifolia (Asteraceae). — Goshen Co., on gravel hills on North Platte, ca. 19 km below Ft. Laramie, 1280 m, 30 Jul 1858, Engelmann s.n. (MO) (determined by D. J. Keil, OBI); Converse Co., top of prom- inent red clinker hill ca. 25 air km nnw. of Bill, common in coarse gravel (T40N R71W S12 e.'/2 of ne.V4), 1458 m, 30 Jul 1978, Guidinger 324 (RM); ca. 25 air km 128 MADRONO [Vol. 32 n. of Bill, scoria slopes with Artemisia and Eriogonum (T40N R70W S6 sw.'A and S7 nw.'/4), 1463 m, 3 Sep 1980, Dorn 3673 (RM). Significance. First records for WY, a range extension of ca. 1 50 km n. from Weld Co., CO. Physalis hederaefolia a. Gray var. comata (Rydb.) Waterfall (Solanaceae).— Goshen Co., 6.4 km n. of Lingle, few plants in coarse gravel soil on prairie hillside (T26N R62W), 1371 m, 10 Jul 1974, Stephens 77428 (KANU). Significance. Reported from Goshen and Sheridan Cos., WY (Barkley 1977), but without location. A short range extension w. from adjacent NE. The record for Sheridan Co. was based on a specimen {Stephens 69841, KANU) of P. heterophylla Nees. PoTENTiLLA HOOKERiANA Lchm. (Rosaceae). —Albany Co., alpine summits. Tele- phone Mines (T16N R79W), 3048-3353 m, 31 Jul 1900, A. Nelson 7877 (RM); Johnson Co., Big Horn Mts., ca. 4 km e. of Cloud Pk., alpine tundra (T51N R85W SI 7 and 18), 3108 m, 9 Jul 1979, Odasz 1647a (RM); Park Co., Beartooth Range, ridge above US 212 at MT border, alpine tundra (T58N R104W SI 5), ca. 3100 m, 5 Jul 1958, Johnson 18 (RM); Teton Co., vicinity of Hoback Canyon, rocky moist hillside (T38N R115W), 2133 m, 24 Jun 1932, Williams and Pierson 704 (RM); Hoback Canyon, s. exposure (T38N Rl 15W), 2286 m, 24 Jun 1933, Williams and Pierson 1164 (RM). (All except Odasz 1647a determined by B. C. Johnston, COLO.) Significance. First records for WY, a range extension of ca. 300 km se. from Deer Lodge Co., MT. RoRiPPA TRUNCATA (Jcps.) R. Stuckcy (Brassicaceae). —Goshen Co., Springer Res- ervoir, ca. 2.4 air km s. of Yoder, frequent in dry sandy clay on margin of reservoir with Chenopodium, Leptochloa, and Spergularia (T22N R62W SIO se.'A), ca. 1310 m, 28 Sep 1978, Nelson 2492 (RM); Hawk Springs Reservoir, ca. 10.5 air km se. of Hawk Springs, common in dry sandy clay on margin of reservoir with Cyperus, Eragrostis, and Gnaphalium (T20N R6 1 W S9 se.'A), ca. 1 360 m, 28 Sep 1978, Nelson 2515 (RM). (Verified by R. L. Stuckey, OS.) Significance. Reported from Laramie Co., WY (Barkley 1977), but without location. A short range extension w. from adjacent NE. SciRPUs heterochaetus Chase (Cyperaceae). — Sheridan Co., ca. 21 km e. of Leiter, abundant around prairie farm pond (T54N R76W), 1158 m, 2 Aug 1973, Stephens 69820 (KANU). Significance. Reported from Sheridan Co., WY (Barkley 1977), but without loca- tion. A range extension of ca. 380 km wnw. from Jackson Co., SD. Thellungiella salsuginea (Pall.) O. E. Schulz (Brassicaceae).— Carbon Co., ca. 38 air km sse. of Wamsutter (T16N R92W S 19), 201 1 m, 21 Jun 1979, Colony Oil III A-E8 (COLO). (Determined by W. A. Weber, COLO.) Significance. First record for WY, a range extension of ca. 300 km nw. from Park Co., CO. Veratrum tenuipetalum Heller (Liliaceae). — Carbon Co., Sierra Madre, Hog Park Reservoir, ca. 2 1 air km sw. of Encampment, frequent in clay in a grassy park with Erigeron (T12N R84W S6 sw.'A), 2560 m, 3 Jul 1977, Nelson and Nelson 1479 (RM); ca. 26.5 air km sw. of Encampment, springy slope (T12N R86W SI), ca. 2530 m, 18 Jul 1979, Nelson and Blunt 4034 (RM). Significance. First records for WY, a short range extension n. from adjacent Routt Co., CO. — Ronald L. Hartman, B. E. Nelson, and Keith H. Dueholm, Rocky Mountain Herbarium, Dept. of Botany, University of Wyoming, 3165 University Station, Laramie 82071. ANNOUNCEMENT Eight Dollar Mountain, Josephine County, Oregon, USA. Eight Dollar Mountain, composed of ultramafic parent material and laterite soils, is a significant natural landmark in the Illinois River Valley of sw. Oregon. The mountain has become a source of conflict among land use planners because the soils support populations of rare plants and contain nickel and other heavy metals. Botanists want to protect the plants, miners want to surface mine, the timber industry wants to log, and ranchers want to graze cattle. The major property owners, the United States Forest Service, the Bureau of Land Management, the State of Oregon, and Josephine County, are presently attempting to develop a management plan for the area that will satisfy these diverse interests. The mountain is the type locality for two taxa, Hastingia bracteosa Wats, and Aster paludicola Piper, and possibly a third, Senecio hesperius Greene. In addition, many Illinois Valley and sw. Oregon endemics (Calochortus howellii Wats., Lewisia op- positifolia (Wats.) Robins., and Gentiana bisetacea Howell, among others) grow on the drier slopes of the mountain or in the many fine interior- valley Darlingtonia bogs around its base. Eight Dollar is a botanical treasure. To ensure that Eight Dollar Mountain remains a botanical rather than a mineral treasure, a strong case must be made for its botanical importance to the scientific community. To accomplish this I need some information. If you knew of Eight Dollar Mountain before reading this note: 1) How did you learn about the mountain? 2) Have you ever visited the mountain? If so, what was the purpose and number of your visits? This information will be summarized and passed on to the government agencies. At least 51 taxa of plants have been described from the relatively small Illinois Valley area. Most of the species were collected by Thomas Howell between 1884 and 1887. Are there other places with that concentration of type localities, or is the Illinois Valley unique? Please send your responses to Frank A. Lang, Department of Biology, Southern Oregon State College, Ashland, OR 97520, USA. ANNOUNCEMENT International Organization of Plant Biosystematists The Executive and Council of the International Organization of Plant Biosyste- matists (lOPB) will meet during the Third International Congress of Systematics and Evolutionary Biology, University of Sussex, Brighton, U.K., 4-10 July 1985. Anyone wishing to place an item on the agenda for discussion should write to Dr. Liv Borgen, Secretary, lOPB, Botanical Garden and Museum, Trondheimsveien 23B, N-Oslo 5, Norway. lOPB is holding a Symposium "Differentiation Patterns in Higher Plants," Zurich, Switzerland, July 13-18, 1986. Information on participation may be obtained from the Chairperson, Dr. Krystyna Urbanska, Geobotanisches Institut, ETH, Stiftung Rubel, Zurichbergstrasse 38, CH-8044 Zurich, Switzerland. lOPB publishes the lOPB Newsletter. Information for the lOPB Newsletter may be sent to the Editor, Dr. Krystyna Urbanska. Madrono, Vol. 32, No. 2, pp. 129-130, 26 April 1985 130 MADRONO [Vol. 32 Application forms for membership in lOPB may be obtained from the President of lOPB, Dr. William F. Grant, Department of Plant Science, P.O. Box 282, Mac- donald College of McGill University, Ste. Anne de Bellevue, Quebec, Canada H9X ICO, or by sending US $25 (for the period 1983-1987) directly to the Secretary- Treasurer of lOPB, Dr. L. Borgen. ANNOUNCEMENT The Shrub Research Consortium is sponsoring the fourth wildland shrub sympo- sium August 7-9, 1985, at Snowbird Resort, near Salt Lake City, Utah. The sym- posium, "Plant/Herbivore Interactions," will feature invited and contributed papers on aspects of plant-animal interactions, emphasizing but not limited to vertebrate herbivores and shrub ecosystems. Contributed presentations will be 20 minutes long. The proceedings will be published. If you would like to present a paper, send a title and abstract by 15 May 1985, to: Dr. F. D. Provenza, Department of Range Science, College of Natural Resources, UMC 52, Utah State University, Logan, UT 84322. For further information about the symposium and facilities, please contact: Theresa A. Bigbie, Conferences and Workshops, Brigham Young University, 297 CONF, Provo, UT 84602, (801) 378-4903. ERRATUM The name Polygonum douglasii E. Greene subsp. austiniae (E. Greene) E. Murray (Kalmia 12:23. 1982) has precedence over the identical combination proposed by Hickman (Madrono 31:250. 1984). Subscriptions— Membership Membership in the CaHfomia Botanical Society is open to individuals ($ 1 8 per year; students $ 10 per year for a maximum of seven years). Members of the Society receive Madrono free. Family memberships ($20) include one ten-page publishing allot- ment and one journal. Emeritus rates are available from the Corresponding Secretary. Institutional subscriptions to Madrono are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Madrono should be sent to the Corresponding Secretary. Information for Contributors Manuscripts submitted for publication in Madrono should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For ar- ticles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed ending with figures, and number each page. Do not use a separate cover page, "erasable" paper, or footnotes. Manuscripts prepared on dot matrix printers may not be considered. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including cap- tions) to the Madrono page, and designed for reduction to % original size. Scales should be included in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff card- board and sent flat. No illustrations larger than 22 x 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in Madrono 22:288. 1974. In- stitutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Herbariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Periodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted ten pages in Madrono each year or 20 pages over a two-year period. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $60.00 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing Madrono on a reasonable schedule, with equity among all members for access to its pages. Printer's fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author's changes after typesetting @ $3.00 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. CALIFORNIA BOTANICAL SOCIETY VOLUME 32, NUMBER 3 JULY 1985 IS3 MADRONO A WEST AMERICAN JOURNAL OF BOTANY Contents The Specific Status of Lasthenia maritima (Asteraceae), an Endemic of Seabird-Breeding Habitats Michael C. Vasey 1 3 1 Observations on Chamaesyce (Euphorbiaceae) in the Galapagos Islands Michael J. Huft and Henk van der Werjf 143 Post-Fire Seedling Establishment of Adenostoma fasciculatum and Ceanothus greggii in Southern California Chaparral Jochen Kummerow, Barbara A. Ellis, and James N. Mills 148 Spartina (Gramineae) in Northern California: Distribution and Taxonomic Notes Douglas Spicher and Michael Josselyn 1 58 The Systematic Relationship of Asarina procumbens to New World Species in Tribe Antirrhineae (Scrophulariaceae) Wayne J. Elisens 168 Mimulus norrisii (Scrophulariaceae), A New Species from the Southern Sierra Nevada Lawrence R. Heckard and James R. Shevock 179 NOTES AND NEWS Notes on the Genus Burroughsia (Verbenaceae) James Henrickson 186 New Combinations in California Chamaesyce (Euphorbiaceae) Daryl L. Koutnik 187 Rediscovery and Reproductive Biology of Pleuropogon oregonus (Poaceae) Paul P. H. But, Jimmy Kagan, Virginia L. Crosby, and J. Stephen Shelly 189 NTOTE WORTHY COLLECTIONS Arizona 191 California 191 Colorado 191 Mexico 191 New Mexico 192 Wyoming 1 92 REVIEWS 193 ANNOUNCEMENTS 142, 157, 167, 178, 185 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Madrono (ISSN 0024-9637) is published quarterly by the California Botanical So- ciety, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to James R. Shevock, Botany Dept., California Academy of Science, San Francisco, CA 941 18. ^■^//Yor— Christopher Davidson Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1985 — Sterling C. Keeley, Whittier College, Whittier, CA Arthur C. Gibson, University of California, Los Angeles 1986— Amy Jean Gilmartin, Washington State University, Pullman Robert A. Schlising, California State University, Chico 1987— J. Rzedowski, Instituto Politecnico Nacional, Mexico Dorothy Douglas, Boise State University, Boise, ID 1988 — Susan G. Conard, USDA Forest Service, Riverside, CA William B. Critchfield, USDA Forest Service, Berkeley, CA 1989 — Frank Vasek, University of Cahfomia, Riverside Barbara Ertter, University of Texas, Austin CALIFORNIA BOTANICAL SOCIETY, INC. Officers for 1985-86 President: Charles F. Quibell, Department of Biology, Sonoma State University, Rohnert Park, CA 94928 First Vice President: Rodney G. Myatt, Department of Biological Sciences, San Jose State University, San Jose, CA 951 14 Second Vice President: David H. Wagner, Herbarium, Department of Biology, University of Oregon, Eugene, OR 97403 Recording Secretary: Virgil Thomas Parker, Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: James R. Shevock, Department of Botany, California Academy of Sciences, San Francisco, CA 94 1 1 8 Treasurer: Wayne Savage, Department of Biological Sciences, San Jose State Uni- versity, San Jose, CA 95192 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, Rolf W. Benseler, Department of Biological Sciences, California State University, Hayward, CA 94542; the Editor of Madrono; three elected Council Members: James C. Hickman, Department of Botany, Uni- versity of Cahfomia, Berkeley, CA 94720; Thomas Fuller, 171 Westcott Way, Sacramento, CA 95864; Annetta Carter, Department of Botany, University of California, Berkeley, CA 94720; and a Graduate Student Representative, Joseph M. DiToMASO, Department of Botany, University of California, Davis, CA 95616. THE SPECIFIC STATUS OF LASTHENIA MARITIMA (ASTERACEAE), AN ENDEMIC OF SEABIRD-BREEDING HABITATS Michael C. Vasey Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Abstract The taxonomic status of Lasthenia maritima has been a subject of some disagree- ment largely because of the poor representation of this taxon in herbaria. The present study, based on field observation and morphological analyses of numerous additional collections, supports specific status. The closely related, self-incompatible L. minor occurs on mainland central California whereas the autogamous L. maritima primarily occupies off-shore rocks and islands inhabited by breeding and roosting seabirds from central California to northern British Columbia. Morphological comparison indicates that the two taxa differ significantly in nine out of the 12 quantitative characters that were investigated. Although the two taxa can produce relatively fertile hybrids under greenhouse conditions, they maintain a high level of reproductive isolation under natural conditions. Lasthenia (Asteraceae: Heliantheae) is a genus of predominantly annual herbs confined largely to western North America (Omduff 1966). It has received the benefit of a comprehensive biosystematic study (Omduff 1966, 1976, Omduff'et al. 1974, Bohm et al. 1974, Altosaar et al. 1 974) and relationships within the genus are generally well defined. One member of this genus, originally described as a species, has been relegated to subspecific status as Lasthenia minor (DC) Omduff* subsp. maritima (Gray) Omduff*. This taxon has been collected rarely and was poorly understood as a consequence of a lack of adequate material for comparative analysis. This paper pre- sents new evidence that this taxon should be recognized at the spe- cific level as Lasthenia maritima (Gray) M. Vasey. Lasthenia maritima was originally described at the specific level by Gray (1868) and was accepted as a species by Greene (1894) and Jepson (1925). In a discussion conceming the biota of the Farallon Islands, however, Blankenship (Blankenship and Keeler 1892) ob- served that "Only one ^XdinX— Baeria maritima [=L. maritima] — has been long enough isolated to show variation for which specific rank has been claimed, and it is seriously questioned whether it has departed far enough from B. uliginosa [=L. minor] to be considered even a variety." These doubts conceming the status of this taxon foreshadowed its later relegation by Ferris (1955) to subspecific sta- Madrono, Vol. 32, No. 3, pp. 131-142, 19 August 1985 132 MADRONO [Vol. 32 tus, a status accepted in the most recent revision of the genus by Omduff(1966). OmdufF(1965, 1966), using field and experimental evidence, con- cluded that L. maritima (as subsp. maritima) is apparently confined to seabird-breeding habitats in California and, presumably, else- where throughout its distribution. At that time, L. maritima was known from only twelve localities over a distance of ca. 2400 km, ranging from the Farallon Islands, California (37°4r53"N, 123°00'05"W) to Triangle Island (50°52'00"N, 129°05'00"W) near the northwestern tip of Vancouver Island, British Columbia; whereas L. minor was known from numerous collections from both the in- terior and coast of mainland central California (Fig. 1). Omduff determined that L. minor and L. maritima have n = A and are interfertile after artificial hybridization. Lasthenia minor is self-in- compatible; L. maritima is predominantly autogamous. One of the primary morphological distinctions between the two taxa, reduced length of ray floret ligules in L. maritima, was believed to be as- sociated with its autogamous breeding system. Omduff* (1966) hy- pothesized that L. maritima represents a relatively recent derivative of L. minor and speculated further that L. maritima has achieved its much more extensive geographical distribution via dispersal by seabirds. The present study was motivated by the findings and hypotheses raised by Omduff*(l 966). The relationship of L. maritima to seabird- breeding habitats was examined and these habitats were explored throughout the potential range of L. maritima in order to survey its actual distribution. As a consequence, the number of known pop- ulations of L. maritima has increased to seventy-four. Material of L. minor was collected in order to provide an adequate basis for comparing the two taxa. These collections and other available ma- terial were used to analyze several key morphological characters. The results of this survey and other lines of evidence support the recognition of L. maritima at the specific level. Methods Nine field trips were made to Southeast Farallon Island at different times of the year in order to gather life history data for Lasthenia maritima and its relationship to the seabird-breeding habitat. Po- tential seabird-breeding habitats were explored between Santa Bar- bara Island in Southern California and Barkley Sound on the west coast of Vancouver Island, resulting in the collection and/or obser- vation of L. maritima at 43 sites. Collections of L. minor were made at 2 1 coastal and eight interior localities. Where possible, collections were made of up to ten or more individuals per population and voucher specimens were deposited at CAS, UC, and SFSU. Field 1985] VASEY: LASTHENIA MARITIMA 133 Fig. 1 . Distribution of Lasthenia minor and L. maritima. Symbols indicate the sites sampled in the morphological analysis as well as the sites known for L. maritima prior to 1970 as contrasted to those that have been discovered since 1970. information recorded included features of the habitat, associated vegetation, and, in the case of L. maritima, associated seabirds and their breeding or roosting activity. Loans of both taxa from 22 her- baria were also examined, providing supplemental material for mor- phological analysis. An analysis of 25 morphological characters and nine derived ratios was used to survey 1 0 individuals per population in six populations of L. minor and 1 2 populations of L. maritima. These characters encompassed vegetative as well as ray floret, pap- pus, and achene traits. Twelve basic and eight ratio characters were then selected and the analysis was expanded to include 19 popula- tions of L. minor (n = 5-10), totaling 165 individuals, and 41 pop- ulations of L. maritima (n = 1-10), totaling 249 individuals. Mean values were computed for these characters in each population and submitted to discriminant function analysis and cluster analysis us- ing Statistical Analysis System programs. The discriminant function analysis provided mean and variance data for each of these taxa and the 20 characters used in the survey. Standard error determinations were computed and the character means were compared by multiple group-comparison -tests (Schefler 1980). 134 MADRONO [Vol. 32 The genetic basis of these quantitative characters was evaluated by planting achenes from 7 L. minor populations and 26 L. maritima populations under uniform conditions at the San Francisco State University greenhouse during winter, 1981. Progenies (n = 1-10) were obtained from 21 populations of L. maritima and phenetic data concerning growth habit and flowering and fruiting times were recorded. Progenies were later scored for the same characters as the parent populations. The mean values for the progeny and parent populations were then compared by correlation analysis (Schefler 1980). Results and Discussion Seabird-breeding habitats include islands, offishore rocks, and iso- lated coastal headlands that annually host colonies of breeding and/ or roosting coastal and oceanic seabirds. Such habitats are charac- terized by trampling, burrowing, and plant predation for nest ma- terials as a result of the activities of these birds, by stringent edaphic conditions resulting in part from guano deposition, and by exposure to persistent wind and salt spray. Lasthenia maritima is currently known to inhabit 74 localities (Fig. 1). Forty five of these sites occur along the California coast, where the most intensive exploratory efforts have been concentrated. It is possible that many undiscovered localities exist along the coasts of Oregon, Washington, and British Columbia. Although L. maritima is listed as rare and endangered in California (Smith et al. 1980), it is now recognized that this status represents an artifact of the collection record and, as a result of this study, L. maritima will be removed from this list. Of the 74 pop- ulations that have been identified, 70 are found on islands or offshore rocks and only four occur on coastal headlands. All of these sites demonstrated differing degrees of activity by breeding and/or roost- ing seabirds. The most common avian associates of L. maritima are the Western Gull {Larus occidentalis) and, north of northwestern Washington, also the Glaucous-winged Gull {Larus glaucescens), which collectively were noted to be breeding at 55 of the 69 L. maritima sites for which data and/or observations were available. Other seabirds often associated with L. maritima sites include Pe- lagic Cormorants {Phalacrocorax pelagicus). Pigeon Guillemots (Cepphus columba), and Brandt's Cormorants (Phalacrocorax pen- icillatus). Over two-thirds of the 43 observed populations of L. maritima were estimated to be small in numbers of individual plants (fewer than 1000) and yet in 34 of these 43 sites, L. maritima either dom- inated or locally dominated the vegetation occupying these habitats (Vasey, unpubl. data). This underscores the observation that few plant species appear to be capable of tolerating seabird-breeding habitats and, of those that can, these habitats are often so limited 1985] VASEY: LASTHENIA MARITIMA 135 in size that only small populations occur. By contrast, when large areas of habitat are available (e.g., Southeast Farallon Island), L. maritima occurs in dense colonies containing hundreds of thousands of individuals. The few mainland localities for L. maritima include the only known mainland site for this species in California (on a granite ledge at the tip of Tomales Point, Point Reyes Peninsula, Marin County), and three headland populations along the central Oregon coast (Seal Rock State Park, Yaquina Head, and Otter Crest) in Lincoln County. All of these populations were included in the morphological analysis. The range of habitats for L. minor is much more varied than for L. maritima, but these do not include seabird-breeding sites. Las- thenia minor occurs along canyon bottoms in the Sierra Nevada foothills that border the San Joaquin Valley, along pond margins, disturbed areas, and in openings in the alkali scrub and grassland in the San Joaquin Valley, in arid valley flats in the Inner South Coast Ranges, and along the immediate coast from San Luis Obispo through Mendocino County. The most extensive populations were observed in the halophytic communities of the southern San Joaquin Valley and San Luis Obispo County. South of San Francisco Bay, L. minor occurs sporadically at the margins of coastal terraces, bluffs, and disturbed sites near the ocean. North of San Francisco Bay, L. minor additionally occupies stabilized dunes as well as coastal bluffs and disturbed areas near the ocean. Nine out of the twelve quantitative morphological characters sur- veyed differ significantly between L. minor and L. maritima, one at the p < 0.05 level and eight at the p < 0.001 level (Table 1). All eight ratio characters also differ significantly. This analysis helps to identify a number of characters that, in combination, consistently distinguish these two taxa. As recognized by Omduff' (1966), ligules of the ray florets for L. maritima are usually short, ca. 2 mm long, whereas in L. minor rays are significantly larger, averaging close to 5 mm long. This distinction is even more consistently expressed as the ratio between ray floret/phyllary length, which is less than 0.7 in L. maritima and greater than 0.7 in L. minor. Pappus characters are also particularly helpful in distinguishing these two species. Las- thenia maritima frequently has ca. 4-12 awns and short, narrow scales each with a deeply laciniate superior margin. Lasthenia minor generally possesses ca. 2-3 awns and long, wide scales with a shal- lowly fimbriate margin. Diflerences in achene size and pubescence pattern also characterize L. minor and L. maritima. Although L. minor achenes are usually 1.9-2.6 mm long and pubescent along achene margins, the majority of L. maritima populations have larger achenes (2.6-3.3 mm) that are densely pubescent throughout; how- ever, occasional populations of L. maritima have much shorter achenes (2. 1-2.4 mm) and certain populations have achenes that are 136 MADRONO [Vol. 32 111 II q g . V * II ^ II S c5 3 ^ II -53 ^* ^ < S g ^ V c II c 7 Ol ^ ^li S <^ S ■2 &X « O S;^ oi TO ^ T3 7 1^ S N C3 W >. C H S c U C« Q Z 5 — 2 »-■ 2 < W Oh o 53 U J3 a; N ^ ^ i , O T3 ^ j= & ca < 3 H o; « ««« « «««« O r^^ w-^ w-^ O On en o O c5 O ^ O — < friooinosoo^Ooor~-fn(^ O ^ t~-- O rsi in Tf r~- ^ vo r^j Tt p r--' iri (N tN fN ^ d d cN d d d d d d cn m -H (SI ^ ^ ^ ^ ^ ^ ^cnvo^tN^fN-^r^inr-oo^ 0'^'^rs)0^o^r-omvo'<4"roosTt^0^or~~ Q^^(N(NVO^'^OOOOSTt^fN(Nr-voin'^Tl-Os^ ^Ttr-osin-j-Tt-^irii^sooos^^inos^- wwwww ^oo oooooooo^ os^^u->ooooos<^-^inos(N^ooo^r~-r-oo O-^tNOmTfOOOOmOOOOOOOOOs roddddddddddddddddddd +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 vo -H (» CO r--_ rn ON t--- rt <^ r^l d 1 n Q 1 n 8. Awn length (mm) 21 0.602 3.015* 9. Scale length (mm) 18 0.44 1.919 10. Scale width (mm) 17 0.678 4.025** 1 1 . Number of scale divisions 16 0.619 2.946* 12. Length of scale division (mm) 16 0.376 1.520 13. Ray ligule/phyllary length 21 0.736 4.739** 14. Awn/achene length 21 0.546 2.840* 15. Achene/ray ligule length 21 0.686 4.110** 16. Scale/awn length 18 0.205 0.842 17. Scale/achene length 18 0.428 1.890 18. Scale division/scale length 16 -0.204 0.780 19. Mid/marginal achene hairs 21 0.072 0.314 20. Total achene hairs 21 0.552 2.893 Sparsely pubescent. Similarly, while L. minor has densely pubescent peduncles (as well as other vegetative features), L. maritima pop- ulations are highly variable with respect to vegetative pubescence, ranging from subglabrous to highly pubescent. Other morphological differences reported by Omduff (1966) include reduced, obtuse an- ther tips and reduced stigmatic hairs in L. maritima as opposed to L. minor. The cluster analysis supported the placement of the Tomales Point and Oregon coast populations in L. maritima. Tomales Point is the only sympatric site known for these two species since one of the largest coastal populations of L. minor occurs on the uplifted dunes that border abruptly the lower granite ledge that hosts L. maritima. The Tomales Point L. maritima population does show some inter- mediate characteristics, but both taxa retain their identity to a great extent despite this proximity. Furthermore, parapatric populations at three other localities (Bird Rock and nearby bluffs, Pt. Reyes Peninsula; stack near Chimney Rock and nearby bluffs, Pt. Reyes Peninsula; and San Pedro Rock and Point San Pedro, San Mateo Co.) maintain a high level of morphological distinctness. The parent-progeny correlation analysis was limited to 21 pop- ulations of L. maritima ranging from Southeast Farallon Island, 138 MADRONO [Vol. 32 California to Triangle Island, British Columbia (Table 2). Because of the high level of morphological inter-population variation in L. maritima, these results are particularly useful for providing insight into the heritability of these quantitative characters. Two of the key characters separating L. minor and L. maritima (ray ligule length and awn number) were found to be highly correlated when popu- lation means were compared to first generation progeny means. This suggests a strong genetic component to these two characters and attests to their reliability in distinguishing L. maritima from L. minor. Although there also appears to be a significant genetic com- ponent to vegetative and achene characters surveyed, these are not as useful in separating the two species because of the amount of inter-population variation in L. maritima and the consequent degree of overlap between L. minor and L. maritima in these characters. Pappus scale characters are also highly variable within individuals and populations and presented some of the lowest correlation coef- ficients. Nevertheless, the scales of L. maritima are generally so qualitatively different from those of L. minor that they are still quite useful in separating the two taxa. The combination of multiple awns, reduced laciniate scales, and densely pubescent achenes in L. maritima may represent an adaptive shift towards enhanced dispersability by seabirds. Achenes of L. maritima were found embedded in the feather matrix of ten out of 15 recently dead Western Gulls examined on Southeast Farallon Island during late summer, 1979 (Vasey 1984). Gulls use large amounts of achene-laden L. maritima during the breeding season on Southeast Farallon Island and, at summer's end, thousands of these birds disperse rapidly from the island to favorite "vacation spots" as far north as the Washington coast (Spear 1982). OmdufFs (1966) seabird dispersal hypothesis for L. maritima thus seems high- ly plausible on the basis of these and other data (Vasey, unpubl. data) and it is bolstered further by the additional distributional information for L. maritima. This distributional pattern would be inexplicable except for dispersal by seabirds. In conjunction with reproductive, ecological, distributional, and morphological differences between L. minor and L. maritima, there are also certain physiological and biochemical distinctions. Altosaar et al. (1974) found distinctively different albumin and globulin pro- files in the dormant achenes of L. minor and L. maritima. Bohm et al. (1974) determined that the flavonoid profiles of these two species also differ. Whereas greenhouse studies suggest that L. minor is not tolerant of guano-modified soils (Vasey, unpubl. data), L. maritima thrives under such conditions. Gilham (1956) and Goldsmith (1973) have proposed that the chief limiting factor for most vegetation in guano-modified soils is the water stress that results from the high organic ion concentrations in such soils. Omduff (1965) found that L. maritima accumulates nitrates in its foliage whereas L. minor 1985] VASEY: LASTHENIA MARITIMA 139 does not. Possibly L. maritima has diverged from L. minor in the development of an organic solute (Dainty 1979) that allows L. ma- ritima to accumulate nitrate ions within its cell vacuoles and avoid this form of water stress. Species within Lasthenia are generally self-incompatible and the divergence of selfing taxa has occurred with important consequences in three indpendent lines (Omduff 1966). One line, comprising L. glaberrima and L. kunthii, is differentiated at the sectional level (sect. Lasthenia) with no known close relationship to other sections in the genus (Omduff 1966). A second line is the putative amphi- diploid L. microglossa in sect. Burrielia. The third line is L. ma- ritima in sect. Ptilomeris. A common denominator between all three lines is that L. glaberrima, L. kunthii, L. microglossa, and L. ma- ritima represent four of the most geographically widespread species in the genus, especially compared to known progenitors in the cases of L. microglossa and L. maritima. Each has been highly successful at colonizing unusual habitats requiring specialized adaptations. They also share in common reduced ray floret ligules and relatively large, trichomed achenes with an awned or awn-like pappus. The present study has demonstrated that there is a high level of reproductive isolation between natural populations of L. minor and L. maritima. It is reasonable to expect that this condition will persist through time because of the ecological divergence between these two species and the resultant spatial isolation that is the rule between seabird-breeding habitats and the mainland coast. The potential for cross-pollination between the two taxa is and presumably will con- tinue to be inhibited by this spatial isolation, by a lack of typical pollinator activity on off-shore islands (Moldenke 1971, Vasey, un- publ. data), and by their different breeding systems. As a conse- quence, L. minor and L. maritima should continue to diverge until the close relationship between these two taxa eventually becomes obscured. It is, however, the close relationship between this species pair that currently offers the potential for a variety of interesting studies. These include a comparison of the genetic consequences of changes in breeding system, evolution of edaphic endemism, and the biology of colonizing species. Crawford et al. (unpubl. data) recently addressed some of these issues in a survey of allozyme variation within and between these two taxa. Unfortunately, a prac- tical deterrent to such studies is the isolation and inaccessibility of the scattered coastal rocks and islands that harbor the great majority of sites occupied by Lasthenia maritima. Taxonomic Treatment Lasthenia maritima (A. Gray) M. Vasey comb, now.— Burrielia ma- ritima A. Gray, Proc. Amer. Acad. Arts 7:358. 1868.— Type: USA, California, San Francisco Co., "On the Farallones, rocky 140 MADRONO [Vol. 32 islets off San Francisco." F. Gruber s.n. 1868. (Holotype: GH!; isotype: US!) Decumbent or prostrate annual, diffusely branching from a short, thick taproot. Leaves narrow to broadly ligulate, succulent with thick, blunt, occasionally compound lobes and prominent veins on adaxial surface. Variably wooly pubescent at nodes and on pedun- cles. Phyllaries lance-ovate, 3.8-7.2 mm long, ciliate along margins and mid-rib. Ray flowers 7-12, ligules usually 1-3 mm long and less than 0.7 times as long as phyllaries. Anther tips variable, occasion- ally oblong and obtuse, about 0.2 mm long; stigmas often lacking apical and subapical hairs. Achenes generally 2.6-3.2 mm long (oc- casionally shorter), sparsely to densely pubescent with retrorse tri- chomes. Pappus dimorphic, rarely absent, consisting of multiple, unequal awns (ca. 4-12) and fewer narrow scales (rarely absent) ca. 0.25-0.33 as long as awns, each scale with a deeply laciniate superior margin. Distribution. Offshore rocks and islands, rarely headland margins, from Lion and Pup Rocks, San Luis Obispo Co., California north- ward along the California, Oregon, Washington, and western Van- couver Island coasts to Triangle Island near the northwestern tip of Vancouver Island, British Columbia. Representative specimens. USA: CALIF.: San Luis Obispo Co., Pup Rock, Vasey 8119; San Mateo Co., San Pedro Rock, Vasey 8345; San Francisco Co., Southeast Farallon Is., Hovanitz s.n., CAS; Contra Costa Co., West Brother Is., Vasey 809; Marin Co., Bird Rock, Vasey 8111; Mendocino Co., Anchor Rock, Vasey 8079; Humboldt Co., Sugarloaf Rock, Osborne 26469, HSC; Del Norte Co., Castle Is., Vasey 8082; OREGON: Curry Co., Hunter's Is., Boekelhyde and Vasey 8217; Coos Co., Table Rock, Vasey 8086; Lincoln Co., Yaquina Head, Kaplan s.n.; Tillamook Co., Pyramid Rock, Vasey 8089; WASH.: Clallam Co., Carroll Is., Vasey 8090; Seal Rock, Vasey 8093; BR. COL.: Vancouver Is., Baeria Rocks, Vasey 8096; Triangle Is., Foster and Vasey 8095. Key to Lasthenia minor and Lasthenia maritima Awns usually 2 or 3, scales broad and to V2 as long as awns, scale margins shallowly fimbriate; ligules of ray florets about 4-6 mm long, greater than 0.7 times as long as phyllaries Lasthenia minor Awns usually 4-12, scales narrow and ^4 to V3 as long as awns, scale margins deeply laciniate; ligules of ray florets about 1.3-3.0 mm long, less than 0.7 times as long as phyllaries Lasthenia maritima 1985] VASEY: LASTHENIA MARITIMA 141 Acknowledgments I especially thank Dr. R. Omduff for his seminal ideas and support during this project. Drs. R. Patterson, R. Omduff, and J. Strother were particularly helpful in the preparation of this manuscript and Drs. V. T. Parker, G. L. Stebbins, S. Jain and J. R. Sweeney provided much useful guidance. G. S. Lester provided invaluable field collections and considerable field support. Additional collections and locality data were provided by M. Sheehan, R. Boekelhyde, B. Foster, B. Henneman, and J. Siddall. Personal collection efforts would not have been possible except for the assistance of M. Nakagawa, M. Harms, J. L. Groulx, L. V. Painter, K. Culligan, L. C. Bemer, and E. Ryde. I also thank Dr. T. Duncan for his assistance with the U.C. Berkeley computer and Dr. G. Hunt for cooperation in the California Channel Island survey. Thanks are also due to the following agencies and organizations for their support during this project: U.S. Fish and Wildlife, California Dept. of Fish and Game, and Pt. Reyes Bird Observatory. I also particularly thank California Dept. of Fish and Game for a grant supporting the distribution work and the California Native Plant Society for a mini-grant. Finally, I am thankful to the following herbaria for their generous loans of specimens: UC, JEPS, CAS, WTU, ORE, DSC, MO, ND, NY, OBI, PH, RSA, UCSB, SD, US, V, DAVH, HSC, GH, F, UBC, ND. Literature Cited Altosaar, I., B. A. BoHM, and R. Ornduff. 1 974. Disc-electrophoresis of albumin and globulin fractions from dormant achenes of Lasthenia. Biochem. Syst. Ecol. 2:67-72. Blankenship, J. W. and C. A. Keeler. 1 882. On the natural history of the Farallon Islands. Zoe 3:144-165. BoHM, B. A., N. A. M. Saleh, and R. Ornduff. 1974. The flavonoids of Lasthenia (Compositae). Amer. J. Bot. 61:551-561. Crawford, D. J., R. Ornduff, and M. C. Vasey. (In Press). Allozyme variation within and between Lasthenia minor and its derivative species L. maritima (Asteraceae). Amer. J. Bot. Dainty, J. 1 979. The ionic and water relations of plants which adjust to a fluctuating saline environment. In R. L. Jeffries and A. J. Davy, eds.. Ecological processes in coastal environments. Blackwell Sci. Publ., UK. Ferris, R. S. 1955. The identity of Monolopia minor DC. Contr. Dudley Herb. 4:331-334. GiLHAM, M. E. 1956. Ecology of the Pembrokeshire Islands. V. Manuring by the colonial seabirds and mammals with a note on seed distribution by gulls. J. Ecol. 44:429-454. Goldsmith, F. B. 1973. The vegetation of exposed sea cliffs at South Stack, An- glesey. I. The multivariate approach. J. Ecol. 61:787-818. Gray, A. 1868. Characters of new plants of California and elsewhere, principally of those collected by H. N. Bolander in the State Geological Survey. Proc. Amer. Acad. Arts 7:327-401. Greene, E. L. 1894. Manual of the botany of the region of San Francisco Bay. Cubery and Co., San Francisco, CA. Jepson, W. L. 1925. A manual of the flowering plants of California. University of California, Berkeley. MoLDENKE, A. R. 1971. Studies on the species diversity of California plant com- munities. Ph.D. dissertation, Stanford Univ., Stanford, CA. Ornduff, R. 1965. Omithocoprophilous endemism in Pacific Basin angiosperms. Ecology 46:864-867. . 1 966. A biosystematic survey of the goldfield genus Lasthenia (Compositae: Helenieae). Univ. Calif. Publ. Bot. 40:1-92. 142 MADRONO [Vol. 32 . 1976. Speciation and oligogenic differentiation in Lasthenia (Compositae). Syst. Bot. 1:91-96. , B. A. BoHM, and N. A. M. Saleh. 1974. Flavonoid races in Lasthenia (Compositae). Brittonia 26:411-420. ScHEFLER, W. C. 1980. Statistics for the biological sciences. Addison-Wesley, MA. Smith, J. P., R. J. Cole, J. O. Sawyer, and W. R. Powell. 1980. Inventory of rare and endangered vascular plants of California. Spec. Publ. No. 1 (2nd ed.). Calif Native PI. Soc. Spear, L. 1982. Dispersal patterns of Farallon Western Gulls. Point Reyes Bird Observatory Newsletter 60:1-1 1. Vasey, M. C. 1981. Status report on bird rock goldfields {Lasthenia minor ssp. maritima). Rare Plants Division, California Dept. of Fish and Game, Sacra- mento, California. . 1984. Vernal pools, seabird rocks, and a remarkable species of Lasthenia. In S. Jain and P. Moyle, eds.. Vernal pools and intermittent streams: a symposium sponsored by the Institute of Ecology, Univ. California, Davis, 9 and 10 May 1981. Institute of Ecology Publ. No. 28. (Received 17 Jul 1984; accepted 18 Jan 1985.) ANNOUNCEMENT Southern California Botanists was established in 1927 and is a non-profit associ- ation of professional botanists and interested lay persons. Its past presidents include Mildred Mathias, Peter Raven, and Robert Thome. It is an active organization of about 500 members. It publishes "Crossosoma," a bi-monthly journal. It conducts 10-15 field trips per year, including at least one trip in Baja California. It presents or co-sponsors one all-day Symposium each year on a subject of botanical interest. Southern California Botanists operates an active book sales activity for its members, providing the opportunity to purchase at a discount botanical books often difficult to locate. Inquiries may be sent to Southern California Botanists, % Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, California 91711. ANNOUNCEMENT The International Organization of Plant Biosystematists encourages members and other botanists to send information to the lOPB Newsletter. Send changes of address, personal news, promotions, society appointments, publications, projects started, re- quests for research materials, short articles, controversy, reports on research confer- ences, etc. Send information to Dr. K. Urbanska, Editor, lOPB Newsletter, Geobo- tanishes Institut, Stiftung Rubel, Zurichbergstrasse 38, CH-8044 Zurich, Switzerland. OBSERVATIONS ON CHAMAESYCE (EUPHORBIACEAE) IN THE GALAPAGOS ISLANDS Michael J. Huft and Henk van der Werff Missouri Botanical Garden, P.O. Box 299, St. Louis 63166 Abstract All collections of the endemic species of Chamaesyce in the Galapagos Islands were reexamined. Results indicate that hybridization has taken place in several localities, notably on Isla Bartolome. Plants previously referred to C. nummularia and C. bindloensis from Bartolome are considered to be of hybrid origin, and C. bindloensis is placed in synonymy under C punctulata, together with two other names that had been overlooked in the treatment of Chamaesyce in the Flora of the Galapagos Islands. The genus Chamaesyce has recently been revised for the Gala- pagos Islands (Burch 1969, 1971). While carrying out ecological studies on the islands, the junior author noticed several aberrant populations that he was unable to identify. We therefore decided to study all available specimens of the native, endemic species in an attempt to place these populations and to clear up some other prob- lems that had become apparent. Isla Bartolome is a tiny islet (1.24 km^) that is situated about 200 m from the east coast of San Salvador (572 km^). The island con- sists of lava flows and a few cinder and tuff cones and supports a sparse vegetation of low shrubs dominated by several species of Chamaesyce dind Tiquilia {=Coldenia, Boraginaceae, see Richardson 1976). The island seems unusually rich in species of Chamaesyce for its size, and Burch (1971) reports five species from here: C amplexicaulis, C punctulata, C viminea, C. bindloensis, and C nummularia var. nummularia. Of these, the first three also occur on San Salvador, as does C recurva, which has not yet been reported from Bartolome. John Thomas Howell collected numerous speci- mens of Chamaesyce on Bartolome in 1932 (CAS, many duplicates in MO) that had not yet been incorporated into the herbarium at CAS at the time the "Flora of the Galapagos Islands" (Wiggins and Porter 1971) was in preparation and consequently were not seen by Burch. Most of these specimens do not clearly fit into one or the other of the species, but exhibit a full range of intermediate char- acteristics between the species C. amplexicaulis, C punctulata, and C. recurva in leaf shape, pubescence, cyathial appendages, and seed shape and surface. The following note by Howell is attached to each of these sheets: "Part of the hybrid complex involving EE. amplex- Madrono, Vol. 32, No. 3, pp. 143-147, 19 August 1985 144 MADRONO [Vol. 32 icaulis, nummularia, and punctulata that grew at the northwest point of Bartholomew Island. The plants grew in volcanic ash among lava rocks in a narrow belt less than 1 00 m long. The individuals varied in habit, vestiture, gland-appendages, and seeds, while the leaves varied in size, shape, margin, and color. Chamaesyce bindloensis may have been one of the elements in the complex." We agree with Howell's note that extensive hybridization occurs on Bartolome, as is clear from his own collections as well as several others made by Wiggins & Porter and Fagerlind & Wibom. It seems much more likely to us, however, that C. nummularia is not in- volved in the hybrid complex, but that C. amplexicaulis is. The collection from Bartolome that has been identified as C nummularia {Wiggins and Porter 307, DS) differs from that species by the sessile leaves and more angular and slender seeds. It is better interpreted as part of the hybrid complex. True C. nummularia appears to be confined to the southern islands of Santa Fe, Santa Maria, San Cris- tobal, and Espanola. Howell evidently considered C nummularia to be a contributor to the hybrid complex because of the presence of hybrid individuals with hirsutulous indument similar to that of C. nummularia. How- ever, although C amplexicaulis is usually completely glabrous, oc- casional plants from such islands as Marchena {v.d. Werff 2132, CAS), Pinta {Stewart 1847, CAS), San Salvador {Stewart 1853, CAS; Howell 10013, CAS, MO), and Bartolome {Howell 10063, CAS; Wiggins and Porter 308, CAS) are hirsutulous. Thus, C amplexi- caulis occurs near the hybrid zone, as C. nummularia does not, and all of the variation in the complex can be explained without in- volving C nummularia. Furthermore, C. amplexicaulis is the only species that can provide the cordate-clasping leaf-bases that are ob- served in several of the hybrid individuals. Chamaesyce punctulata is also represented on San Salvador, just across from Bartolome, by several hirsutulous collections {Howell 10054, 10055, 10056, CAS), although this species is normally glabrous. Table 1 gives the relevant characters of the three species contrib- uting to the hybrid complex, as well as of C. nummularia; the same characters for eight hybrid individuals chosen to illustrate the full range of variability are also given. The other species that Burch reported from Bartolome, Chamae- syce bindloensis, is problematical. We have examined the type {Stew- art 1868, GH, lectotype, chosen by Burch 1969; CAS, isolectotype), and it is our opinion that this species, originally described as a wide- leafed variety of what is now called C. punctulata, falls within the range of variation exhibited by C. punctulata. The prominent gland appendages that are said to characterize C. bindloensis are not found on all cyathia. Furthermore, specimens that have the characteristic 1985] HUFT AND VAN der WERFF: CHAMAESYCE 145 short, broad leaves and short stipules of C. bindloensis but that have the ridged seeds of C punctulata are known from other locations (e.g., east coast of Santa Cruz, v.d. Werff 2094, CAS, U; Plaza, Fagerlind and Wibom 3361, 3386, S). Because there is no clear-cut distinction between the two concepts, we do not hesitate to place C bindloensis in synonymy under C punctulata. Other specimens that have been identified as C bindloensis are C punctulata (e.g., A. and H. Adsersen 1144, C), C. recurva (e.g., M. and O. Hamann 981, C) and part of the hybrid complex on Bartolome (e.g., Fagerlind and Wibom 3482, 3505, S; Wiggins and Porter 300, DS; 304, CAS, GH and 311, MO). Thus in addition to the hybrid populations, Barto- lome supports three species of Chamaesyce, C. amplexicaulis (e.g., Howell 10065, CAS, MO; Marling 5369, S), C punctulata (e.g., Wiggins and Porter 294, CAS, MO; Fagerlind and Wibom 3483, S), and C viminea (e.g., A. and H. Adsersen 1887, C). Reports of Chamaesyce nummularia from Isla Wolf in the far northwestern part of the archipelago (Snodgrass and Heller 11, DS; Dawson s.n., DS; Stewart 1855, CAS; Fosberg 44967, MO) are based on specimens that appear morphologically intermediate between C amplexicaulis and C nummularia. These specimens are probably of hybrid origin, but it is doubtful that C nummularia is involved, since that species is known with certainty only from the four south- easternmost islands. Further collections as well as more detailed field observations will be needed in order to elucidate the nature of these puzzling plants. Another probable hybrid population occurs on Isla Pinzon, rep- resented in herbaria by Howell 9845 (CAS— 4 sheets, MO) and 9846 (CAS), and noted by the collector to be variable in the field. These specimens are similar to C punctulata, but have conspicuous cy- athial appendages, and the seeds are, for the most part, smooth, unlike the ridged seeds of typical C. punctulata. The leaves exhibit a prominent dimorphy (large on main stems, small on laterals) that is suggestive of C. recurva, as are the large stipules. Again, further collections are needed to make a definitive interpretation. Chamaesyce abdita was known to Burch only from the type col- lection on Santa Fe. It is now represented by several collections from Santa Cruz (A. and H Adsersen 248, C; Howell 9128, 9129, CAS, MO; v.d Werff 1121, CAS), Baltra (Howell 9955, CAS, MO), Es- pafiola {A. and H. Adsersen 666, C), and Champion near Santa Maria {A. and H. Adsersen 1459, C; v.d. Werff 2059, U). Some of these collections, especially those from Champion, have glabrous capsules and herbage and thus would not key out properly in the Flora. Howell 9130 (CAS, MO) from Santa Cruz has the pubescence, leaf shape, and leaf size of C abdita, but is perennial. Some of the seeds are typical of C. abdita, but a few are ridged like those of C recurva. 146 MADRONO [Vol. 32 P3 u > < X h < X H bs o 6 h O CQ 2 9 Z oa O >- UK O Q y < ;;5 I ^5 <: 3 2 g CQ < < PQ ^ z o o < So g .Spo .Sptv .^0^ 0^ 1985] HUFT AND VAN der WERFF: CHAMAESYCE 147 and therefore our identification of this specimen as C abdita is with some hesitation. Following is a list of native species in the Galapagos that we accept, along with synonymy that differs from that given by Burch (1969, 1971), including two published names that were not accounted for by Burch. These are Euphorbia bisulcata Howell, Proc. Calif Acad. Sci., 4th sen, 21:330, 1935; type: Howell 9880 (CAS, holotype; MO, isotype), which differs from typical C. punctulata only in that the back of each carpel on the mature capsule is broadly bisulcate; and Euphorbia howellii Wheeler, Contr. Gray Herb. 124:42, 1939, no- men novum for E. diffusa Hook, f , non Jacq., a synonym of E. punctulata. Chamaesyce abdita Burch C. AMPLEXiCAULis (Hook. f ) Burch C. GALAPAGEiA (Robins. & Greenm.) Burch C. NUMMULARiA (Hook. f ) Burch var. nummularia C. NUMMULARIA var. GLABRA (Robins. & Greenm.) Burch C. PUNCTULATA (Andcrss.) Burch C bindloensis (Stewart) Burch Euphorbia articulata Anderss. var. bindloensis Stewart E. bisulcata Howell E. diffusa Hook, f , non Jacq. E. howellii Wheeler C. RECURVA (Hook f ) Burch C. viMiNEA (Hook, f ) Burch One additional adventive species, Chamaesyce lasiocarpa (Kl.) Arthur, has been found on the islands since the Flora was published (San Cristobal, v.d. Werff2171, U, van der Werff*, 1977). Acknowledgments We thank the curators of C, CAS, DS, F, GH, K, MO, S, and U for making their collections available to us. This is contribution no. 373 from the Charles Darwin Foundation. Literature Cited Burch, D. 1969. Notes on the Galapagos Euphorbieae (Euphorbiaceae). Ann. Mis- souri Bot. Gard. 56:173-178. . 1971. Chamaesyce. In I. L. Wiggins and D. M. Porter, Flora of the Galapagos Islands, pp. 575-582. Stanford University Press, Stanford, CA. Richardson, A. 1976. Reinstatement of the genus Tiquilia (Boraginaceae: Ehre- tioideae) and descriptions of four new species. Sida 6:235-240. VAN DER Werff, H. 1977. Vascular plants from the Galapagos Islands: new records and taxonomic notes. Bot. Not. 130:89-100. Wiggins, I. L. and D. M. Porter. 1971. Flora of the Galapagos Islands. Stanford University Press, Stanford, CA. (Received 17 Feb. 1984; 30 Oct. 1984.) POST-FIRE SEEDLING ESTABLISHMENT OF ADENOSTOMA FASCICULATUM ANU CEANOTHUS GREGGII IN SOUTHERN CALIFORNIA CHAPARRAL JocHEN KuMMEROw, BARBARA A. Ellis, and James N. Mills Department of Biology, San Diego State University, San Diego, CA 92182 Abstract Mortality of Ceanothus greggii and Adenostoma fasciculatum seedlings during the first growing season after a bum in the southern California chaparral was 92 and 90%, respectively. Survival of these shrub seedlings was not affected by the presence of stump sprouts, which grew at a rate of 2 cm in height per week, or by herbaceous plants (65% cover). However, the presence of annuals reduced the growth of the shrub seedlings significantly. Watering increased seedling growth, but additional fertilizer had no significant effect. Stump sprouting and abundant seedling establishment after fire appear to be main factors that insure maintenance of the shrub species compo- sition of the frequently fire-disturbed chaparral vegetation. Chaparral in southern California bums at 25- to 40-year intervals, although lower or higher fire frequencies occur (Philpot 1977, Keeley 1982). In spite of these repeated perturbations, the vegetation re- generates rapidly. This phenomenon of fast chaparral re-establish- ment after fire has been described in detail for a number of different California localities (e.g., Sampson 1944, Horton and Kraebel 1955, Hanes 1971, Keeley and Keeley 1982). From these observations, it can be concluded that the shrub species establishing during the first growing season after a fire are the same ones that composed the pre- fire chaparral community (Hanes 1971). However, the quantitative composition of the shrub cover may change. Assessment of the cover values for Ceanothus crassifolius and Adenostoma fasciculatum in a 15 -year-old stand in the San Gabriel Mountains in 1934 and the re-assessment in 1982, 22 years after a bum, showed that the cover value for the former species had increased by the factor of two while the latter species had declined correspondingly (Jacks 1984). Although these observations indicate the importance of stump sprout and seedling densities after fire for mature stand composition, little is known about the factors that influence shrub seedling estab- lishment. The importance of the post-fire physical environment has been emphasized (Sauer 1977). Competing seedlings in the early establishment phase can modify the environment, for example through rapid water and nutrient uptake. It is unclear to what degree Madrono, Vol. 32, No. 3, pp. 148-157, 19 August 1985 1985] KUMMEROW ET AL.: CHAPARRAL SEEDLINGS 149 shrub seedling establishment is influenced by spatially variable fac- tors such as the presence or absence of nearby stump sprouts. The competitive action of the sometimes dense but short-lived herba- ceous vegetation on shrub seedling establishment is likewise unclear. Selective herbivory by rabbits on shrub seedlings has recently been documented (Mills 1983). The objective of the present study was to obtain quantitative information on shrub seedling mortality and establishment. The data may improve our predictive capacity of mature chaparral stand composition. Materials and Methods Research site. The study site was the Sky Oaks Biological Field Station at 1500 m elevation about 15 km north of Warner Springs, San Diego County, California. The climate of the area is Mediter- ranean; and the site receives annually about 550 mm of precipitation between November and May. Some thundershowers interrupt the long summer drought. In most winters some snow remains on the ground for a few days. Minimum temperatures of — 8°C and summer maxima of 38°C have been observed (Bowman 1984). The research site was covered by a dense, 54-year-old shrub vege- tation (aged by year ring counts, P. Zedler, pers. comm.). Pre-bum analysis showed a density of 1.01 shrubs per m^; Adenostoma fas- ciculatum and Ceanothus greggii were the most important species, with 0.42 and 0.40 individuals per m^, respectively. We observed the spotty occurrence of Cercocarpus betuloides Nutt., and Quercus dumosa Nutt. mostly on north-facing slopes and the tree Quercus agrifolia Nee. in ravines and washes. The soil is a loamy sand with abundant stones and pebbles. The soil layer, varying between 30 and 50 cm depth, rests on bedrock of a micaceous schist (USDA 1973). In December 1981, a 2-ha site in this area was burned with the help of the California Department of Forestry and the U.S. Forest Service. The bum was complete and temperatures at the soil surface reached about 350°. Shrub seedling establishment. Seedling establishment of A. fascic- ulatum and C. greggii was assessed post-fire on eight 8 x 2-m plots randomly distributed over the area. In each plot the shrub seedlings of ten 0.25-m^ subquadrats were counted on 8 May and again on 11 December (i.e., 6 and 13 mo post-fire). Repeated counts from mid- April to the beginning of May indicated no further increases in the number of shrub seedlings; by mid-December the initiation of sustained rainfall made further drought-induced seedling mortality for the current year improbable. 150 MADRONO [Vol. 32 Effect of stump sprouts and annuals on shrub seedling growth. To assess the effect of stump sprouts and annuals on growth of shrub seedUngs, twelve 1-m^ plots, each with a stump- sprouting ^. yi25dc- ulatum in the center were distributed randomly over the 6 mo-old bum area in May 1982. Four replicate plots were established for each of the following treatments: (A) all stump sprouts removed, (B) all stump sprouts and annual herbs removed, and (C) control. Shrub seedlings were counted and their heights recorded biweekly. Stump sprouts and annuals in treatments (A) and (B) were removed with each observation date. The stump sprouts were carefully broken off by hand in order to avoid major wounding of the burls. New sprouts appeared during the entire year after the bum. Irrigation and fertilizer application. On the bum site, twelve 1 -m^ plots without stump- sprouting shmbs and a minimum of five sec- ond-year C. greggii and A. fasciculatum seedlings each were fenced in the spring of 1983 with 0.5-m tall chicken wire to prevent small- mammal herbivory. Shoots that grew from adjacent stump sprouts into the plots were removed. Three treatments, consisting of four replicates each, were established. (1) Fertilized and watered: 10 liters of full-strength Hoagland's solution, containing 2 g N, 0.3 g P, and the other nutrients in corresponding amounts, were applied to the plots with sprinkler cans. The fertilizer application was repeated four times in two-week intervals between the beginning of June and the end of July. Water and fertilizer application was initiated in early June when soil moisture in the 10-20-cm depth layer had declined to about 3% (g H2O g~^ dry weight; Fig. 1) and herbaceous plants showed symptoms of water stress. From January to June 1983, a total of 650 mm of precipitation fell. Thus, soil moisture remained relatively high until the end of May. Each fertilizer plot received 8 g N and 1.2 g P. (2) Watered only: Irrigation consisted of 10 liters m"^ of de-ionized water applied at the same times as the nutrient solution. In addition, the "fertilized and watered" and the "watered only" plots received a total of 80 liters of de-ionized water (20 liters per application) between the end of June and the beginning of Sep- tember. (3) Unwatered plots. In September and November 1983 the heights and crown diam- eters (maximum and minimum lengths between distal branch tips) of five representative seedlings per plot of A. fasciculatum and C. greggii were recorded. The crown diameter values were more mean- ingful than height values because branches elongated relatively more than the main leader shoots. Biweekly gravimetric soil measurements in depth layers of 0-10, 10-20, and 20-30 cm, about one meter distant from the control plots of this experiment, provided basic information on the seasonal fluctuation of soil moisture in the rooting zone of the shmb seedlings. 1 985] KUMMEROW ET AL.: CHAPARRAL SEEDLINGS 1 5 1 18r Mar. Apr. May June July Aug Sept Oct Nov. Dec Jan. Feb. m 1 983 ► 1 984— ► Fig. 1 . Soil moisture one meter from the control plots of the irrigation and fer- tilizer experiment. Each data point is the mean of four samples. Vertical bars indicate standard errors. Five representative seedlings each of both the species were ex- cavated in bimonthly intervals. Although some root breakage ap- peared to be unavoidable especially in late summer, we think that the main roots were mostly recovered to their full extension. Results Shrub seedling establishment. Abundant rainfall occurred during the first post-fire period (566 mm fell from July 1981-June 1982, Ellis et al. 1983). Thus the conditions for seed germination were favorable. In May 1982, an average of 78 shrub seedlings (34 A. fasciculatum and 44 C. greggii) per m^ were counted (Table 1). In December 1982, 9.8% of the A. fasciculatum and 7.8% of the C greggii seedlings were still alive. Periodic counts showed that most of the seedlings had died in May and June (Fig. 2). Effect of stump sprouts and herbaceous plants on shrub seedling growth. Survival of C. greggii and A. fasciculatum seedlings during the first post-fire growing season was not affected by the presence of stump sprouts or herbaceous plants (the latter with ca. 65% cover). However, seedling growth was reduced by about ^3 under the influ- 152 MADRONO [Vol. 32 Table 1. Mean Number of A. fasciculatum and C. greggii Seedlings at Be- ginning AND End of the First Post-fire Growing Season. Values are means ± standard errors of eighty 0.25-m^ plots randomly staked on eight 2 x 8-m observation areas. A. fasciculatum C. greggii % % Mean no. of seedlings m"^ surv. Mean no. of seedlings m^^ surv. May 8 Dec. 1 1 May 8 Dec. 1 1 33.6 ± 12.96 3.3 ± 1.34 9.8 43.8 ± 7.84 3.4 ± 0.97 7.8 ence of the herbaceous vegetation (Table 2). Removal of stump sprouts did not enhance seedling growth, although these stump sprouts grew vigorously at an elongation rate of about 2 cm per week from May to October (Fig. 3). This indicated an adequate water supply for stump-sprouting A. fasciculatum shrubs during the summer of 1982. Irrigation and fertilizer application. Biweekly gravimetric soil moisture measurements in the vicinity of the experimental plots demonstrated that by mid- June 1983, the soil in the upper layer had already dried out and only the 20- and 30-cm layer was still holding a small moisture reserve (Fig. 1). Our periodic seedling excavations Fig. 2. Survival of Ceanothus greggii and Adenostoma fasciculatum seedlings which had germinated between February and March 1982, following a bum in De- cember 1981. The highest observed seedling density was considered 1 00%. Each point is the mean number of seedlings in eight 1 -m^ plots. There was no significant difference in seedling survival between species. The hatched bars indicated rainfall events in mm. 1985] KUMMEROW ET AL.: CHAPARRAL SEEDLINGS 153 Table 2. Height (mm ± S.E.) of C. greggii and A. fasciculatum Seedlings at THE Initiation of the Observations (4/28/82) and after 10 Weeks under Treat- ments A (Stump Sprouts Removed), B (Stump Sprouts and Herbaceous Plants Removed), and C (Unchanged Control). Significance of differences among treat- ments indicated with different letters. Number of observed seedlings in parentheses. Statistical treatment: Basic statistics, ANOVA on log-transformed data, Newman- Keuls multiple comparison tests, 'p < 0.01. ^p < 0.05. C. greggii A. fasciculatum Initial height Final height Initial height Final height 4/28 7/9 4/28 7/9 A. Stump spr. 10.5 ± 0.6a 21.0 ± 1.3a 9.8 ± 0.5a 27.8 ± 2.0a removed (32) (26) (31) (21) B. Stump spr. 10.1 ± 0.6a 35.5 ± 5.3b' 10.4 ± 0.6a 33.1 ± 3.7b2 + herbs (30) (11) (40) (24) removed C. Control 10.4 ± 0.6a 25.7 ± 3.7a 7.6 ± 0.4b2 21.9 ± 2.3a (30) (11) (40) (14) showed that at this time the main roots of these 2nd-year seedhngs had reached a length of 20-30 cm. By the end of August and again at the beginning of October 1983, heavy thunderstorms produced enough rain to remoisten the soil profile down to 30 cm (Fig. 1). On 9 September, the seedlings of both shrub species were larger in both the watered and the watered and fertilized plots. Six weeks later (2 1 October) this difference was still visible, although not sta- tistically significant in the case of A. fasciculatum (Table 3). Discussion More than 90% of the seedlings that germinated in March and April 1982 after a fire in December 1981 died during the first growing season. Our biweekly observations had shown that May and June were the months with the highest mortality. These values seem high when compared with others (Musick 1972, Keeley and Zedler 1978, Horton and Kraebel 1955), although exact shrub seedling mortality rates in the spring following a fall fire are poorly quantified (Schles- inger et al. 1982). We can deduce from precipitation data that drought conditions became severe in May and more so in June (Fig. 1). Excavation of five seedlings each of A. fasciculatum and C greggii in July 1982 showed mean tap root lengths of 5 cm and 8 cm, respectively (Ellis 1983). Thus, lack of soil moisture is probably one of the factors causing the high seedling mortality. A second cause for the death of many seedlings was herbivory. By November 1982, 25% of the A. fasciculatum and 43% of the C. greggii seedlings on this specific research site had been killed by rabbits (Mills 1983). 154 MADRONO [Vol. 32 May June July Aug. Sept. Oct Nov. 1 982 Fig. 3. Growth in height of Adenostoma fasciculatum stump sprouts during the 1982 growing season after a bum in December 1981. Ten average-sized shoots of 4 randomly chosen A. fasciculatum stumps were measured. A typical standard error is indicated for the first (May) value. Note the linear height increase during the dry summer months. Very little is known about the effects of the herbaceous post-fire vegetation on shrub seedling establishment. We could not find evi- dence for increased survival of shrub seedlings when competing stump sprouts and herbaceous plants were eliminated, although roots of the stumps must have remained active. This result contrasts sharply with data shown by Schultz et al. (1955). A grass density of 85% eliminated all shrub seedlings, and a grass density of 1 8% reduced the number of surviving shrub seedlings by 50%. An explanation for these conflicting results may be that grasses were virtually absent from our research site, whereas the above-mentioned observations were made on a burned and ryegrass re-seeded area. Ryegrass plant- ings have been shown to inhibit herb establishment and chaparral seedling growth (Corbett and Green 1965). In our site the estimated herbaceous cover was about 65% and consisted mostly of the native fire followers, Phacelia brachyloba, Streptanthus heterophyllus, and Gilia caruifolia, all with much less aggressive tap root systenis than the fibrous root system of ryegrass (Rice and Green 1964). Although seedling survival was not affected by stump sprouts of A. fasciculatum and the herbaceous vegetation, we found that herbs caused a significant growth reduction of the shrub seedlings. The vigorously growing A. fasciculatum stump sprouts did not affect shrub seedling growth. These stump sprouts, consisting of clusters with about 150 shoots per burl, grew at a rate of about 2 cm per month from April to October, when they reached a final mean height 1985] KUMMEROW ET AL.: CHAPARRAL SEEDLINGS 155 Table 3. Relative Size (Height of the Main Shoot x Average Diameter) of C. greggii and A. fasciculatum Second-year Seedlings Which Were not Watered, Watered, and Watered + Fertilized, Three and Four Months after Treatment Initiation. For each treatment n = 19. Statistical treatment: One-way ANOVA on log-transformed data followed by Newman-Keuls multiple range test. Significant differences among treatments for each observation date indicated with different letters (p < 0.05). Standard errors given in parentheses. Observation date C. greggii A. fasciculatum 9/9/83 10/21/83 9/9/83 10/21/83 Unwatered 24.3 (3.7)a 30.0 (4.7)a 72.1 (22.0)a 91.1 (12.7)a Watered 62.1 (11.7)b 71.4(13.5)b 100.4 (18.8)ab 127.9 (26.1)a Fertilized + watered 53.0 (9.2)b 61.6(10.6)b 117.3(9.6)b 133.4 (13.9)a of ca. 50 cm (Fig. 3). The fine root density around the burls was several times higher than that around the stems of unbumed control shrubs (Kummerow and Lantz 1983). The fine roots of resprouting burls were predominantly located at about 20-30 cm depth, and thus were deeper than the roots of the V2-yr shrub seedlings, which had hardly reached a depth of 10 cm (Ellis 1983). However, roots of annuals and shrub seedlings occupied the same depth zone of the soil and might have competed for the same resources. Continued observations are needed to establish if 2nd- and 3rd-year seedling mortality is higher among seedlings whose growth was retarded by competing annuals and stump sprouts. The fertilizer and irrigation experiment tested if resources might limit shrub seedling growth in the second post-fire year. The shrub seedlings reacted with a significant growth increase to fertilizer and water addition. The differences between the irrigation only and ir- rigation + fertilizer treatment in A. fasciculatum, although sugges- tive, were not statistically significant (Table 3). Thus, no clear in- dication for nutrient deficiency was found. Overall, the results of this study show that drought is a main factor for shrub seedling mortality in the first year after a bum. During the second year, the seedlings responded with enhanced growth to fer- tilizer and water addition. The results suggest that the sprouting burls of A. fasciculatum plus the abundant seedlings of this species and of C. greggii provide more than enough plants to insure per- petuation of these shrubs even after an initial mortality of more than 90% during the first post-fire growing season. Mortality during the second year was insignificant. Acknowledgments This study was supported by NSF-Grant No. DEB-8025977-01. Dr. David Rayle helped in revising a first draft; assistance from Ms. Mildred Johnson and Ms. Robin McQuitty with the manuscript preparation is gratefully acknowledged. 156 MADRONO [Vol. 32 Literature Cited Bowman, W. D. 1984. Seasonal and diurnal changes in water relations parameters in evergreen chaparral shrubs. M.S. thesis, San Diego St. Univ., San Diego, CA. CoRBETT, E. S. and L. R. Green. 1965. Emergency revegetation to rehabilitate burned watersheds in southern California. USDA Forest Service, Pac. Southw. For. Range Expt. Sta., Res. Paper PSW-22. Ellis, B. A. 1983. Seedling mortality and reestablishment in an early post-fire chaparral community. (Abstract.) Bull. Ecol. Soc. Amer. 64:147. , J. R. Verfaillie, and J. Kummerow. 1983. Nutrient gain from wet and dry atmospheric deposition and rainfall acidity in southern California chaparral. Oecologia 60:118-121. Hanes, T. L. 1971. Succession after fire in the chaparral of southern California. Ecol. Monogr. 41:27-52. HoRTON, J. S. and C. J. Kraebel. 1955. Development of vegetation after fire in the chamise chaparral of southern California. Ecology 36:244-262. Jacks, P. M. 1984. The drought tolerance of Adenostoma fasciculatum and Cea- nothus crassifolius seedlings and vegetation change in the San Gabriel chaparral. M.S. thesis, San Diego St. Univ., San Diego, CA. Keeley, J. E. 1982. Distribution of lightning- and man-caused wildfires in Cali- fornia. In C. E. Conrad and W. C. Oechel, techn. coord.. Dynamics and man- agement of mediterranean-type ecosystems, pp. 431-437. USDA For. Serv., Techn. Rep. No. PSW-58. Pac. Southw. For. Range Expt. Sta., Berkeley, CA. and P. Zedler. 1978. Reproduction of chaparral shrubs after fire: a com- parison of sprouting and seedling strategies. Amer. Midi. Naturalist 99:142-161. Keeley, S. C. and J. E. Keeley. 1982. The role of allelopathy, heat, and charred wood on the germination of chaparral herbs. In C. E. Conrad and W. C. Oechel, techn. coord.. Dynamics and management of mediterranean-type ecosystems, pp. 128-134. USDA For. Serv., Techn. Rep. No. PSW-58. Pac. Southw. For. Range Expt. Sta., Berkeley, CA. Kummerow, J. and R. K. Lantz. 1 983. Effect of fire on fine root density in redshank {Adenostoma sparsifolium Torr.) chaparral. PI. Soil 70:347-352. Mills, J. N. 1983. Herbivory and seedling establishment in post-fire southern California chaparral. Oecologia 60:267-270. MusiCK, H. B. 1 972. Post-fire seedling ecology of two Ceanothus species in relation to slope exposure. M.A. thesis, Univ. California, Santa Barbara. Philpot, C. W. 1977. Vegetative features as determinants of fire frequency and intensity. In H. A. Mooney and C. E. Conrad, techn. coord., Proceedings of the symposium on the environmental consequences of fire and fuel management in mediterranean ecosystems, pp. 12-16. USDA For. Serv., Gen. Techn. Rep. WO-3, Washington, D.C. Rice, R. M. and L. R. Green. 1 964. The effect of former plant cover on herbaceous vegetation after fire. J. Forest. (Washington) 62:820-821. Sampson, A. W. 1944. Plant succession of burned chaparral lands in northern California. Calif Agric. Exp. Sta. Bull. No. 635, 145 pp. Sauer, J. D. 1977. Fire history, environmental patterns, and species patterns in Santa Monica mountain chaparral. In H. A. Mooney and C. E. Conrad, techn. coord.. Proceedings of the symposium on the environmental consequences of fire and fuel management in mediterranean ecosystems, pp. 383-386. USDA For. Serv., Gen. Techn. Rep. WO-3, Washington, D.C. ScHLESiNGER, W. H., J. T. Gray, D. S. Gill, and B. E. Mahall. 1982. Ceanothus megacarpus chaparral: a synthesis of ecosystem processes during development and annual growth. Bot. Rev. 48:71-1 17. Schultz, a. M., J. L. Launchbaugh, and H. H. Biswell. 1955. Relationship between grass density and brush seedHng survival. Ecology 36:226-238. 1985] KUMMEROW ET AL.: CHAPARRAL SEEDLINGS 157 USDA. 1973. Soil survey, San Diego area, California. Part L USDA, Soil Conserv. Serv., San Diego Co. Planning Dept., San Diego, CA. (Received 26 Oct 1984; accepted 15 Feb 1985.) ANNOUNCEMENT The first segment of a flora for Butte County, California, on the Boraginaceae, initiates the series "Publications from the Herbarium, California State University, Chico." This 35-page pamphlet contains keys to the 3 1 taxa occurring in Butte County, brief descriptions of the plants, their habitats and (from the literature), their repro- ductive biology. Thirty detailed range maps are included. For a copy please send $2 to Rob Schlising, Department of Biological Sciences, California State University, Chico, CA 95929. ANNOUNCEMENT A regional reference herbarium has recently been established at Deep Springs Col- lege (located at the south end of the White Mts. in Inyo Co., CA). As curator in absentia and a recent alumnus, I am seeking contributions of duplicate specimens from the White Mts., Deep Springs Valley, and the adjacent basins and ranges. A few more exotic species would also be welcomed for teaching purposes. Very common species are already well represented and should be avoided. The collections can be made accessible to outside workers through prior arrangements with the college, and a standard herbarium designation will be obtained when sufficient size is reached. Inquiries or unmounted specimens may be sent to: Professor of Biology, Attn: HER- BARIUM, Deep Springs College, CA, via Dyer, NV 89010. Specimens received will be acknowledged as tax-deductible contributions; correspondence will, if necessary, be forwarded to me.— James D. Morefield, NAU Box 6201, Northern Arizona Univ., Flagstaff" 86011. SPARTINA (GRAMINEAE) IN NORTHERN CALIFORNIA: DISTRIBUTION AND TAXONOMIC NOTES Douglas Spicher and Michael Josselyn Paul F. Romberg Tiburon Center for Environmental Studies, San Francisco State University, P.O. Box 855, Tiburon, CA 94920 Abstract In addition to the native Spartina foliosa, four species of Spartina have been established in San Francisco Bay by human introduction. One species, Spartina patens, has been reported previously and appears to have been introduced acciden- tally. Three species, S. alterniflora, S. anglica, and S. densiflora, have been introduced in attempts to establish cordgrass within marsh restoration projects. Only S. alter- niflora and S. densiflora have spread beyond their original sites of introduction. The latter species has been introduced from Humboldt Bay, where it was previously included in the taxon S. foliosa. Morphological and ecological data support the con- clusion that the species occurring in Humboldt Bay should be referred to as Spartina densiflora and was probably introduced to northern California from South America during the mid-nineteenth century. Mobberley (1956), in his monograph of the genus Spartina, cites two species in California: Spartina foliosa Trin., found in coastal salt marshes, and Spartina gracilis Trin., found along inland alkali lakes and streams. The distribution of S. foliosa is given as Baja CaHfomia to Humboldt and Del Norte Counties by Mobberley (1956), Mason (1957), Munz (1973), and Macdonald and Barbour (1974), whereas Jepson (1925) and Hitchcock (1935) cite San Francisco Bay as being the northern limit. Since these accounts, new information has been gathered on the occurrence of this and several additional Spartina species in the salt marshes of northern California. Coastal Spartina in California. Spartina foliosa (California cord- grass) is the dominant Spartina in southern and central California and San Francisco Bay. Its northern coastal limit occurs north of San Francisco Bay at Bodega Bay. The single patch (ca. 20 m x 30 m) suggests its presence there is recent. Spartina foliosa is also pres- ent at Bolinas Lagoon and Drakes Estero, but is absent at Tomales Bay even though suitable habitat seems to occur. Macdonald and Barbour (1974) note its "conspicuous absence" here and in several other estuaries and lagoons in California. No Spartina occurs north of Bodega Bay until Humboldt Bay and the nearby Eel River delta. In the past, Spartina at these two locations was regarded as an ecotype of S. foliosa (Mobberley 1956, Gerish 1979, Rogers 1981, MADRof^o, Vol. 32, No. 3, pp. 158-167, 19 August 1985 1985] SPICHER AND JOSSELYN: SPARTINA IN CALIFORNIA 159 30^ 15! 12^00^ Fig. 1. Locations of introduced Spartina spp. in San Francisco Bay. A— South- ampton Bay {S. patens); B— Creekside Park {S. densiflora, S. anglica) and Corte Madera Creek {S. densiflora); C— Muzzi Marsh (5. densiflora); D— Greenwood Cove {S. densiflora); E— Alameda Creek Flood Control Channel {S. alter niflora). Claycomb 1983). However, as discussed in further detail below, ecological and taxonomic investigations have shown it to be a dis- tinct species, Spartina densiflora Brong. Despite reports that Spar- tina occurs in Del Norte County (Mason 1957, Munz 1973), we have not seen it north of Humboldt Bay as far as and including Coos Bay, Oregon. Introduced Spartina in San Francisco Bay. Until 1973, Spartina foliosa was the only Spartina described for San Francisco Bay. Since then, four more species have been introduced either accidentally or intentionally: Spartina patens (Ait.) Muhl., Spartina alterniflora Lois., Spartina anglica C. E. Hubbard, and Spartina densiflora. 160 MADRONO [Vol. 32 Fig. 2. Introduced Spartina alterniflora near the mouth of the Alameda Creek Flood Control Channel. It is taller than 5. foliosa which is in the foreground. Munz (1973) reported Spartina patens (saltmeadow cordgrass) for Southampton Bay (A— Fig. 1). We found an existing patch, but this species does not appear to have spread from its original location. The second species, S. alterniflora (smooth cordgrass), occurs at the mouth of the Alameda Creek Flood Control Channel (E— Fig. 1; Fig. 2) and along the shoreline approximately 3 km to the south. Both of these species are endemic to salt marshes of the eastern United States. The method and precise date of their introduction into San Francisco Bay are unknown. The third species, Spartina anglica (common cordgrass), was in- troduced at Creekside Park Marsh (B— Fig. 1) from Puget Sound, Washington in 1977 (K. Floyd, pers. comm.). These particular plants have been renamed internationally and misidentified locally in the past, so the use of S. anglica requires clarification. Locally in San Francisco Bay, they have been called Spartina maritima (K. Floyd, pers. comm., Hedgpeth 1980, Josselyn and Buchholz 1984). Taxo- nomic descriptions (Mobberley 1956, Hubbard 1968) and herbar- ium specimens [374220, 466912 (CAS)] clearly indicate these plants are not S. maritima; their oversized culms, leaves, and spikelets are among deciding features (Table 1) that place them in S. anglica.The name S. anglica was coined when two forms of S. townsendii (Town- send's cordgrass) were separated nomenclaturally. Spartina town- sendii, discovered in England in 1870, was regarded as a sterile 1985] SPICHER AND JOSSELYN: SPARTINA IN CALIFORNIA 161 Table 1 . Comparison of Morphological Characteristics Between Spartina maritima and S. anglica as Described by Hubbard (1968) and S. anglica from Creekside Park Marsh, Kentfield, CA. Species S. anglica Feature S. maritima S. anglica (Creekside Park) Culms to 50 cm tall to 130 cm tall to 1 26 cm tall Blades 2-18 cm long 10-45 cm long 36-46 cm long to 6 mm wide 6-15 mm wide 1 1-13 mm wide Ligules 0.2-0.6 mm long 2-3 mm long to 2.5 mm long Inflorescence 4-10 cm long 1 2-40 cm long 27-33 cm long Spikes 1-5 in number 2-12 in number 8-11 in number Spikelets 1 1-15 mm long 14-21 mm long 16-20 mm long Anthers 4-6 mm long 8-13 mm long 8-10 mm long hybrid resulting from the natural hybridization between the alien S. alterniflora from America and the endemic S. maritima (Hubbard 1968, Ranwell 1967, 1972). In 1892, a fertile form appeared, ap- parently a result of natural chromosome doubling (Hubbard 1968, Ranwell 1972). This fertile form remained unnamed until 1968, when Hubbard (1968) gave it the binomial, Spartina anglica C. E. Hubbard. The male-sterile hybrid is now Spartina x townsendii H. and J. Groves (Hubbard 1968). Because of its aggressive colonization and effective sediment-ac- creting abilities, Spartina anglica (and perhaps S. x townsendii) ramets were distributed worldwide upon request for creating salt marshes and controlling shoreline erosion (Mobberley 1956, Ran- well 1972, Chung 1983). In 1961 or 1962, as H. M. Austenson noted on a specimen [Ml 55990 (UC)], Washington State University and the U.S. Department of Agriculture introduced S. townsendii in Puget Sound, Washington (Snohomish County, Stillaguamish Es- tuary, near Stanwood). These plants are now known to be S. anglica because ramets of these plants introduced at Creekside Park Marsh flowered and produced 20% viable seeds in 1983. No flowering occurred in 1984. Spartina densiflora (Humboldt cordgrass) is the fourth Spartina introduced in San Francisco Bay. As mentioned previously, this species was introduced at Creekside Park Marsh in 1977, and was thought to be an ecotype of S. foliosa. Its present distribution in San Francisco Bay is limited to Marin County: at Creekside Park Marsh and Corte Madera Creek, Muzzi Marsh, and Greenwood Cove (Fig. 1). Taxonomy of the Humboldt Bay Spartina. In 1932, the identity of the Spartina growing in Humboldt Bay was questioned when Saint- Yves (1932) annotated a specimen identified earlier as S. fo- MADRONO [Vol. I- ^ 2 Q C O O 2 2 •S (U o o ^ ^ > d 3 2 .S S .S ^ -S OA I 6 ^4 2.S S 6 o N 2 =3 (u g 3 g "2 "2 S .3 ^ c« a 0:3 (u > o c O O ^ ^ <-> .2 .2 o •O ^ ^ 6 o I I 3 00 00 I— J (U !/5 9J C/5 1 ^ Eli CO X 1985] SPICHER AND JOSSELYN: SPARTINA IN CALIFORNIA 163 Fig. 3. Individual tussocks of Spartina densijlora at Creekside Park Marsh occupy slightly higher elevations (A) while Spartina foliosa forms meadow-like stands nearer channels (B). liosa by Hitchcock to be Spartina densijlora Brong. forma acuta St. Y. Mobberley (1956) rejected Saint- Yves' reidentification, stating that Saint- Yves based his decision only on the smaller spikelet lengths of the Humboldt Bay species. However, Saint- Yves (1932) based his opinion on three features: difference in spikelet lengths, difference in foliar structure, and strongly keeled glumes in the Humboldt Bay Spartina. Mobberley (1956) subdivided Spartina into three species com- plexes. The first contains species with numerous short, closely im- bricate spikes, hard slender culms, and short (or even lacking) rhizomes (e.g., S. spartinae). Complex two is characterized by species with thick, succulent, fleshy culms that grow from solitary bases or \ in small clumps; spikelets are usually less closely imbricate. These plants rarely show purple coloration (e.g., S. foliosa). The third complex contains species with indurate culms, more or less spreading spikes with closely imbricate spikelets, and very often are streaked or tinted with purple color. Spartina patens and S. densijlora are members of this group. A comparison of some morphological, phenological, and ecolog- ical characters of the Humboldt Bay Spartina with those of S. foliosa and S. densijlora were made from living and herbarium specimens (Table 2). The caespitose habit of the Humboldt Spartina, which 164 MADRONO [Vol. 32 differs from the solitary, evenly-spaced culms of S. foliosa, is the most visible difference between the two species (Fig. 3). The Hum- boldt Bay Spartina possesses all the characteristics of Mobberley's third complex {S. densiflora) except for its usually appressed-im- bricate spikes. Mobberley (1956) amends his general rule for S. densiflora, however, which possesses appressed spikes. There was speculation that the Spartina in Humboldt Bay was S. spartinae (Gerish 1979). Spartina densiflora does share some char- acteristics with S. spartinae, but Mobberley (1956) distinguished S. densiflora and S. spartinae in South America as follows: 1) spikelets of S. densiflora exceed 8 mm, whereas those of S. spartinae do not exceed 7 mm (some N. American specimens to 10 mm) 2) trichomes of S. densiflora are short, rigid, and slender; they are about one-half as long as the thicker trichomes of S. spar- tinae 3) the first glume of S. densiflora is about one-half as long as the second; rarely is the first shorter by more than 2 mm in any of the other Spartina spp., including S. spartinae The differences between herbarium specimens (CAS, UC) (Table 3) of Spartina spartinae and S. densiflora were found generally to be true. Not all characteristics are necessarily found in every spikelet, but the smaller spikelets and longer, thicker trichomes on the spike- lets of the S. spartinae inflorescence give it a tighter and more pu- bescent appearance than in the inflorescence of S. densiflora. The spikelets and inflorescences of the Humboldt Bay Spartina closely resemble those of S. densiflora. Gerish (1979) found the chromosome number of the Humboldt Bay Spartina to be In = 60, the same number counted for S. foliosa by Pamell (1976). Gerish inferred that the Humboldt Bay Spartina was from S. foliosa genetic stock and that any morphological dif- ferences were caused by genotypic or phenotypic processes. Although the chromosome numbers match, this single common denominator does not demonstrate conclusively that they are the same species. Many species in the genus have identical chromosome numbers (Moore 1973, Goldblatt 1981). Spartina densiflora introduction to North America. Spartina densiflora is almost certainly not native to Humboldt Bay. Its dis- tribution was reported previously only in South America below the 23rd parallel (Mobberley 1956). If it were a North American native, it would be expected to occur more extensively than in just one location. Therefore, Spartina densiflora was probably introduced into Humboldt Bay, as were many organisms in other estuaries of Cal- 1985] SPICHER AND JOSSELYN: SPARTINA IN CALIFORNIA 165 Table 3. Identification Numbers of Herbarium Specimens Studied in Comparing Three Spartina Species at the University of California Herbarium, Berkeley and at the California Academy of Sciences, San Francisco. Locations where specimens were collected are abbreviated in parentheses [California (CA), Texas (TX), Louisiana (LA), Florida (FL), Mexico (MX), Brazil (BR), Uruguay (UR), Ar- gentina (AR), Costa Rica (CR)]. Specimen identification number Herbarium Spartina Spartina Spartina location foliosa Trin. densijlora Brong. spartinae Trin. UC Berkeley M260502 (CA) 298388 (UR) MO47062 (BR) M027317 (BR) 627472 (AR) 627546 (AR) M025678 (UR) Ml 53237 (TX) 821629 (TX) 35760 (MX) California 444653 (MX) 101351 (AR) 303562 (CR) Academy of 368772 (CA) 686493 (MX) Sciences 440197 (CA) 386343 (CA) 418931 (CA) 274331 (CA) 101332 (CA) 382866 (FL) 182083 (LA) ifomia in modem times. In San Francisco Bay after 1850 for ex- ample, organisms were introduced unintentionally by ships from foreign lands. Among these organisms were many marsh plant species, including Atriplex semibaccata (L.) Presl. (Australia) and Cotula coronopifolia L. (South Africa) (Munz 1973, Atwater et al. 1979). Similarly, Spartina densiflora may have been introduced from Chile. During the 1 850s and early 1 860s, Chile experienced a period of rapid economic growth that created a demand for processed lum- ber, much of which was supplied from the northern California coast and Humboldt Bay (Cox 1974, Carranco 1982). Many company- owned lumber ships returned from South America without heavy cargo. For stabilization these ships often took on solid ballast gath- ered from the shoreline. The Chilean beachhopper, Orchestia chi- liensis, was introduced to San Francisco Bay in this manner, by the "Discharge of shingle ballast (stones, algae, and debris gathered from beaches) by lumber ships returning from Chile in or before 1900 . . (Carlton 1975). Similarly, we propose that seeds of S. densiflora were brought to Humboldt Bay from Chile. Spicher (1984) showed that the seeds of this species are tolerant of long periods of storage in either dry or moist conditions. In addition, Mobberley (1956) found S. densijlora spikes to shorten in length and increase in num- ber on inflorescences of plants from north to south along the east coast of South America and across to Chile. The greater number 166 MADRONO [Vol. 32 and shorter spikes of the Humboldt Bay Spartina (Table 2) reflect what might be expected in S. densiflom from Chile. Acknowledgments We acknowledge the discovery of S. foliosa in Bodega Bay by J. W. Buchholz and S. alternijlora at points south of the Alameda Flood Control Channel by T. E. Harvey. We thank M. Barkworth and an anonymous reviewer for their comments on an earlier draft of this paper, and Dr. Stanley Williams and Dr. Robert Patterson for their suggestions. This research was supported in part by a grant from the San Francisco Foundation. Literature Cited Atwater, B. F., S. G. Conrad, J. N. Dowden, C. W. Hedel, R. L. MacDonald, and W. Savage. 1979. History, landforms, and vegetation of the estuary's tidal marshes. In T. J. Conomos, ed., San Francisco Bay: the urbanized estuary, pp. 347-385. Pacific Division, Amer. Assoc. Advanc. Sci., San Francisco, CA. Carlton, J. T. 1975. Introduced intertidal invertebrates. In R. I. Smith and J. T. Carlton, eds.. Light's manual: intertidal invertebrates of the central California coast, pp. 17-25. Univ. Calif. Press, Berkeley. Carranco, L. 1982. Redwood lumber industry. Golden West Books, San Marino, CA. Chung, Chung-Hsin. 1983. Geographical distribution of Spartina anglica C. E. Hubbard in China. Bull. Marine Sci. 33:753-758. Claycomb, D. W. 1983. Vegetational changes in a tidal marsh restoration project at Humboldt Bay, California. M.A. thesis, Humboldt State Univ., Areata, CA. Cox, T. R. 1974. Mills and markets: a history of the Pacific coast lumber industry to 1900. Univ. Washington Press, Seattle. Gerish, W. 1979. Chromosomal analysis of a previously unidentified Spartina species. M.A. thesis, Long Island Univ., Long Island, NY. GoLDBLATT, P. 1981. Index to plant chromosome numbers 1975-1978. Missouri Bot. Gard., St. Louis. Hedgpeth, J. W. 1980. The problem of introduced species in management and mitigation. Helgolander Meeresunters. 33:662-673. Hitchcock, A. S. 1935. Manual of the grasses of the United States. USDA Misc. Publ. No. 200, Washington, D.C. Hubbard, C. E. 1968. Grasses: a guide to their structure, identification, uses, and distribution in the British Isles. Penguin Books Ltd., Middlesex, England. Jepson, W. L. 1925. Manual of the flowering plants of California. Assoc. Students Bookstore, Univ. Calif., Berkeley. Josselyn, M. N. and J. W. Buchholz. 1984. Marsh restoration in San Francisco Bay: a guide to design and planning. Techn. Report No. 3. Tiburon Center for Environmental Studies, San Francisco State Univ., Tiburon, CA. Kasapligil, B. 1 976. A synoptic report on the morphology and ecological anatomy of Spartina foliosa Trin. Appendix C in U.S. Army Corps of Engineers, Dredge Disposal Study, San Francisco Bay and Estuary, Appendix K. San Francisco, CA. Macdonald, K. B. and M. G. Barbour. 1974. Beach and salt marsh vegetation of the North American Pacific Coast. In R. J. Reimold and W. H. Queen, eds., Ecology of halophytes, pp. 175-233. Academic Press, NY. Mason, H. L. 1957. A flora of the marshes of California. Univ. Calif. Press, Berkeley. MoBBERLEY, D. G. 1956. Taxonomy and distribution of the genus Spartina. Iowa State J. Sci. 30:71-574. Moore, R. J. 1973. Index to plant chromosome numbers 1967-1971. Oosthoek's Uitgeversmaatschappij B. V., Utrecht, Netherlands. 1985] SPICHER AND JOSSELYN: SPARTINA IN CALIFORNIA 167 MuNZ, p. A. 1973. A California flora and supplement. Univ. Calif. Press, Berkeley. Parnell, D. R. 1976. Chromosome numbers in growth forms of Spartina foliosa Trin. Appendix D in U.S. Army Corps of Engineers, Dredge Disposal Study, San Francisco Bay and Estuary, Appendix K. San Francisco, CA. Ranwell, D. S. 1967. World resources of Spartina townsendii (sensu lato) and economic use of Spartina marshland. J. Appl. Ecol. 4:239-256. . 1 972. Ecology of salt marshes and sand dunes. Chapman and Hall, London. Rogers, J. D. 1981. Net primary productivity of Spartina foliosa, Salicornia vir- ginica, and Distichlis spicata in salt marshes at Humboldt Bay, California. M.A. thesis, Humboldt State Univ., Areata, CA. Saint-Yves, A. 1932. Monographia Spartinarem. CandoUea 5:19-100. Spicher, D. p. 1 984. The ecology of a caespitose cordgrass {Spartina sp.) introduced to San Francisco Bay. M.A. thesis, San Francisco State Univ., San Francisco, CA. (Received 24 Oct 1984; accepted 19 Feb 1985.) ANNOUNCEMENT The Executive Council of the California Botanical Society is pleased to announce that Mr. Wayne R. Ferren, Jr., has been appointed Editor of Madrono to follow the term served by Dr. Christopher Davidson. Dr. J. Robert Haller has been appointed Associate Editor. All manuscripts to be submitted to Madrono for review, and all inquiries concerning manuscripts submitted previously, should be directed to the Editor or Associate Editor, Department of Biological Sciences, University of Cali- fornia, Santa Barbara, CA 93106. THE SYSTEMATIC RELATIONSHIP OF ASARINA PROCUMBENS TO NEW WORLD SPECIES IN TRIBE ANTIRRHINEAE (SCROPHULARIACEAE) Wayne J. Elisens Department of Botany and Microbiology, University of Oklahoma, Norman 73019 Abstract A broad-based examination of Asarina s. str. has been undertaken to elucidate its systematic and phytogeographic relationship to New World species in tribe Antir- rhineae. Asarina procumbens is differentiated from Old World and New World species by a combination of distinctive characters: procumbent stems, opposite leaves, or- biculate to reniform laminas, solitary flowers in leaf axils, large personate corollas, and globose capsules. Although the pollen morphology and seed coat anatomy of Asarina s. str. are shared with some taxa in the New World, A. procumbens is cross- incompatible with purported congeneric species and differs from native American species by three unique features: opposite leaves with orbiculate to cordiform laminas, a chromosome base number of nine, and a buUate-corrugate seed coat ornamentation. It is hypothesized that A. procumbens is more closely related to Old World species, that Asarina sensu Pennell delimits an unnatural and heterogeneous assemblage of species, and that Asarina, unlike Antirrhinum, does not represent a genus with a North American— European Mediterranean disjunction. Asarina procumbens Mill, is an herbaceous perennial in tribe An- tirrhineae confined to calcareous habitats from 300 to 800 m in the Pyrenees Mountains of southern France and northeastern Spain. This species has been recognized in most twentieth century treat- ments of the Scrophulariaceae (Rothmaler 1943; but not in Melchior 1964) and the European flora (Hartl 1965, Webb 1972) as the sole species constituting the genus Asarina Mill. Although many authors (Bentham 1846, Wettstein 1891, Rothmaler 1943) have recognized Asarina s. str. as a monotypic taxon, its ranking at the level of genus has not been uniform. Asarina has been either accorded generic rank (e.g., Quer y Martinez 1762, Moench 1802) or recognized as a sub- genus (Reichenbach 1828, Rouy 1909) or section (Chavannes 1833,Bentham 1846, 1876, Wettstein 1891) in Antirrhinum L. As- arina procumbens has been segregated from Old World species in Antirrhinum and from species in New World Antirrhinum sect. Saerorhinum A. Gray (sensu Rothmaler 1 956) on the basis of several distinctive vegetative, floral, and fruit characters (Rothmaler 1943). Contrary to most supraspecific concepts in the tribe, Asarina's taxonomic boundaries were greatly expanded when Pennell (1947) transferred 1 5 North American species into it and did not delineate Madrono, Vol. 32, No. 3, pp. 168-178, 19 August 1985 1985] ELISENS: ASARINA 169 any infrageneric groups. The species included in Pennell's amplified genus had previously been treated in several supraspecific taxa by other authors. For example, Bentham (1876) and Wettstein (1891) had recognized the species in Asarina sensu Pennell in five sections in two genera {Antirrhinum sects. Asarina and Maurandella A. Gray; Maurandya Ort. sects. Eumaurandya (A. Gray) I. M. Johnst., Epi- xiphium (Engelm. ex A. Gray) A. Gray, and Lophospermum (D. Don) A. Gray); Rothmaler (1943) had placed the taxa in six genera {Asarina, Neogaerrhinum Rothm., Maurandella (A. Gray) Rothm., Maurandya, Epixiphium (Engelm. ex A. Gray) Munz, and Lopho- spermum D. Don). The rationale behind Pennell's mass transfer was "the form of the foliage and also the large flaring corollas . . (Pennell 1947, p. 174) supposedly shared by A. procumbens and some native American taxa. The floral diversity encompassed in Pennell's expanded Asa- rina was dismissed as adaptation to specific pollinators; polymor- phism among other key characters (e.g., capsule shape, seed orna- mentation, phyllotaxy, lamina outline and venation, stem type) was not addressed. Although certain authors regarded Pennell's (1947) expanded Asarina as unnatural (Johnston 1950, Munz 1959), Asa- rina sensu Pennell has been followed in some recent taxonomic, floristic, and horticultural treatments (DeWolf 1956, Shreve and Wiggins 1968, St. John 1973, Bailey and Bailey 1976, Wiggins 1980). In order to assess critically the taxonomic and phylogenetic re- lationships of Asarina s. str., an examination of A. procumbens has been undertaken that incorporates morphological, geographical, an- atomical, chromosomal, palynological, and crossability data. The present study represents the first report of seed coat microsculpturing and anatomical pattern, pollen morphology, and artificial hybrid- ization with New World species for A. procumbens. It also presents the first published chromosome count for a species in Neogaerrhi- num. The primary aims of the investigation have been to gather new systematic information on A. procumbens, assess the data to eluci- date the tribal affinities of Asarina s. str., and evaluate Pennell's (1947) expanded generic concept of Asarina. The nomenclature used in the text has followed Rothmaler (1943) or Elisens (1985a). Materials and Methods Comparative macromorphological studies of Asarina procumbens have been based on examination of herbarium specimens from F, MO, NY, PH, TEX-LL, and US as well as material cultivated from seed supplied by the Barcelona, Dijon, and Leipzig botanical gar- dens. Voucher specimens for all descriptive and experimental studies are on deposit at the University of Texas Herbarium (TEX); col- lection data are listed in the Appendix. 170 MADRONO [Vol. 32 Mature seeds for morphological and anatomical studies were based on samples supplied by botanical gardens. Each sample was 1) pre- pared for and examined with scanning electron microscopy (SEM) and 2) paraffin-embedded, stained, sectioned, and observed using light microscopy. Preparative procedures and materials were similar to those reported in Elisens and Tomb (1983) and Elisens (1985b). Bud material for meiotic chromosome counts was obtained from glasshouse-grown individuals propagated from seed supplied by bo- tanical gardens (A. procumbens) or the author's field collections {Neogaerrhinum). Buds were fixed in freshly mixed chloroform, ab- solute ethanol, and glacial acetic acid (4:3:1, v/v/v). Root-tips for mitotic counts were obtained from germinating seeds treated in a saturated 8-hydroxyquinoline solution. Chromosomes were stained with aceto-orcein. Pollen grains, obtained from fresh anthers, were dehydrated in acetic acid, acetolysed, and prepared using procedures and materials outlined in Elisens (1985a). The glycerin jelly-mounted grains were measured using light microscopy and observed and photographed using the SEM facilities at the Kansas Agricultural Experiment Sta- tion. Terminology for the exo- and endomorphology is that of Moore and Webb (1978). Investigations of reproductive biology were conducted on plants from 1 3 populations representing 1 2 species in Asarina sensu Pennell that were grown from field-collected seed (North American taxa) or seed supplied by botanical gardens {Asarina procumbens) in the glasshouse facilities at the University of Texas and Miami Univer- sity. Crossability and compatibility studies were undertaken on emasculated flowers with hand pollinations performed with fresh pollen using forceps dipped in 95% ethanol after each pollen transfer. Artificial pollinations that resulted in capsule development and seed set were considered successful crosses. Estimates of pollen fertilities were determined using cotton blue in lactophenol; 300 grains were scored for each flower and 2-3 flowers were examined per individual. Results Within tribe Antirrhineae, Asarina s. str. is distinctive because of its procumbent stems, opposite leaves, orbiculate to cordiform, glan- dular-pubescent laminas and palmate venation (Fig. 1); solitary, axillary flowers; lanceolate, apically-distinct calyx segments and large personate corollas (Fig. 2); and globose symmetric capsules (Fig. 3). No other taxon in the Antirrhineae has this combination of macro- morphological characters. An opposite leaf arrangement with or- biculate to cordiform laminas is not found among any New World taxon. The bullate-corrugate seed coat ornamentation pattern (Fig. 4) of 1985] ELISENS: ASARINA 171 Figs. 1-3. Photographs of distinctive morphological features of Asarina procum- bens, Elisens 799. Scale lines equal 5.0 mm. Fig. 1. Leaf attachment, outline, and vestiture; left leaf abaxial surface, right leaf adaxial surface. Fig. 2. Calyx and corolla. Fig. 3. Mature capsule; three calyx segments removed. A. procumbens is different from the seed morphology of any New World species in the tribe. The expanded crests on the seed surface form a characteristic reticulated pattern and are covered with in- terconnected ridges of low relief made up of radial cell walls (Fig. 5). The outer epidermal cell walls lack any protuberances (Fig. 5) similar to those found on expanded crests and tubercles in Mau- randya and Lophospermum (Elisens and Tomb 1983). Mean seed lengths within A. procumbens are 1.52 mm long (SD = 0.16 mm, n = 60); the seeds weigh an average 0.20 mg (n = 100). The seed coat anatomical pattern of A. procumbens is an epidermis with 90% to 95% of the cells radially elongate (Fig. 6) and a hypodermis of one to three flattened layers (Fig. 7). Reticulated thickenings are present in the epidermal cells (Figs. 6, 7) and are responsible for the ribbed appearance of the radial cell walls visible on the seed surface (Fig. 5). The differentially elongated epidermal cells are solely re- sponsible for the exotestal relief (previously noted by Bachmann 1880). The pollen morphology of A. procumbens was examined from one collection. The grains are subspheroidal, trizonocolporate (Fig. 8), and have a perforate-tectate exine pattern (Fig. 9) with the perfo- rations less than 1 ^im in diameter. The mean equatorial diameter 172 MADRONO [Vol. 32 Figs. 4-9. Photomicrographs of seed and pollen of Asarina pwcumbens. Key to labeling: cot, cotyledon; cw, cell wall; emb, embryo; end, endosperm; epi, epidermis; etm, endothelium; hyp, hypodermis. Figs. 4-5. Scanning electron micrographs, seed; Elisens 613. Scale lines equal 0.05 mm. Figs. 6-7. Light micrographs, transversely- sectioned seed; Elisens 613. Scale lines equal 0.05 mm. Figs. 8-9. Scanning electron micrographs, pollen; Elisens 799. Scale lines equal 1 jxm. 1985] ELISENS: ASARINA 173 of the pollen is 19.8 1 fim with a polar diameter/equatorial diameter (P/E) ratio of 1.13. The chromosome base number of A. procumbens isx=9. Meiotic chromosome number determinations for A. procumbens were ob- tained from two collections. These counts, as well as Jackson's (1971) unpublished ones from four European botanical garden seed sam- ples, verify the 2« = 18 reported by Heitz (1927). A mitotic count of In = 30 was also obtained for one population of Neogaerrhinum filipes (see Appendix). Crossability studies have been undertaken between A. procumbens and 1 1 American species in Asarina sensu Pennell. Even though pollen fertilities (93%-98%) were uniformly high for all plants, the twenty directional crosses attempted (162 hand pollinations) re- sulted in no capsule or seed set. In all instances, gynoecia from emasculated flowers of A. procumbens aborted within two weeks after anthesis whether they received no pollen or pollen from the other species. Reciprocal crosses also resulted in ovary abortion. Untreated flowers had successful capsule and seed set (25/25) in- dicating that A. procumbens is self-compatible and autogamous. Be- cause unpollinated emasculated flowers or those pollinated from different species did not set capsules, A. procumbens exhibited no evidence of apomixis. Purported congeners in Pennell's (1947) ex- panded Asarina are uniformly self-compatible, autogamous (except for Mabrya geniculata (Rob. & Femald) Elisens), and also showed no evidence of apomixis (Elisens 1985a). Discussion The systematic information reported in this study supports the taxonomic segregation of Asarina s. str. from Old World and New World genera in tribe Antirrhineae. Among the New World taxa, A. procumbens can be readily distinguished from each native genus by a combination of distinctive characters. Additionally, A. procum- bens has several unique characters not found in any native American species: an opposite phyllotaxy with orbiculate to reniform laminas, a bullate-corrugate seed surface sculpturing pattern, and a chro- mosome base number of nine. The pollen morphology and testal anatomy are shared with some New World taxa. Among seeds of New World taxa, the bullate-corrugate exotestal ornamentation pattern is most similar to the tumid tuberculate/ cristate pattern found on seeds in Maurandya subg. Maurandya and some Mabrya Elisens species (Elisens and Tomb 1983, Elisens 1 985a). The seed surfaces of the last two taxa have minute protuberances on the outer tangential and radial epidermal cell walls; similar pro- tuberances are lacking on the seeds of A. procumbens. Mean seed lengths and weights are in the range of the nonalate seeds of Mau- 174 MADRONO [Vol. 32 randya and Mabrya (Elisens and Tomb 1983). The seed coat ana- tomical pattern of A. procumbens is similar to testae of Maurandya subg. Maurandya, Mabrya, and Holmgrenanthe Elisens (Elisens 1985b). The testal anatomy of very few Old World species has been examined (Bachmann 1880). Both the seed coat morphological and anatomical pattern of Antirrhinum sect. Saerorhinum, Neogaerrhi- num, Pseudorontium (A. Gray) Rothm., Gambelia Nutt., and other Antirrhinum segregates in the New World are different from Asarina procumbens (Elisens and Tomb 1983, Elisens 1985b). The pollen exine pattern of A. procumbens is found also in Mau- randya, Mabrya, Lophospermum (Elisens 1985a), Neogaerrhinum, Galvezia Dombey ex Juss., and some Old World and New World Antirrhinum species (Elisens, unpubl. data). Similar mean equatorial pollen diameters to A. procumbens are found among New World species (15.74 to 25.54 fim). The pollen P/E ratio of A. procumbens (1.13) is slightly outside the range of the American species (0.96 to 1.10) except for Linaria texana Scheele which has a P/E of 1 .34 and some Old World species in Antirrhinum and Chaenarrhinum Reichb. (Elisens, unpubl. data). Because the pollen exine pattern, dimensions, and shape of A. procumbens are widespread in the Antirrhineae, pollen morphology appears to be of limited utility in elucidating the taxonomic or phylogenetic relationships of Asarina s. str. No New World species in tribe Antirrhineae has a base chro- mosome number of nine, although Anarrhinum Desf. and Kickxia Dumort. are Old World genera with x = 9. Other Old World base numbers are x = 8, 7, and 6 (Fedorov 1969). New World tribal base numbers are 15 {Galvezia, Mohavea A. Gray, some Antirrhinum speciQS, Neogaerrhinum), 13 (Pseudorontium), 12 (Maurandya, Ma- brya, Lophospermum), 8 (Antirrhinum), and 6 (Linaria Mill.) (Giin- ther and Rothmaler 1963, Raven et al. 1965, Jackson and Spellen- berg 1973). The New World and Old World base numbers suggest that aneuploidy has been important in trans-specific evolution in the Antirrhineae (Elisens 1985a). If this is the case, the base number of nine for Asarina s. str. is clearly anomalous in the New World. The Old World distribution of A. procumbens does not preclude automatically the segregation of Asarina s. str. from New World genera. Within the Antirrhineae, Antirrhinum and Linaria have Old World and New World species, although New World species in these genera usually are placed in different sections. Several North Amer- ican species in Antirrhinum occur in habitats similar to A. procum- bens'. Antirrhinum also has a North American-European Mediter- ranean disjunct pattern (Rothmaler 1956, Raven 1973). The marked morphological and chromosomal differentiation of A. procumbens from New World Antirrhineae suggest, however, that Asarina s. str. is phylogenetically distant from New World species in Antirrhinum and other native American taxa. 1985] ELISENS: ASARINA 175 The complete cross-incompatibility of Asarina procumbens with eleven species from Neogaerrhinum {= Antirrhinum sect. Mauran- della p.p.), Maurandya, Mabrya, and Lophospermum further rein- forces the view that Asarina s. str. is evolutionarily very distant from New World genera in the Antirrhineae. Species from these four genera were transferred into Asarina by Pennell (1947). As noted by Grant (1981), the general interspecific crossability pattern in the Scrophulariaceae (and many perennial herbs with prominent species- to-species differences in floral mechanism) is for congeneric species to be interfertile within wide limits (e.g., Garber and Gorsic 1956, Yeo 1966, Vickery 1978). This pattern occurs within the Antirrhi- neae in y4«^z>r/z/«wm (Baur 1932, Mather 1947), Linaria (Viano 1978), and subtribe Maurandyinae (Elisens 1982). Baur (1914, 1932) has previously demonstrated that Asarina procumbens is cross-incom- patible with several Old World Antirrhinum species as well. Asarina sensu Pennell (1947) is extremely polymorphic and in- corporates three chromosome base numbers (15, 12, 9), suggesting that Penneirs expanded genus is an unnatural assemblage of species. The heterogeneity within the amplified Asarina is in distinct contrast to the variation pattern characteristic of most generic concepts in the tribe (e.g., Wettstein 1891, Rothmaler 1943, EUsens 1985a, D. Sutton, unpubl. data). Other than solitary flowers in the leaf axils (found in other taxa in the tribe), most potentially unifying characters in Asarina sensu Pennell (1947) are those generally used to char- acterize the tribe or family. Not even Linaria or Antirrhinum, the largest genera in the Antirrhineae, encompass the morphological diversity of stem, leaf, floral, and fruit characters found in Asarina s. lat. Pennell based his expanded genus concept on very few mor- phological characters: stem type, lamina outline, corolla type, and, to a lesser extent, capsule shape; none of these "key" characters is monomorphic within his boundaries of Asarina. The expanded ge- nus can be divided into several chromosomally- and morphologi- cally-coherent segregate taxa, such as the four genera constituting subtribe Maurandyinae (Elisens 1985a) and the genus Neogaerrhi- num {Rothmsder 1943). In summary, the findings of the present study indicate that Asarina i procumbens is morphologically, chromosomally, and geographically different from its purported New World congeners or any New World Antirrhineae species. Furthermore, it is recommended that neither the generic concept of Pennell (1 947) nor the purported North Amer- ican—European Mediterranean disjunction in Asarina be recog- nized. Considering its chromosome base number, shared by other Old World Antirrhineae, and a restricted distribution in the Med- iterranean region, where generic diversity among Old World Antir- rhineae is the greatest, A. procumbens evidently is more closely allied to taxa in the Old World. Even though its relationship among Old 176 MADRONO [Vol. 32 World species in tribe Antirrhineae is obscure, it seems unnecessary and incorrect to look in the southwestern United States and Mexico for its relatives. Acknowledgments The author thanks T. G. Lammers and J. J. Furlow for their comments on an earlier draft of the manuscript. Thanks are also extended to A. S. Tomb for his help in obtaining access to the SEM facilities at the Kansas Agricultural Experiment Station and to G. L. Floyd for use of his darkroom facilities. Literature Cited Bachmann, E. 1880. Die Entwickelungsgeschichte und der Bau der Samenschalen der Scrophularinee. Nova Acta Acad. Caes. Leop. -Carol. German Nat. Cur. 43: 1-179. Bailey, L. H. and E. Z. Bailey (compilers). 1976. Hortus third (revised by the stalf of Liberty Hyde Bailey Hortorium). Macmillan, NY. Baur, E. 1914. Einfiihrung in die experimentelle Vererbungslehre. Bomtraeger, Berhn. . 1 932. Artumgrenzung und Artbildung in der Gattung Antirrhinum, Sektion Antirrhinastrum. Z. Indukt. Abstammungs- Vererbungsl. 63:256-302. Bentham, G. 1846. Scrophulariaceae. In A. DeCandolle, Prodromus systematis naturahs regni vegetabilis 10:186-586. . 1876. Scrophularineae. /« G. Bentham and J. D. Hooker, Genera plantarum 2:913-980. Reeve and Co., London. Chavannes, E. 1833. Monographic des Antirrhinees. Treuttel and Wurtz, Paris. DeWolf, G. p. 1956. Notes on cultivated Scrophulariaceae 2. Antirrhinum and Asarina. Baileya 4:55-68. Elisens, W. J. 1982. A taxonomic monograph of the Maurandyinae. Ph.D. disser- tation, Univ. Texas, Austin. . 1 985a. Monograph of the Maurandyinae (Scrophulariaceae— Antirrhineae). Syst. Bot. Mongr. vol. 5. . 1 985b. The systematic significance of seed coat anatomy among New World species in tribe Antirrhineae (Scrophulariaceae). Syst. Bot. 10: in press. and A. S. Tomb. 1983. Seed coat morphology in New World Antirrhineae: systematic and phylogenetic implications. PI. Syst. Evol. 142:23-47. Fedorov, a. a., ed. 1969. Chromosome numbers of flowering plants. Acad. Sci. USSR, L. Komarov Bot. Inst., Leningrad. Garber, E. D. and J. Gorsic. 1956. The genus Collinsia IL Interspecific hybrids involving C. heterophylla, C. concolor, and C. sparsifolia. Bot. Gaz. (Crawfords- ville) 118:73-77. Grant, V. 1981. Plant speciation. Columbia Univ. Press, NY. Gray, A. 1868. Characters of new plants of Cahfomia and elsewhere, principally of those collected by H. N. Bolandier in the State's Geological Survey. Proc. Amer. Acad. Arts 7:327-401. GuNTHER, E. and W. Rothmaler. 1963. Chromosomezahlen amerikanischer tirrhinum-Artcn. Biol. Zentralbl. 82:95-99. Hartl, D. 1965. Scrophulariaceae. In G. Hegi, ed., Illustrierte Rora von Mittel- Europa. Carl Hanser Verlag, Munchen. Heitz, E. 1927. iiber multiple und aberrante Chromosomenzahlen. Abh. Naturwiss. Verein. Hamburg 21:47-57. Jackson, D. M. 1971. Biosystematics of the Maurandya erubescens complex. Mas- ter's thesis. Biology Dept., Mankato State Univ., Mankato, MN. and R. Spellenberg. 1973. In lOBP chromosome number reports XLII. Taxon 22:647-654. 1985] ELISENS: ASARINA 177 Johnston, I. M. 1950. Noteworthy species from Mexico and adjacent United States, III. J. Arnold Arbor. 31:188-195. Melchior, H. 1964. Scrophulariaceae. In A. Engler, Syllabus der Pflanzenfamilien 2:424-470. Bomtrager, Berlin. MoENCH, C. 1 802. Supplementum ad methodum plantas. Marburgi Cattorum, Mar- burg. Moore, P. D. and J. A. Webb. 1 978. An illustrated guide to pollen analysis. Hodder and Stoughton, Toronto. MuNZ, P. A. 1959. A California flora. Univ. California Press, Los Angeles. Pennell, F. W. 1947. Some hitherto undescribed Scrophulariaceae of the Pacific states. Proc. Acad. Nat. Sci. Philadelphia 99:155-199. QuER Y Martinez, J. 1762. Flora Espafiola (vol. 3). Joachin Ibarra, Madrid. Raven, P. H. 1973. The evolution of "Mediterranean" floras. In H. Mooney and F. di Castro, eds.. Evolution of mediterranean ecosystems, pp. 2 1 3-224. Springer- Verlag, NY. , D. W. Kyhos, and A. J. Hill. 1965. Chromosome numbers of spermato- phytes, mostly Califomian. Aliso 6:105-1 13. Reichenbach, H. G. L. 1828. Conspectus Regni Vegetabilis. Apud Carolum Cno- bloch, Leipzig. Rothmaler, W. 1 943. Zur Gliederung der Antirrhineae. Feddes Repert. Spec. Nov. Regni Veg. 52:16-39. . 1956. Taxonomische Monographic der Gattung Antirrhinum. Feddes Re- pert. Beih. 136:1-124. RouY, G. 1909. Flore de France. Deyrolle, Paris. Shreve, F. and I. Wiggins. 1968. Vegetation and flora of the Sonoran Desert. Stanford Univ. Press, Stanford, CA. St. John, H. 1973. List and summary of the flowering plants in the Hawaiian Islands. Pacific Trop. Bot. Gard. Mem. no. 1. Lawai, Kauai, HI. Viano, J. 1978. Croisements experimentaux interspecifiques au sein du genre Linar- ia. Caryologia 31:383-425. Vickery, R. K. 1978. Case studies in the evolution of species complexes in Mimulus. Evol. Biol. 11:405-507. Webb, D. A. 1972. Asarina. In T. G. Tutin, V. H. Heywood, N. A. Burges, D. H. Valentine, S. M. Walters, and D. A. Webb, eds.. Flora Europaea 3:224. Cam- bridge Univ. Press, Cambridge. Wettstein, R. 1891. Scrophulariaceae. In A. Engler and K. Prantl, eds.. Die na- tiirlichen Pflanzenfamilien IV (3b): 3 9- 107. Wiggins, I. L. 1980. Flora of Baja California. Stanford Univ. Press, Stanford, CA. Yeo, P. 1966. The breeding relationships of some European Euphrasiae. Watsonia 6:216-245. (Received 19 Nov 1984; accepted 19 Feb 1985.) Appendix Vouchered collections of 1) Asarina procumbens used in seed, pollen, and chro- mosomal studies, 2) Neogaerrhinum species whose chromosome numbers were de- termined, and 3) taxa used in the crossability studies. Voucher specimens are de- posited at TEX-LL unless otherwise indicated. Seed Coat Morphology and Anatomy. Asarina procumbens. France: Dijon Bot. Gard., 1981; Elisens 613. Pollen Morphology. Asarina procumbens. Spain: Barcelona Bot. Gard., 1981; Elisens 799. 178 MADRONO [Vol. 32 Chromosome Number Determination. Asarina procumbens, 2n = 18. France: Dijon Bot. Gard., 1981; Elisens 613. Spain: Barcelona Bot. Gard., 1981; Elisens 799. Neogaerrhinum filipes, 2n = 30. United States: Nevada, Elisens 617. Crossability Studies. Asarina procumbens. France: Dijon Bot. Gard., 1981; Elisens 613. Spain: Barcelona Bot. Gard., 1981; Elisens 799. Neogaerrhinum filipes. United States: Nevada, Elisens 617. Maurandya. M. antirrhiniflora. United States: Texas, Travis Co., Elisens 528. M. barclaiana. Mexico: Nuevo Leon, Turner and Danes A- 13. M. scandens. Mexico: Oaxaca, Elisens 655. M. wislizeni. United States: Texas, Ward Co., Elisens 530. Mabrya. M. acerifolia. United States: Arizona, Maricopa Co., Elisens 584. M. erecta. Mexico: Coahuila, Gordon 777. M geniculata. Mexico: Sonora, Gordon 763. Lophospermum. L. atrosanguineum. Mexico: Oaxaca, Elisens 665. L. purpusii. Mex- ico: Oaxaca, Elisens 549. L. scandens. Mexico: Morelos, Elisens 652. ANNOUNCEMENT On Saturday, 28 September 1985, Huntington Botanical Gardens will host its Second Symposium on Succulent Plants. Featured will be the following speakers and their topics: A. Gibson, Univ. California, Los Angeles, The classification of cacti above the species level; M. Kimnach, Huntington Bot. Gard., The origins of epiphytic cacti; P. Nobel, Univ. California, Los Angeles, Environmental influences on agaves— Implications for establishment, tolerances, and productivity; C. Uhl, Cornell Univ., Polyploidy in Echeveria (Crassulaceae); G. Webster, Univ. California, Davis, Evo- lution and systematics of neotropical Jatropha and Cnidoscolus (Euphorbiaceae); A. Zimmerman, Univ. Texas, Systematics of the genus Coryphantha (Cactaceae). Included in the day's events will be special tours of the recently opened Desert Garden Conservatory, an auction of rare plants, an optional luncheon and dinner, and an evening panel discussion. For information concerning registration and the schedule of events, please write: Succulent Plant Symposium, Huntington Botanical Gardens, 1151 Oxford Road, San Marino, CA 91 108. MIMULUS NORRISII (SCROPHULARIACEAE), A NEW SPECIES FROM THE SOUTHERN SIERRA NEVADA Lawrence R. Heckard Jepson Herbarium, University of California, Berkeley 94720 James R. Shevock Department of Botany, California Academy of Sciences, San Francisco 94118 Abstract Mimulus norrisii, a new cliff-dwelling species from the Sierra Nevada foothills primarily in Sequoia National Park, Tulare County, California, is described and illustrated. The new species is ecologically similar and morphologically closest to M. dudleyi in sect. Paradanthus but differs in its moister habitat requirements, leaves with attenuate bases and less serrate margins, and in particular the smaller calyces that develop conspicuous enlarged and rounded ribs. This attractive new species of Mimulus was discovered by the second author and Larry Norris, Naturalist in Sequoia National Park. Review of Mimulus specimens in California herbaria (CAS, DS, JEPS, POM, RSA, UC) failed to locate a collection of the new species. That it has been overlooked by botanists remains a mystery, but most plant collections from the Park have been from the co- niferous forests and alpine environments during the summer months and the lower elevations within the chaparral and blue oak com- munities have not been systematically surveyed. Other additions to the Park flora (Draba cuneifolia, Notholaena jonesii, and Parietaria hespera var. hespera) also come from the marble outcrops associated with the new Mimulus. Mimulus norrisii Heckard & Shevock sp. nov. Planta annua dense glanduloso-villosa, caulibus adscendentibus. Foliorum paginae ovatae leniter repando-denticulatae, palmato- ) venosae, basibus attenuatis vel cuneatis. Calyx apud florem 3.5-5.0 mm longus campanulatus, dentibus deltatis ca. 1.5 mm longis, api- cibus obtusis vel leviter mucronatis. Calyx apud fructus 5-6 mm longus, urceolatus, costis ampliatis, rotundatis. Corolla 1 5-30 mm longa, infundibuliformis faucis brevi-expansa et limbo quasi rotato (Figs. 1, 2). Madrono, Vol. 32, No. 3, pp. 179-185, 19 August 1985 180 MADRONO [Vol. 32 Fig. 1 . Mimulus norrisii. Top: Marble outcrop habitat at Comb Rocks, the type locality. Shrubs in foreground are Toxicodendron. Bottom: Habit of plants in rock crevice. 1985] HECKARD AND SHEVOCK: NEW M/Ml/LC/S 181 Fig. 2. Mimulus norrisii. Lateral view of flower showing pedicel and calyx and their indument, and the widely flaring corolla. Annual with diffuse roots; stems ascending, 3-15(-25) cm long with intemodes reaching 6-7 cm, sometimes branched from lower nodes, often floriferous from near base, the longer stems often ge- niculate; herbage densely glandular- villous (particularly in nodal re- gions) with trichomes mostly under 1 mm, occasionally to 2 mm. Leaves usually 2-5 pairs per stem, the blades ovate with attenuate to cuneate bases, mostly 2.0-3.5 cm long and 1-2 cm wide, weakly repand-denticulate with 3-5 pairs of small teeth on upper % of blade, reduced and often narrowed upwards on stem, palmately 3-5 veined or upper pair (and occasionally 1-2 additional veins) diverging pin- nately from parallel unfused veins of midrib, typically the veins running distinct into the petiole; petiole 0.5-1 .5 cm long, not sharply delineated from the tapering blade, reduced upwards on stem. Flow- ers axillary on slender ascending pedicels 2-5 cm long, the pedicels reflexed in fruit and sometimes hooked at the apex; calyx in anthesis 3.5-5.0 mm long, narrow-campanulate, sulcate between rounded ribs, rib region strongly glandular-pilose, usually infused or spotted with purplish red, sulca paler with less spotting and sparser indu- ment; calyx teeth ca. 1.5 mm long, rounded deltate with apices obtuse to slightly mucronulate, inwardly concave; calyx in fruit elongating to 5-6 mm, becoming somewhat urceolate with incurving of lobes and expansion of thinner and paler inter-rib region during capsule enlargement, the ribs enlarged and rounded; corolla caducous, fun- 182 MADRONO [Vol. 32 nelform, 1 5-30 mm long, the tube gradually widening to about mid- point, expanding to form short, open flaring throat and spreading to almost rotate limb made up of nearly equal lobes (6-10 mm long) that are broader than long, rounded to truncate and often retuse to emarginate, centrally grooved; corolla yellow, marked on throat cen- trally below each lobe with bilobed to irregular maroon-purple blotches (mostly one per lobe except several smaller ones on central lobe of lower lip), usually with two white spots on throat beneath the two sinuses forming the central lobe, weakly puberulent and sometimes glandular on exterior, the inner surface of lobes with scattered yellow clavate hairs that become smaller and denser on palatal folds of throat and down tube; stamens glabrous, included in lower one-half and attached near tube base, the upper filaments 3.5-4.0 mm long, the lower filaments 5-6 mm long; anthers ex- planate, longitudinally oriented, ca. 1 mm long; gynoecium ca. 10 mm long, style ca. 6 mm long, the stigma ca. 1.0 mm long, exceeding anthers by ca. 1-3 mm, bilamellate with equal, spreading lobes that are rounded and fimbriolate; capsule narrow-ovoid, 4-6 mm long, about equalling calyx, often unequally developed on opposite sides of style base, the style eventually breaking near the base leaving short, curved apiculation, the placentae adherent to apex, dehiscing full length along both sutures; seeds many (up to 100/capsule), el- lipsoid-oblong, longitudinally minutely rugose-striate, tawny-col- ored, ca. 0.5 mm long. Chromosome number In = 32. Type: USA, CA, Tulare Co.: Comb Rocks above Washburn Cove, 2 mi n. of Three Rivers, T17S R28E SI, 2800 ft. (854 m), 1 May 1983, L. L. Norris 389. (Holotype: JEPS; isotypes: CAS, FSC, K, MO, NY, RSA, US.) Paratypes: USA, CA, Tulare Co.: West ridge of Blossom Pk., South Fork Kaweah River, Three Rivers, T17S R28E S25, 19 Mar 1984, Norris 627 (JEPS); Comb Rocks, 2 mi n. of Three Rivers, T17S R28E SI, 1 May 1983, Shevock 10353 (CAS); Sequoia Nat'l. Park: Divide between Elk Cr. and Marble Fork Kaweah River, T16S R29E S23, Norris 351 (JEPS) and Shevock 10165 (CAS); 19 Mar 1983, Norris 354 (RSA) and Shevock 10187 (CAS, JEPS); 18 Apr 1983, Shevock 10330 (CAS, JEPS, RSA); 12 May 1983, Bacigalupi 9350 (JEPS, OSC, SD); and 24 Apr 1984, Norris 638 (CAS, FSC, JEPS, MO, RSA); Generals Highway 0.7 mi e. of Ash Mtn., T16S R29E S34, 30 Mar 1983, Norris 363 (SBBG, UC); and 24 Apr 1984, Norris 637 (JEPS); Yucca Point just n. of Ash Mtn., T16S R29E S34, 18 Apr 1983, Norris 372 (JEPS) and Shevock 10329 (CAS, RSA); 12 May 1983, chromosome voucher, « = 16, Bacigalupi 9344 (CHSC, JEPS, OBI, WTU); Clough Cave, South Fork Kaweah River, T18S R30E SI 9, 19 Mar 1984, Norris 626 (JEPS); Above Alder Cr. near Ash Mtn., T16S R29E S34, 23 Mar 1984, Norris 631 (THRI). 1985] HECKARD AND SHEVOCK: NEW MIMULUS 183 Distribution, habitat and phenology. Marble outcrops in chamise chaparral or blue oak woodland, Kaweah River drainage, 610-1310 m, southern Sierra Nevada within Tulare Co., California; most pop- ulations located within Sequoia National Park. Flowers March-May. The plants find rootholds in soil pockets, moss covered ledges, cracks and fractures in the marble outcrops, primarily in areas with con- centrations of beige-colored deposits of calcium carbonate. Steep east- or west-facing outcrops have the densest concentrations of plants, although plants do occur sparingly on south-facing cliffs. Light regimes vary during the day from full sun to full shade. The most robust plants occur where dripping water from mossy over- hangs keeps the cliff face moist. Associated species. Asterella californica, Anacolia menziesii var. baueri, Bryum pseudotriquetrum var. bimum, Encalypta vulgaris, Selaginella hanseni, Aspidotis californica, Cheilanthes cooperae, Notholaena jonesii, Pellaea mucronata, Pityrogramma triangularis, Dudleya cymosa subsp. cymosa, Eriogonum nudum subsp. muri- num, Lithophragma bolanderi, Parietaria hespera var. hespera, Pte- rostegia drymarioides. Toxicodendron diversilobum and Yucca whip- plei subsp. caespitosa. Populations of Mimulus norrisii are fairly common on all marble outcrops investigated in the Kaweah River drainage. The total area of the marble habitat, however, is estimated at only 200 hectares. The number of plants occupying the available habitat can only be grossly estimated because of the inaccessibility of the rugged cliff habitat on which the species occurs. We expect population size to vary markedly from year to year. Following an exceedingly wet winter (1983) for the southern Sierra Nevada, we estimated a total population of 7000 individuals. Essentially all populations are free from disturbances by man. Mimulus norrisii belongs to the sect. Paradanthus, an assemblage of about 70 species that Grant (1924) proposed to accommodate "a collection of groups not necessarily related to one another and in all probability most of them have been derived from members of other sections" (p. 1 17). The relationship of M. norrisii within this poorly understood and possibly polyphyletic section is equivocal. In most features, M. norrisii is closest to M. dudleyi Grant of the M. flori- bundus Dougl. ex Lindl. alliance, but the smaller, campanulate (to urceolate in fruit) calyx with thickened and rounded ribs is quite unlike that in the M. floribundus group. Thickened calyx ribs are found in only two other species of the section— M bicolor Hartw. ex Benth. and M. filicaulis Wats, (the latter includes M. biolettii Eastw. according to Bacigalupi [1 98 1])— but these species differ from M. norrisii in several well-marked features, including size and shape 184 MADRONO [Vol. 32 of calyx as well as the type of thickening itself. Thus their calyx is cylindric-oblong with pointed lobes and the "corky ribs," as Grant (1924) and Pennell (1951) described the strongly developed rib an- gles, are composed of softer tissue than in M. norrisii. Based on these differences in the calyx, it seems likely that thickening of the ribs has evolved independently in M. norrisii and the M. bicolor-filicaulis group. Our opinion that the nearest relative of Mimulus norrisii is M. dudleyi is based on their basic similarity in nearly all morphological respects, except the calyx and a few minor traits. The calyx of M. dudleyi is narrow-campanulate (with spreading pointed lobes) and ridge-angled in flower, becoming narrowly oblong with erect or spreading lobes in fruit in contrast to the shorter calyx of M. norrisii, which changes from campanulate (with rounded lobes) in flower to urceolate with thickened ribs in fruit. Although the plants of both species are petrophilous, their habits differ in that M. dudleyi is prostrate-ascending over granitic rocks, while M. norrisii is loosely erect or hanging from marble cliffs. Other differences are that the leaves of M. dudleyi have obtuse to truncate bases and much more pronounced serrate margins than those of M. norrisii, in which the bases are tapered and the margins are sparingly denticulate. Although both species have stems that are geniculate at times, only M. norrisii has (fruiting) pedicels that are reflexed and sometimes hooked at the tips, which facilitates very local dispersal of seeds. The corollas of the two species appear remarkably similar in shape and color, as are the short stamens and style-stigma that are included well within the tube. There are slight differences in corolla markings between the two species: M. dudleyi lacks the two white spots between the lower lobes and has the maroonish markings at the base of the lobes as flecks that are in less discrete blotches than in M. norrisii. The same chromosome number {2n = 32) has been found by Dr. T.-I. Chuang in M. norrisii (cited above) and M. dudleyi (CA, Tulare Co.: 10 mi se. of Porterville, Heckard and Chuang 4003). Although Mimulus norrisii and M. dudleyi both occur in the foot- hills of the southern Sierra Nevada, they occupy different habitats and their elevation ranges rarely overlap. Thus plants of M. norrisii are in damp and mostly shaded situations on cliffs of metamorphic rock in chamise chaparral (above 600 m). Mimulus dudleyi, on the other hand, is mostly at lower elevations in valley grassland or blue oak savanna and occurs on granitic outcrops that are in full sun most of the day and are wet only for short periods following rains. The liverworts, mosses, and ferns typically associated with M. nor- risii are absent in this latter habitat. Acknowledgments We thank Larry L. Norris for specimens, information, photographs, and field as- sistance, Rimo Bacigalupi and Robert J. Meinke for sharing their knowledge of 1985] HECKARD AND SHEVOCK: NEW MIMULUS 185 Mimulus, Jim Hickman for critical comment, and T.-I. and Fei-mei Chuang for kindly supplying chromosome counts. Literature Cited Bacigalupi, R. 1981. The identity of Mimulus filicaulis. The Changing Seasons 1(3, suppl.):3-5. Grant, A. L. 1924 (publ. Jan 1925). A monograph of the genus Mimulus. Ann. Missouri Bot. Gard. 1 1 . Pennell, F. W. 1951. Mimulus. In Abrams, Illustrated flora of the Pacific States 3:688-731. Stanford Univ. Press, Stanford. (Received 17 Dec 1984; accepted 16 Jan 1985.) ANNOUNCEMENT The Graduate Student Meetings, sponsored by the California Botanical Society, will be held this year at the University of California, Santa Barbara, on October 1 9 and 20, 1985. Graduate students wishing to present papers should prepare abstracts to be submitted when the call for abstracts is announced in August. Please contact Ms. Kathy Rindlaub, Department of Biological Sciences, University of California, Santa Barbara 93 1 06, or Mr. Joseph M. DiTomaso, Department of Botany, University of California, Davis 95616, for information on registration, schedule of events, field trips, and the banquet. NOTES AND NEWS Notes on the Genus Burroughsia (Verbenaceae).— The genus Burroughsia was erected by Moldenke (Phytologia 1:411, 1 940), based entirely on the presence of filament-like extensions of the connectives of the distal, abaxial pair of stamens (Fig. ld-0- These appendages have many multicellular glands at the tip and are slightly exserted from the corolla tube throat. Their function is unknown, but their glandular tips may serve to attract insects to the corolla tube. As erected, the genus consists of two species: Burroughsia appendiculata (Robins. & Greenm.) Moldenke (=Lippia appendiculata Robins. & Greenm.), of Chihuahua, Coahuila, eastern Durango, and northern San Luis Potosi and B.fastigiata (Brandegee) Moldenke (=L. fastigiata T. S. Brandegee), of Baja California del Sur. Except for the unique connective extensions the species fit well within the Lippia-Aloysia alliance. Both species are strigose, low subshrubs with opposite, small, ovate, lobed, deeply impressed- veined leaves; and both have flowers in short, cylindric, spike-like racemes borne on long axillary peduncles. The flowers have 2-lobed calyces, lavender to white, 5 -lobed, zygomorphic corollas with broad, cylindric tubes, and didynamous stamens; Fig. 1. Lippia appendiculata Robins. & Greenm.— a. Habit, showing section of thick, corky basal rhizome and erect ascending stems with long axillary peduncles and spicate inflorescences. —b. Leaf, showing lateral veins that extend to tooth sinuses. Veins are deeply impressed on upper surface.— c. Calyx is two-lobed, hirsute with spreading hairs and has many conspicuous orange glands.— d. Flower, lateral view, showing subtending bract, calyx (without vestiture), and corolla lobe orientation. Note exserted filament-like extensions of the distal anthers. —e. Diagrammatic "trans- parent" top view of flower, showing position of ovary, stamens, and corolla lobes. — f. Lateral view of stamens of a pre-anthesis flower showing the point of origin of the gland-tipped, filament-like extension of the distal anther connective. All from D. S. Correll and I. M. Johnston 21557 (LL). Magnifications as indicated. Drawing by Kathleen Cook. Madrono, Vol. 32, No. 3, pp. 186-190, 19 August 1985 1985] NOTES AND NEWS 187 the fruits are obovoid, 2-locular and dry. Burroughsia appendiculata is a relatively short plant, 1-1.5 dm tall, with pencil-thick, corky, horizontal rhizomes, rather strong- ly lobed leaves, uniform, antrorse, strigose stem vestiture accompanied by orange- red glands and a distinct yellow corolla eye (Fig. 1). Burroughsia fastigiata, on the other hand, is usually a taller, twiggier plant, with smaller, more crowded, fewer- lobed leaves, and generally denser, more curved vestiture that is retrorse on stems, antrorse on leaves, with light yellowish to colorless glands and corollas that lack a yellow eye (see illustration in Wiggins' Flora ofBaja California, p. 527, 1980). Several notes on the genus have been published by Moldenke (Phytologia 30:186-189, 1975; 40:423, 1978; 46:402, 1980) and these cite a number of additional references. Routine study of the two species in connection with the Chihuahuan Desert flora, however, showed that B. fastigiata lacks the filament-like extensions on the distal anther connectives — the very character upon which the genus was erected. It is perhaps surprising that this error was not noticed for over 45 years, but the species is restricted to Baja California and is seldom collected. Moldenke (in Shreve and Wiggins, Veget. Flora Sonoran Desert 2:1246-1247, 1964) separated Burroughsia in the key on the basis of "anthers appendaged" but did not mention the structures in the species description, though it is noted in the generic description. Wiggins (Fl. Baja California 1980) separates the genus in the key on the basis of the anther appendages, but correctly omits the appendages in his accompanying species illustration. Clearly B. fastigiata must be returned to the genus Lippia, as L. fastigiata T. S. Brandegee. One can argue phenetically to retain Burroughsia as a monotypic genus based on the character of a distinct filament-like connective extension. Cladistically, however, one sees that the genus is based entirely on a single, apomorphic feature and that its generic segregation cannot be supported. Within Lippia, the relationship of taxon appendiculata is not entirely clear. It shares a number of vegetative and floral characteristics with Lippia fastigiata and may indeed be most closely related to that taxon. In the Verbenaceae the presence of anther connective extensions is not restricted to B. appendiculata: similar, though less well defined extensions occur on the distal, abaxial filaments in some species of Glandularia. This, however, is hardly a synapomorphic character, because the genera differ in many other basic features. At the present time I can see no reason to retain Burroughsia as a distinct genus based on a single apomorphic character and suggest the two species be returned to Lippia as Lippia appendiculata Robins. & Greenm., of the Chihuahuan Desert region, and Lippia fastigiata T. S. Brandegee ofBaja California. Perhaps the time has come to look also into the validity of other generic segregates of Lippia such as Aloysia A. L. Juss., which is based on the presence of elongate inflorescences, and the low- growing Phyla Lour.— James Henrickson, Department of Biology, California State University, Los Angeles 90032. (Received 19 Nov 1984; accepted 16 Mar 1985.) New Combinations in California Chamaesyce (Euphorbiaceae).— As an addi- tional installment of nomenclatural changes (Hickman, Madrono 31:249-252. 1984) for a revision of W. L. Jepson's Manual (Jepson, Man. fl. pis. Calif 1925), four new combinations in the genus Chamaesyce are made for California taxa. Attention is also drawn to a nomenclatural change in the taxonomy of the genus from that of Wheeler (Rhodora 43:97-154, 168-286. 1941). Euphorbia s.l. encompasses a group of plants ranging from small temperate annuals to ten-meter-tall tropical trees. Within this diverse collection, several natural assem- blages can be recognized and are variously treated at generic, subgeneric or sectional 188 MADRONO [Vol. 32 levels. The generic delimitation of the tribe Euphorbieae to be followed in the new Jepson's Manual will be that of Webster (Taxon 24:593-601. 1975) in which Cha- maesyce is segregated from Euphorbia. Koutnik (S. African J. Bot. 3:262-264. 1984) recently reviewed the characteristics distinguishing Chamaesyce from Euphorbia. Of primary importance is the sympodial growth habit of Chamaesyce, which is not found in Euphorbia. Although the stem anatomy is not fully understood at present (Rosengarten and Hayden, Virginia J. Sci. 34:142. 1983), sympodial growth arises from the abortion of the apical meristem of the main stem after the first true leaves have formed. Subsequent growth is from lateral branches originating in the region of the cotyledonary nodes. Sympodial growth continues throughout the life of the plant: each terminal bud of a branch aborts and is alternately replaced by a bud from either side of the stem apex. The morphology of this pattern is explained in some detail (Verdus, Bull. Soc. Hist. Nat. Toulouse 99: 1 38-1 56. 1 964). There are no other species in the Euphorbiaceae that display this growth form. Another feature of Chamaesyce not found in Euphorbia is the occurrence of the C4 photosynthetic pathway (Downton, Photosynthetica 9:96-105. 1975). Associated with the C4 photosynthetic pathway is the Kranz anatomy of large chlorenchymatous cells forming a sheath surrounding the vascular bundle, also displayed by Chamaesyce species but not by Euphorbia. Additional distinguishing characters of Chamaesyce are the typically prostrate to ascending plant habit, the alternate arrangement of stem- branches, the opposite leaves, each with a discemibly asymmetric base, the presence of stipules, the frequent presence of white to pink petaloid appendages on the four (rarely five) involucral glands of the cyathium, and the ecarunculate seeds. All of these characters taken collectively should easily place an unknown specimen in the correct genus. One important departure from the nomenclature of Chamaesyce by Wheeler (1941) is the correct application of C. maculata (L.) Small (see Burch, Rhodora 68: 155-166. 1966). Chamaesyce maculata is the correct name for the common garden weed called "spotted spurge." This is a prostrate annual plant that is frequently given the name E. supina Raf (e.g., Munz, Calif fl. 1959; R. S. Calif 1974). The plant described under C. {E. ) maculata by Munz and Wheeler has ascending branches and is properly identified as C. nutans (Lag.) Small. The following four new combinations complete the generic assignment to Cha- maeysce for all the currently accepted taxa in California. The subspecific rank is used to indicate geographic unity of the taxon and to maintain uniformity within the existing taxonomy for the California members of Chamaesyce. Chamaesyce abramsiana (Wheeler) Koutnik comb, now.— Euphorbia abramsiana Wheeler, Bull. S. CaHf Acad. Sci. 33:109. 1934. - Euphorbia pediculiferaEngelm. var. abramsiana Ewan in Jeps., Fl. Calif 2:427. 1936.— Type: CA, Imperial Co., Heber, Imperial Valley, Jun 1904, Abrams 4097 (DS). Chamaesyce hooveri (Wheeler) Koutnik comb, now .—Euphorbia hooveri Wheeler, Proc. Biol. Soc. Wash. 53:9. 1940. -Type: CA, Tulare Co., Yettem, 30 Jun 1937, Hoover 2583 (GH). Chamaesyce ocellata subsp. rattanii (S. Watson) Koutnik comb, now .—Euphorbia rattanii S. Watson, Proc. Amer. Acad. Arts 20:372. \^^5. — Chamaesyce rattanii Millsp., Publ. Field Columbian Mus. Bot. Ser. 2:41 1. 1916.— Euphorbia ocellata var. rattanii Wheeler, Bull. S. CaHf Acad. Sci. 33:107. 1934.-Type: CA, Glenn Co., Stony Cr., Jun 1884, Rattan 57 (GH). Chamaesyce serpyllifolia subsp. hirtula (Engelm. ex S. Watson) Koutnik comb. nov. — Euphorbia hirtula Engelm. ex S. Watson, Bot. Calif 2:74. ISSO. — Chamaesyce hirtula Millsp., Publ. Field Columbian Mus. Bot. Ser. 2:409. \9\6. -Euphorbia serpyllifolia var. hirtula Wheeler, Proc. Biol. Soc. Wash. 53:11. 1940.— Type: CA, San Diego Co., Talley's, Cuyamaca Mts., 1875, Palmer 451 (GH). 1985] NOTES AND NEWS 189 The following is a list of the currently accepted Chamaesyce taxa in California: C. abramsiana (Wheeler) Koutnik C. albomarginata (Torrey & A. Gray) Small C. arizonica (Engelm.) Arthur C. fendleri (Torrey & A. Gray) Small C. glyptosperma (Engelm.) Small C. hooveri (Wheeler) Koutnik C. maculata (L.) Small C. melanadenia (Torrey) Millsp. C. micromera (Boiss.) Wooton & Stan- dley C. nutans (Lag.) Small C. ocellata (E. M. Durand & Hilgard) Millsp. subsp. ocellata C. ocellata subsp. arenicola (Parish) Thome C. ocellata subsp. rattanii (S. Watson) Koutnik C. parishii (Greene) Millsp. C. parryi (Engelm.) Rydb. C. pediculifera (Engelm.) Rose & Stan- dley C. platysperma (Engelm. ex S. Watson) Shinners C. polycarpa (Benth.) Millsp. var. poly- carpa C. polycarpa var. hirtella (Boiss.) Millsp. C. prostrata (Alton) Small C. revoluta (Engelm.) Small C. serpens (H.B.K.) Small C. serpyllifolia (Pers.) Small subsp. ser- pyllifolia C. serpyllifolia subsp. hirtula (Engelm. ex S. Watson) Koutnik C. setiloba (Engelm. ex Torrey) Millsp. C. vallis-mortae Millsp. I thank the two anonymous reviewers and the editor for their helpful suggestions. — Daryl L. Koutnik, Research Assistant, Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166. (Received 22 Oct 1984; accepted 19 Apr 1985.) Rediscovery and Reproductive Biology of Pleuropogon oregonus (Poaceae).— Pleuropogon oregonus Chase (Oregon semaphore grass) was first collected in 1886 by W. C. Cusick in Hog Valley, probably near Union, in northern Oregon. In 1901, another collection of P. oregonus was made by A. B. Leckenby in Union, Oregon; and in 1936 M. E. Peck found it again, but in swampy ground 25.8 km west of Adel, Lake County, Oregon. Because P. oregonus has not been collected for nearly half a century and is reported as extinct or endangered (Smithsonian Rept. to Congress, Serial No. 94-A, 1975; Aysenu and DeFilipps, Endang. Threat. PI. U.S., Smithsonian Inst, and World Wildlife Fund, Wash., D.C., 1978; Siddall, Chambers and Wagner, Rare, Threat. Endang. Vase. PI. Oregon, Oregon Nat. Area Preserves Advisory Com- mittee, 1979; U.S. Fish Wildlife Serv., Fed. Reg. 45(242):82480-82569, 1980), its recollection is worthy of note. Oregon, Lake Co., ca. 25 km w. of Adel on Hwy. 140, T39S, R22E, Sec. 5 nw.'A and T38S, R22E, Sec. 32 sw.iA. J. Kagan 60482 (ORE), 4 Jun 1979. Very probably the same locality where Peck made the last previous collection, 47 years ago. Habitat. Restricted to sluggish water in depressions and sloughs fed by Mud Cr. on both sides of Hwy. 140, on gravelly silt loam or clay. It grows in association with various grasses and sedges, including Beckmannia syzigachne, Deschampsia dan- thonioides, Glyceria borealis, Hordeum brachyantherum, Poa nevadensis, Carex an- throstachya, C. nebraskensis, and Eleocharis palustris. The meadow area, including the portion occupied by P. oregonus, has been used for years for fall grazing. Reproductive biology. Oregon semaphore grass blooms from early June to late July and fruits from late July to mid-August. Its inflorescence is a simple, erect raceme. 190 MADRONO [Vol. 32 1 3-20 cm long, bearing 6-7 spikelets. Pedicels are 2-5(-l 2) mm long. Spikelets spread toward one side of the raceme, 2-4(-5) cm long, each bearing 7-14 florets. Bentham and Hooker f. ( 1 883, Genera Plantarum) described the florets of the genus Pleuwpogon as "hermaphroditis v. summo masculo." However, the uppermost floret of P. ore- gonus is usually reduced. The upper florets are pistillate, whereas the lower ones are perfect. Anthesis within each gynomonoecious spikelet is protogynous, starting with the upper pistillate flowers and then progressing to the lowest protandrous, her- maphroditic flowers, then upward. Gynomonoecy and overall protogyny in spikelets but protandry in hermaphroditic florets found in P. oregonus are also observed in P. californicus (Bui, Systematics of Pleuwpogon R.Br. (Poaceae), Ph.D. diss., U.C. Berke- ley, 1977). Connor (1979, Breeding systems in the grasses: a survey. New Zealand J. Bot. 17:547-574) noted that gynomonoecism is uncommon among the Gramineae. Tests of the pollen, using four enzyme systems (malate dehydrogenase, isocitrate dehydrogenase, succinate dehydrogenase, and monoamine oxidase) showed 87% vi- ability (I. Baker, pers. comm. 1983). Low fecundity may contribute to its rarity. Of 4645 florets inspected, only 494 bore caryopses. Germinability test of a random sample of 30 caryopses (8 months old) with 0.1% tetrazolium salt solution showed 85% viability. Although P. oregonus should no longer be considered 'extinct,' we suggest that it should remain classified as endangered. We thank Lincoln Constance and Lawrence Heckard for assistance and encour- agement in this study, Irene Baker for the pollen viability tests, and Tammy But and Lawrence But for field assistance.— Paul P. H. But, The Chinese Univ. of Hong Kong, Shatin, N.T. Hong Kong; Jimmy Kagan, The Nature Conservancy, 1234 Northwest 25th Avenue, Portland, OR 97210; Virginia L. Crosby, U.S.D.I., Bureau of Land Management, Denver Service Center, Denver Federal Center, Bldg. 50, Denver, CO 80225; and J. Stephen Shelly, Dept. Botany, Oregon State Univ., CorvaHis 97331. (Received 25 Oct 1984; accepted 25 Mar 1985.) NOTEWORTHY COLLECTIONS Arizona CoRCHORUS HIRTUS L. (Tiliaceae). — Cochisc Co., San Bernardino Ranch, at T24S R30E S14 SW.V4, 1 160 m, grassy w.-facing slope of mesa near inlet to House Pond, deep alluvium; 21, 22 Aug 1981. Marrs-Smith 855, 945 (ASU). Significance. First definite record for AZ. Pringle's 1 884 collection gives the locality only as sandy plains near the Mexican Boundary. Ibervillea tenuisecta (Gray) Small (Cucurbitaceae).— Chochise Co., 0.65 km n. of Geronimo Trail, 7.7 km e. of Douglas, just n. of Douglas Hill at T24S R28E SIO se.V4, 1325 m, ne.-facing slope of rocky hmestone hillside, under Larrea divaricata; 21 Aug 1981, T. Van Devender and J. B. Iverson s.n. (ARIZ). San Bernardino Ranch, 20 m n. of ranch at T24S R30E S12 sw.^A, 1158 m, under Larrea divaricata in Scleropogon brevifolius grassland, deep alluvium; 26 Oct 1981, Marrs-Smith 1196 (ASU). (Verified by D. J. Pinkava, ASU.) Significance. New records for AZ. Seeds show some intermediacy in size and texture toward /. sonoreae. —Gwue Marrs-Smith, Biological Sciences Center, Desert Re- search Institute, P.O. Box 60220, Reno, NV 89506. California JuNCus CYPEROiDES Laharpc (Juncaceae).— CA, Butte Co., Forbestown Ridge ca. 18 km e. of Oroville and ca. 5.3 km sw. of Forbestown, e. of and adjacent to Black Bart Rd. ca. 0.8 km sw. of its junction with Oroville-Forbestown Hwy. (T19N, R6E, SI 8 SW.V4 se.'A). Shaded n.-nw. slope of a low-elevation yellow pine forest. Known from this locality since 1981. Plants produce viable seed and viviparous vegetative shoots occasionally arise in the inflorescence. Several nearby seepages and meadows have been searched but no other populations were found. Ahart 2925, 3058 and 3527 (CAS, CHSC); Jokerst and Ahart 1754 (CHSC). Determination confirmed by J. T. Howell. Significance. First known collection from North America (consulted CAS, MO, NY) representing ca. 5600 km range extension nw. from populations in the Andes of Colombia. It is also known from Argentina, Chile, Equador, and Peru, where elevations range from ca. 2000-4000 km, except in Chile where populations extend down to sea level (Balslev 1982, A monograph of neotropical Juncaceae, Ph.D. diss.. City Univ. New York). The method of introduction is unknown and it has not expanded its range or population boundaries since 1981.— James D. Jokerst, Route 7, Box 3 12c, Oroville, CA 95965. Colorado Draba apiculata C. L. Hitchcock (Brassicaceae).— Lake Co., Sawatch Mts., Mt. Campion Basin, granitic rock outcropping, 3750 m, 27 Jul 1 984, Hartman & Rottman 6025 (COLO, UC). Significance. First report for CO. The species is known from mountainous areas of n. WY and the Uintah Mts. of UT. We concur with Rollins (Contr. Gray Herb. 214:6, 1984) in treating D. apiculata as distinct from D. densifolia Nutt. and D. daviesiae (HiXchc.) Rollins.— Robert A. Price, Botany Dept., Univ. California, Berkeley 94720; Mary Lou Rottman and Emily Hartman, Biology Dept., Univ. Colorado, Denver 80202. Mexico PiNus PATULA var. longepedunculata Loock (Pin aceae)— Mexico, Oaxaca, Sierra Madre del Sur, 2600 m, near 16°30'N, 97°10'W, 17 Feb 1984, Perry Mex. 3884 (NCS, Perry herbarium, to be distributed). Forest on 60% ne. slope with Pinus ayacahuite Madrono, Vol. 32, No. 3, pp. 191-193, 19 August 1985 192 MADRONO [Vol. 32 Ehrenb., Pinus pseudostrobus hind., Pinus montezumae Lamb., Abies sp., and Quercus spp. Verified by J. P. Perry, Jr., Feb. 1984. Previous knowledge. Known from mountains near the village of Guajimaloyas, Oaxaca, Mex. 2800 m (near 17°15'N, 96°15'W) and e. and w. of Hwy. 175 near the town of Ixtlan. Also from mountains in area around (10-20 km) the city of San Cristobal de las Casas, Chiapas, Mex. at 2200-2400 m. (Herbaria consulted: MEXU, A, CHIP, NCS; published sources: M. Martinez, Los Pinos Mexicanos, 1948; E. E. M. Loock. The pines of Mexico and British Honduras, 1950; N. T. Mirov, The Genus Pinus, 1967; W. H. G. Barrett, Variacion de caracteres morfologicas en poblaciones naturales de Pinus patula Schlecht. et Cham, en Mexico, 1972; W. B. Critchfield and E. L. Little, Jr., Geographic distribution of the pines of the world, 1966.) Significance. First record in the Sierra Madre del Sur, ca. 100 km disjunction from populations found in the mountain ranges north of the city of Oaxaca. Special thanks are expressed here to W. S. Dvorak, Dir. CAMCORE (Central America and Mex. Conifer Resources Coop.) and the field staff J. Donahue and Miguel Munoz for providing transportation and help in locating this isolated population.— J. P. Perry, Jr., Assoc. Dir. Agri. Scs. Program, The Rockefeller Foundation (Retired), 306 Front St., Hertford, NC 27944. New Mexico Grayia brandegei a. Gray (Chenopodiaceae). — San Juan Co., "bad lands" along highway [NM Hwy. 57, formerly 56] w. of Otis Trading Post [T24N RlOW, ca. 12 km by air nw. of Nageezi], 29 Aug 1932, Nelson and Nelson 296 (RM); Kutz Canyon near junction of East and West Forks (T27N RlOW SI 8), uncommon on steep shale slopes with Atriplex and Juniperus, 1800 m, 23 Jul 1976, Levin 993 (ARIZ, DAV); Cottonwood Arroyo, 4 km w. of NM Hwy. 170, T30N R14W S24, n. slope of sand/ clay badlands, 19 Aug 1984, Porter 4665 (SD). Significance. First records for NM. This rarely collected species was previously known from sw. CO, se. UT, and ne. AZ, generally on shale. The long time span between collections probably reflects the relative inaccessibility of the species' habitat and the plant's strong resemblance to the much more abundant Atriplex canescens, which the Nelsons had determined their specimen to be. I thank Ken Heil for calling my attention to the Porter specimen.— Geoffrey A. Levin, Botany Dept., San Diego Natural History Museum, San Diego, CA 921 12. Rhynchelytrum repens (Willd.) C. E. Hubb. (Poaceae).— Luna Co., Florida Mts., Copper Kettle Spring (T26S R8W SI 3 ne.V4 se.'A), frequent on gentle sw.-facing gravelly rhyolitic slope, 1450 m, 9 Nov 1984, Mcintosh and Bevacqua 1637 (NMC, NY, RSA). Significance. First record for New Mexico. Closest known records are in se. Arizona and w. Texas.— Laird McIntosh, Bureau of Land Management, Las Cruces, NM 88004. Wyoming Draba spectabilis var. oxyloba (Greene) Gilg & Schulz (Brassicaceae).— Carbon Co., w. slope of Sierra Madre w. of Encampment, spruce-fir forest along Battle Creek, 2560 m, 15 Jul 1966, C. L. and M. W. Porter 10216 (UC). Significance. First report for WY. Hitchcock (Univ. Wash. Publ. Biol. 1 1:44, 1941) cited a collection from the same area (Battle, Carbon Co., Tweedy 4475; NY, US), but omitted the name of the state and listed the range of the species as UT and sw. CO. Subsequent collections from WY were misidentified as D. aurea M. Vahl and the species was omitted from the recent Flora of Wyoming. The species occurs in the Abajo and La Sal Mts. of e. UT (var. spectabilis), the Lukachukai Mts. of ne. AZ and more broadly on the w. slope of the Rocky Mts. in CO. It is apparently quite uncommon in nw. CO (occurring in Routte and Garfield cos.) and reaches the extreme 1985] NOTEWORTHY COLLECTIONS 193 n. limit of its range in WY. — Robert A. Price, Botany Dept., Univ. Calif., Berkeley 94720. REVIEWS Vascular Plants of Montana. By Robert D. Dorn, illustrations by Jane L. Dorn. Mountain West Publishing, P.O. Box 1471, Cheyenne, WY 82003. 1984. $7.95 + $1.00 shipping. Dom's is a new and needed flora to satisfy both the herbarium and, particularly, the field botanist. The book (in paper covers) is small and light, well and pertinently, but sparsely, illustrated with nice line drawings, fully indexed to species, families and genera within families (arranged alphabetically), and has a well-illustrated glossary. It has a map of Montana counties and their combination into six areas used to describe plant distributions, and a very valuable map of the scattered distributions of the six kinds of floras found in Montana (Alpine, Rocky Mt., Pacific Northwest, Palouse Prairie, Great Basin, and Great Plains). The latter scheme is more meaningful, but the former is used in the text. Dot maps would offer a unique opportunity to document the latter, floristic classification. The book is a masterpiece of condensation and composition. Does a word processor do better what a press once did? The book has only 276 pages, with a format of 13.7 X 21.6 X 1,4 cm and a corresponding light weight. Anyone who has carried Munz's California flora or Hulten's Alaska flora for several hundred miles on foot will appreciate Dom's successful effort. Keys are full and serve as descriptions. They seem excellent. Aquatic and woody plants have their own separate keys, and for the Brassicaceae, Apiaceae, and Astragalus there are both flowering and fruiting keys. The characters used are simple, explained, evident, and separable. There are concise morphological descriptions for families and genera so a check can be made before keying to the species level. References in concise form are given for almost all genera. English names are supplied, but invention is not run into the ground. The habitat notes are concise and informative, not schematic. The illustrations are excellent in quality, informative, and helpful. Again, plants know no political boundaries. The literature on many "North Amer- ican" plants includes much published outside North America. Numerous Montana plants are also Alaskan, yet Alaska was physically and biologically a part of north- easternmost Asia and separated by vast ice sheets from North America during each glacial. The Soviet botanical literature is a rapidly expanding, very valuable mine of information on many North American plants. Hulten's Alaskan flora has been, and still is, a source of name corrections that are a prerequisite to an improved ecological and taxonomic understanding of Rocky Mountain plants, including several in Mon- tana. The literature on the distribution, ecology, uses, etc. of several Montana plants is hidden by some of the names Dorn uses (Kobresia bellardii for K. myosuroides, Eriophorum polystachion for E. angustifolium, Carex stenophylla or C. eleocharis for C. duriuscula, Calamagrostis canadensis for C. purpures are examples). However, this is a carping criticism. Dorn has brought the nomenclature and taxonomy of Montana plants up to date for the most part (cf. Alnus viridus and A. incana, Artemisia tridentata Madrono, Vol. 32, No. 3, pp. 193-195, 19 August 1985 194 MADRONO [Vol. 32 and its relatives, Sphaewmeria, Triticeae, Leucopoa, Salix, and Heracleum, among others). A similarly excellent flora of the Black Hills of Wyoming and South Dakota was published earlier by the Doms (1977) and is a real bargain ($1.50 if ordered with the Montana flora). This "local" flora of 1260 species contains 80% of the flora of South Dakota and 59% of the flora of Wyoming.— Jack Major, Botany Department, Uni- versity of California, Davis 95616. Aven Nelson of Wyoming. By Roger L. Williams, xii + 407 pp. Colorado Associated University Press, Boulder, CO 80309. 1984. $29.50. ISBN 0-87081-147-9. This scholarly book is more than a biography of Aven Nelson. It is a brief history of systematic botany from about 1 895 to 1 945. It describes university life, particularly in the West, for this same period, and it touches on state politics. It also includes abbreviated biographies of many of Nelson's students. The author, a historian specializing in French history, has taken the time to become familiar with the science of systematic botany to the point of personally collecting and identifying plants. This familiarity is reflected in his writing, yet he frequently provides helpful comments for those who are not taxonomists. Those who know, or know of, the personalities in systematic botany from Nelson's time will easily relate to the book. Those who do not may find parts somewhat dry reading but should find the author's periodic reflections of interest. One example: "The posture of perfectionism, in academia, usually masks either indolence or fear." Teachers and researchers who read the book will see how well off" they really are. Students can discover the qualities for success, qualities that are not taught in the university. Aven Nelson was one of the first six faculty members at the newly-established University of Wyoming in 1 887. His formal training was in English, but the University mistakenly hired two English professors. He had an interest in natural history, so he was appointed Professor of Biology. He founded the Rocky Mountain Herbarium and developed it in his spare time to be the largest in the interior West and one recognized world-wide. Nelson was instrumental in founding the Colorado- Wyoming Academy of Science and the American Society of Plant Taxonomists and served as the first president of each. He was the first interior westerner to be elected president of the Botanical Society of America. He was forced to retire at age 83 when the university instituted a mandatory limited service plan. Nelson was rewarded for his 55 years of service to the school, including 5 years as its president, with a pension of $ 1 500 per year. The book also gives insight into personalities of other systematic botanists of Nelson's time including E. L. Greene, P. A. Rydberg, B. L. Robinson, M. L. Femald, N. L. Britton, J. N. Rose, M. E. Jones, J. M. Coulter, L. M. Underwood, W. Trelease, T. S. Brandegee, and A. A. Heller. It provides a basic history of the development of an International Code of Botanical Nomenclature, a valuable asset for students ma- joring in systematics. The controversies are portrayed better here by Williams than they are, for example, in G. H. M. Lawrence's treatment in Taxonomy of Vascular Plants. The battle between the conservatives at the Gray Herbarium and the New York "radicals" frequently surfaces. There is also the struggle of western botanists trying to break free from the dominance of the eastern schools, just as the eastern botanists struggled with Europeans at an earlier date. Once, when M. L. Femald strongly suggested to Nelson that he should come back East for study before describing new species and offered an invitation. Nelson wrote back, "I wish it were possible for me to accept it at once and to spend much time within the walls of the Gray Herbarium .... One cannot always do as he would, but must do as he can." In compiling the New Manual of Botany of the Central Rocky Mountains (1909), Nelson was forced to rethink the matter of what constitutes a species for all practical 1985] REVIEWS 195 purposes. He moved toward conservatism by reducing to synonymy 1788 species names. Most botanists were becoming concerned with the excessive splitting of species that was taking place, so Nelson's lead was well received. The book is concluded with two appendices listing Nelson's proposed new genera, species, and varieties along with their current status and his botanical and horticultural publications. Biographies of earlier botanists like Thomas Nuttall and Asa Gray help us under- stand botanical activities of the 19th century. Williams has given us an excellent addition for a succeeding period. — Robert D. Dorn, P.O. Box 147 1 , Cheyenne, WY. Subscriptions— Membership Membership in the California Botanical Society is open to individuals ($ 1 8 per year; students $ 10 per year for a maximum of seven years). Members of the Society receive Madrono free. Family memberships ($20) include one ten-page publishing allot- ment and one journal. Emeritus rates are available from the Corresponding Secretary. Institutional subscriptions to Madrono are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Madrono should be sent to the Corresponding Secretary. Information for Contributors Manuscripts submitted for publication in Madrono should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For ar- ticles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed ending with figures, and number each page. Do not use a separate cover page, "erasable" paper, or footnotes. Manuscripts prepared on dot matrix printers may not be considered. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including cap- tions) to the Madrono page, and designed for reduction to % original size. Scales should be included in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff card- board and sent flat. No illustrations larger than 22 x 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in Madrono 22:288. 1974. In- stitutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Herbariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Periodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted ten pages in Madrono each year or 20 pages over a two-year period. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $60.00 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing Madrono on a reasonable schedule, with equity among all members for access to its pages. Printer's fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author's changes after typesetting @ $3.00 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. CALIFORNIA BOTANICAL SOCIETY VOLUME 32, NUMBER 4 OCTOBER 1985 t. MADRONO A WEST AMERICAN JOURNAL OF BOTANY Contents Sand Dune Flora of the Great Basin and Mojave Deserts of California, Nevada, and Oregon Bruce Michael Pavlik Addenda to the Vascular Flora of San Luis Obispo County, California David J. Keil, Robert L. Allen, Joy H. Nishida, and Eric A. Wise A Revised Vascular Flora of Tumamoc Hill, Tucson, Arizona Janice E. Bowers and Raymond M. Turner The Alpine Vascular Flora of Three Cirque Basins in the San Juan Mountains, Colorado Emily L. Hartman and Mary Lou Rottman NOTES AND NEWS Notes on the Salvia leucophylla Complex (Lamiaceae) of California and Baja California Norte Kurt R. Neisess 197 214 225 253 REVIEWS ANNOUNCEMENTS REVIEWERS OF MANUSCRIPTS EDITORS' REPORT FOR VOLUME 32 INDEX TO VOLUME 32 MEETING PROGRAM FOR 1985-1986 273 276 224, 252, 272, 280, 281 282 283 284 289 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRofio (ISSN 0024-9637) is published quarterly by the California Botanical So- ciety, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to James R. Shevock, Botany Dept., California Academy of Science, San Francisco, CA 941 18. Editor— V^AYTsiE R. Ferren, Jr. Associate Editor— J. Robert Haller Department of Biological Sciences University of California Santa Barbara, CA 93106 Board of Editors Class of: 1985— Sterling C. Keeley, Whittier College, Whittier, CA Arthur C. Gibson, University of California, Los Angeles 1986— Amy Jean Gilmartin, Washington State University, Pullman Robert A. Schlising, California State University, Chico 1987— J. Rzedowskj, Instituto de Ecologia, A.C., Mexico Dorothy Douglas, Boise State University, Boise, ID 1988— Susan G. Conard, USDA Forest Service, Riverside, CA William B. Critchheld, USDA Forest Service, Berkeley, CA 1989— Frank Vasek, University of California, Riverside Barbara Ertter, University of California, Berkeley CALIFORNIA BOTANICAL SOCIETY, INC. Officers for 1985-86 President: Charles F. Quibell, Department of Biology, Sonoma State University, Rohnert Park, CA 94928 First Vice President: Rodney G. Myatt, Department of Biological Sciences, San Jose State University, San Jose, CA 951 14 Second Vice President: David H. Wagner, Herbarium, Department of Biology, University of Oregon, Eugene, OR 97403 Recording Secretary: V. Thomas Parker, Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: James R. Shevock, Department of Botany, California Academy of Sciences, San Francisco, CA 941 18 Treasurer: Wayne Savage, Department of Biological Sciences, San Jose State Uni- versity, San Jose, CA 95192 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, Rolf W. Benseler, Department of Biological Sciences, California State University, Hayward, CA 94542; the Editor of Madroisio; three elected Council Members: James C. Hickman, Department of Botany, Uni- versity of California, Berkeley, CA 94720; Thomas Fuller, 171 Westcott Way, Sacramento, CA 95864; Annetta Carter, Department of Botany, University of California, Berkeley, CA 94720; and a Graduate Student Representative, Joseph M. DiToMASO, Department of Botany, University of California, Davis, CA 95616. SAND DUNE FLORA OF THE GREAT BASIN AND MOJAVE DESERTS OF CALIFORNIA, NEVADA, AND OREGON Bruce Michael Pavlik Department of Biology, Mills College, Oakland, CA 94613 Abstract A floristic survey of 20 stabilized and unstabilized sand dunes in the Great Basin and Mojave deserts was conducted over a 9-year period. Plants were tallied as present only if they were observed growing on deep, windblown accumulations of sand. A total of 166 taxa was documented, representing 88 genera and 27 families of vascular plants. Relative to other desert areas in Nevada and California, the dune flora was enriched with members of Asteraceae, Fabaceae, Chenopodiaceae, and Polemonia- ceae and somewhat deficient in Poaceae. Approximately 95% of the taxa were in- digenous. In comparison with Death Valley and eastern California, the dunes were overrepresented with annuals and geophytes and underrepresented with hemicryp- tophytes, chamaephytes, and stem succulents. Only a third of the taxa can be con- sidered as strictly psammophytic and one-sixth as halophytic. Five taxa are endemic to one or a few dunes; 1 2 are sand-restricted but less limited geographically than the endemics. Ecological studies on the deserts of North America have concen- trated on habitats that are widespread or readily accessible. Spe- cialized habitats, such as limestone outcrops, lava flows, mountain- sides, wetlands, and sand dunes, have been largely ignored. These may, in fact, constitute separate ecosystems— islands isolated by their physical and biological properties or by factors related to the evolution of the landscape through geologic time. Desert sand dunes, defined as windblown accumulations of sand that exist independent of small-scale, fixed-wind obstructions (Bag- nold 1933), can be distinguished ecologically from surrounding hab- itats by a unique set of extreme physical characteristics. These in- clude an intense energy environment (Pavlik 1980), a mobile and abrasive substrate (Bagnold 1941, Evans 1962, Sharp 1966) having coarse texture and high porosity (MacDonald 1970, Dean 1978, Pavlik 1979a) and relatively mesic substrate conditions (Went and Westergaard 1949, Norris and Norris 1961, DeDecker 1976). Plants growing on dunes can maintain a higher internal water status than non-dune plants under the same microclimatic conditions (Pavlik 1980, Toft and Pearcy 1982). Desert dunes are associated invariably with the playas, intermittent water courses, and remnant lakes of arid, lowland basins. These basins and sand dunes are often isolated from one another by high mountain ranges that coalesce and divide MadroSio, Vol. 32, No. 4, pp. 197-213, 20 December 1985 198 MADRONO [Vol. 32 in complex physiographic patterns, especially in Nevada and eastern California. The insular nature of dune habitats, as demonstrated by studies of dune-dwelling insects (D. Giuliani, pers. comm.) and mammals (Brown 1973), has yet to be established rigorously from comprehensive floristic data. Vasek and Barbour (1977) concluded that very few data are avail- able on the floristics and plant ecology of desert sand dunes. The information is confined either to descriptions of single dunes (Rem- pel 1936, DeDecker 1976, Pavlik 1979a, Thome et al. 1981) or studies in relation to off-road vehicle impact (Westec 1977, Pavlik 1979b, Bury and Luckenbach 1983). The present study documents the dune flora at 20 locations in the Great Basin and Mojave deserts of California, Nevada, and Oregon as part of an ongoing contribution to the plant ecology and phytogeography of this habitat. Methods Approximately 34 major sand dune systems are known to occur in the desert regions of California, Nevada, and Oregon (Fig. 1). This tally does not include systems dominated by sand sheet, ob- stacle or shrub-coppice formations (Dean 1978), or small, senescent dunes in advanced stages of stabilization. (Stabilization is indicated by decreasing substrate and dune mobility, high concentrations of silt and clay particles in a saline-alkaline substrate, and uniformity of vegetation between dune and non-dune sites.) Only 20 of the locations shown in Fig. 1 were included in this survey, but many of the others were visited once. Excluded were dunes with a strong floristic affinity to sites in the Colorado and Sonoran deserts (see Johnson 1977, 1982) and several low, semi-stabilized or stabilized dunes found commonly throughout the Great Basin. With respect to regional vegetation, seven of the dunes included in the study are associated with Great Basin, ten with Mojave Desert, and three with transitional community types (Beatley 1976) (Table 1). Twelve dunes may be described as massive and unstabilized, whereas the remain- der are relatively small, low, and semi-stabilized. At all sites the dune substrate is a deep, coarse-grained, quartz-dominated, and finely-sorted aeolian sand. The negative impact of off-road vehicles on the native flora is probably significant at three sites, although some undisturbed habitat with plant cover could be found at each of these. Further information on desert sand dunes in California may be found in Dean (1978). Each site was surveyed intensively on at least three occasions from 1976 to 1984. This period included several years with average or above average amounts of fall and winter precipitation (1975-76, 1977-78, 1979-80, 1982-83) and, consequently, an abundance and diversity of annuals. Only species found growing on dunes sensu 1985] PAVLIK: DESERT SAND DUNE FLORA 199 Fig. 1, Locations of major desert sand dune systems in California, Nevada, and Oregon. Dunes included in this study (solid circles) are numbered according to Table 1. 200 MADRONO [Vol. 32 1^ PQ ^ .2^ *3 N g s - '5b »-i 5 sag >-) . t« "I^ (U «-i c/5 N a p ^ 1^ + c + II ■s + U5 W H iS "-^ u- S £2 i-: S |§ = H 13 <<< uuu < «1 >.>.^>. ^ b >^ S § S J3 o 1/3 1/5 C/3 C/5 C/D :3 _ O 03 1/5 ^ o 00 d d 1985] PAVLIK: DESERT SAND DUNE FLORA 201 stricto (in areas having enough accumulated sand to affect local topography) were noted as present and collected. This insured that taxa were related ecologically to dune conditions rather than to ecotonal or non-dune conditions. Voucher specimens were deposited with the Herbarium of the University of California, Davis (DAV), except those of Gilia (deposited at CAS). In addition to the field survey, collection data were obtained from searches conducted at CAS, DAV, RSA, UC, UNR, and the private herbarium of M. DeDecker (Independence, California). These records were used to add taxa to the flora only if the label contained sufficient information to establish that the collection was made on a specific dune. Oth- erwise, the specimen was simply noted and used to describe the substrate fidelity and general distribution of the taxon. Plant lists for individual dune systems may be obtained from the author. Un- confirmed taxa that may occur on the dunes in this study have been listed in Appendix I. These were not included in subsequent analyses of the flora. Nomenclature is based on Kartesz and Kartesz (1980) and the authorities to whom specimens were sent (see Acknowledg- ments). Subsequent analyses of the dune flora were made by annotating each taxon with respect to 1 ) relative occurrence among the twenty dunes (rare = present at 1 or 2 dunes, uncommon = present at 3-5 dunes, common = present at 6 or more dunes); 2) Raunkier life form (annual, Geo = geophyte, Hemi = hemicryptophyte, Cham = chamaephyte, P-shrub = phanerophyte shrub, Succ = stem succu- lent); 3) fidelity to substrate texture (Ps = sand-restricted psam- mophyte, Ct = a plant of coarse-textured substrates including rock, gravel, and sand. At = a plant of many substrate textures including rock, gravel, sand, silt, and clay); 4) occurrence on saline or alkaline substrates (indicated by -S following the texture designation, e.g. At- S); and 5) geographic center of distribution among the dunes of this study (GB = dunes of the Great Basin Desert, M = dunes of the Mojave Desert, W = widespread among GB and M dunes, T = dunes that lie in a transition zone between the Great Basin and Mojave deserts). Raunkier life forms are defined in terms of the relationship between perennating buds and the soil surface (Radford et al. (1974) present the criteria used in this study). This relationship is complicated by sand movement in geophytes and hemicrypto- phytes that produce buds near (usually ±10 cm) the dune surface. For this reason, taxa had to possess a specialized, subterranean organ (e.g. rhizome, creeping rootstock, bulb) in order to be designated as geophytes, and hemicryptophytes were identified by a rosette or tussock growth form. The overlapping categories of substrate fidelity attempt to segregate sand-restricted plants (those Ps taxa that are dependent on the moisture and ionic properties of coarse-textured substrates, yet tolerant of sand movement, intense solar radiation, 202 MADRONO [Vol. 32 and low nutrient availability) from the wider ranging Ct plants (de- pendent on moisture and ionic conditions of coarse substrates but less tolerant of sand movement, intense solar radiation, and low nutrient availability) and At plants (independent of moisture and ionic conditions, and mostly intolerant of the aforementioned char- acteristics of unstabilized dunes). It should be emphasized that sub- strate fidelity may vary within the range of widely distributed species, especially those not confined to deserts. The substrate designations were assigned to reflect the principal occurrences of taxa as observed in the Great Basin and Mojave deserts. Similarly, the geographic descriptors (GB, M, W, and T) were used to describe the distribution of species within the study area and do not indicate the entire range. This was done to facilitate the analysis of phytogeographic patterns in subsequent research. All annotations were determined by the concurrence or consensus of many floras, taxonomic monographs, ecological studies, herbarium labels, and field observations. Discussion One hundred and sixty-six taxa were found among the 20 dunes included in this study, with 88 genera and 27 families represented. The 159 indigenous taxa were collected within a total area of about 310 km^, resulting in an index of taxon density that is low (0.51 taxa/km^) when compared to other plant assemblages in eastern California (0.36-1.56, Thome et al. 1981). The dune flora was en- riched with members of Asteraceae, Fabaceae, Chenopodiaceae, and Polemoniaceae and somewhat deficient in Poaceae (21%, 1 1%, 8%, 7%, and 7% of the taxa, respectively. Table 2). This is in comparison with larger, more ecologically diverse desert floras, such as that of the Nevada Test Site (17%, 6%, 4%, 4%, and 9%, Beatley 1976), eastern California (16%, 4%, 2%, 5%, and 7%, Thome et al. 1981) and the entire Califomia Desert Conservation Area (15%, 8%, 3%, 4%, and 9%, BLM unpublished data 1980). The diversity of the Asteraceae was due to a large number of mono- or bi-typic genera, whereas that of the Fabaceae and Polemoniaceae was due to a few genera with large numbers of species and infraspecific taxa (e.g. Astragalus and Gilia). Although only 7 introduced taxa were ob- served, these sometimes constituted a significant proportion of the vegetation cover (especially Salsola) at any one dune. The life form spectrum of desert dune plants (Table 3) was very different from other desert spectra in the literature, thus supporting the recognition of a separate psammophytic vegetation type (Thome et al. 1981 ). Dunes were overrepresented with annuals and geophytes and underrepresented with hemicryptophytes, chamaephytes, and stem succulents when compared with the desert vegetation of Death Valley (42, 2, 19, 13, and 3% respectively, Raunkier 1934) and 1985] PAVLIK: DESERT SAND DUNE FLORA 203 Table 2. The 8 Largest Families and 9 Largest Genera of the Sand Dune Flora of the Great Basin and Mojave Deserts. Family Taxa Genus Taxa Asteraceae 34 Astragalus 10 Fabaceae 19 Gilia 8 Chenopodiaceae 13 Eriogonum 7 Poaceae 12 At rip lex 6 Polemoniaceae 12 Camissonia 6 Onagraceae 11 Cryptantha 5 Polygonaceae 11 Chrysothamnus 4 Boraginaceae 8 Mentzelia 4 Oenothera 4 eastern California (42, 4, 19, 13, and 3% respectively, Thome et al. 1981). The proportion of dune phanerophytes was the same as in eastern California deserts (17%) but lower than observed in Death Valley (23%). The relative abundance of annuals and geophytes in this flora indicates that these life forms offer an advantage in coping with the extreme conditions found on dunes, particularly with ref- erence to sand movement. The physiological characteristics of an- nuals, including high rates of carbon assimilation, growth and de- velopment, minimize the time that vulnerable vegetative tissues are exposed to burial, deflation, and abrasion (Pavlik 1980, Toft and Pearcy 1982). The resource allocation patterns of annuals emphasize seed production, and seeds can comply with the forces of wind and sand (by moving), unlike rooted plants that must tend to resist those forces. Thus, annuals, either in the form of ephemeral vegetative plants or seeds, are less likely to act as fixed wind obstructions and to be subjected to abrasion or self-induced burial. Although geo- phytes are perennial, they frequently produce ephemeral shoots and always possess rhizomes or rootstocks that can maintain buds near the dune surface in spite of sand accumulation. The perennating buds of hemicryptophytes and chamaephytes, however, are not formed on an elongate axis capable of extensive subterranean growth. These "stationary" buds, produced at or near the dune surface, would readily be covered by sand, especially on unstable slopes that commonly oscillate by as much as one meter per year (Sharp 1966). The slow growth of stem succulents and phanerophytes further re- stricts or excludes these species from unstabilized dunes where the rates of sand movement exceed the elongation rates of shoots and roots. Working on coastal beaches and sand dunes, Barbour et al. (1 985) also reported that annuals and geophytes were the most abun- dant life forms and emphasized exposure (wind) and sand movement to explain this pattern in the vegetation. 204 MADRONO [Vol. 32 The dune taxa were divided almost equally among the three classes of substrate fidelity (Table 3). Typical psammophytes, such as Abro- nia spp., Chaetadelpha wheeleri, Cleome sparsifolia, Dicoria spp., Lupinus shockleyi, Psoralea lanceolata var. purshii, Tiquilia plicata and Tripterocalyx spp. have such a high fidelity to sand that they accentuate the insular nature of dune vegetation. Additional restric- tions on the distribution of these plants also may be imposed by the proximity of many dunes to saline-alkaline playas. Few taxa of well- drained sandy and coarse-textured substrates appear to tolerate sa- linity (Table 3), although this is complicated by the fact that playa soils are fine-textured in addition to being highly ionic. The impli- cation is that psammophytes lack physiological mechanisms of tol- erance to low soil water potentials and, therefore, are dependent on the moisture release characteristics and low ion content of sand accumulations (Pavlik 1980, Toft and Pearcy 1982). The current and historical distributions of such insular taxa would be limited to dunes ("islands") and intermittent patches of sandy habitat ("isles and archipelagos"). In contrast, taxa that could tolerate the moisture and ionic conditions of fine-textured or saline soils would be "ocean- ic" (to complete the metaphor) and distributed widely between dunes. Clearly, the phytogeographic analysis of dunes will be served only by an examination of the distribution patterns of insular, psam- mophytic taxa. Nearly half of the taxa in this flora are associated commonly with dunes of the Mojave Desert and another quarter with those of the Great Basin (Table 3). Geographic transitionals are taxa that occur within a broad area between the two major deserts and are not readily assignable to either region. Included here are the Eureka Dune en- demics (see below), several Death Valley endemics, and a few other taxa found in a zone extending from Inyo County, California, through southern Nye and Clark counties, Nevada. Ecologically, however, all of the geographical transitionals are associated with Mojavean vegetation. It is possible to recognize two categories of endemism in this flora. Dune endemics are those psammophytes found only on deep ac- cumulations of windblown sand. These include Astragalus lentigi- nosus var. micans, Oenothera avita subsp. eurekensis, and Swallenia alexandrae from a single dune complex in Eureka Valley, and As- tragalus pseudiodanthus and Psorothamnus kingii from several dunes of the Lahontan, Columbus-Tonopah, and Mono basins. The fluvial endemics that appear restricted to sandy substrates associated with drainage courses and basins are more widespread geographically than the dune endemics. These fluvial endemics include Astragalus lentiginosus var. borreganus, A. I. var. macrolobus, Chaetadelpha wheeleri, Chamaesyce vallis-mortae, Cleome sparsifolia, Eriastrum densifolium subsp. mohavense, Eriogonum nummularae, Gilia 1985] PAVLIK: DESERT SAND DUNE FLORA 205 Table 3. Sand Dune Flora Analyzed by Life Form, Fidelity to Substrate Texture, Salt Tolerance, and Geographic Distribution. Introduced taxa are included. No. of taxa % of total flora Raunkier life form Annual 83 50.0 Geophyte 30 18.1 Hemicryptophyte 7 4.2 Chamaephyte 16 9.6 Phanerophyte— shrub 28 16.9 Stem succulent 2 1.2 Substrate fidelity— salt tolerance Sand restricted Ps 49 29.5 Ps-S 3 1.8 Coarse textures Ct 56 33.7 Ct-S 4 2.4 All textures At 33 19.9 At-S 21 12.6 Geographic distribution Great Basin Desert 43 25.9 Mojave Desert 80 48.2 Transitional 7 4,2 Widespread 29 17.5 Introduced 7 4.2 campanulata, Machaer anther a leucanthemifolia, Penstemon acu- minatus var. latebracteatus, Stanleya pinnata subsp. inyoensis, and Tripterocalyx crux-maltae. Other dunes (not included in this study) also are known to support endemic, psammophytic taxa, including those in central Nevada {Astragalus callithrix, Cymopterus ripleyi), southwestern Utah {Argemone corymbosa var. arenicola, Asclepias welshii, Astragalus striatiflorus, and the "gigas" form of Atriplex canescens), and southeastern California/northern Baja California (Ammobroma sonorae, Croton wigginsii, Dicoria argentea, Eriogo- num deserticola, Helianthus niveus subsp. tephrodes, and Palafoxia arida var. gigantea). It seems reasonable to conclude that a high degree of plant endemism is often associated with desert sand dunes. This initial analysis of the flora of sand dunes in the Great Basin and Mojave deserts has presented four lines of evidence to support the insular concept of the dune habitat and its vegetation: 1) the taxonomic composition of the dune flora differs from that of the desert as a whole; 2) dune vegetation has a distinctive life form spectrum that may be related to sand movement; 3) a subset of the 206 MADRONO [Vol. 32 flora appears to be edaphically restricted to dunes and patches of sand habitat; and 4) the presence of endemic taxa at several dunes indicates some degree of geographic and ecologic isolation through time. Additional studies will test the insular hypothesis more rig- orously and attempt to analyze the patterns of dune plant distri- bution in relation to the recent geologic history of the Great Basin and Mojave deserts. Sand Dune Flora of the Great Basin and Mojave Deserts Gnetophyta Ephedraceae Ephedra nevadensis. Uncommon; P-shrub, At, W. AnTHOPH YTA — DiCOTYLEDONEAE Amaranthaceae Tidestromia oblongifolia. Rare; Cham, Ct, M. Asclepiadaceae Asclepias erosa. Rare; Geo, At, M. Asteraceae Acamptopappus sphaerocephalus. Rare; Cham, At, M. Ambrosia acanthicarpa. Uncommon; Annual, Ct, W. Ambrosia dumosa. Uncommon; Cham, At, M. Artemisia spinescens. Uncommon; Cham, At-S, GB. Artemisia tridentata var. tridentata. Uncommon; P-shrub, At, GB. Atrichoseris platyphylla. Uncommon; Annual, Ct, M. Baileya pleniradiata. Uncommon; Annual, Ps, M. Chaenactis fremontii. Uncommon; Annual, Ct, M. Chaenactis stevioides var. stevioides. Uncommon; Annual, Ct, W. Chaetadelpha wheeleri. Common; Geo, Ps, GB. Chrysothamnus nauseosus subsp. consimilis. Uncommon; P-shrub, At, GB. Chrysothamnus nauseosus subsp. hololeucus. Uncommon; P-shrub, At, GB. Chrysothamnus viscidiflorus subsp. pumilus. Uncommon; Cham, At, GB. Chrysothamnus viscidiflorus subsp. viscidiflorus. Uncommon; Cham, At, GB. Dicoria canescens subsp. canescens. Common; Annual, Ps, M. Dicoria canescens subsp. clarkae. Common; Annual, Ps, GB. 1985] PAVLIK: DESERT SAND DUNE FLORA 207 Glyptopleura marginata. Rare; Annual, Ps, GB. Haplopappus acradenius. Rare; P-shrub, At-S, M. Hymenoclea salsola var. salsola. Uncommon; P-shrub, Ct-S, M. Machaeranthera arida. Rare; Annual, Ct, M. Machaer anther a leucanthemifolia. Rare; Cham, Ps, M. Malacothrix californica var. glabrata. Uncommon; Annual, Ct, W. Malacothrix sonchoides. Uncommon; Annual, Ct, GB. Monoptilon bellidiforme. Uncommon; Annual, Ct, M. Palafoxia arida var. arida. Uncommon; Annual, Ps, M. Pluchea sericea. Uncommon; P-shrub, At-S, M. Psathyrotes annua. Uncommon; Annual, Ct-S, W. Raflnesquia californica. Uncommon; Annual, At, M. Rafinesquia neomexicana. Uncommon; Annual, At, M. Solidago spectabilis. Rare; Geo, At-S, GB. Stephanomeria pauciflora. Uncommon; Cham, Ct, M. Tetradymia tetrameres. Rare; P-shrub, Ct, GB. Tetradymia glabrata. Uncommon; P-shrub, At, GB. Tetradymia spinosa. Uncommon; P-shrub, At, GB. Bignoniaceae Chilopsis linearis var. arcuata. Rare; P-shrub, At, M. Boraginaceae Cryptantha angustifolia. Common; Annual, Ct, M. Cryptantha costata. Rare; Annual, Ct, M. Cryptantha maritima. Rare; Annual, Ct, M. Cryptantha micrantha subsp. micrantha. Common; Annual, Ps, W. Cryptantha pterocarya. Rare; Annual, Ct, W. Heliotropium curassavicum var. oculatum. Rare; Geo, At-S, W. Tiquilia nuttallii. Common; Annual, Ps-S, GB. Tiquilia plicata. Common; Geo, Ps, M. Brassicaceae Descurainia pinnata subsp. fllipes. Rare; Annual, Ct, GB. Dithyrea californica. Uncommon; Annual, Ps, M. Lepidium fremontii. Uncommon; Cham, Ct, M. Lepidium lasiocarpum. Rare; Annual, Ct, W. Lepidium perfoliatum. Uncommon and introduced; Annual, At, GB. Stanleya pinnata subsp. inyoensis. Uncommon; P-shrub, Ps, T. Streptanthella longirostris. Rare; Annual, Ct, W. 208 MADRONO [Vol. 32 Cactaceae Opuntia echinocarpa var. echinocarpa. Uncommon; Succ, Ct, M. Opuntia pulchella. Rare; Succ, Ct, GB. Capparaceae Cleome sparsifolia. Common; Annual, Ps, GB. Cleomella obtusifolia. Rare; Annual, At-S, M. Caryophyllaceae Achyronychia cooperi. Uncommon; Annual, Ct, M. Chenopodiaceae Atriplex canescens subsp. canescens. Common; P-shrub, At-S, W. Atriplex confertifolia. Common; P-shrub, At-S, W. Atriplex hymenelytra. Rare; P-shrub, At-S, M. Atriplex parryi. Rare; Cham, At-S, M. Atriplex polycarpa. Uncommon; P-shrub, At-S, M. Atriplex torreyi. Rare; P-shrub, At-S, M. Ceratoides lanata. Uncommon; P-shrub, Ct, W. Corispermum hyssopifolium. Rare and introduced; Annual, Ps, T. Kochia americana. Rare; Cham, At-S, W. Nitrophila occidentalis. Rare; Geo, At-S, W. Salsola spp. (includes both S. australis and S. paulsenii). Common and introduced; Annual, At-S, W. Sarcobatus vermiculatus. Common; P-shrub, At-S, GB. Suaeda torreyana var. ramosissima. Uncommon; P-shrub, At-S, W. Euphorbiaceae Chamaesyce micromera. Uncommon; Annual, Ps, M. Chamaesyce ocellata. Uncommon; Annual, Ps, M. Chamaesyce parryi. Rare; Annual, Ps, M. Chamaesyce vallis-mortae. Rare; Geo, Ps, M. Croton calif ornicus var. mojavensis. Rare; Cham, Ps, M. Fabaceae Acacia greggii. Rare; P-shrub, At, M. Astragalus didymocarpus var. dispermus. Uncommon; Annual, Ps, M. Astragalus geyeri var. geyeri. Uncommon; Annual, Ct, GB. Astragalus lentiginosus var. borreganus. Uncommon; Geo, Ps, M. Astragalus lentiginosus var. fremontii. Rare; Geo, Ct-S, W. 1985] PAVLIK: DESERT SAND DUNE FLORA 209 Astragalus lentiginosus var. kennedyi. Rare; Geo, Ps-S, GB. Astragalus lentiginosus var. macrolobus. Rare; Geo, Ps, GB. Astragalus lentiginosus var. micans. Rare and endemic to Eureka Valley, Inyo Co., CA; Geo, Ps, T. Astragalus lentiginosus var. variabilis. Uncommon; Geo, Ps-S, M. Astragalus pseudiodanthus. Rare and endemic to the Mono Basin and the southernmost extensions of the Lahontan Basin; Geo, Ps, GB. Astragalus sabulonum. Uncommon; Annual, Ps, M. Lupinus pusillus subsp. intermontanus. Common; Annual, Ps, GB. Lupinus shockleyi. Common; Annual, Ps, M. Peteria thompsonae. Rare; Geo, Ct, T. Prosopis glandulosa var. torreyana. Uncommon; P-shrub, Ct-S, M. Prosopis pubescens. Rare; P-shrub, Ct, M. Psoralea lanceolata var. purshii. Uncommon; Geo, Ps, GB. Psorothamnus kingii. Rare and endemic to the Lahontan Basin; Geo, Ps, GB. Psorothamnus polydenius var. polydenius. Uncommon; P-shrub, At, W. Geraniaceae Erodium cicutarium. Uncommon and introduced; Annual, At, W. Hydrophyllaceae Nama demissum var. demissum. Uncommon; Annual, At, W. Phacelia bicolor. Rare; Annual, Ps, GB. Phacelia fremontii. Rare; Annual, At, W. Loasaceae Mentzelia albicaulis. Uncommon; Annual, Ct, W. Mentzelia congesta. Rare; Annual, At, GB. Mentzelia longiloba. Rare; Hemi, Ps, M. Mentzelia nitens. Uncommon; Annual, Ct, W. Pentalonyx thurberi subsp. gilmanii. Rare; P-shrub, Ct, T. Pentalonyx thurberi subsp. thurberi. Common; P-shrub, Ct, M. Malvaceae Eremalche exilis. Rare; Annual, Ct, M. Eremalche rotundifolia. Rare; Annual, Ct, M. Sphaeralcea ambigua subsp. ambigua. Common; Cham, At, M. Sphaeralcea ambigua subsp. monticola. Rare; Cham, At, GB. 210 MADRONO [Vol. 32 Nyctaginaceae Abronia pogonantha. Rare; Annual, Ps, M. Abronia turbinata. Common; Annual, Ps, GB. Abronia villosa var. villosa. Uncommon; Annual, Ps, M. Tripterocalyx crux-maltae. Rare; Annual, Ps, GB. Tripterocalyx micranthus. Rare; Annual, Ps, M. Onagraceae Camissonia boothii subsp. condensata. Uncommon; Annual, Ct, M. Camissonia brevipes subsp. brevipes. Rare; Annual, At, M. Camissonia claviformis subsp. aurantiaca. Uncommon; Annual, Ct, M. Camissonia claviformis subsp. claviformis. Rare; Annual, Ct, M. Camissonia claviformis subsp. funerea. Uncommon; Annual, Ct, T. Camissonia claviformis subsp. integrior. Common; Annual, Ct, GB. Oenothera avita subsp. avita. Rare; Geo, Ct, GB. Oenothera avita subsp. eurekensis. Rare and endemic to Eureka Valley, Inyo Co., CA; Geo, Ps, T. Oenothera deltoides subsp. deltoides. Uncommon; Annual, Ps, M. Oenothera deltoides subsp. piperi. Uncommon; Annual, Ct, GB. Oenothera primiveris subsp. bufonis. Rare; Annual, At, M. Papaveraceae Argemone corymbosa var. corymbosa. Uncommon; Hemi, At, M. Plantaginaceae Plantago insularis var. fastigiata. Uncommon; Annual, Ct, M. Polemoniaceae Eriastrum densifolium subsp. mohavense. Rare; Cham, Ps, M. Gilia campanulata. Uncommon; Annual, Ps, GB. Gilia filiformis. Uncommon; Annual, Ct, M. Gilia latiflora subsp. latiflora. Uncommon; Annual, Ct, M. Gilia latifolia. Rare; Hemi, At, M. Gilia leptomeria subsp. "A." Common; Annual, Ps, GB. Gilia leptomeria subsp. "B." Common; Annual, Ps, GB. Gilia micromeria. Uncommon; Annual, Ps, GB. Gilia sinuata. Rare; Annual, Ct, W. Langloisia matthewsii. Rare; Annual, Ct, M. 1985] PAVLIK: DESERT SAND DUNE FLORA 211 Langloisia schottii. Rare; Annual, Ct, M. Langloisia setosissima. Rare; Annual, Ct, W. Polygonaceae Chorizanthe brevicornu subsp. brevicornu. Uncommon; Annual, Ct, M. Chorizanthe rigida. Uncommon; Annual, At, M. Eriogonum cernuum var. cernuum. Rare; Annual, At-S, GB. Eriogonum deflexum. Rare; Annual, Ct, M. Eriogonum inflatum var. inflatum. Uncommon; Hemi, Ct, M. Eriogonum insigne. Rare; Annual, Ps, M. Eriogonum nummularae M. E. Jones {=E. kearneyi var. kear- neyi). Uncommon; Cham, Ps, GB. Eriogonum reniforme. Uncommon; Annual, Ct, M. Eriogonum trichopes. Uncommon; Annual, Ct, M. Rumex hymenosepalus. Uncommon and introduced; Geo, Ct, W. Rumex venosus. Uncommon; Geo, Ps, GB. Scrophulariaceae Penstemon acuminatus var. latebracteatus. Rare; Geo, Ps, GB. Penstemon thurberi. Rare; Cham, Ct, M. Zygophyllaceae Larrea tridentata. Common; P-shrub, At, M. AnTHOPHYTA — MONOCOTYLEDONEAE Liliaceae Hesperocallis undulata. Uncommon; Geo, Ps, M. Poaceae Bouteloua barbata. Uncommon; Annual, Ct, M. Bromus tectorum. Common and introduced; Annual, At, W. Distichlis spicata var. striata. Uncommon; Geo, At-S, W. Elymus cinereus. Uncommon; Hemi, At, GB. Hilaria jamesii. Uncommon; Geo, At, GB. Hilaria rigida. Uncommon; Geo, Ct, M. Muhlenbergia asperifolia. Uncommon; Geo, At-S, W. Oryzopsis hymenoides. Common; Hemi, Ct, W. Panicum urvilleanum. Rare; Geo, Ps, M. Schismus barbatus. Rare and introduced; Annual, At, M. Sporobolus airoides. Uncommon; Hemi, At-S, W. Swallenia alexandrae. Rare and endemic to Eureka Valley, Inyo Co., CA; Geo, Ps, T. 212 MADRONO [Vol. 32 Acknowledgments I am very grateful to the following individuals who examined the collections made in this study: R. C. Bameby (Astragalus), A. Day (Gilia), M. DeDecker (flora of eastern California), P. Raven (Camissonia), J. Reveal (Eriogonum), and G. Webster (Chamaesyce). I am particularly indebted to Mary and Paul DeDecker for their encouragement. Finally, I thank Derham Giuliani and Enid Larson for many hours of discussion on the biology and beauty of desert dunes. Literature Cited Bagnold, R. a. 1933. A further journey through the Libyan desert. Geog. J. 82: 103-129. . 1 94 1 . The physics of blown sand and desert dunes. William Morrow and Co., NY. Barbour, M. G., T. DeJong, and B. Pavlik. 1985. Marine beach and dune plant communities. In B. F. Chabot and H. A. Mooney, eds., Physiological ecology of North American plant communities, pp. 296-322. Chapman and Hall, Lon- don. Beatley, J. C. 1976. Vascular plants of the Nevada Test Site and central-southern Nevada: ecologic and geographic distributions. Energy Research and Devel. Adm, Nat. Tech. Inform. Serv., VA. Bureau of Land Management (BLM). 1980. California Desert Plan. Final en- vironmental impact statement and proposed plan, appendix volume E. California Desert Plan Staff; Riverside, CA. Brown, J. H. 1973. Species diversity of seed-eating desert rodents in sand dune habitats. Ecology 54:775-787. Bury, R. B. and R. A. Luckenbach. 1 983. Vehicular recreation in arid land dunes: biotic responses and management alternatives. In R. H. Webb and H. G. Wil- shire, eds., Environmental effects of off'-road vehicles, impacts and management in arid regions, pp. 207-224. Springer- Verlag, NY. Dean, L. E. 1978. The California desert sand dunes. BLM-NASA #NSG 7220, California Desert Plan Staff", Riverside, CA. DeDecker, M. 1976. The Eureka Dunes. Fremontia 3:17-20. Evans, J. R. 1 962. Falling and climbing sand dunes in the Cronese ("Cat") Mountain area, San Bernardino Co., California. J. Geol. 70:107-113. Johnson, A. F. 1977. A survey of the strand and dune vegetation along the Pacific and southern Gulf coasts of Baja California, Mexico. J. Biogeogr. 7:83-99. . 1982. Dune vegetation along the eastern shore of the Gulf of California. J. Biogeogr. 9:317-330. Kartesz, J. T. and R. Kartesz. 1980. A synonymized checklist of the vascular flora of the United States, Canada and Greenland. Vol. IL the biota of North America. Univ. North Carolina Press, Chapel Hill. MacDonald, a. a. 1970. The northern Mojave Desert's little Sahara. Calif. Div. Mines and Geology, Mineral Information Service 23:3-6. NoRRis, R. M. and K. S. Norris. 1961. Algodones Dunes of southeastern California. Geol. Soc. America Bull. 72:605-620. Pavlik, B. M. 1979a. A synthetic approach to the plant ecology of desert sand dunes. Eureka Valley, California. M.S. thesis, Univ. Cahfomia, Davis. . 1979b. The biology of endemic psammophytes. Eureka Valley, California, and its relation to off'-road vehicle impact. BLM #CA-060-CT8 -000049, Cah- fomia Desert Plan Staff", Riverside, CA. . 1980. Patterns of water potential and photosynthesis of desert sand dune plants, Eureka Valley, California. Oecologia 46:147-154. 1985] PAVLIK: DESERT SAND DUNE FLORA 213 Radford, A. E., W. C. Dickison, J. R. Massey, and C. R. Bell. 1974. Vascular plant systematics. Harper and Row, NY. Raunkier, C. 1934. The life forms of plants and statistical plant geography. Clar- endon Press, Oxford. Rempel, p. J. 1936. The crescentic dunes of the Salton Sea and their relation to the vegetation. Ecology 17:347-358. Sharp, R. P. 1966. Kelso Dunes, Mojave Desert, California. Geol. Soc. America Bull. 77:1045-1074. Thorne, R. F., B. a. Prigge, and J. Henrickson. 1981. A flora of the higher ranges and the Kelso Dunes of the eastern Mojave Desert in California. Aliso 10:71- 186. Toft, N. L. and R. W. Pearcy. 1 982. Gas exchange characteristics and temperature relations of two desert annuals: a comparison of a winter-active and a summer- active species. Oecologia 55:170-177. Vasek, F. C. and M. G. Barbour. 1977. Mojave desert scrub vegetation. In M. G. Barbour and J. Major, eds.. Terrestrial vegetation of California, pp. 835-867. Wiley Interscience, NY. Went, F. W. and W. Westergaard. 1949. Ecology of desert plants III. Develop- ment of plants in the Death Valley National Monument, California. Ecology 30: 26-38. Westec. 1977. Survey of sensitive plants of the Algodones Dunes. Unpublished report. Bureau of Land Management, Riverside, CA. (Received 20 Dec 1984; accepted 21 Jun 1985.) Appendix L Plants that may occur but could not be confirmed at the dunes included in this study. Antirrhinum kingii Nama aretiodes Bailey a pauciradiata N. depressum Chamaesyce polycarpa Oligomeris linifolia Cleome lutea Orobanche cooperi Cryptantha circumscissa Pedis papposa Cycloloma atriplicifolia Phacelia ivesiana Eriastrum eremicum Pholisma arenarium E. wilcoxii Plagiobothrys kingii Eriogonum brachyanthum Psoralea castorea E. maculatum Psorothamnus mollis Erioneuron pulchellum Sarcobatus vermiculatus var. baileyi Iva nevadensis Stephanomeria exigua Krameria parvifolia Stillingia spinulosa ADDENDA TO THE VASCULAR FLORA OF SAN LUIS OBISPO COUNTY, CALIFORNIA David J. Keil, Robert L. Allen*, Joy H. Nishida^, and Eric A. Wise^ Biological Sciences Department California Polytechnic State University, San Luis Obispo, CA 93407 Abstract Documentation is provided for the occurrence of 76 species of flowering plants not reported previously from San Luis Obispo County, California. Recent collections confirm the presence of three additional species that were reported from San Luis Obispo County but not included in The Vascular Plants of San Luis Obispo County, California (Hoover 1970). Of the addenda to the county's flora, 21 are native to California and 58 are introduced. The additions include members of 32 families, four of which were not represented previously in the county's flora: Basellaceae, Halo- ragaceae, Molluginaceae, and Hydrocharitaceae. The Vascular Plants of San Luis Obispo County, California (Hoo- ver 1970) has provided a sound foundation for subsequent floristic studies in the county. The thoroughness of Hoover's research on the county flora and his careful scholarship have been demonstrated repeatedly by users of his flora. As can be expected with any flora, subsequent investigations have documented the presence in the county of species not reported by Hoover. In this paper we are adding 76 taxa to the known flora of San Luis Obispo County. Another three species are listed that were reported from San Luis Obispo County in Munz (1968) but overlooked or deliberately omitted in the preparation of the final drafts of Hoover's flora. Recent collec- tions of these species are noted. The addenda to the San Luis Obispo County flora include mem- bers of 32 families. Four of these, Basellaceae, Haloragaceae, Mol- luginaceae, and Hydrocharitaceae, were not reported previously from the county. Page numbers from Hoover (1970) are indicated for those families already known from San Luis Obispo County. The following list is based on specimens deposited in OBL For most taxa we have followed the nomenclature of Munz and Keck (1959) and Munz (1968). For some weedy taxa, particularly those ' Present address: Museum of Systematic Biology, Univ. of California, Irvine, CA 92717. 2 Present address: Section of Botany, Carnegie Museum of Natural History, 4400 Forbes Ave., Pittsburgh, PA 15213. 3 Present address: Biological Sciences Dept., Chabot College, Hayward, CA 94545. MadroRo, Vol. 32, No. 4, pp. 214-224, 20 December 1985 1985] KEIL ET AL.: ADDENDA TO SAN LUIS OBISPO CO. FLORA 2 1 5 of European origin, the names used by Munz and Keck differ from those used by Tutin et al. (1964-1980) and by Kartesz and Kartesz (1980). For these taxa, we have noted in brackets the names used by Munz and Keck. Authorities are from Munz and Keck (1959), Munz (1968), and Kartesz and Kartesz (1980), except as noted. The additions to the San Luis Obispo County flora represent 2 1 native and 58 introduced species. Most of the native taxa were probably present when Hoover did his field work but were over- looked. [Sagittaria latifolia is native to California but known to have been introduced to San Luis Obispo County.] Some native taxa, such as Elatine rubella and Limosella aquatica, are very inconspic- uous plants. Others occur in very restricted areas that Hoover never visited. Of the introduced taxa, some are well-established and prob- ably were present during Hoover's studies. Others appear to be recent introductions. The documentation of their occurrence at this time will allow later studies to determine whether the taxa have become permanent members of the county's flora. In the species list, intro- duced taxa are indicated by an asterisk (*). Some of the additions are a result of intensive floristic surveys. Local areas covered in these floristic studies include the Arroyo de la Cruz region in the northwestern comer of the county, American Canyon in the La Panza Mountains, Black Lake Canyon on the Nipomo Mesa, Laguna Lake Park in San Luis Obispo, and the Huer- huero Creek drainage near Creston. A survey of aquatic habitats throughout the county by Wise (1 984) produced several new records. Addenda to the Vascular Flora AnTHOPHYTA — DiCOTYLEDONEAE Anacardiaceae (p. 188) Rhus integrifolia Benth. & Hook, f ex Brewer & S. Wats. Between Arroyo Grande and San Luis Obispo along Hwy 227, ca. 3.2 km south of southern jet. with Corbett Canyon Rd., coastal scrub hillside {Keil 13688). This population is the northernmost known occurrence for R. integrifolia. Apiaceae (p. 208) *Apium leptophyllum (Pers.) F. MueU. ex Benth. & Muell. Common lawn weed on campus of California Polytechnic State Univ., San Luis Obispo {Keil 13677). Perideridia kelloggii (A. Gray) Mathias. Grassy slope in heavy clay soil on south side of Arroyo de la Cruz {Keil 17394). *Torilis arvensis (Huds.) Link subsp. purpurea (Ten.) Hay- ek. Along channel of Arroyo de la Cruz {Keil 14881)\ lawn 216 MADRONO [Vol. 32 weed on campus of California Polytechnic State Univ., San Luis Obispo {Keil 14931); Lopez Wilderness Area, abundant along trail in upper Lopez Canyon in open woods (Keil and Webster 18188). Asteraceae (p. 273) *Chondrilla juncea L. Reported from the county by Munz (1968). This species was first collected in the county by R. F. Hoover in 1965 from the summit of Cuesta Grade 9.7 km north of San Luis Obispo (Hoover 9634) and 1 .6 km west of San Luis Obispo along Hwy 1 {Hoover 9645). The two infestations were treated chemically by the San Luis Obispo County Agricultural Com- missioner's office in 1966 (Fuller 1966) and Hoover did not include the species in his flora. The eradication efforts were apparently unsuccessful and the species was collected at several locations in the San Luis Obispo area in the early 1980s. This species is locally common along the right-of-way of the Southern Pacific Railroad and on roadsides from Santa Margarita to San Luis Obispo, and has been collected from the following sites: Cuesta Summit along unpaved fire road in chaparral area {Keil s.n.)\ San Luis Obispo {Keil s.n., Burdett and Burden s.n., Allen A 5 00). Doug Barbe of the State Department of Food and Ag- riculture states (pers. comm.) that county personnel are once again attempting to eradicate this invasive weed. *Eclipta prostrata (L.) L. [=E. alba (L.) Hassk.]. East end of Lopez Lake below confluence of Arroyo Grande Cr. and Phoenix Cr., drying mudflats and sandbars {Keil 13644)\ ]usX south of Creston along Middle Branch of Huerhuero Cr., sandy dry stream bed {Keil 17976). *Erechtites glomerata (Poir.) DC. [=E. arguta (A. Rich.) DC. (Bark- ley 1981)]. Reported from the county by Munz (1968) but not listed by Hoover (1970); recent collection from lower slopes of Cypress Mountain, in a live oak woodland, roadside on mesic north-facing slope {Keil et al. 15942). *Erechtites minima (Poir.) DC. [=E. prenanthoides (A. Rich.) DC.]. Arroyo de la Cruz in bed of unused dirt road on steep oak-wooded slope {Keil 17401). Lasthenia glabrata Lindl. Baywood Park at Sweet Springs Marsh. Locally common in salt marsh {Cardwell 238). Lasthenia maritima (A. Gray) M. Vasey [=L. minor (DC.) Omduff* subsp. maritima (A. Gray) Omduff']. Pup Rock, near Lion Rock, offshore of mouth of Diablo Canyon; in crevices and among loose, guano-soaked rocks in western gull breeding area {Vasey and Harms 8119). The Diablo Canyon site is disjunct from the nearest known population in San Mateo County by over 300 km (M. C. Vasey, pers. comm.). 1 985] KEIL ET AL.: ADDENDA TO SAN LUIS OBISPO CO. FLORA 2 1 7 "^Leontodon taraxacoides (Vill.) Merat subsp. taraxacoides [=L. leysseri (Wallr.) G. Beck]. Along Prefumo Canyon Rd., 1.6 km west of Los Osos Valley Rd. {Keil 11830). Basellaceae *Anredera cordifolia (Ten.) Steenis. San Luis Obispo, in shrubs next to San Luis Cr. {Keil s.n.). Brassicaceae (p. 142) *Coronopus didymus (L.) Sm. Sand dunes at south end of IVIorro Bay {Schwartz and Long 32); garden weed in Los Osos {Keil s.n.); 2.9 km north of Arroyo de la Cruz {Keil 17387). *Erophila verna (L.) Chev. [=Draba verna L.]. Nipomo Mesa along Black Lake Canyon, locally abundant along trail {Jones and Keil 15812); American Canyon campground on slight slope in blue oak woodland {Nishida and Allen 213). *Lepidium latifolium L. Along Hwy 41, 1.6 km northeast of jet. with Hwy 1 in ]VIorro Bay, roadside {McLeod s.n.). *Lepidium oblongum Small. La Panza Mts., American Canyon in blue oak woodland {Nishida 231, 287, 551); Los Osos in cracks of sidewalk {Keil 15887); Ridge between Arroyo de la Cruz and Arroyo del Oso {Keil 16995). Caryophyllaceae (p. 129) *Sagina apetala Ard. North of Arroyo de la Cruz and west of Hwy 1, locally common in damp soil of dune slack at edge of trail {Keil 16924); south end of Santa Lucia Mts., on rd. to Stony Cr. campground, 0.6 km from jet. with Avenales-Agua Escon- dido Rd. (T31S R16E S27), 700 m elev., very local on moss- covered rocks in shaded ravine with other minuscule herbs and Anthoceros {Keil and Riggins 18211). Chenopodiaceae (p. 120) Chenopodium chenopodioides (L.) Aellen. 4.8 km south of Creston along Hwy 229, along Middle Branch of Huerhuero Cr. {Keil 14259). ^Chenopodium multifidum L. Baywood Park-Los Osos area on roadsides {Keil 17407, 18301, 18433). Elatinaceae (p. 196) Elatine rubella Rydb. Just south of Creston in damp sand at edge of Creston Lake {Keil 17996). 218 MADRONO [Vol. 32 Euphorbiaceae (p. 186) * Euphorbia serpens H.B.K. Campus of California Polytechnic State Univ., San Luis Obispo, common weed in ornamental plantings {Keil 18004). Fabaceae (p. 165) *Lathyrus japonicus Willd. Intersection of South Bay Blvd. and Turri Rd. in sandy soil just above salt marsh {Meredith 23). Lotus oblongifolius (Benth.) Greene. American Canyon in La Pan- za Mts., locally common among rocks in stream channel {Keil et al 18136). "^Spartium junceum L. Well-established and spreading in the sandy channel of Arroyo de la Cruz {Keil 13992); occasional on brushy slopes on campus of California Polytechnic State Univ., San Luis Obispo {Keil 18299); along Foothill Blvd. ca. 0.4 km from Los Osos Valley Rd., grassy roadside in agricultural area {Keil 18319). *Trifolium campestre Schreb. [=T. procumbens sensu auct. non L. ] . 1.4 km north of San Simeon along Hwy 1 , locally an aspect dominant growing in dense colonies {Keil 16950); ridge south of Arroyo de la Cruz, locally common in grassy areas {Keil 18123). *Trifolium dubium Sibthorp. Campus of California Polytechnic State Univ., San Luis Obispo, lawn weed {Keil s.n.). *Trifolium fragiferum L. Campus of California Polytechnic State Univ., San Luis Obispo, weed in lawn {Keil 17163); ca. 4.8 km west of Cuesta College along Chorro Cr. near Highway 1 {Keil 12508); Laguna Lake Park, San Luis Obispo {Smeltzer and Turnquist 208); south of San Luis Obispo, 3.7 km from Johnson Ave. on Orcutt Rd. in rocky streamlet {Wise 1206); weed of lawn areas on campus of Los Osos Junior High School {Keil 18431). *Trifolium glomeratum L. Just south of highway bridge over Ar- royo de la Cruz, very local on roadside {Keil 17072). *Trifolium pratense L. Locally common weed of summer-irrigated lawns and playground areas on campus of Bay wood Elementary School, Baywood Park {Keil 18432). * Vicia sativa L. subsp. nigra (L.) Ehrh. [= V. angustifolia L.]. Ridge system between Arroyo de la Cruz and Arroyo de los Chinos on grassy slope {Keil 16962). * Vicia villosa Roth subsp. varia (Host.) Corb. [=V. dasycarpa Ten.]. Well-established in western half of San Luis Obispo County. Adelaida {Jackman and Truesdale 11); 12.3 km north- east of Santa Margarita {Arnold and Allen 280); Reservoir Can- yon {Burrows 68; Berry and Wilson 55); Laguna Lake Park, San 1985] KEIL ET AL.: ADDENDA TO SAN LUIS OBISPO CO. FLORA 2 1 9 Luis Obispo (Smeltzer and Turnquist 55); San Luis Obispo (Dettloff 01); Indian Knob {Vanderwier 150); Black Lake Can- yon (Keil and Wise 16268); near Rinconada Mine, south of Santa Margarita (Keil et al 18224). Fumariaceae (p. 142) *Fumaria parviflora Lam. San Luis Obispo in cultivated field southwest of Madonna Rd. Shopping Plaza {Ashley s.n.). Haloragaceae *Myriophyllum spicatum L. Lopez Lake, locally common in shal- low water, particularly around inlets {Keil 13672 [voucher de- termined by O. Ceska]; Wise 1065, 1259). Smith (1976) listed Myriophyllum spicatum subsp. exalbescens (Fern.) Jeps. from wetland areas of the Nipomo Dunes. He cited no specimens and did not indicate the source of his information. We have been unable to verify the occurrence of this taxon in San Luis Obispo County. Hypericaceae (p. 196) "^Hypericum perforatum L. South of San Luis Obispo on Orcutt Rd. {Keil 18318). Linaceae (p. 186) *Linum bienne P. Mill. [=L. angustifolium Huds.]. Locally com- mon on grassy roadsides along Hwy 1 from San Carpoforo Cr. {Brown 2023) south to Arroyo de la Cruz {Jones s.n.; Keil 16247). Malvaceae (p. 193) *Abutilon theophrasti Medic. Atascadero, weed in cultivated field {Dempsey s.n.); campus of California Polytechnic State Univ., San Luis Obispo, locally common weed in cultivated fields ( Wil- genburg 8). *Lavatera arborea L. Scattered in coastal areas from Morro Bay south into Santa Barbara County. Morro Bay, locally common on sand dunes {Keil 18309); Bay wood Park on damp roadside near edge of marsh {Keil 18289); Arroyo Grande, agricultural area {Keil 18314); just north of Santa Maria River on Hwy 1 {Keil 18315). *Modiola caroliniana (L.) G. Don. Lawn weed at Laguna Lake Park {Turnquist and Smeltzer 184); San Luis Obispo {McLeod 1390). 220 MADRONO [Vol. 32 Molluginaceae *Glinus lotoides L. Reported from the county by Munz (1968) but not listed by Hoover (1970); recent collections from Creston {Keil 17971, 18015) and east of Atascadero in Rocky Canyon Cr. {Keil 18035). *Mollugo verticillata L. Just south of Creston in bed of Middle Fork of Huerhuero Cr. and in adjacent grassy areas {Keil 17989); ca. 2.4 km south of Creston at Beck Lake {Wise 1865). Onagraceae (p. 200) Clarkia rubicunda (Lindl.) Lewis & Lewis subsp. rubicunda. On grassy slope above Arroyo de los Chinos east of BM 77 {Keil 17144). Plantaginaceae (p. 266) *Plantago arenaria Waldst. & Kit. [=P. indica L.]. Nipomo Mesa, locally common along Pomeroy Rd. ca. 1.6 km north of Willow Rd. in area of open dune chaparral {Wise and Keil 16251). Polemoniaceae (p. 224) Allophyllum divaricatum (Nutt.) A. & V. Grant. Scattered about open disturbed ground and roadside on eastern slope of Pine Mountain in full sun on sandstone soil, 760 m {Miller 581-44). Polygonaceae (p. 109) *Emex australis Steinh. In stabilized coastal dunes at south end of Alexander St. in Los Osos {Keil 15692); along Pecho Rd. in Montana de Oro State Park {Keil et al. s.n.). ^Polygonum argyrocoleon Steud. ex Kunze. Occasional lawn weed at Los Osos Junior High School in Baywood Park {Keil s.n.). *Reynoutria sachalinense (F. S. Petrop.) Nakai in Mori [=Polygo- num sachalinense F. S. Petrop. (Webb 1964)]. A 27-m^ infes- tation was found to the east of El Camino Real, just south of Santa Ysabel Ave. in Atascadero in 1967 {Fuller 15919). Be- cause there have been no reports of this infestation since 1967, it is presumed to have been eradicated (Doug Barbe, pers. comm.). Subsequently, the site in Atascadero where this species was collected has been developed commercially. *Rumex kerneri Borbas. Black Lake Canyon in Eucalyptus grove in partial shade {Keil and Jones 15811); San Luis Obispo on roadside in partial shade of Eucalyptus {Keil 18459). Rumex occidentalis S. Wats. var. fenestratus (Greene) Lepage [=R. 1985] KEIL ET AL.: ADDENDA TO SAN LUIS OBISPO CO. FLORA 221 fenestratus Greene]. Local in freshwater marsh in lee of dunes just north of Arroyo de la Cruz {Keil 17068). Scrophulariaceae (p. 255) *Bellardia trixago (L.) All. In grassland of Laguna Lake Park, San Luis Obispo {Smeltzer and Turnquist s.n.\ Keil 13060); along Turri Rd. (Walters s.n.); Arroyo de la Cruz, on flood plain (Keil 17075); Poly Canyon, locally abundant on grazed slope on north side of creek (Riggins 1493). *Kickxia elatine (L.) Dumort. East of Arroyo Grande along rd. to Lopez Lake, ca. 1.6 km east of Orcutt Rd., edge of cultivated field at border of woods {Keil 13638); along channel of San Luis Cr. in San Luis Obispo {Keil 16314). Limosella aquatica L. Canyon Ranch south of Shandon on Shell Cr. Rd., Sinton Middle Pond {Wise 1339, 1493). "^Linaria vulgaris P. Mill. Arroyo Grande at comer of Hwy 1 and Grand Ave. in wet area of roadside ditch {Byrum and Koivisto 9). * Veronica persica Poir. American Canyon, uncommon in small patch along road {Nishida and Luckow 555); Shandon, weed in vegetable garden {Keil 17167); lawn weed on campus of Cali- fornia Polytechnic State Univ., San Luis Obispo {Keil 18300). Solanaceae (p. 253) Solanum cornutum Lam. [=S. rostratum Dunal]. Atascadero in 3-F Meadows area {Cunningham s.n.). Verbenaceae (p. 246) * Verbena brasiliensis Veil. On roadcut just south of highway bridge over Arroyo de la Cruz {Keil 17967). AnTHOPH YTA — MONOCOT YLEDONEAE Alismataceae (p. 51) Echinodorus rostratus (Nutt.) Engelm. [=E. berteroi (Spreng.) Fas- sett]. Campus of California Polytechnic State Univ., San Luis Obispo at edge of large pond (^par/Zw^yi/y, 1372, 1373; Ashley s.n.; Wise 936); Santa Margarita Lake {Wise 916, 1712); east end of Lopez Lake below confluence of Arroyo Grande Cr. and Phoenix Cr. {Keil 13664). *Sagittaria latifolia Willd. Canyon Ranch on Shell Cr. Rd. south of Shandon where introduced in the 1970s as an ornamental (Norma Sinton, pers. comm.). It has spread locally to nearby 222 MADRONO [Vol. 32 reservoirs: Canyon Ranch {McKei s.n.)\ Sinton South Pond {Wise 1487, 1507); Sycamore Reservoir (Wise 1508). Araceae (p. 83) *Peltandra virginica (L.) Schott. Farm ponds on Canyon Ranch south of Shandon {Wise 1353, 1496, 1509). This species was introduced in the 1970s to a small ornamental pool (Norma Sinton, pers. comm.) and has spread subsequently to nearby reservoirs, probably through the activities of waterfowl. Cyperaceae (p. 76) *Cyperus esculent us L. Along Middle Branch of Huerhuero Cr., 4.8 km south of Creston along Hwy 229, damp area in sandy stream channel {Keil 14247). Hydrocharitaceae *Egeria densa Planch. Canyon Ranch south of Shandon, Sinton South Pond {Wise 1317, 1484). Iridaceae (p. 98) */m pseudacorus L. On damp banks of San Luis Cr. {Wise 2060) and Meadow Lane {Wise 1452). Juncaceae (p. 84) Juncus rugulosus Engelm. Pine Canyon at the Cuyama River ( Wise 781). Hoover (1970) indicated that the occurrence of this taxon [as /. dubius Engelm. forma rugulosus (Engelm.) Hoover] in San Luis Obispo County was to be expected. Liliaceae (p. 88) Calochortus weedii Wood var. vestus Purdy. Along road to Hearst Springs on Pine Mountain, 850 m, exposed ocean-facing slope with soil derived from serpentine mixed with rhyolite {Miller 783-l-A,B). Fritillaria ojaiensis Davidson. Reservoir Canyon just north of San Luis Obispo in foothills of Santa Lucia Range {Hrusa 121, 123). *Leucojum aestivum L. Escaped from cultivation on wet bank at Atascadero Lake {Wise 1673) and in San Luis Obispo {Wain- wright 18). Poaceae (p. 52) *Alopecurus pratensis L. Campus of California Polytechnic State 1985] KEIL ET AL.: ADDENDA TO SAN LUIS OBISPO CO. FLORA 223 Univ., San Luis Obispo, at base of railroad overpass on High- land Ave. (Keil 14048). *Arundo donax L. Occasional to locally abundant along creeks and in ruderal areas: San Luis Cr. {Wise 1766, 2059); Cuesta College Rd. (Wise 2063); Cambria (Wise 2064); along Turri Rd. (Wise 2061); observed along Hwy 1 near California Men's Colony, at Atascadero, east of Arroyo Grande, and near Santa Margarita Lake. *Chloris gayana Kunth. Weed in flower bed on campus of Cali- fornia Polytechnic State Univ., San Luis Obispo (Keil 16310). *Eleusine tristachya (Lam.) Lam. Canyon Ranch south of Shandon on wet bank of Sinton South Pond (Wise 1325). *Eragrostis curvula (Schrad.) Nees. Scattered for several kilome- ters along Lopez Lake Rd. east of Arroyo Grande (Keil 14286, 18310); near San Luis Obispo just north of jet. with Los Osos Valley Rd. on Foothill Blvd., grassy roadside (Keil 18298). *Festuca arundinacea Schreb. In Los Osos Valley (Hoover 9023) and in dry woods near Rocky Butte Lookout (Hoover 9055). Misidentified by Hoover (1970) as F. elatior L. *Panicum dichotomiflorum Michx. Campus of California Poly- technic State Univ., San Luis Obispo, weed at edge of cultivated field (Ashley s.n.). *Panicum hillmanii Chase. Hwy 166 at jet. with US 101 just north of Santa Maria River (Brown 2055). *Panicum miliaceum L. Los Osos, weed in residential areas (Keil 18370; 18430). *Secale cereale L. Common along old Hwy 101 between Santa Margarita and Atascadero (Brown 2050). Sporobolus contractus A. S. Hitchc. Locally common on grassy roadside along Hwy 101 at San Miguel (Keil 18434). Potamogetonaceae (p. 50) Potamogeton illinoensis Morong. Santa Margarita Lake (Wise 908). Potamogeton nodosus Poir. Santa Margarita Lake (Wise 1723). Potamogeton pusillus L. Shepperd's Reservoir on the campus of CaHfomia Polytechnic State Univ., San Luis Obispo ( Wise 1 745); Lopez Treatment Plant near Arroyo Grande (Wise 626-A). Acknowledgments We thank the collectors who have provided their personal collection records. We are also indebted to Doug Barbe, plant taxonomist with the State of California De- partment of Food and Agriculture, for providing information regarding Reynoutria sachalinense and Chondrilla juncea. We also wish to thank the following individuals for specimen determinations: R. R. Haynes {Potamogeton), D. O. Santana {Fritillaria ojaiensis), and O. Ceska (Myriophyllum). 224 MADRONO [Vol. 32 Literature Cited Barkley, T. M. 1981. Senecio and Erechtites (Compositae) in the North American Flora: supplementary notes. Brittonia 33:523-527. Fuller, T. C. 1 966. Skeleton Weed: a serious threat to California agriculture. Dept. Ag. Bull. 55(1):20-31. Hoover, R. F. 1970. The vascular plants of San Luis Obispo County, California. Univ. Calif Press, Berkeley. Kartesz, J. T. and R. Kartesz. 1980. A synonymized checklist of the vascular flora of the United States, Canada and Greenland. Univ. North Carolina Press, Chapel Hill. MuNz, p. A. 1968. Supplement to a California flora. Univ. Calif Press, Berkeley. and D. D. Keck. 1959. A California flora. Univ. Calif Press, Berkeley. Smith, K. A. 1976. The natural resources of the Nipomo Dunes and wetlands. Calif Dept. Fish and Game Coastal Wetlands Series 15. TuTiN, T. G. et al., eds. 1964-1980. Flora Europaea. Cambridge Univ. Press. 5 vols. Webb, D. A. 1964. Reynoutria. In T. G. Tutin et al., eds.. Flora Europaea. Vol. 1. Lycopodiaceae to Platanaceae, p. 8 1 . Cambridge Univ. Press. Wise, E. A. 1 984. A flora of freshwater vascular plants of San Luis Obispo County, California. M.S. thesis, California Polytechnic State Univ., San Luis Obispo. (Received 16 Nov 1984; accepted 26 Jan 1985) ANNOUNCEMENT The 1985 Jesse M. Greenman Award The 1985 Jesse M. Greenman Award has been won by George K. Rogers for his publication ""Gleasonia, Henriquezia, and Platycarpum (Rubiaceae)" (Flora Neotro- pica Monograph No. 39). This monographic study is based on a Ph.D. dissertation from the University of Michigan Herbarium under the direction of William R. An- derson. The Greenman Award, a cash prize of $250, is presented each year by the Missouri Botanical Garden. It recognizes the paper judged best in vascular plant or bryophyte systematics based on a doctoral dissertation that was published during the previous year. Papers published during 7955 are now being considered for the 18th annual award, which will be presented in the summer of 1 986. Reprints of such papers should be sent to: Greenman Award Committee, Department of Botany, Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166-0299, U.S.A. In order to be considered for the 1986 award, reprints must be received by 1 July 1986. A REVISED VASCULAR FLORA OF TUMAMOC HILL, TUCSON, ARIZONA Janice E. Bowers and Raymond M. Turner U.S. Geological Survey, Research Project Office, 300 West Congress -FB Box 44, Tucson, AZ 85701 Abstract Tumamoc Hill, a 352-ha preserve near Tucson, Arizona, was the site of the Desert Laboratory of the Carnegie Institution of Washington from 1 903 to 1 940. The present flora of Tumamoc Hill comprises 346 specific and infraspecific taxa of vascular plants compared with 238 listed in a 1909 flora of the hill. Forty-nine of the new additions to the flora are introduced species, many of which colonized disturbed habitats created on the study area after 1940. Many of the species added may have dispersed to Tumamoc Hill from the nearby Santa Cruz River floodplain as a result of artificial wetland habitats created on the hill in recent years. Two species apparently have become locally extirpated since 1909. In 1903 the Carnegie Institution of Washington established a Des- ert Laboratory on Tumamoc Hill two miles west of Tucson, Arizona. The climate, geology, and vegetation of the hill and environs were first described by Spalding (1909); Thomber (1909) prepared the first flora of the hill. The purpose of this paper is two-fold: first, to update the plant list for a site possessing considerable significance in the history of American plant ecology and, second, to assess changes in the flora over the past 75 years. Few authors of local floras in Arizona have examined short-term floristic changes in their study areas. Amberger (1947) listed 151 species for Walnut Canyon, and six years later Spangle (1953) added 82 species to the list. A 1976 study of the vegetation of Walnut Canyon National Monument (Joyce 1976) added another 93 species to the flora but did not speculate on floristic changes that might have occurred since 1947. Similarly, Reeves (1966) listed 687 taxa for Chiricahua National Monument with no discussion of possible losses from or additions to an earlier checklist (Clark 1940). One of the few authors to assess floristic change in local floras in Arizona was Schaack (1983). He noted previous floristic work by Little (1941) and Moore (1965) in the alpine zone of San Francisco Mountain and discussed recent additions to the flora. Another was Bowers (1984), who discussed probable local extirpation between 1909 and 1983 of at least six species in the Rincon Mountains. Her assessment of floristic change was based not on an earlier flora but on collections made in 1909 by J. C. Blumer. Madrono, Vol. 32, No. 4, pp. 225-252, 20 December 1985 226 MADRONO [Vol. 32 Study Area Environment. Tumamoc Hill, an outlier of the nearby Tucson Mountains, reaches an elevation of 948 m above sea level and rises 245 m above the surrounding plain. The hill is composed of large blocks of dark brown Tertiary basalt that have weathered to a fine clay soil, forming a matrix between the rocks. Precipitation is biseasonal: from 1907 to 1983, 27.3% of the an- nual average rainfall (299 mm) fell in the winter months (December- March), and 50.8% fell in the summer (July-September). April, May, and June, called the arid foresummer by Shreve (1 9 1 1), are the driest months and a time of great moisture stress for all elements of the vegetation. Temperatures frequently exceed 38°C in the summer and occasionally drop below freezing in winter. The lowest temperature recorded on Tumamoc Hill was — 9.4°C in 1913 (Tumage and Hinckley 1938), but such low temperatures are rare. Freezing tem- peratures seldom last longer than 1 5-20 hours. Vegetation of Tumamoc Hill fits into Shreve's Arizona Upland subdivision of the Sonoran Desert (Shreve 1951). Dominant species on the rocky, basaltic slopes include Cercidium microphyllum, Car- negiea gigantea, Fouquieria splendens, Hyptis emoryi, Opuntia phaeacantha, Encelia farinosa, Lycium berlandieri, and Acacia con- stricta. The level or gently rolling plains west of the hill are char- acterized by Cercidium microphyllum, Carnegiea gigantea, Larrea divaricata, Ambrosia deltoidea, Opuntia fulgida, O. phaeacantha, O. versicolor, Fouquieria splendens, and Calliandra eriophylla. Broad washes on the plain are dominated by Cercidium floridum and Pro- sopis velutina and also support Acacia greggii, Celtis pallida, Zizy- phus obtusifolia, and other shrubs. A more detailed discussion of the vegetation of Tumamoc Hill and vegetation changes during the first part of this century can be found in Shreve (1929) and Shreve and Hinckley (1937). History. Spalding (1909) regarded the "Desert Laboratory do- main" to be Tumamoc Hill (his Zone I), the fenced plain west of the hill (his Zone II), and the Santa Cruz River floodplain and streambed (his Zones III and IV). Our definition of Tumamoc Hill includes the hill itself and the fenced plain to the west, that is, Spalding's Zones I and II. To the north, west, and south, our study area is bounded by Anklam Road, Grease wood Road, and 22nd Street, respectively; the eastern boundary is irregular. Our total area is 352 ha. The U.S.D.A. Forest Service took over the laboratory buildings and land when the Desert Laboratory closed in 1940. Under Forest Service management, and later under that of the University of Ar- izona, various incursions took place on the property, although the 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 227 area had not been disturbed since the grounds were fenced in 1907. These incursions, which included a clay quarry, a sanitary landfill, electric powerlines, gas pipelines, access roads, and a booster pump for the city water system, have had a significant effect on the flora of Tumamoc Hill, as we will show. Flora Methods. In 1968 and 1969, R. M. Turner collected plants on Tumamoc Hill to document additions and losses to the flora since 1909. Collection of Tumamoc Hill plants was resumed in 1977 and continued through 1984. Thomber's 1909 flora comprised 429 specific and infraspecific taxa of vascular plants in 68 families and 269 genera. Of these 276 occurred in Spalding's Zones I and II, the areas we examine in this report. Problems that arose in comparing Thomber's list with our own included changes in nomenclature, misapplied names, mis- identified specimens, and species listed under two or more names. Not all species listed by Thomber were documented by voucher specimens at ARIZ. We searched for vouchers for the 49 species listed by Thomber that we did not collect between 1968 and 1984, and were able to locate Thomber vouchers collected on Tumamoc Hill, "mesas, Tucson," or "mesas, Tucson Mountains," for 1 8 species. Of the 429 taxa listed by Thomber for the Desert Laboratory do- main, we eliminated those listed only for Zones III or IV (153 in all), those listed for Zones I and II but not documented by herbarium vouchers nor collected during the present study (3 1 in all), and those listed under two or more names (7 in all). Thus, Thomber's recon- structed flora consists of 238 taxa. Florist ic change. The number of taxa has apparently increased from 238 to 346 over the past 75 years. Although it is difficult to argue from negative evidence (i.e., just because Thomber did not list these "new" species does not mean they did not occur on the study area), we find meaningful pattems that suggest substantial floristic changes have occurred at Tumamoc Hill since 1909. Many of the recent additions to the Tumamoc Hill flora resulted from changes in habitat, especially from disturbance associated with construction of roads, pipelines, a clay quarry, and a sanitary landfill on the property. Although Thomber listed 52 introduced species in all, most of these were restricted to the Santa Cruz River floodplain. In contrast, 40% of the 1 26 taxa we added to the flora are not native, and the majority are closely associated with disturbance. Some of the introduced species in the fiora—Lantana horrida, Phacelia par- ryi, Molucella laevis, Melia azederach, Opuntia lindheimeri var. lin- guiformis, Dimorphotheca aurantiaca, Pennisetum ruppelii, Cypems 228 MADRONO [Vol. 32 alternifolius, and Cupressus sempervirens—SiTQ common in culti- vation in and around Tucson. Recent development of the land sur- rounding Tumamoc Hill has no doubt facilitated their spread onto our study area. Not all of the introduced species collected on the Desert Laboratory domain have become established. Bromus tec- torum, a European species common in the Great Basin, was collected on Tumamoc Hill in 1979, but has not been collected since and apparently did not become permanently established. Changes in habitat have also been responsible for the migration of some species from the Santa Cruz River floodplain (Zones III and IV) to Tumamoc Hill (Table 1). Wetland and riparian species that formerly occupied the seasonally wet bed of the Santa Cruz River now find suitable habitat at several locations on Tumamoc Hill. Spalding (1909) noticed this process occurring with Cynodon dactylon as early as 1908. Currently, artificial wetland habitats on Tumamoc Hill include the seasonal ponds at the sanitary landfill and clay quarry; the overflow from a water tank southeast of the laboratory buildings and from the booster pump on Anklam Road; the septic tank installed northwest of the laboratory buildings; and a moist ditch at a broken water main near the eastern boundary of the property. A few of the apparent "migrants," such as Poa bigelovii and Bromus arizonicus, are annuals characteristic of rocky slopes and gravelly flats and may have been overlooked by Thomber. The majority, however, are recently adventive to our study area, having capitalized upon the availability of new, suitable habitat. Of the 48 species listed in Table 1, 20 are introduced. In addition to species that may have migrated to our study area from the Santa Cruz River floodplain, several other moisture-loving species not listed by Thomber are found in artificial wetland habitats on Tumamoc Hill: Scirpus maritimus var. paludosus, Diplachne fascicularis, Cyperus alternifolius, Tamarix pentandra, Phalaris minor, Molucella laevis, Conyza bonariensis, Typha domingensis, and Cupressus sempervi- rens. Certain apparent additions to the flora since 1909 are not easily explained. No doubt Thomber overlooked more than a few species when preparing his flora, and this may account for the recent ad- dition of characteristic desert species such as Matelea parvifolia. Astragalus wootonii, Eriogonum thurberi, Oenothera primiveris, Thamnosma texana, Yabea microcarpum, Eucrypta micrantha. Eu- phorbia micromera, Pectocarya recurvata. Ambrosia dumosa, Filago arizonica, Filago depressa, and Tillaea erecta. A few taxa added to the list were probably not overlooked by Thomber but are new to the flora. One of these, Polanisia dodecandra subsp. trachysperma, was first collected in a wash near Anklam Road in 1980 and ap- parently occurs nowhere else on the study area. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 229 Table \. Plants of Tumamoc Hill Listed by Thornber (1909) only for the Santa Cruz River or Its Floodplain. * = introduced. Koeberlinia spinosa Forty-nine species listed by Thornber for Zones I or II were not collected by us. Eighteen of these are documented by ARIZ voucher specimens collected on Tumamoc Hill, on "mesas, Tucson," or on "mesas, Tucson Mountains." The remaining 31 species were not documented by voucher specimens at ARIZ, and we did not include them in our reconstruction of Thomber's list. It is Ukely that some of these species still occur on the study area but were overlooked. Two species, Olneya tesota and Simmondsia chinensis, are of more interest, because they may have been locally extirpated. (Forestiera shrevei might be included here, since Thornber collected it on Tu- mamoc Hill and we did not; however, it still occurs within one- quarter mile of the boundary of our study area.) Although we may have overlooked these species, both are large, woody plants that are not easily missed. Spalding noted that the Olneya growing near the east edge of his permanent plot # 1 2 was the only individual known to occur on the Desert Laboratory grounds (Spalding, unpubl. notes, 1906). This individual was shown on maps of permanent plot #12 made by Shreve in 1929 and 1936, but had disappeared by 1948 when the plot was mapped again. Olneya is frost-sensitive, and the Amaranthus palmeri Sarcostemma cynanchoides var. Teucrium cubense *Malva parvijlora Sphaeralcea coulteri Boerhaavia coccinea *Avena fatua Bromus arizonicus *Bromus rubens *Bromus willdenowii *Cynodon dactylon *Echinochloa colonum *Eragrostis cilianensis Eriochloa lemmonii var. gracilis *Hordeum murinum Hordeum pusillum Poa bigelovii *Polypogon monspeliensis Setaria macrostachya * Sorghum halepense Androsace Occident alis Clematis drummondii Maurandya antirrhinijlora *Nicotiana glauca Physalis acutifolia *Tribulus terrestris cynanchoides Aster subulatus var. ligulatus Baccharis salicifolia *Centaurea melitensis Conyza canadensis Erigeron divergens Gutierrezia microcephala Heterotheca psammophila Hymenothrix wislizenii ^Matricaria matricarioides *Sonchus oleraceus Verbesina encelioides *Matthiola bicornis Sambucus mexicana Atriplex canescens Chenopodium fremontii *Chenopodium murale *Medicago polymorpha var. vulgaris *Melilotus indicus Corydalis aurea *Erodium cicutarium Nama hispidum 230 MADRONO [Vol. 32 single individual on Tumamoc Hill may have died following a severe freeze such as the one that occurred in 1937. Alternatively, it may have died after senescence. Simmondsia chinensis was collected on Tumamoc Hill in 1905 {Thornber 2576), and although it is common in the Tucson Mountains, it has failed to reoccupy the hill. Perhaps individuals of Simmondsia were so few that the level of reproduction fell below that necessary to maintain the population. If the few remaining individuals in the population were all of one sex, repro- duction would have been impossible, and the population would have died out eventually. Although adults are hardy, seedlings are sus- ceptible to freezing, drought, and predation by rodents (Sherbrooke 1977). Annotated checklist. The annotated checklist includes 346 specific and infraspecific taxa, in 67 families and 241 genera, known either to occur presently on the Tumamoc Hill property or to have occurred historically and for which vouchers exist. Habitat, local distribution, and relative abundance are noted for most. Species not listed by Thornber for Zones I and II are denoted by an asterisk. Species collected by Thornber and others for which vouchers exist, but which we did not collect, are denoted by a dagger. Names applied by Thornber are listed in brackets where appropriate. Nomenclature follows Lehr (1978) and Lehr and Pinkava (1980, 1982). Nomen- clature for cultivated species not listed in Lehr or Lehr and Pinkava follows Bailey and Bailey (1976). A full set of our vouchers has been deposited at ARIZ. Additional vouchers have been deposited at ASU, BRI, ENCB, HUF, LIL, MEXU, MICH and MNA. Vascular Plants of Tumamoc Hill^ Pterophyta Adiantaceae Cheilanthes wootonii Maxon. Rocky slopes; under trees; rare. Cheilanthes wrightii Hook. Rocky, north-facing slopes; rare. Notholaena cochisensis Goodding. Rocky, north-facing slopes; oc- casional. Notholaena standleyi Maxon. Rocky slopes; occasional to com- mon. Pellaea truncata Goodding [Pellaea wrightiana Hook.]. Rocky, north-facing slopes; occasional. See text for explanation of symbols. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 231 CONIFEROPHYTA Cupressaceae *Cupressus sempervirens L. Local, along moist ditch; tree com- monly cultivated in Tucson, probably spreading onto our area from nearby housing developments. Ephedraceae Ephedra trifurca Torr. Gravelly flats and along washes; occasional to common. AnTHOPH YTA — DiCOTYLEDONEAE Acanthaceae Anisacanthus thurberi (Torr.) Gray. Along washes; common; usu- ally flowering in the spring. Carlowrightia arizonica Gray. Rocky slopes; occasional. Ruellia nudiflora (Engelm. & Gray) Urban. Banks of washes, in shade of trees; locally common. Siphonoglossa longiflora (Torr.) Gray. Rocky slopes, often in shade of trees; common. Aizoaceae Trianthema portulacastrum L. Disturbed sites, abundant on san- itary landfill; introduced. Amaranthaceae Amaranthus fimbriatus (Torr.) Benth. Washes and sandy flats; oc- casional summer annual. * Amaranthus palmeri Wats. Washes and roadsides; common sum- mer annual. Tidestromia lanuginosa (Nutt.) Standi. Gravelly slopes; locally common summer annual. Anacardiaceae *Rhus lancea L. f Moist soil, local, ditch at broken water main; an ornamental common in cultivation in Tucson; probably spreading to our area from nearby housing developments. Apiaceae Bowlesia incana Ruiz & Pa v. Rocky slopes and gravelly flats, often under shrubs, trees or rocks; common spring annual. 232 MADRONO [Vol. 32 Daucus pusillus Michx. Rocky slopes or gravelly flats, often under shrubs, trees, or rocks; common spring annual. Spermolepis echinata (Nutt.) Heller. Rocky slopes and gravelly flats; common spring annual. *Yabea microcarpum (Hook. & Am.) K.-Pol. Rocky slopes; com- mon spring annual. Apocynaceae Haplophyton crooksii L. Rocky slopes, flowering in spring; com- mon. Aristolochiaceae Aristolochia watsonii Woot. & Standi. Disturbed sites on flats; ap- parently uncommon. Asclepiadaceae *Asclepias nyctaginifolia Gray. Along washes; apparently uncom- mon. "fCynanchum arizonicum (Gray) Shinners. Thomber 4855, 8989. *Matelea parvifolia (Torr.) Woods. Climbing on cacti, trees, and shrubs; gravelly flats; rare. *Sarcostemma cynanchoides Decne. var. cynanchoides. Along washes; climbing on shrubs; apparently rare. *Sarcostemma cynanchoides Decne. var. hartwegii (Vail) Shin- ners. Along washes; climbing on trees and shrubs; common. Asteraceae Acourtia nana (Gray) Reveal & King. Gravelly flats, often under shrubs; flowering in spring. Acourtia wrightii (Gray) Reveal & King. Banks of washes and rocky slopes; common. Ambrosia confertiflora DC. Washes and dirt roads, often in dis- turbed areas; locally common. Ambrosia deltoidea (Torr.) Payne. Rocky slopes and gravelly flats; a common dominant. * Ambrosia dumosa (Gray) Payne. Gravelly flats and rocky slopes; near clay quarry and on south slopes of hill; locally common; probably overlooked by Thomber. Aster subulatus Michx. var. ligulatus Shinners. Moist soil; local, near pond at sanitary landfill and along ditch below water tank; apparently recently adventive to our study area. Baccharis brachyphylla Gray [Baccharis wrightii Gray]. Gravelly flats; local; occasional. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 233 *Baccharis salicifolia (Ruiz & Pav.) Pers. Low-lying, disturbed sites; apparently recently adventive to our study area. Baccharis sarothroides Gray [Baccharis emoryi Gray]. Washes and disturbed sites; locally common. Bahia absinthifolia Benth. Rocky slopes and gravelly flats; com- mon on edges of old roadways and on soils containing caliche. Baileya multiradiata Harv. & Gray. Along washes and on sandy flats; common; flowering sporadically throughout the year. Bebbia juncea (Benth.) Greene. Gravelly flats, often along washes; occasional. *Brickellia californica (Torr. & Gray) Gray. Rocky flats, in shade of trees; rare. Brickellia coulteri Gray. Along washes and on rocky slopes, often under trees; common; flowering sporadically throughout the year. Calycoseris wrightii Gray. Gravelly flats and rocky slopes; a com- mon and showy spring annual. *Centaurea melitensis L. Disturbed sites; common on sanitary landfill; introduced; apparently recently adventive to our study area. Chaenactis stevioides Hook. & Am. Gravelly flats and rocky slopes, often under shrubs and trees; a common and showy spring annual. *Cirsium neomexicanum Gray. Rocky slopes; rare. "^Conyza bonariensis (L.) Cronq. Moist soil; local, along moist ditch at broken water main. *Conyza canadensis (L.) Cronq. Disturbed sites; local along moist ditches and other damp spots; perhaps recently adventive to our study area. Conyza coulteri Gray. Moist soil; local, near pond in sanitary land- fill and at moist ditch at broken water main. *Dimorphotheca aurantiaca DC. Gravelly flats and along washes; occasional; apparently adventive from nearby housing develop- ments; an introduced ornamental common in cultivation in and around Tucson. Dyssodia pentachaeta (DC.) Robins. Rocky flats; common on soils containing caliche and on disturbed sites such as dirt roads and scraped ground. Dyssodia porophylloides Gray. Rocky slopes and gravelly flats; un- common. Encelia farinosa Gray. Rocky slopes; a common dominant; flow- ering in late winter and early spring. Ericameria laricifolia (Gray) Shinners. Rocky slopes; rare; only a few individuals known from the study area; a marginal popu- lation at the lower limit of its range. 234 MADRONO [Vol. 32 *Erigeron divergens Torr. & Gray. Rocky slopes and gravelly flats; common; flowering sporadically throughout the year. Eriophyllum lanosum Gray. Rocky slopes and gravelly or rocky flats; common spring annual. Evax multicaulis DC. Gravelly or rocky flats; locally common spring annual. *Filago arizonica Gray. Rocky slopes and gravelly or rocky flats; common spring annual. Filago californica Nutt. Rocky slopes; common spring annual. *Filago depressa Gray. Gravelly flats, often under shrubs; rare. Gaillardia arizonica Gray. Sandy flats and washes; locally com- mon; flowering in late spring. Gutierrezia arizonica (Gray) Lane. Gravelly flats; rare; flowering in late spring. *Gutierrezia microcephala (DC.) Gray. Rocky slopes; rare; perhaps recently adventive to our study area. *Heterotheca psammophila Wagenkn. Disturbed sites and low-lying areas; locally common; perhaps recently adventive to our study area. *Hymenothrix wislizenii Gray. Often on disturbed sites, common along roads; perhaps recently adventive to our study area. Isocoma tenuisecta Greene. Gravelly flats, often on disturbed sites; common on scraped ground and along roads. *Lactuca serriola L. Along washes; usually under shrubs; occa- sional; introduced. '\Lasthenia californica DC. ex Lindl. Thornber 5307. ^Machaeranthera gracilis (Nutt.) Shinners. Thornber 2028. Machaer anther a pinnatifida (Hook.) Shinners. Gravelly flats, often on disturbed sites; common. *Machaeranthera tagetina Greene. Disturbed sites; along roads and near buildings; locally common. Malacothrix californica DC. var. glabrata Eaton. Gravelly flats; apparently rare; showy spring annual. Malacothrix clevelandii Gray. Rocky slopes, under trees and shrubs; rare spring annual. ^Malacothrix coulteri Gray. Thornber 387, 4621. ^Matricaria matricarioides (Less.) Porter. Disturbed sites; intro- duced; perhaps recently adventive to our study area; Schoen- wetter T-28. Microseris linearifolia (DC.) Schultz-Bip. Rocky slopes; common spring-flowering annual. Monoptilon bellioides (Gray) H. M. Hall. Rocky or gravelly flats; common spring annual. Parthenium incanum H.B.K. Rocky slopes and gravelly flats; lo- cally common. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 235 Pectis papposa Harv. & Gray. Gravelly flats, often in low-lying areas; locally common summer annual. Porophyllum gracile Benth. Rocky slopes and gravelly flats; locally common. Psilostrophe cooperi (Gray) Greene. Gravelly flats; abundant; flow- ering after winter and summer rains. Raflnesquia neomexicana Gray. Rocky slopes; a common and showy spring annual. *Senecio douglasii DC. var. douglasii. Often along washes; com- mon spring- flowering annual. Senecio lemmonii Gray. Rocky slopes; common spring annual, rarely flowering in late fall. *Sonchus oleraceus L. Rocky slopes and gravelly flats; locally com- mon in moist areas; introduced. Apparently recently adventive to our study area. Stephanomeria pauciflora (Torr.) A. Nels. Along washes, also on rocky slopes; occasional. *Stylocline gnaphaloides Nutt. Gravelly flats, perhaps rare, but easily overlooked. Stylocline micropoides Gray. Rocky slopes and gravelly or rocky flats; common spring annual. Trixis calif ornica Kellogg. Rocky slopes; common. *Verbesina encelioides (Cav.) Benth. & Hook. Locally common in low-lying areas; perhaps recently adventive to our study area. Zinnia acerosa (DC.) Gray [Zinnia grandiflora Nutt.]. Gravelly flats; locally common; often on soils containing caliche. Boraginaceae Amsinckia intermedia Fisch. & Mey. Rocky slopes and gravelly flats; a common spring annual. Amsinckia tessellata Gray. Disturbed sites; apparently only locally common. Cryptantha angustifolia (Torr.) Greene. Gravelly or sandy flats; an uncommon spring annual. Cryptantha barbigera (Gray) Greene. Rocky slopes and sandy or gravelly flats; a common spring annual. Cryptantha micrantha (Torr.) Johnst. Sandy washes; apparently rare. Cryptantha nevadensis Nels. & Kenn. [Cryptantha intermedia (Gray) Greene]. Rocky slopes; common spring annual. Cryptantha pterocarya (Torr.) Greene. Rocky and gravelly slopes and flats, often under shrubs; a common spring annual. Harpagonella palmeri Gray. Rocky slopes and flats; typically sprawling over and between rocks; a common spring annual. 236 MADRONO [Vol. 32 Lappula redowskii (Homem.) Greene var. redowskii. Gravelly flats, often in disturbed areas; common spring annual. Lappula redowskii (Hornem.) Greene var. cupulatum (Gray) Jones. Gravelly flats; rare. Pectocarya heterocarpa Johnst. Gravelly flats; often growing with the next two species; common. Pectocarya platycarpa Munz & Johnst. Gravelly flats; occasional. "^Pectocarya recurvata Johnst. Rocky and gravelly flats; locally abundant. Plagiobothrys arizonicus (Gray) Greene. Gravelly flats; an uncom- mon spring annual. '\ Plagiobothrys pringlei Greene. Thornber 533, 2206. Tiquilia canescens (DC.) A. Richardson. Gravelly flats and dirt roads; especially common on soils containing caliche. Brassicaceae Arabis perennans Wats. Rocky, north-facing slopes; local and un- common. *Brassica tournefortii Gouan. Disturbed ground; locally common along roads; introduced. Caulanthus lasiophyllus (Hook. & Am.) Payson. Rocky slopes and gravelly flats, often under shrubs and trees; common spring annual. Descurainia pinnata (Walt.) Britt. Rocky slopes and gravelly flats; common spring annual. Draba cuneifolia Nutt. Gravelly and rocky flats, often under trees; common; flowering early in the spring. *Dryopetalon runcinatum Gray. Rocky, north-facing slopes; un- common. Lepidium lasiocarpum Nutt. Rocky slopes and gravelly flats; com- mon spring annual. "^Lepidium oblongum Small. Disturbed sites; common on sanitary landfill; introduced. Lesquerella gordoni (Gray) Wats. Rocky slopes and gravelly or rocky flats; common spring annual. *Matthiola bicornis (Sibth. & Smith) DC. Disturbed sites; com- mon on and near sanitary landfill; introduced; perhaps recently adventive to our study area. * Sisymbrium altissimum L. Along washes; apparently uncommon; introduced. Turner 78-11 is unusual in its soft pubescence and rather wide leaf segments. ''^Sisymbrium irio L. Rocky slopes and gravelly flats, often on dis- turbed sites; introduced. Streptanthus arizonicus Wats. Rocky slopes and gravelly or rocky flats; common spring annual. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 237 Thysanocarpus curvipes Hook. Rocky slopes, often under shrubs and trees; common spring annual. Cactaceae Carnegiea gigantea (Engelm.) Britt. & Rose. Rocky slopes and gravelly flats; a common dominant. *Echinocereus fasciculatus (Engelm.) L. Benson. Gravelly flats; oc- casional. Echinocereus fendleri Engelm. Gravelly flats; common. Ferocactus wislizenii (Engelm.) Britt. & Rose. Rocky slopes and gravelly flats; occasional; flowering in August. Mammillaria microcarpa Engelm. [Cactus grahamii (Engelm.) Kuntze]. Gravelly flats, often under trees, shrubs, and large cacti; common. *Opuntia ficus-indica (L.) Mill. Gravelly flats; common in culti- vation in and around Tucson; apparently spreading to our study area from nearby housing developments. Opuntia fulgida Engelm. Gravelly flats west of hill; locally abun- dant. *Opuntia kleiniae DC. var. tetracantha (Toumey) Marshall. Rocky flats; widely scattered; locally common. Possibly of hybrid or- igin between O. leptocaulis and O. versicolor. Opuntia leptocaulis DC. Rocky slopes and gravelly flats; some- times forming impenetrable thickets. *Opuntia lindheimeri Engelm. var. linguiformis (Griffiths) L. Ben- son. Gravelly flats; common in cultivation in and around Tuc- son; apparently spreading to our area from nearby housing de- velopments. Opuntia phaeacantha Engelm. var. discata (Griffiths) Benson & Walkington. Rocky slopes and gravelly ffats; common; inter- grading with O. p. var. major. Opuntia phaeacantha Engelm. var. major Engelm. Rocky slopes and gravelly flats; common; intergrading with O. p. var. discata. Opuntia spinosior (Engelm.) Toumey. Gravelly flats; uncommon. Opuntia versicolor Engelm. Rocky slopes and gravelly flats; com- mon. Peniocereus greggii (Engelm.) Britt. & Rose. Gravelly flats, usually under trees; occasional. Campanulaceae Nemacladus glanduliferus Jeps. var. orientalis McVaugh [Nemacla- dus ramosissimus Nutt.]. Rocky slopes and gravelly flats; un- common spring annual. I 238 MADRONO [Vol. 32 \ Caprifoliaceae j *Sambucus mexicana Presl. Local, wet area by septic system near laboratory buildings; apparently recently adventive to our study area. Caryophyllaceae *Herniaria cinerea DC. Gravelly flats; rare; introduced. Loeflingia squarrosa Nutt. Gravelly flats; an uncommon spring annual. Silene antirrhina L. Rocky slopes and gravelly and rocky flats; common spring annual. Chenopodiaceae *Atriplex canescens (Pursh) Nutt. Along washes and on rocky slopes; occasional. Atriplex elegans (Moq.) D. Dietr. Disturbed sites; common on san- itary landfill. *Chenopodium fremontii Wats. Gravelly flats and rocky slopes; often under trees; common spring annual. *Chenopodium murale L. Disturbed site near laboratory buildings; introduced. fMonolepis nuttalliana (Schult.) Greene. Gravelly flats and wash- es; occasional spring annual; P. S. Martin 1053. *Salsola iberica Sennen & Pau. Disturbed sites; abundant on san- itary landfill; introduced. Cleomaceae *Polanisia dodecandra (L.) DC. var. trachysperma (Torr. & Gray) litis. Washes and roadsides; localized in somewhat disturbed sites; apparently recently adventive to our study area. Convolvulaceae *Ipomoea barbatisepala Gray. Climbing on shrubs and trees in rocky slopes, often in shallow ravines. Crassulaceae *Tillaea erecta Hook. & Am. Gravelly flats; perhaps local, but easily overlooked. Cucurbitaceae '\Apodanthera undulata Gray. Thornber 5259. Cucurbita digitata Gray. Gravelly flats and low-lying spots; un- common. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 239 Tumamoca madougalii Rose [Maximowiczia tripartita Cogn. var. tenuisecta Wats.]. Gravelly flats, climbing on shrubs and trees; occasional on our study area, but rare in Arizona. Described in 1912 from specimens collected on Tumamoc Hill. Euphorbiaceae Argythamnia neomexicana Muell.-Arg. Rocky slopes and gravelly flats; occasional; flowering after summer rains. Euphorbia capitellata Engelm. Rocky slopes; common; flowering sporadically throughout the year. Euphorbia Jlorida Engelm. Gravelly flats and shallow, sandy wash- es; common summer annual. ^Euphorbia heterophylla L. Along washes and in low-lying areas; locally common summer annual. "^Euphorbia hyssopifolia L. Gravelly flats and shallow, sandy wash- es; common summer annual. "^Euphorbia micromera Boiss. Gravelly flats and sandy washes; common summer annual. Euphorbia pediculifera Engelm. Gravelly flats and sandy washes; common; flowering in summer. Euphorbia setiloba Engelm. Gravelly and rocky flats; flowering af- ter winter and summer rains; occasional. '\ Euphorbia serrula Engelm. Thornber 47, 8948. Jatropha cardiophylla (Torr.) Muell.-Arg. Rocky slopes; occasion- al. *Tragia nepetaefolia Cav. Rocky slopes; uncommon. Fabaceae Acacia constricta Benth. Gravelly flats, rocky slopes, and along washes; common. Acacia greggii Gray var. arizonica Isely. Washes, gravelly flats, and rocky slopes; common. Astragalus nuttallianus DC. Rocky slopes and gravelly flats; com- mon spring annual. * Astragalus wootonii Sheldon. Gravelly flats; occasional. Calliandra eriophylla Benth. Gravelly flats and rocky slopes; com- mon. Cercidium floridum Benth. Along the larger washes; a common dominant. Cercidium microphyllum (Torr.) Rose & Johnst. Gravelly flats and rocky slopes; a common dominant. "fHoffmanseggia glauca (Ort.) Eifort. Thornber s.n. (1904). Lotus humistratus Greene. Gravelly flats; common spring annual. Lotus tomentellus Greene [Hosackia humilis Greene]. Gravelly flats; occasional spring annual. 240 MADRONO [Vol. 32 Lupinus concinnus Agardh. Along washes and on gravelly flats; occasional spring annual. Lupinus sparsiflorus Benth. Rocky slopes; common spring annual. Marina parryi (Torr. & Gray) Bam. Rocky slopes, often along paved roads; locally common. *Medicago polymorpha L. var. vulgaris (Benth.) Shinners. Local, moist site near buildings; introduced; apparently recently ad- ventive to our study area. *Melilotus indicus (L.) All. Pond at sanitary landfill; locally com- mon; introduced; apparently recently adventive to our study area. Nissolia schottii (Torr.) Gray. Rocky slopes; climbing on trees and shrubs; occasional. -fOlneya tesota Gray. Spalding (1909). *Parkinsonia aculeata L. Disturbed sites; along roads and on san- itary landfill; an introduced ornamental common in cultivation in and around Tucson. Prosopis velutina Woot. Gravelly flats and along washes; a com- mon dominant. Senna bauhinioides (Gray) Irwin & Bameby. Thornber s.n.{\ 903). Senna coves ii (Gray) Irwin & Bameby. Gravelly flats and rocky slopes; common; flowering after summer rains. *Sphinctospermum constrictum (Wats.) Rose. Rocky slopes; rare summer annual; not listed by Thomber, although collected by him {Thornber 4851) on Tumamoc Hill in August 1906. Vicia ludoviciana Nutt. Rocky slopes, climbing on shrubs and an- nuals; common spring annual. Fouquieriaceae Fouquieria splendens Engelm. Gravelly flats and rocky slopes; a common dominant. Fumariaceae *Corydalis aurea Willd. Along washes; rare spring annual. Geraniaceae *Erodium cicutarium (L.) L'Her. Rocky slopes and gravelly flats; locally common spring annual; introduced. Listed by Thomber only for the Santa Cruz River floodplain, but occurring on Tu- mamoc Hill according to Spalding (1909). Erodium texanum Gray. Gravelly and rocky flats; locally common spring annual. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 241 Hydrophyllaceae Eucrypta chrysanthemifolia (Benth.) Greene. Rocky slopes, often under trees; occasional spring annual. "^Eucrypta micrantha (Torr.) Heller. Rocky slopes and gravelly flats, often under trees and shrubs; common spring annual. *Nama hispidum Gray. Along washes; occasional spring annual. *Phacelia affinis Gray. Along washes; rare spring annual. ^Phacelia arizonica Gray. Thornber 4013. Phacelia crenulata Torr. Rocky slopes and gravelly flats; common spring annual. Phacelia distans Benth. Rocky slopes and gravelly flats; often re- clining on shrubs and other annuals; common spring annual. "^Phacelia parryi Torr. Gravelly flats; local. An exotic species in Arizona, native to California, cultivated nearby at St. Mary's Hospital and adventive to our study area. Koeberliniaceae *Koeberlinia spinosa Zucc. Gravelly flats and banks of washes; locally common. Krameriaceae Krameria grayi Rose & Painter. Gravelly flats and rocky slopes; occasional. Krameria parvifolia Benth. Gravelly flats and rocky slopes; occa- sional. Lamiaceae Hyptis emoryi Torr. Rocky slopes; common. *Molucella laevis L. Disturbed sites, low-lying areas; locally com- mon; an introduced ornamental cultivated in and around Tuc- son. Salvia columbariae Benth. Gravelly flats and along washes; oc- casional spring annual. *Teucrium cubense Jacq. Along washes; locally common; perhaps recently adventive to our study area. Linaceae Linum lewisii Pursh. Rocky slopes and gravelly flats; uncommon. Although L. lewisii is described by Kearney and Peebles (1960) as a perennial herb, it is an annual on Tumamoc Hill and on other desert mountain ranges in southern Arizona. It can be distinguished from L. usitatissimum (cultivated flax), which is 242 MADRONO [Vol. 32 also an annual, by its capitate stigmas and ovate sepals. L. usitatissimum has longitudinal stigmas and acuminate, ciliate sepals. Loasaceae Mentzelia albicaulis Dougl. Rocky slopes and gravelly flats; oc- casional spring annual. Mentzelia multiflora (Nutt.) Gray. Rocky slopes; occasional. Loranthaceae Phoradendron californicum Nutt. Gravelly flats and along washes; parasitic on a variety of trees and shrubs; common. Malpighiaceae Janusia gracilis Gray. Rocky slopes and gravelly flats; common. Malvaceae Abutilon incanum (Link.) Sweet subsp. pringlei (Hochr.) Felger & Lowe. Rocky slopes; occasional. Anoda pentaschista Gray. Rocky slopes; rare; flowering in the sum- mer. fEremalche exilis (Gray) Greene. Thornber 4884, 5326. Herissantia crispa (L.) Brizicky. Rocky slopes; common. Hibiscus coulteri Harv. Rocky slopes; often among shrubs; occa- sional. Hibiscus denudatus Benth. Rocky slopes and gravelly flats; locally common. *Malva parviflora L. Disturbed sites, low-lying areas; locally com- mon; introduced; apparently recently adventive to our study area. Rhynchosida physocalyx (Gray) Fryxell. Disturbed sites and on banks of washes; locally common. Sida procumbens Swartz. Sandy or gravelly flats; occasional sum- mer-flowering perennial herb. Sphaeralcea angustifolia (Cav.) G. Don. var. cuspidata Gray. Disturbed sites and along washes; local and uncommon. ^Sphaeralcea coulteri (Wats.) Gray. Gravelly flats; uncommon spring annual. *Sphaeralcea emoryi Torr. var. californica (Parish) Shin- ners. Rocky, north-facing slopes; rare and local. Sphaeralcea laxa Woot. & Standi. [Sphaeralcea pedata Torr.]. Rocky slopes and gravelly flats; common. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 243 Martyniaceae Proboscidea altheaefolia (Benth.) Decne. Gravelly flats and sandy washes; occasional; flowering in summer. "^Proboscidea parviflora (Wool.) Woot. & Standi. Disturbed sites; apparently local, near sanitary landfill. Meliaceae *Melia azedarach L. Disturbed sites; local, at sanitary landfill; an ornamental commonly cultivated in and around Tucson. Nyctaginaceae Allionia incarnata L. Gravelly flats, rocky slopes and along washes; common; flowering sporadically throughout the year. *Boerhaavia coccinea Mill. Banks of washes; locally common. Boerhaavia coulteri (Hook, f.) Wats. Sandy flats; occasional sum- mer annual. Boerhaavia intermedia Jones. Rocky slopes, low-lying areas, often on disturbed sites; common. '\ Boerhaavia megaptera Standi. Thornber 161, 4863. ^Boerhaavia spicata Choisy. Rocky slopes and washes; uncommon summer annual; B. Fink s. n. *Boerhaavia wrightii Gray. Disturbed sites; locally common on roadbanks; not listed by Thornber, although collected by him (Thornber 2617) on Tumamoc Hill in September 1903. Commicarpus scandens L. Along washes, scandent on trees and shrubs; occasional. Oleaceae '\Forestiera shrevei Standi. Thornber s.n. (1906). Menodora scabra Gray. Rocky slopes and gravelly flats; common; flowering after spring and summer rains. Onagraceae Camissonia californica (Nutt. ex Torr. & Gray) Raven. Rocky slopes and flats; common spring annual. Camissonia chamaenerioides (Gray) Raven. Gravelly flats; occa- sional spring annual. Camissonia clavaeformis (Torr. & Frem.) Raven. Gravelly or sandy flats; locally common spring annual. "fOenothera caespitosa Nutt. Shreve s.n. (1931). ^Oenothera primiveris Gray. Gravelly flats and rocky slopes; oc- casional spring annual. 244 MADRONO [Vol. 32 Orobanchaceae *Orobanche cooperi (Gray) Heller. Rocky slopes and disturbed sites, particularly favoring berms along dirt roads; uncommon. Papaveraceae Argemone pleicantha Greene subsp. pleicantha. Gravelly flats, often in disturbed areas; occasional. Eschscholzia californica Cham, subsp. mexicana (Greene) C. Clark. Rocky slopes; locally common spring annual. Plantaginaceae Plantago insularis Eastw. Gravelly flats and rocky slopes; com- mon; flowering early in spring. Plantago patagonica Jacq. Rocky slopes and gravelly flats; com- mon spring annual. Plantago rhodosperma Decne. [Plantago virginica L.]. Moist soil on rocky slopes; local. Polemoniaceae Eriastrum diffusum (Gray) Mason [Gilia floccosa Gray]. Rocky slopes and gravelly flats; common spring annual. Gilia stellata Heller [Gilia glutinosa Benth.; Gilia inconspicua (Small) Dougl. var. sinuata Gray]. Rocky slopes and gravelly flats; occasional spring annual. flpomopsis longiflora (Torr.) V. Grant. Thornber 4439, 4988. Linanthus bigelovii (Gray) Greene. Rocky slopes; occasional spring annual. Polygalaceae Polygala macradenia Gray. Rocky slopes and gravelly flats, often on soil containing caliche; occasional. Polygonaceae Chorizanthe brevicornu Torr. Gravelly flats; common spring an- nual. Chorizanthe rigida (Torr.) Torr. & Gray. Gravelly flats; locally common spring annual. Eriogonum abertianum Torr. Gravelly flats, disturbed sites; lo- cally common. Eriogonum deflexum Torr. Along washes; common summer an- nual. Eriogonum maculatum Heller. Gravelly flats; occasional spring annual. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 245 "^Eriogonum polycladon Benth. Along washes; locally common. *Enogonum thurberi Torr. Gravelly flats; occasional. Eriogonum trichopes Torr. Gravelly flats and sandy washes; com- mon. Portulacaceae Calyptridium monandrum Nutt. Sandy flats; locally common. Primulaceae *Androsace occidentalis Pursh. Rocky, north-facing slopes; un- common spring annual. Ranunculaceae Anemone tuberosa Rydb. Rocky slopes; common; flowering in spring. '^'Clematis drummondii Torr. & Gray [Clematis ligusticifolia Nutt.]. Along washes, climbing on trees and shrubs; occa- sional. Delphinium scaposum Greene. Rocky slopes and gravelly flats; common. Resedaceae Oligomeris linifolia (Vahl) Macbr. Gravelly flats, often in dirt roads; locally common. Rhamnaceae Condalia warnockii M. C. Johnst. var. kearneyana M. C. Johnst. Gravelly flats and borders of washes; occasional. Zizyphus obtusifolia (Hook, ex Torr. & Gray) Gray var. canescens (Gray) M. C. Johnst. Gravelly flats and along washes; occa- sional. Rubiaceae Galium proliferum Gray. Rocky slopes; locally common spring annual. Galium stellatum Kellogg. Rocky slopes; rare. Rutaceae *Thamnosma texana (Gray) Torr. Gravelly flats, often on banks of washes and under trees; locally common. 246 MADRONO [Vol. 32 Scrophulariaceae *Maurandya antirrhiniflora Humb. & Bonpl. Along washes, climbing on trees; occasional to common. Orthocarpus purpurascens Benth. Rocky slopes; locally abundant spring annual. Penstemon parryi Gray [Penstemon wrightii Hook.]. Gravelly flats, rocky slopes and along washes; occasional; flowering in spring. Simmondsiaceae "fSimmondsia chinensis (Link.) Schneid. Thornber 2576. Solanaceae *Datura discolor Bemh. Sandy flats, disturbed sites; locally com- mon. Lycium berlandieri Dunal. Rocky slopes; common dominant. Lycium exsertum Gray. Rocky slopes and along washes; occasion- al. *Nicotiana glauca Graham. Rocky slopes and along moist ditch; locally common; introduced; apparently recently adventive to our study area. Nicotiana trigonophylla Dunal. Rocky slopes; occasional. *Physalis acutifolia (Miers) Sandw. Rocky slopes, moist soil; rare; perhaps recently adventive to our study area. Physalis crassifolia Benth. Gravelly flats and rocky slopes; uncom- mon; flowering mostly in summer. Quincula lobata (Torr.) Raf. Gravelly and sandy flats, occasionally in washes; locally common; flowering after spring and summer rains. Solarium elaeagnifolium Cav. Disturbed sites, near buildings and along roads; occasional. Sterculiaceae *Ayenia compacta L. Rocky slopes and flats, often under shrubs and trees; locally common. Not listed by Thornber although collected by him {Thornber 2561) on Tumamoc Hill in March 1905. Ayenia microphylla Gray. Rocky slopes and gravelly flats, often under trees; occasional. ^Hermannia pauciflora Wats. Thornber 2281. Tamaricaceae *Tamarix pentandra Pall. Disturbed sites, moist soil; common near ponds at sanitary landfill and clay quarry; introduced. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 247 Ulmaceae Celtis pallida Torr. Rocky slopes and along washes; common. Urticaceae Parietaria hespera Hinton [Parietaria debilis Forst. f.]. Rocky slopes, usually in recesses under rocks and boulders; common spring annual. Verbenaceae Aloysia wrightii (Gray) Heller. Rocky, north-facing slopes and along washes; locally common. Glandularia gooddingii (Briq.) Solbrig [Verbena ciliata Benth.]. Rocky slopes; common; flowering sporadically throughout the year. *Lantana horrida H.B.K. Disturbed sites; occasional; an orna- mental commonly cultivated in and around Tucson. Tetraclea coulteri Gray. Disturbed sites, often in low-lying areas; locally common. Zygophyllaceae Kallstroemia grandiflora Torr. Rocky and gravelly flats and along washes; occasional summer annual. * Kallstroemia hirsutissima Vail. Disturbed sites and along roads; occasional. Larrea divaricata Cav. subsp. tridentata (Sesse & Moc. ex DC.) Felger & Lowe. Gravelly flats and rocky slopes; common dom- inant; flowering sporadically throughout the year. *Tribulus terrestris L. Disturbed sites; occasional; introduced; ap- parently recently adventive to our study area. AnTHOPHYTA — MONOCOTYLEDONEAE I Agavaceae "^Agave americana L. Perhaps local; under tree along wash; an ornamental commonly cultivated in and around Tucson, prob- ably spreading onto our area from nearby housing develop- ments. Yucca elata Engelm. Gravelly flats; rare; only one individual known from the study area. Spalding (1 909) also found only one plant; this was not the same individual currently found on the study area. 248 MADRONO [Vol. 32 Cyperaceae *Cyperus alternifolius L. Moist soil; local; along ditch near broken water main; an ornamental commonly cultivated in and around Tucson. *Scirpus maritimus L. var. paludosus (A. Nels.) Kukenthal. Moist soil; local; in pond at sanitary landfill. Liliaceae Allium macropetalum Rydb. Gravelly flats; not uncommon local- ly; flowering in the spring. Calochortus kennedyi Porter. Rocky slopes; occasional; flowering in the spring. Dichelostemma pulchellum (Salisb.) Heller. Rocky slopes and gravelly flats; common; flowering in the spring. Poaceae Aristida adscensionis L. Rocky slopes; common along roads; flow- ering in spring and summer. "^Aristida parishii Hitchc. Rocky slopes; occasional. Aristida purpurea Nutt. var. glauca (Nees) A. Holmgren & N. Holm- gren. Rocky slopes; common along paved road. Aristida ternipes Cav. [Aristida scheidiana Trin. & Rupr.]. Rocky slopes and gravelly flats; common. *Avena fatua L. Along washes, in shade of trees; occasional; in- troduced; probably recently adventive to our study area. Bothriochloa barbinodis (Lag.) Herter. Rocky slopes and gravelly flats; locally common; flowering after summer rains. Bouteloua aristidoides (H.B.K.) Griseb. Gravelly flats and shallow, sandy washes; locally abundant summer annual. Bouteloua barbata Lag. var. barbata. Gravelly flats; common on disturbed sites; summer-flowering annual. Bouteloua barbata Lag. var. rothrockii (Vasey) Gould. Gravelly flats; rare; flowering after summer rains. Bouteloua curtipendula (Michx.) Torr. Rocky slopes; locally com- mon; flowering after summer rains. Bouteloua repens (H.B.K.) Scribn. & Merr. [Bouteloua bromoides (H.B.K.) Lag.]. Rocky slopes; common along paved road. Bouteloua trifida Thurb. Gravelly flats; probably occasional. *Bromus arizonicus (Shear) Stebbins. Rocky slopes and gravelly flats, also banks of washes, usually under trees and shrubs; com- mon spring annual. *Bromus rubens L. Rocky slopes and gravelly flats, often on dis- turbed sites; common; introduced; apparently recently adven- tive to our study area. 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 249 *Bromus willdenowii Kunth. Disturbed sites; local and rare; intro- duced; probably recently adventive to our study area. Chloris virgata Swartz. Gravelly flats; occasional summer annual. *Cortaderia selloana (Schult. & Schult.) Asch. & Graebn. Moist soil; local; ditch near broken water main; an ornamental com- mon in cultivation in Tucson; recently adventive to our study area. Cottea pappophoroides Kunth. Rocky slopes; locally common; flowering after summer rains. ^Cynodon dactylon L. Banks of washes; occasional; introduced. Listed by Thomber only for the Santa Cruz River floodplain, but noted by Spalding (1909) to occur near buildings on the hill. *Diplachne fascicularis (Lam.) Beauv. Moist soil; local, around ponds at clay quarry and sanitary landfill. *Echinochloa colonum (L.) Link. Disturbed sites; local, moist soil below water tank; introduced; probably recently adventive to our study area. Enneapogon desvauxii Beauv. Gravelly flats and rocky slopes; oc- casional; flowering after summer rains. *Eragrostis barrelieri Daveau. Disturbed sites; local and uncom- mon; introduced; summer annual. *Eragrostis cilianensis (All.) Mosher. Disturbed sites and sandy flats; locally common summer annual; introduced; probably recently adventive to our study area. *Eragrostis echinochloidea Stapf Moist soil; local, below water tank; introduced. *Eragrostis lehmanniana Nees. Gravelly flats and disturbed sites; locally common; introduced. *Eragrostis pectinacea (Michx.) Nees. Along washes and in moist soil; occasional; summer annual. *Eriochloa lemmonii Vasey & Scribn. var. gracilis (Foum.) Gould. Moist soil; rare; apparently recently adventive to our study area. Erioneuron pulchellum (H.B.K.) Tateoka. Gravelly flats and rocky slopes; common. Heteropogon contortus (L.) Beauv. Rocky slopes, ravines and road- ways; common. Hilaria belangeri (Steud.) Nash. Rocky, north-facing slopes and gravelly flats; occasional. Hilaria mutica (Buckl.) Benth. Sandy flats and rocky slopes; locally common. *Hordeum murinum L. Disturbed sites, gravelly flats, and rocky slopes; common; introduced. Listed by Thomber only for the Santa Cruz River floodplain, but noted by Spalding (1909) to occur on Tumamoc Hill. 250 MADRONO [Vol. 32 *Hordeum pusillum Nutt. Gravelly flats, low-lying areas; locally common. Leptochloa filiformis (Lam.) Beauv. Along washes, on rocky slopes and in moist soil at disturbed sites; common summer annual. Muhlenbergia microsperma (DC.) Kunth. Rocky slopes and along washes; locally common spring annual. Muhlenbergia ported Scribn. Gravelly flats, typically growing among shrubs; common. Panicum arizonicum Scribn. & Merr. Sandy flats and shallow washes; occasional; flowering in the summer. Panicum hirticaule Presl. Disturbed sites; along roads and near buildings. Pappophorum vaginatum Buckl. Along washes and on gravelly flats; locally common. *Pennisetum ciliare (L.) Link. Rocky slopes and disturbed sites, along roads and on sanitary landfill; locally common; intro- duced. *Pennisetum setaceum (Forssk.) Chiov. Disturbed sites, near buildings; an introduced ornamental common in cultivation in and around Tucson. *Phalaris minor Retz. Moist soil; local, at pond in sanitary landfill; introduced. *Phragmites australis (Cav.) Trin. ex Steud. Moist soil; rare; local, near booster pump on Anklam Road; apparently recently ad- ventive to our study area. *Poa bigelovii Vasey & Scribn. Rocky slopes and gravelly flats; common spring annual. *Polypogon monspeliensis (L.) Desf. Moist soil; local, at pond in sanitary landfill; introduced; apparently recently adventive to our study area. *Schismus arabicus Nees. Rocky slopes and gravelly flats; com- mon spring annual; introduced. *Schismus barbatus (L.) Thell. Rocky slopes and gravelly flats; common spring annual; introduced. *Setaria liebmannii Foum. Rocky slopes; occasional; summer- flowering annual. *Setaria macrostachya H.B.K. Rocky slopes; occasional. *Sitanion hystrix (Nutt.) J. G. Smith. Rocky slopes; occasional. "^Sorghum halepense (L.) Pers. Disturbed sites, moist soil; locally common; introduced; apparently recently adventive to our study area. *Sporobolus airoides Torr. var. wrightii (Munro ex Scribn.) Gould. Along washes; locally common; listed by Thomber only for the Santa Cruz River floodplain, but collected by F. Shreve in 1908 "near wash northwest of Desert Laboratory." 1985] BOWERS AND TURNER: TUMAMOC HILL FLORA 251 "^Sporobolus contractus Hitchc. Gravelly or sandy flats and rocky slopes; occasional. Sporobolus cryptandrus (Torr.) Gray. Gravelly flats, low-lying areas; occasional. Trichachne californica (Benth.) Chase. Rocky slopes and gravelly flats; common; flowering after summer rains. Tridens muticus (Torr.) Nash. Rocky slopes; occasional. *Trisetum interruptum Buckl. Moist soil and at roadsides; uncom- mon. Vulpia octoflora (Walt.) Rydb. Rocky slopes and gravelly flats; common spring annual. Typhaceae *Typha domingensis Pers. Moist soil; locally common along wet ditch, pond at clay quarry and below water tank. Acknowledgments We thank T. L. Burgess, C. T. Mason, Jr., and S. P. McLaughlin for reviewing the manuscript; J. R. Reeder and L. J. Toolin for identifying several grasses; and O. M, Grosz for compiling the rainfall data. Literature Cited Arnberger, L. p. 1947. Flowering plants and ferns of Walnut Canyon. Plateau 20: 29-36. Bailey, L. H. and E. Z. Bailey. 1976. Hortus Third. Macmillan Publishing Co., Inc., New York. Benson, L. 1982. The cacti of the United States and Canada. Stanford Univ. Press, Stanford, CA. Bowers, J. E. 1984. Woodland and forest flora and vegetation of Saguaro National Monument. Unpublished report on file at Saguaro National Monument, Route 8, Box 695, Tucson, AZ. Clark, O. M. 1940. Checklist of the flora of Chiricahua National Monument. Southwestern Monuments Monthly Reports (Sept.):201-215. Joyce, J. F. 1976. Vegetational analysis of Walnut Canyon, Arizona. J. Arizona- Nevada Acad. Sci. 11:127-135. Kearney, T. H. and R. H. Peebles. 1960. Arizona flora, 2nd ed. Univ. CaHf Press, Berkeley. Lehr, J. H. 1978. A catalogue of the flora of Arizona. Desert Botanical Garden, Phoenix, AZ. and D. J. Pinkava. 1980. A catalogue of the flora of Arizona: supplement L J. Arizona-Nevada Acad. Sci. 15:17-32. and . 1982. A catalogue of the flora of Arizona: supplement IL J. Arizona-Nevada Acad. Sci. 17:19-26. Little, E. L., Jr. 1 94 1 . Alpine flora of San Francisco Mountain, Arizona. Madrono 6:65-81. McGiNNiES, W. G. 1981. Discovering the desert. Univ. Arizona Press, Tucson. Moore, T. C. 1965. Origin and disjunction of the alpine tundra on San Francisco Mountain, Arizona. Ecology 46:860-864. 252 MADRONO [Vol. 32 Reeves, T. 1976. Vegetation and flora of Chiricahua National Monument, Cochise County, Arizona. M.S. thesis, Arizona State Univ., Tempe. ScHAACK, C. G. 1983. The alpine vascular flora of Arizona. Madrono 14 (supple- ment): 7 9-8 8. Sherbrooke, W. C. 1977. First year seedling survival of jojoba (Simmondsia chi- nensis) in the Tucson Mountains, Arizona. Southw. Naturalist 22:225-234. Shreve, F. 1911. Estabhshment behavior of the palo verde. PI. World 14:289-296. . 1929. Changes in desert vegetation. Ecology 10:364-373. . 1951. Vegetation of the Sonoran Desert. Publ. Carnegie Inst. Wash., no. 591. and A. L. Hinckley. 1937. Thirty years of change in desert vegetation. Ecology 18:463-478. Spalding, V. M. 1 909. Distribution and movements of desert plants. Publ. Carnegie Inst. Wash., no. 113. Spangle, P. F. 1953. A revised checklist of the flora of Walnut Canyon National Monument. Plateau 26:86-88. Thornber, J. J. 1909. Vegetation groups of the Desert Laboratory domain. In V. M. Spalding, Distribution and movements of desert plants, pp. 103-1 12. Publ. Carnegie Inst. Wash., no. 1 13. TuRNAGE, W. V. and A. L. Hinckley. 1938. Freezing weather in relation to plant distribution in the Sonoran Desert. Ecol. Monogr. 8:530-550. (Received 23 Jul 1984; accepted 16 Jan 1985) ANNOUNCEMENT Additional authors are sought for the revision of JEPSON'S MANUAL OF CAL- IFORNIA PLANTS. If you have expertise or particular interest in any of the groups listed below and are willing to contribute to this project, or know of those we might invite to participate, or would like more information, please write or call James C. Hickman, Botany Dept., Univ. of Cahfomia, Berkeley, CA 94720, (415)642-2465. Groups available: Apocynaceae; Aristolochiaceae; Asclepiadaceae; Asteraceae (some genera); Betulaceae; Boraginaceae (esp. Cryptantha, Hackelia, Plagiobothrys); Cac- taceae; Callitrichaceae; Capparidaceae; Caprifoliaceae; Convolvulaceae; Crassulaceae (esp. Sedum); Elatinaceae; Garryaceae; Gentianaceae; Haloragaceae; Hydrophyllaceae (esp. Phacelia); Hypericaceae; Lamiaceae (esp. Monardella, Scutellaria, Stachys); Polygalaceae; Portulacaceae (esp. Calyptridium, Lewisia); Resedaceae; Rhamnaceae (esp. Ceanothus, Rhamnus); Salicaceae (Populus); Sterculiaceae; Urticaceae; Verben- aceae; Violaceae; Vitaceae; Liliaceae (esp. Brodiaea [+ Dichelostemma, Triteleia], Erythronium, Fritillaria, Lilium, Yucca, Zigadenus); Poaceae (some genera). THE ALPINE VASCULAR FLORA OF THREE CIRQUE BASINS IN THE SAN JUAN MOUNTAINS, COLORADO Emily L. Hartman and Mary Lou Rottman Department of Biology, University of Colorado at Denver, Denver 80202 Abstract The San Juan Mountains are located along the Continental Divide in southwestern Colorado. The only previous floristic study of this range was done in the Needle Mountains of the San Juan Range by Michener (1964). Three cirque basins, repre- sentative of the San Juan tundra, were analyzed floristically. A vascular flora of 197 species in 92 genera and 3 1 families is reported. Eight species are Colorado endemics. The phytogeographic distribution of the flora is primarily alpine and western North American. The San Juan Mountains are a discontinuous section of the south- em Rocky Mountains situated along the Continental Divide in southwestern Colorado. These mountains are considered to be a quadrilateral block with dimensions of 77.5 km from east to west and 65.1 km from north to south (Larsen and Cross 1956), resulting in an area of approximately 20,000 km^, 2000 km^ of which are in the alpine zone (Carrara et al. 1984). The relief of the area varies from sharp pinnacles, rounded crests, serrate ridges, and broad up- land surfaces of the alpine to the foothills, plateaus, and canyons of the lowlands. The elevation ranges from 1524 m in the southwest comer to 4358 m at the summit of Uncompahgre Peak. The San Juan Mountains, a youthful part of the southem Rocky Mountains, are composed largely of Tertiary volcanic tuffs and lavas that lie unconformably over metamorphic, sedimentary, and vol- canic intmsive rocks of Precambrian age, as well as sediments of Paleozoic, Mesozoic, and early Cenozoic age (Casadwall and Ohmo- to 1977). During the Pleistocene, the San Juan Mountains were gla- ciated by broad regional ice fields and transection glaciers. Glacial effects are evident in the present alpine landscape as cirques, basins, tams, hanging valleys, and broad U-shaped valleys. Palynological investigations suggest that the San Juan Mountains were free of glaciers prior to 10,000 b.p. (Andrews et al. 1975). The climate of the San Juan Mountains is montane continental. Winters are long and severe; the first snowfall usually occurs by mid- to late September and snowstorms continue to late May or early June. Maximum snow depth at higher elevations (over 3680 m) has been estimated to reach 1 1.6 m. Summers are cool and short with Madro5Jo, Vol. 32, No. 4, pp. 253-272, 20 December 1985 254 MADRONO [Vol. 32 an estimated maximum of 75 frost-free days in the alpine tundra. Periglacial features such as active patterned ground and active rock glaciers indicate the occurrence of sporadic or discontinuous perma- frost (Ives and Fahey 1971, Barsch 1978). Previous investigations in the San Juan Mountains include one floristic study by Michener (unpubl. thesis 1964) and an ecological study of snowpack augmen- tation by Webber et al. (1976). Collections were made in the San Juan Mountains during the summers 1981-1983. Nomenclature in the checklist follows Kartesz and Kartesz (1980). Partial voucher sets are deposited in COLO, CS, and CU-Denver. Phytogeographic abbreviations used in the annotated checklist of vascular species are identified in the discus- sion section. Description of Study Basins Three alpine cirque basins, representative of the San Juan Moun- tain tundra, were selected for floristic analysis. The basins are within a 1 0 km radius of each other. American Basin, 1 5 km southwest of Lake City, contains the headwaters of the Lake Fork of the Gunnison River. It is separated from Bums Basin, 8 km north of Silverton, by Jones Mountain. Stony Basin is located 4.5 km northeast of Silverton. The drainage of Bums and Stony basins is the Animas River. American Basin. American Basin, T42N R6W, Hinsdale County, is characterized by a well-developed, moist, turf mantle intermpted by areas of bedrock outcrops, talus deposits, a tongue-shaped rock glacier, and pattemed ground features. The elevational range of the basin is 3536-3962 m. The vegetation in this basin reflects a more moist climatic regime than in the other two basins. This mesic environment is attributable to several factors: a north-south orientation, massive headwall on the south, windward pass on the southwest and high peaks on the east and west sides of the basin. These features contribute collectively to a heavy accumulation and retention of snow in the winter months and adequate substrate moisture in the summer months. The moist meadow is the predominant community type in the basin. The moist ledges on the north side of the basin yield a high rep- resentation of two families, Caryophyllaceae and Saxifragaceae. This habitat seems to be favorable for members of these families, which normally are not too successful in the highly competitive moist meadow turf The plant species on the top surface of the rock glacier, although representative of a modified fellfield community, consist of several mesophytic species, e.g., Chionophila jamesii, which is overwhelmingly dominant. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 255 Dry habitats are poorly represented in the basin. There are no dry meadows, and fellfield communities are restricted to two high-ele- vation sites. The vegetation on dry ledges as well as on the unstable slopes of the rock glacier, although not limited to xerophytic species, provides the best example of this type of vegetation in the basin. The high diversity of species represented in the krummholz com- munity reflects the expected blending of the subalpine and alpine vegetation in an ecotone. Most of the subalpine species present in this community are not found higher in the basin. Burns Basin. Bums Basin, T42N R6W, San Juan County, has a northwest-southeast orientation. The elevational range is 3634-3932 m. The convex slopes forming the perimeter of the main basin present an interesting contrast of moisture regimes. The southeast- facing slope is characterized by a dry meadow, dominated by Carex elynoides, Geum rossii var. turbinatum, and Hymenoxys grandi- flora, alternating with fellfields and unvegetated talus. The com- munities on this slope are adapted to high solar insolation, steep slope, and relatively poor soil development. The northwest-facing slope consists of a series of tiers of massive, moist and wet ledges with adjacent moist meadows. The ledge com- munity is dominated by Salix reticulata subsp. nivalis. A number of rare species occur within the ledge complex. An isolated solitary meadow, dominated by Kobresia myosuroides, is present on the lower part of the slope below the ledges. Another indication of the effects of slope aspect and differing moisture regimes on vegetation is reflected by the discrete occurrence of a community dominated by Salix brachycarpa on the southeast- facing slope and a community dominated by Salix planifolia on the northwest-facing slope. Both communities show a segregation of dominants and associated species along a moisture gradient. The openness of the mid-section of the basin, created by the lack of headwall protection, promotes the occurrence of strong winds, as shown by the uniform upslope shearing of the krummholz conifers and the presence of many typical feUfield communities in this part of the basin. Bums Basin also contains a rock glacier complex composed of tongue and lobate units. Well-defined communities are present in limited areas of fines on the unstable slopes. A highly-localized moist meadow occurs on the top surface of the rock glacier. Stony Basin. Stony Basin, T41N R6W, San Juan County, is formed of three broad turf-mantled steps, each separated by a bedrock es- carpment. The elevational range of the basin is 3764-3926 m. This northwest-facing basin is continually bufleted by strong wind, re- sulting in a more severe climatic regime than is found in the other 256 MADRONO [Vol. 32 two basins. Islands of dry meadow and fellfield interrupt the moist meadow of the upper two steps of the basin. The latter have an abundance of frost-associated features including frost boils, subnival boulder pavement, ephemeral ponds, and rock debris islands. The lower step is characterized by a shallow lake and hummocky wet meadow. Moist meadows are the most extensive type of vegetation in the basin; the dominant and some of the associated species, however, vary according to aspect and elevation. The dry meadow dominated by Carex elynoides matures earlier than do adjacent moist meadow communities. The subnival boulder pavement, a periglacially-related habitat, supports a community of high diversity including several rare species. The ridgetop vegetation is exposed to the most severe environmental extremes found in the Basin. However, the species in these communities show no reduction in stature as contrasted with the many dwarfed species found in moist meadows dominated by Salix reticulata subsp. nivalis and located in more protected sites than the dry meadows. Discussion The alpine flora of the basins selected as representative of the San Juan Mountains consists of 191 species representing 86 genera of angiosperms, three genera and three species of gymnosperms, and three genera and three species of pteridophytes. Twenty-eight taxa included in this study were unreported by Mich- ener (1964) and Webber et al. (1976). These taxa are: Arabis divaricarpa Artemisia campestris ssp. bo- realis Botrychium lunaria Carex incurviformis Carex norvegica Draba carta Draba nivalis Draba spectabilis var. spectabilis Erigeron compositus var. gla- bratus Erigeron grandiflorus Erigeron vagus Erysimum capitatum var. amoenum Hierochloe hirta ssp. arctica Kobresia sibirica Minuartia rossii Moehringia lateriflora Oxytropis podocarpa Poa leptocoma Potentilla hookeriana Potentilla subjuga var. minuti- folia Salix glauca var. villosa Saxifraga chrysantha Senecio porteri Silene kingii Silene uralensis Taraxacum lyratum Townsendia rothrockii Vaccinium scoparium. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 257 Habitats and Communities. The moist meadow is the predomi- nant habitat in the study basins. This contrasts with the predominant dry meadows of the Front Range and correlates generally with the higher moisture regime characteristic of the San Juan Mountains. The moist meadow may be regarded as a complex of several com- munities, each with a distinct spatial occurrence within the complex. A Deschampsia caespitosa-Geum rossii var. turbinatum element is found in lower sites on basin slopes and in concavities; a Carex nigricans-Sibbaldia procumbens element is located in flat areas at mid-slope; and a Salix reticulata ssp. nivalis element is present on the highest sites of this complex. The dry meadow community is virtually absent in American Basin and is of minor importance in Bums and Stony basins. Carex ely- noides is the most frequent dominant in the dry meadow commu- nity. Throughout all three basins only two, small, isolated meadows dominated by Kobresia myosuroides occur. This is of interest when compared to the Front Range where Kobresia has long been rec- ognized as the climatic climax community (Cox 1933, Bamberg 1961, Marr 1961, and Eddleman and Ward 1984). The highly re- stricted occurrence of Kobresia myosuroides in this study contrasts sharply with the Kobresia meadows in the Eldorado Lake and Wil- liams Lakes basins of the San Juan Mountains as reported by Webber et al. (1976), who extrapolate that Kobresia meadows are one of the two community types that account for perhaps 50 percent of the alpine vegetation in the San Juan Mountains. Michener (1964), on the other hand, reports only one occurrence of Kobresia in the Needle Mountains in a soil-filled depression site near timberline. Dryas octopetala subsp. hookeriana, a dominant species in dry meadows and mat shrub communities in the Front Range (Cooper 1908, Cox 1933, and Eddleman and Ward 1984), is conspicuously absent or rare in occurrence in the San Juan Mountains. This species was found by Michener (1964) only on north-facing ledges on the ridge and also to the east of Ruby Pass, outside our study areas, and was reported as very rare by Webber et al. (1976). Habitats dominated by rocks are varied and common in occur- rence in all three basins. Some of these, such as moist and wet ledges, fellfields, ridgetops, and surfaces on rock glaciers support commu- nities where dominants can be recognized. Others, such as rock crevices, subnival boulder pavement, blockfield, and certain pat- terned ground forms show a lack of dominants. In general, diversity or species richness of a community increases as the amount of rock material in a habitat increases. Rare species also may increase in predominantly rock habitats, especially on ledges, subnival boulder pavement, and on unstable slopes of the rock glacier. Major and Bamberg (1968) and Komarkova (1976) have discussed the occur- 258 MADRONO [Vol. 32 rence of rare species and conclude that they are most Hkely to occur where competition is low or where the development of a climax vegetation is continuously disrupted. Fellfield and ridgetop habitats reflect a higher degradation of rock material and greater amount of mineralized soil than the other rock- predominating habitats. Some of the rare species such as Artemisia campestris ssp. borealis, Erigeron compositus var. glabratus, E. va- gus, Townsendia rothrockii, Poa epilis, and Anemone multifida var. globosa are found only on the ridgetop sites. The typical fellfield cushion plant community, dominated by Paronychia pulvinata and Phlox caespitosa ssp. condensata, which is frequently found on windswept sites of the Front Range (Cox 1933, and Eddleman and Ward 1984), is absent in the San Juan Mountains (Michener 1964, Webber et al. 1976, and Rottman 1984). Although we have emphasized some of the major differences in community structure and dominants between this study and other studies in the San Juan Mountains and Front Range, much similarity is evident in the floristic inventories from the respective areas. For example, there is a 76.5 percent similarity in the vascular plant species found in this study and those of the Williams Lakes and Eldorado Lake basins as reported by Webber et al. (1976). They report a greater diversity of grasses, including Calamagrostis pur- purascens, Danthonia intermedia, Poa pattersonii, and P. fendler- iana, and shrub species, inchxding Actaea rubra, Arctostaphylos uva- ursi, and Gaultheria humifusa. They also found two Carex species that were not observed during this study: Carex scopulorum, a wet meadow species, and C rupestris var. drummondiana, a. dry mead- ow species. There is a 73.6 percent similarity between species found in this study and those reported by Michener (1964) for the Needle Mountains of the San Juan Range. This lesser degree of similarity may be due, in part, to substrate differences. The Needle Mountains are a discrete metamorphic unit within the larger volcanic San Juan Range. Michener (1964) found Pinus flexilis and Populus tremu- loides at a timberline elevation of 3749 m, both of which were absent in this study. She also reports some grasses not found during this study, including Agrostis filicumis, Poa interior, P. longiligula, Dan- thonia intermedia, and Koeleria cristata, and shrubs, including Ac- taea rubra, Gaultheria humifusa, Lonicera involucrata, Ribes la- custre, and R. wolfii. In comparing the vascular plant inventories of this study and that of the Indian Peaks area of the Front Range, a still lower (70.0%) similarity results. This reffects differences in the phytogeographic distributions of species from the northern and southern Colorado alpine tundra. Phytogeography. Table 1 shows the phytogeographic distribution of the flora. Four elements are recognized, each of which may be 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 259 Table I. Affinities of the Flora Elements in Three High Basins of the Northern San Juan Mountains, Colorado. Abbreviations following each unit are cited in the annotated checklist. Element Abbreviation Percent of taxa Element Boreal montane BM 19.8 Montane M 6.0 Arctic alpine AA 31.5 Alpine A 42.6 Geographic subelement Circumpolar C 22.8 North American NA 10.2 Western North American WNA 25.8 Rocky Mountains RM 13.7 Southern Rocky Mountains SRM 11.6 Colorado CO 4.0 North American— Asiatic NAA 9.6 North American— European NAE 2.0 combined with more specific geographic subelements (Komarkova, 1976). As may be seen from the percentages given, the largest part of the vascular flora is made up of alpine species (42.6%) and Western North American species (25.0%). The circumpolar subelement (22.8%), which is largely identified with the arctic-alpine element, is a second important component of the flora. The higher percentage of North American-Asiatic species (9.6%) in relation to North American-European species (2.0%) indicates a stronger aflinity of the San Juan alpine flora to the Asiatic alpine flora than to the European alpine flora. A comparison of phytogeographical analyses of the San Juan Mountains shows a lower boreal-montane and montane represen- tation and a concomitantly higher alpine and arctic-alpine repre- sentation in this study than in that of Webber et al. (1976). This reflects the higher base elevations of the basins in this study and the I greater distance from treeline. There is close agreement between the two studies in the percentages of North American-Asiatic, Colorado, western North American, and Rocky Mountains subelements. In comparing the phytogeographic analysis of this study with that of the vascular flora of the Indian Peaks area, northern Colorado (Ko- markova 1976), a decrease in circumpolar and North American- Asiatic subelements and an increase in Rocky Mountain, southern Rocky Mountain, and Colorado subelements is evident. A similar north to south trend is reported by Webber et al. (1976). The number of Colorado endemics in the San Juan Mountains is greater than the 260 MADRONO [Vol. 32 number reported for the Indian Peaks area. This agrees with Major and Bamberg (1967), who state that endemism increases in a south- erly direction. The endemic species found in this study include: Besseya ritteriana, Draba streptobrachia, Minuartia macrantha, Penstemon harbourii, Potentilla subjuga var. subjuga, P. subjuga var. minutifolia, Senecio soldanella, and Townsendia rothrockii. Annotated Checklist of Vascular Plant Species Pterophyta Selaginellaceae Selaginella densa Rydb. Common; dry and moist meadows, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. A/WNA. Ophioglossaceae Botrychium lunaria (L.) Sw. Very rare; fellfield. BM/NA. Aspleniaceae Cystopteris fragilis (L.) Bemh. Infrequent; shrub tundra, rock ledge, and rock debris habitats. AA/C. CONIFEROPHYTA Pinaceae Abies lasiocarpa (Hook.) Nutt. Very rare; krummholz. BM/WNA. Juniperus communis L. Very rare; krummholz. BM/C. Picea engelmannii Parry ex Engelm. Rare; krummholz. BM/WNA. AnTHOPH YTA — DiCOT YLEDONEAE Adoxaceae Adoxa moschatellina L. Very rare; rock crevice. BM/C. Apiaceae Angelica grayi Coult. & Rose. Infrequent; moist meadow, rock ledge, rock crevice, and rock debris habitats. A/SRM. Oreoxis bakeri Coult. & Rose. Ubiquitous; dry, moist and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. A/SRM. Oxypolis fendleri (Gray) Heller. Very rare; wet meadow and rock ledge. M/SRM. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 261 Pseudocymopterus montanus (Gray) Coult. & Rose. Very rare; krummholz. M/SRM. Asteraceae Achillea millefolium L. var. lanulosa (Nutt.) Piper. Very rare; krummholz. A/WNA. Agoseris aurantiaca (Hook.) Greene. Very rare; wet meadow. BM/ NA. Agoseris glauca (Pursh) Raf. Very rare; krummholz. BM/NA. Antennaria alpina (L.) Gaertn. Infrequent; moist meadow, shrub tundra, krummholz, and rock debris habitats. AA/NAE. Arnica cordifolia Hook. Very rare; wet meadow and krummholz. BM/WNA. Arnica mollis Hook. Very rare; krummholz. BM/NA. Artemisia campestris L. subsp. borealis (Pallas) Hall & Clem- ents. Rare; moist meadow and fellfield. AA/C. Artemisia scopulorum Gray. Ubiquitous; dry, moist and wet meadows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/RM. Cirsium scopulorum (Greene) Cockerell. Infrequent; moist mead- ow, krummholz, and rock ledge. A/RM. Dugaldia hoopesii (Gray) Rydb. Very rare; krummholz. M/RM. Erigeron compositus Pursh var. glabratus Macoun. Very rare; rock ledge. /WNA. Erigeron grandiflorus Hook. Very rare; dry meadow and rock ledge. /WWNA. Erigeron melanocephalus A. Nels. Common; moist and wet mead- ows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/SRM. Erigeron pinnatisectus (Gray) A. Nels. Infrequent; dry meadow, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. A/SRM. Erigeron simplex Greene. Ubiquitous; dry, moist and wet mead- ows, shrub tundra, krummholz, fellfield, rock crevice, and rock debris habitats. A/WNA. Erigeron vagus Payson. Rare; rock ledge and rock debris habitats. A/WNA. Haplopappus pygmaeus (Torr. & Gray) Gray. Rare; dry meadow and fellfield. /VRM. Hymenoxys grandiflora (Torr. & Gray ex Gray) Parker. Infre- quent; dry and moist meadows, shrub tundra, krummholz, fellfield, rock ledge, and rock crevice. A/RM. Senecio amplectens Gray var. amplectens. Infrequent; moist meadow, fellfield, and rock debris habitats. M/RM. 262 MADRONO [Vol. 32 Senecio amplectens Gray var. holmii (Greene) Harring- ton. Infrequent; moist meadow, fellfield, rock ledge, and rock debris habitats. A/WNA. Senecio atratus Greene. Very rare; krummholz. A/SRM. Senecio dimorphophyllus Greene. Infrequent; moist and wet meadows, shrub tundra, rock ledge, and rock crevice. M/RM. Senecio porteri Greene. Very rare; rock debris habitats. A/RM. Senecio soldanella Gray. Infrequent; moist meadow, shrub tundra, fellfield, rock crevice, and rock debris habitats. A/CO. Senecio triangularis Hook. Very rare; rock ledge. BM/WNA. Senecio werneriifolius Gray. Infrequent; dry and moist meadows, shrub tundra, rock crevice, and rock debris habitats. M/RM. Taraxacum ceratophorum (Ledeb.) DC. Infrequent; moist mead- ow and fellfield. AA/C. Taraxacum lyratum (Ledeb.) DC. Rare; fellfield and rock debris habitats. AA/NAA. Townsendia rothrockii Gray ex Rothrock. Very rare; fellfield. A/ CO. Boraginaceae Mertensia bakeri Greene. Infrequent; moist meadow, shrub tun- dra, krummholz, fellfield, rock ledge, rock crevice and rock debris habitats. /VSRM. Mertensia ciliata (James ex Torr.) G. Don. Very rare; moist mead- ow and krummholz. BM/WNA. Brassicaceae Arabis divaricarpa A. Nels. Infrequent; moist meadow, rock ledge and rock debris habitats. BM/NA. Arabis drummondii Gray. Very rare; krummholz. BM/NA. Arabis lemmonii S. Wats. Rare; fellfield and rock debris habitats. A/WNA. Cardamine cordifolia Gray. Infrequent; moist and wet meadows and rock ledge. BM/WNA. Draba aurea Vahl. Infrequent; dry and moist meadows, shrub tun- dra, krummholz, fellfield, and rock ledge. AA/C. Draba cana Rydb. Infrequent; dry and moist meadows, shrub tun- dra, krummholz, fellfield, and rock ledge. AA/C. Draba crassa Rydb. Common; dry, moist and wet meadows, fell- field, rock ledge, rock crevice, and rock debris habitats. A/RM. Draba crassifolia Graham. Ubiquitous; moist and wet meadows, krummholz, fellfield, rock crevice, and rock debris habitats. AA/ NAE. Draba fladnizensis Wulfen. Infrequent; dry and moist meadows. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 263 shrub tundra, krummholz, fellfield, rock ledge, and rock debris habitats. AA/C. Draba nivalis Lilj. Rare; fellfield and rock ledge. AA/C. Draba spectabilis Greene var. spectabilis. Very rare; rock ledge. M/ RM. Draba streptobrachia Price. Rare; fellfield, rock ledge and rock debris habitats. A/CO. Erysimum capitatum (Dougl.) Greene var. amoenum (Greene) R. J. Davis. Infrequent; dry and moist meadows, shrub tundra, krummholz, fellfield, and rock debris habitats. A/SRM. Rorippa curvipes Greene var. alpina (S. Wats.) R. Stuckey. Very rare; wet meadow. A/RM. Smelowskia calycina (Steph.) C. A. Mey. ex Ledeb. Common; dry and moist meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAA. Thlaspi montanum L. Common; dry, moist and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. A/C. Campanulaceae Campanula uniflora L. Rare; dry and moist meadows, rock ledge, and rock debris habitats. AA/C. Caryophyllaceae Cerastium earlei Rydb. Ubiquitous; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. A/RM. Minuartia macrantha (Rydb.) House. Rare; fellfield. A/CO. Minuartia obtusiloba (Rydb.) House. Common; dry and moist meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAA. Minuartia rossii (R. Br.) Graebn. Infrequent; dry and moist mead- ows, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NA. Minuartia rubella (Wahlenb.) Hiem. Common; dry, moist and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/C. Moehringia lateriflora (L.) Fenzl. Infrequent; dry meadow, shrub tundra, krummholz, fellfield, and rock ledge. AA/C. Sagina saginoides (L.) Karst. Infrequent; moist meadow, shrub tundra, fellfield, rock ledge, and rock debris habitats. AA/C. Silene acaulis (L.) Jacq. var. subacaulis (F. N. Williams) Fern. & St. John. Ubiquitous; dry and moist meadows, shrub tundra. 264 MADRONO [Vol. 32 krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAA. Silene drummondii Hook. Very rare; wet meadow and krumm- holz. BM/NA. Silene kingii (S. Wats.) Bocquet. Rare; shrub tundra and rock ledge. A/SRM. Silene uralensis (Rupr.) Bocquet subsp. uralensis. Infrequent; rock ledge, rock crevice, and rock debris habitats. AA/C. Stellaria irrigua Bunge. Very rare; rock debris habitats. A/NAA. Stellaria umbellata Turcz. ex Kar. & Kir. Common; moist and wet meadows, shrub tundra, fellfield, rock ledge, and rock debris habitats. A/NAA. Crassulaceae Sedum integrifolium (Raf.) A. Nels. ex Coult. & A. Nels. Common; moist and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAA. Sedum lanceolatum Torr. Infrequent; dry and moist meadows, shrub tundra, krummholz, fellfield, and rock ledge. A/WNA. Sedum rhodanthum Gray. Rare; moist meadow and rock ledge. A/RM. Ericaceae Vaccinium caespitosum Michx. Rare; moist and wet meadows, shrub tundra, and fellfield. BM/NA. Vaccinium myrtillus L. subsp. oreophilum (Rydb.) Love, Love & Kapoor. Very rare; krummholz. BM/C. Vaccinium scoparium Leib. Rare; moist meadow and fellfield. BM/ WNA. Fabaceae Astragalus alpinus L. Rare; moist meadow and rock ledge. AA/C. Oxytropis podocarpa Gray. Very rare; dry meadow. AA/C. Trifolium attenuatum Greene. Infrequent; dry and moist mead- ows, fellfield, rock ledge, rock crevice, and rock debris habitats. A/SRM. Trifolium dasyphyllum Torr. & Gray. Very rare; rock ledge. A/ RM. Trifolium nanum Torr. Common; dry and moist meadows, shrub tundra, krummholz, fellfield, rock ledge, and rock debris hab- itats. A/RM. Trifolium parryi Gray. Ubiquitous; moist and wet meadows, shrub tundra, fellfield, rock ledge, and rock debris habitats. A/RM. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 265 Gentianaceae Gentiana algida Pallas. Rare; dry and moist meadows. AA/NAA. Gentiana prostrata Haenke ex Jacq. Infrequent; dry and moist meadows, fellfield and rock ledge. AA/NAA. Gentianella amarella (L.) Bomer. Rare; dry meadow and rock ledge. BM/C. Gentianella tenella (Rottb.) Bomer. Rare; dry and moist meadows and rock ledge. AA/C. Swertia perennis L. Very rare; rock ledge. A/C. Hydrophyllaceae Phacelia sericea Hook. Infrequent; dry meadow, shrub tundra, krummholz, fellfield, rock crevice, and rock debris habitats. A/ WNA. Onagraceae Epilobium anagallidifolium Lam. Rare; wet meadow, shrub tun- dra, and rock ledge. AA/C. Epilobium angustifolium L. Very rare; rock ledge. BM/C. Epilobium latifolium L. Very rare; rock debris habitats. A/C. Polemoniaceae Polemonium delicatum Rydb. Rare; krummholz and rock ledge. M/SRM. Polemonium viscosum Nutt. Common; dry and moist meadows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/WNA. Polygonaceae Oxyria digyna (L.) Hill. Common; moist and wet meadows, rock ledge, rock crevice, and rock debris habitats. AA/C. Polygonum bistortoides Pursh. Ubiquitous; dry, moist and wet meadows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/WNA. Polygonum viviparum L. Infrequent; dry, moist and wet meadows, krummholz, rock ledge, rock crevice, and rock debris habitats. AA/C. Portulacaceae Claytonia megarhiza (Gray) Parry ex S. Wats. Common; moist meadow, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/RM. 266 MADRONO [Vol. 32 Lewisia pygmaea (Gray) B. L. Robins. Rare; moist meadow and rock ledge. A/WNA. Primulaceae Androsace septentrionalis L. Ubiquitous; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/C. Primula parryi Gray. Infrequent; moist and wet meadows. A/RM. Ranunculaceae Anemone multiflda Poir. var. globosa (Nutt.) Torr. & Gray ex Fritz. Very rare; fellfield. BM/NA. Aquilegia coerulea James. Infrequent; wet meadow, shrub tundra, krummholz, fellfield, rock crevice, and rock debris habitats. M/RM. Caltha leptosepala DC. Common; moist and wet meadows, shrub tundra, krummholz, and rock ledge. A/WNA. Delphinium barbeyi (Huth) Huth. Very rare; moist meadow. M/ SRM. Ranunculus eschscholtzii Schlecht. Infrequent; dry and moist meadows, krummholz, and rock ledge. AA/NAA. Ranunculus macauleyi Gray. Infrequent; moist meadow, fellfield, rock ledge, and rock debris habitats. A/SRM. Thalictrum alpinum L. Infrequent; dry and moist meadows, fell- field, rock ledge, and rock debris habitats. A/SRM. Thalictrum fendleri Engelm. ex Gray. Rare; fellfield and rock ledge. BM/WNA. Trollius laxus Salisb. subsp. albiflorus (Gray) Love, Love & Ka- poor. Very rare; shrub tundra. BM/WNA. Rosaceae Geum rossii (R. Br.) Ser. var. turbinatum (Rydb.) C. L. Hitch. Ubiquitous; dry, moist, and wet meadows, shrub tun- dra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAA. Potentilla diver sifolia Lehm. Ubiquitous; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAA. Potentilla gracilis Dougl. ex Hook. var. pulcherrima (Lehm.) Fern. Very rare; krummholz. BM/WNA. Potentilla hookeriana Lehm. Very rare; fellfield. AA/NAA. Potentilla nivea L. Infrequent; dry and moist meadows, shrub tun- dra, fellfield, rock ledge, and rock debris habitats. AA/C. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 267 Potentilla rubricaulis Lehm. Infrequent; dry meadow, shrub tun- dra, fellfield, and rock ledge. AA/NA. Potentilla subjuga Rydb. var. minutifolia Rydb. Infrequent; dry and moist meadows, shrub tundra, and fellfield. A/CO. Potentilla subjuga Rydb. var. subjuga. Infrequent; dry and moist meadows, shrub tundra, fellfield, and rock ledge. A/CO. Potentilla uniflora Ledeb. Very rare; rock ledge. AA/NAA. Sibbaldia procumbens L. Ubiquitous; moist and wet meadows, shrub tundra, krummholz, fellfield, and rock crevice. AA/C. Salicaceae Salix arctica Pallas. Common; dry, moist, and wet meadows, shrub tundra, and rock ledge. A/WNA. Salix brachyphylla Nutt. Rare; shrub tundra and rock ledge. BM/ NA. Salix glauca L. var. villosa (Hook.) Anderss. Very rare; shrub tun- dra. BM/WNA. Salix planifolia Pursh. Infrequent; wet meadow, shrub tundra, and rock ledge. BM/NA. Salix reticulata Hook, subsp. nivalis (Hook.) Love, Love & Ka- poor. Common; moist meadow, shrub tundra, fellfield, and rock ledge. A/WNA. Saxifragaceae Heuchera parvifolia Nutt. ex Torr. & Gray. Rare; krummholz, fell- field, and rock ledge. A/SRM. Parnassia fimbriata Koenig. Very rare; rock ledge. BM/WNA. Parnassia kotzebuei Cham. & Schlecht. Very rare; rock ledge. AA/ NAA. Ribes montigenum McClatchie. Rare; krummholz, fellfield and rock ledge. BM/WNA. Saxifraga adscendens L. subsp. oregonensis (Raf.) Ba- cig. Infrequent; dry and moist meadows, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/NAE. Saxifraga bronchialis L. subsp. austromontana (Wieg.) Pip- er. Infrequent; moist meadow, fellfield, rock ledge, and rock crevice. A/WNA. Saxifraga cernua L. Infrequent; dry and moist meadows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris hab- itats. AA/C. Saxifraga cespitosa L. subsp. delicatula (Small) Por- sild. Infrequent; dry and moist meadows, rock ledge, and rock debris habitats. AA/C. 268 MADRONO [Vol. 32 Saxifraga cespitosa L. subsp. monticola (Small) Porsild. Very rare; rock ledge and rock crevice. A/C. Saxifraga chrysantha Gray. Rare; moist meadow, rock ledge and rock debris habitats. AA/NAA. Saxifraga debilis Engelm. ex Gray. Common; moist and wet mead- ows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/RM. Saxifraga flagellaris (Stemb.) Willd. subsp. platysepala (Trautv.) Porsild. Infrequent; dry and moist meadows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/SRM. Saxifraga odontoloma Piper. Rare; wet meadow and rock ledge. BM/WNA. Saxifraga rhomboidea Greene. Ubiquitous; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. A/WNA. Saxifraga rivularis Greene. Rare; moist meadow and rock ledge. AA/C. Scrophulariaceae Besseya alpina (Gray) Rydb. Common; dry, moist, and wet mead- ows, shrub tundra, fellfield, rock ledge, rock crevice, and rock debris habitats. A/SRM. Besseya ritteriana (Eastw.) Rydb. Very rare; rock ledge. A/CO. Castilleja haydenii (Gray) Cockerell. Infrequent; dry and moist meadows, fellfield, and rock ledge. A/SRM. Castilleja miniata Dougl. ex Hook. Very rare; shrub tundra. BM/ WNA. Castilleja occidentalis Torr. Infrequent; moist and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, and rock debris habitats. A/RM. Castilleja rhexifolia Rydb. Rare; moist meadow and rock ledge. BM/WNA. Chionophila jamesii Benth. Infrequent; moist and wet meadows, fellfield, rock ledge, and rock debris habitats. A/SRM. Mimulus guttatus DC. Very rare; rock ledge. BM/NA. Pedicularis groenlandica Retz. Infrequent; moist and wet mead- ows. AA/NA. Pedicularis sudetica Willd. subsp. scopulorum (Gray) Hulten. Very rare; shrub tundra. A/RM. Penstemon harbourii Gray. Rare; rock debris habitats. A/CO. Penstemon whippleanus Gray. Infrequent; krummholz, fellfield and rock ledge. M/RM. Veronica wormskjoldii Roemer & Schultes. Infrequent; dry, moist, and wet meadows, shrub tundra, fellfield, and rock ledge. AA/ NA. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 269 Violaceae Viola adunca Sm. subsp. bellidifolia (Greene) Harrington. Rare; moist meadow and rock ledge. BM/NA. AnTHOPH YTA — MONOCOT YLEDONEAE Cyperaceae Carex albonigra Mackenzie. Infrequent; dry, moist, and wet mead- ows, knimmholz, fellfield, rock ledge, and rock debris habitats. AA/WNA. Carex aquatilis Wahlenb. Rare; wet meadow. AA/C. Carex arapahoensis Clokey. Infrequent; dry and moist meadows, krummholz, fellfield, rock ledge, and rock debris habitats. A/SRM. Carex bella Bailey. Very rare; krummholz. M/RM. Carex capillaris L. Very rare; moist meadow. AA/C. Carex ebenea Rydb. Infrequent; dry, moist, and wet meadows, shrub tundra, fellfield, rock crevice, and rock debris habitats. A/RM. Carex elynoides Holm. Infrequent; dry and moist meadows, shrub tundra, fellfield, and rock debris habitats. A/WNA. Carex hallii Olney. Very rare; moist meadow. BM/NA. Carex haydeniana Olney. Infrequent; moist meadow, fellfield, rock ledge, rock crevice, and rock debris habitats. A/WNA. Carex heteroneura W. Boott var. chalciolepis (Holm) F. J. Herm. Common; moist meadow, shrub tundra, krummholz, fellfield, rock ledge, and rock debris habitats. A/WNA. Carex incurviformis Mackenzie. Infrequent; dry and moist mead- ows and fellfield. A/WNA. Carex misandra R. Br. Rare; moist meadow and rock ledge. AA/C. Carex nardina Fries var. hepburnii (Boott) Kukenth. Rare; moist meadow and rock debris habitats. AA/NAA. Carex nelsonii Mackenzie. Infrequent; moist meadow. A/SRM. Carex nigricans C. A. Mey. Infrequent; moist and wet meadows, shrub tundra, krummholz, and fellfield. A/NAA. Carex norvegica Retz. subsp. norvegica. Rare; dry and moist meadows and shrub tundra. AA/NAE. Carex nova Mackenzie. Infrequent; dry and moist meadows, fell- field, rock ledge, and rock crevice. BM/WNA. Carex perglobosa Mackenzie. Infrequent; dry and moist meadows, fellfield, rock ledge, and rock debris habitats. A/SRM. Carex phaeocephala Piper. Infrequent; dry meadow, fellfield, and rock debris habitats. A/WNA. Carex pseudoscirpoidea Rydb. Infrequent; dry, moist and wet 270 MADRONO [Vol. 32 meadows, shrub tundra, krummholz, fellfield, and rock debris habitats. A/WNA. Carex pyrenaica Wahlenb. Rare; moist and wet meadows and rock debris habitats. A/C. Carex vernacula Bailey. Infrequent; moist and wet meadows and shrub tundra. A/WNA. Kobresia myosuroides (Vill.) Fiori & Paol. Infrequent; dry mead- ow, fellfield and rock debris habitats. AA/C. Kobresia sibirica Turcz. Very rare; moist meadow and rock debris habitats. AA/NAA. Juncaceae Juncus drummondii E. Mey. Infrequent; moist and wet meadows, fellfield, and rock ledge. A/WNA. Juncus mertensianus Bong. Very rare; wet meadow. A/NAA. Luzula parviflora (Ehrh.) Desv. Very rare; shrub tundra. BM/C. Luzula spicata (L.) DC. Common; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, and rock debris habitats. A/RM. Liliaceae Lloydia serotina (L.) Salis. ex Reichenb. Infrequent; dry and moist meadows, shrub tundra, fellfield, rock ledge, and rock debris habitats. AA/C. Veratrum tenuipetalum Heller. Very rare; krummholz. A/SRM. Poaceae Agropyron scribneri Vasey. Infrequent; moist meadow, krumm- holz, fellfield, rock ledge, and rock debris habitats. A/WNA. Agropyron trachycaulum (Link) Malte ex H. F. Lewis var. latiglume (Scribn. & Smith) Beetle. Infrequent; dry meadow, fellfield, and rock ledge. AA/NA. Deschampsia caespitosa (L.) Beauv. Common; dry, moist, and wet meadows, fellfield, and rock ledge. BM/C. Festuca brachyphylla Schultes. Ubiquitous; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/C. Hierochloe hirta (Schrank) Borbas subsp. arctica (Presl.) G. Weis- marck. Very rare; krummholz. AA/C. Phleum alpinum L. Infrequent; moist and wet meadows and krummholz. AA/C. Poa alpina L. Ubiquitous; dry, moist, and wet meadows, shrub tundra, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/C. 1985] HARTMAN & ROTTMAN: SAN JUAN MTNS. ALPINE FLORA 27 1 Poa epilis Scribn. Very rare; fellfield. BM/WNA. Poa leptocoma Trin. Very rare; dry meadow and rock ledge. A/ WNA. Poa reflexa Vasey & Scribn. ex Vasey. Very rare; moist meadow. A/ WNA. Poa rupicola Nash ex Rydb. Common; dry, moist, and wet mead- ows, shrub tundra, krummholz, fellfield, rock ledge, rock crev- ice, and rock debris habitats. A/WNA. Trisetum spicatum (L.) Richter. Ubiquitous; dry, moist, and wet meadows, krummholz, fellfield, rock ledge, rock crevice, and rock debris habitats. AA/C. Literature Cited Andrews, J. T., P. E. Carrara, F. B. King, and R. Stuckenrath. 1975. Holocene environmental changes in the alpine zone, northern San Juan Mountains, Col- orado: evidence from bog stratigraphy and palynology. Quaternary Res. 5:173- 197. Bamberg, S. A. 1961. Plant ecology of alpine tundra areas in Montana and adjacent Wyoming. M.A. thesis, Univ. Colorado, Boulder. Barsch, D. 1978. Rock glaciers as indicators for discontinuous alpine permafrost: an example from the Swiss Alps. In Proceedings of the Third International Conference on Permafrost, pp. 349-352. National Research Council of Canada, Ottawa. Carrara, P. E., W. N. Mode, M. Rubin, and S. W. Robinson. 1984. Deglaciation and post-glacial timberline in the San Juan Mountains, Colorado. Quaternary Res. 21:42-55. Casadwall, T. and H. Ohmoto. 1977. Sunnyside Mine, Eureka Mining District, San Juan County, Colorado: geochemistry of gold and base metal ore deposition in a volcanic environment. Econ. Geol. 72:1285-1320. Cooper, W. S. 1908. Alpine vegetation in the vicinity of Long's Peak. Bot. Gaz. 45:319-337. Cox, C. F. 1933. Alpine plant succession on James Peak, Colorado. Ecol. Monog. 3:299-372. Eddleman, L. F. and R. T. Ward. 1984. Phytoedaphic relationships in alpine tundra, north-central Colorado, U.S.A. Arctic and Alpine Research 16:343-359. Ives, J. D. and B. D. Fahey. 1971. Permafrost occurrence in the Front Range, Colorado Rocky Mountains, United States of America. J. Glaciology 10:105- 111. Kartesz, J. T. and R. Kartesz. 1980. A synonymized checklist of the vascular flora of the United States, Canada, and Greenland. Vol. II: the biota of North America. Univ. North Carolina Press, Chapel Hill. KoMARKOVA, V. 1976. Alpine vegetation of the Indian Peaks area, Colorado Rocky Mountains. Ph.D. dissertation, Univ. Colorado, Boulder. Larsen, E. S., Jr. and C. W. Cross. 1956. Geology and petrology of the San Juan region, southwestern Colorado. U.S.G.S. Prof. Paper 258. Major, J. and S. A. Bamberg. 1967. A comparison of some North American and Eurasian alpine ecosystems. In H. E. Wright, Jr. and W. H. Osbum, eds., Arctic and Alpine environments, pp. 89-1 18. Indiana Univ. Press, Bloomington. Marr, J. W. 1961. Ecosystems of the East Slope of the Front Range in Colorado. Univ. Colorado Stud., Series in Biology 8:1-134. 272 MADRONO [Vol. 32 MiCHENER, M. J. 1964. The high altitude vegetation of the Needle Mountains of southwestern Colorado. M.A. thesis, Univ. Colorado, Boulder. RoTTMAN, M. L. 1984. A floristic analysis of three alpine basins in the northern San Juan Mountains, Colorado. Ph.D. dissertation, Univ. Colorado, Boulder. Webber, P. J., J. C. Emerick, D. C. E. May, and V. Komarkova. 1976. The impact of increased snowfall on alpine vegetation. In H. W. Steinhoff, and J. D. Ives, eds.. Ecological impacts of snowpack augmentation in the San Juan Mountains, Colorado, pp. 201-254. Final Report, San Juan Ecology Project, Colorado State Univ., Fort Collins. (Received 7 Jun 1984; accepted 16 Jan 1985) ANNOUNCEMENT California Botanical Society Award for Graduate Student Research An award of $250 will be given annually to the student member of the California Botanical Society submitting the winning proposal for thesis or dissertation research. The award is intended to help defray costs of research travel, field work, or laboratory supplies for which no other source of support exists. Each year's competition will be announced in MADRONO and at the CBS Graduate Student Meeting. Each applicant must be a student member of the California Botanical Society both in the year of submittal and during the year in which the funds are used. During the year of the award the student must be enrolled in a graduate program leading to the preparation of a thesis. Proposals are to include a brief description of the project, including an assessment of its importance; the need for funds; and a summary of current support. They should not exceed six double-spaced typewritten pages. Proposals should be accompanied by an evaluative cover letter from the major adviser for the project. Five copies are to be sent to the Past President of the Society. Proposals will be reviewed by a Committee consisting of the three elected Council members, the Second Vice President, and, as ex officio convener, the Past President. They will be evaluated on criteria of scientific merit and appropriateness for publi- cation in MADRONO upon completion (no contract is assumed on either side). The Committee will present a small number (usually 2-3) of the best proposals to the Council with a recommendation of one as winner. The final decision will rest with the Council. NOTES AND NEWS Notes on the Salvia leucophylla Complex (Lamiaceae) of California and Baja California Norte.— The Salvia leucophylla complex includes two species, S. leu- cophylla Greene and S. chionopeplica Epling. Salvia leucophylla is a member of the Southern California coastal sage scrub formation in the Coast Range foothills of California, from Monterey County southward to the Santa Ana Mountains of eastern Orange County. Salvia chionopeplica, a little-known Baja California endemic closely related to S. leucophylla, is differentiated by distinct floral and leaf features. Both species share a dendritic pubescence virtually unique to the Salvia section in which they belong, and have almost identical inflorescence structure, calyx morphology, and volatile oil components (Neisess, K. R., 1983, Evolution, systematics, and terpene relationships of Salvia Section Audibertia, Ph.D. diss., Univ. California, Riverside). Salvia leucophylla has a uniformly rose-lavender corolla (sometimes very pale) with pollen ranging in color from dusky-yellow to olive-drab, whereas S. chionopeplica has a distinctly blue-lavender corolla with bright yellow pollen. Vegetatively, the leaf blades of S. leucophylla are usually 3 or more times longer than wide, whereas those of S. chionopeplica are usually less than 2.5 times as long as wide. The range of S. chionopeplica has been stated to encompass the "western slopes of (the) Sierra San Pedro Martir from the vicinity of San Telmo south to San Fernando" (Wiggins, I. L., 1980, Flora of Baja California, Stanford Univ. Press). An extensive review of col- lected material deposited in western herbaria (SD, LAM, RSA, PC, OBI, DAV, UNLV, MACF, OSC, ASUC, TUC, UCR, MO, WTU, UTC, CIC, TEX, LL, BRY) indicates, and uniform garden studies confirm, nomenclatural confusion and distor- tion of the distributional range of both species. The focal point of this confusion involves a population occurring on Mesa el Barrial, along the road from San Telmo to Meling Ranch in the western foothills of the Sierra San Pedro Martir of Baja California. On-site investigations conducted in Au- gust, 1980, established that the leaves of the Barrial plants more closely resembled those of S. leucophylla, although somewhat smaller in size. To examine the exact degree of this relationship, seed was collected at the Barrial locality and 24 seedlings were grown under uniform garden conditions at the University of California, Riv- erside, with like samples of S. chionopeplica from the type locality (about 56 km east of El Rosario, near Rancho El Arenoso) and S. leucophylla from San Luis Obispo, Los Angeles, and Orange counties. Under these conditions, leaves of the Barrial plants clearly approximated the leaf shape characterizing S. leucophylla (Fig. 1). The flowers of these plants matched those of the S. leucophylla populations, although they came into bloom about two months earlier. Hand pollinations, conducted in greenhouse facilities at UC Riverside, proved both species interfertile and self-compatible. Plant association for S. chionopeplica, given by Wiggins as creosote bush scrub, is somewhat misleading. Although the area in which 5. chionopeplica occurs is predominantly creosote bush scrub, the species is usually found in localized, relatively mesic areas (north-facing slopes and summits of hills and small peaks) associated with typical coastal sage scrub species, such as Eriogonum fasciculatum Benth., Lotus scoparius Ottley, and Viguiera laciniata Gray. Evidently, the Mesa el Barrial population represents a southeastward disjunction of approximately 350 km for S. leucophylla, and is not S. chionopeplica as assumed previously (Fig. 2). It appears likely that the Mesa el Barriel population of S. leu- cophylla and the closely related S. chionopeplica are both remnants of Pleistocene assemblages of coastal sage scrub vegetation. — Kurt R. Neisess, Rancho Santa Ana Botanic Garden, 1500 N. College Ave., Claremont, CA 91711. (Received 3 May 1984; accepted 7 Aug 1985.) Madrono, Vol. 32, No. 4, pp. 273-275, 20 December 1985 274 MADRONO [Vol. 32 0 5 Scale cm Fig. 1 . Leaf variation in Salvia leucophylla and S. chionopeplica. Pairs represent the range of largest leaf size from each of three representative population samples in the uniform garden study conducted at the University of California, Riverside. A. El Arenoso, Baja California Norte (type locality of S. chionopeplica). B. Mesa el Barrial, Baja California Norte {S. leucophylla). C. Santiago Canyon, Orange County, California {S. leucophylla). 1985] NOTES AND NEWS 275 Fig. 2. Distribution of the Salvia leucophylla complex. REVIEWS Bibliographies on Chaparral and the Fire Ecology of Other Mediterranean Systems. By Jon E. Keeley. California Water Resources Center. University of California, Davis. Report No. 58. 1984. ISSN 0575-4968. This volume is a collection of separate and non-overlapping bibliographies on ten topics: 1) Chaparral— Evolution and Systematics, 2) Chaparral— Community Struc- ture, 3) Chaparral— Fire and Demography, 4) Chaparral— Morphology and Physi- ology, 5) Chaparral— Soils and Management, 6) Chaparral— Animals, 7) Seed Ger- mination [California species], 8) California Grasslands, 9) California Forests— Fire and Demography, and 10) Mediterranean Systems— Fire and Demography. Within each bibliography, citations are arranged alphabetically by authors' names. The first five of the six bibliographies that focus on chaparral contain an impres- sively comprehensive compilation of the literature. Several major references, how- ever, are missing from the animal bibliography. The inclusion of allelopathy references in the bibliography on seed germination is confusing, especially since the review of allelopathy literature is probably more comprehensive than that on germination (for example, only two of the articles in Seeds of Woody Plants of the United States are cited specifically). Particularly as regards non-chaparral species, there is a wealth of germination literature that is not reflected in this bibliography. The grasslands bib- liography also appears reasonably complete, except for the curious omission of any literature on vernal pools. By far the least thorough of the bibliographies is the one on California forests, whose coverage of the literature is spotty at best. This section seems to contain leftovers from the other bibliographies rather than representing a thorough search of its own. It was difficult to decide just what segment of the literature it was intended to represent. I found it particularly annoying that the literature on tree form oaks was apparently randomly split between this section and the chaparral bibliographies. The final section, which contains references for publications on Med- iterranean ecosystems outside of western North America, should provide easy access to the international literature, much of which might otherwise be ignored. The division of this volume into ten separate bibliographies with no cross-refer- encing is sometimes more of a hindrance than a help. Because there is no overlap between the citation listings, most searches will require paging through a minimum of three or four of the bibliographies to be sure relevant papers are not missed. Although I recognize the effort it would have required, the usefulness of this set of bibliographies would have been enhanced substantially by the addition of a com- prehensive subject index. Despite their weaknesses, these bibliographies should prove an excellent resource for those interested in chaparral. Proofing errors are generally minimal— with the exception that my name is misspelled throughout! The compiler is to be lauded for his attention to frequently ignored early papers and to easily missed master's theses. This completeness should make the bibliographies especially useful— to researchers and to graduate and undergraduate students— as an introduction to a wide body of literature. — Susan G. Conard, USDA Forest Service Forest Fire Laboratory, Riv- erside, CA. Insects and Flowers: The Biology of a Partnership. By Friedrich G. Barth. Trans- lated by M. A. Biederman-Thorson. ix + 297 pp. Princeton University Press, NJ. 1985. $35. ISBN 0-691-08368-1. New pollination books and reviews abound. One's first impression of this lively book. Insects and Flowers, is that it is a twin to another beautiful book published recently by Bastiaan Meeuse and Sean Morris called (unfortunately) The Sex Life of MADRofio, Vol. 32, No. 4, pp. 276-280, 20 December 1985 1985] REVIEWS 277 Flowers. But this is not an identical twin by any means; the greater emphasis in Earth's book is on the insects although the flowers are given a good deal of attention. The illustrations in both books are magnificent (in color and black and white plates and appropriate line drawings) and the texts are accurate and readable in both books. They are synergistic. Insects and Flowers was first published (in 1982) in German by Friedrich Barth, who is Professor of Zoology in Frankfurt. He was a graduate student at UCLA after gaining his first degree at the University of Munich. The book was translated by M. Biederman-Thorson, who is also a biologist as well as a professional translator of scientific works. The result is a book that reads as if it were written originally in English. There is an introductory chapter, followed by 6 chapters on pollination mechanisms and pollinators, followed, in turn, by 4 chapters on the collecting of pollen and nectar. Then there are 1 8 chapters on the senses and behaviors of insects (with most space, naturally, being devoted to bees). A concluding chapter explores the co-evolution that has produced the efficient "orthodox" pollination systems as well as the bizarre stories of the aroids and Orchidaceae and their relations with insects. Barth shows that he means business by starting with the complicated inter- action between figs and fig- wasps, but the going gets easier later on. His book will be useful to professional and amateur biologists for it is simply written, factually con- vincing, and well-backed by references. These references include German works that are not usually considered in books written in English. His explanation of the structure of insect mouth-parts is a marvel of clarity, which will be appreciated by non-ento- mologists. In his evolutionary considerations, Barth is an unashamed Darwinian "phyletic gradualist" (as opposed to being a "punctuated equilibrium" supporter), and he has the advantage of material that fits the gradualist view of insect/plant co-evolution. His discussion of the evolution of the social habit among bees, and the almost incredible senses and communication patterns that these insects show, is considered in relation to the flowers and their phenology, thus making this book appropriate reading for botanists! There are very few misstatements of botanical fact even though the author is a zoologist. Thus, reference is made to "nectar guides" in some taxa that do not produce this liquid, but these are scarcely visible blemishes on a text that is well-informed and most informative. Princeton University Press has brought us a book that will be a leader in its field.— H. G. Baker, University of CaHfomia, Berkeley. California Riparian Systems: Ecology, Conservation, and Productive Management. Edited by Richard E. Warner and Kathleen M. Hendrix. University of California Press, Berkeley. 1984. xxix + 1035 pp. $57.50, $19.95 paper. ISBN-529-05034-7 and -05035-5 (pbk.). This massive, catholic, inclusive, wide-ranging book is the record of, and a mon- ument to, the California Riparian Systems Conference held at the University of California, Davis, 17-19 Sep 1981. The more than 250 authors are outstandingly expert, dedicated, responsible, involved. One hundred twenty-eight papers are printed in 22 sections: Biogeography and change; Structure, status and trends; Hydrology related to structure, function and protection; Aquatic riparian interactions; Riparian/ upland interactions; Economic and social values; legal framework; Classification, inventory, and monitoring; National and regional trends in use; Restoration; Water diversion project conflicts; Levees; Bird populations; Coastal zone; Desert systems; Sustained yield production; Cultural, recreational and aesthetic values; Integrated approaches to management— protection; State versus local control; The Rivers and Harbors Act of 1 899 and conservation; Non-avian wildlife; and Private ownership. This sectional classification is far from absolute. Very many papers burst their bounds to become even more interesting. However, a 39 page, double column index will facilitate reference. Notably it does not include authors' names. 278 MADRONO [Vol. 32 The book is obviously more than a mine of information for California's botanists. Even on the subject of riparian ecosystems it ranges not only throughout California but from Oregon to the Great Basin and the desert Southwest, from the mountains to the sea coast. This vast area is mostly arid, but riparian ecosystems are obviously uniquely mesic in the summer-dry, lowland landscapes of the American West. Judging from the productivity studies available from Soviet work on similar riparian systems in central Asia (tugai), our western riparian plant communities may have been the most productive natural vegetation in the West. Several papers, quite arbitrarily selected, supply intelligible botanical data, contain traces of the field work on which they are based, use more or less standard methods, compare their results with similar studies, produce an inductive rather than a de- ductive classification, and employ few or no ""Salix spp." (Strahan, Whitlow, and Bahre; McBride and Strahan; Laymon, Shanfield, Warner; etc.). Others, which are simply intensely interesting, on the Carmel River, structure of vegetation. Mono Lake, and Prosopis glandulosa productivity, deserve mention (Kondolf and Curry; Stone, Cavallaro, and Stromberg; Stine, Gaines, and Vorster; Nilsen, Rundel, and Shariff; etc.). Holstein's biogeographical paper is a model, with adequate paleobo- tanical and ecological data, including good distribution maps. Other readers will certainly have their own highlights. An only slightly slimmer European book is an interesting supplement and contrast to the Califomian volume (Gehu, J.-M., ed., 1984, La vegetation des forests alluviales, Colloques Phytosociologiques 9 in Strasbourg, 1980. J. Cramer, Vaduz, xiv + 744 pp., tables). Riparian vegetation from Austria, Hungary, Rumania, and Czechoslo- vakia to Spain, and from Italy to W. Germany was discussed by 94 attendees in 46 papers. There were field trips in addition to the presentations. Languages are French, German, English, and Italian. About Vs of the book's thickness is tables of the stands of vegetation studied, the data manipulated and discussed. Braun-Blanquet's methods of studying vegetation and its ecology were used. A classification of riparian vegetation has resulted and is generally agreed upon. Where numerical methods of analysis were employed, the results are also often illustrated by tables of stands. The numerical data can thus be tested by the reader's ecological experience with the species in the field, greenhouse, or laboratory. A stand's position in an ordination or in a classification of vegetation is due to its species composition. Which species? The unique value of Braun-Blanquet's method of studying vegetation is that the result is a table of individual species occurrences in particular stands. Much can be done with such data; no interpretation of vegetation and particularly no interpretation of the ecology of vegetation is of value without such data. If the reader can tear himself away from this volume, including a paper that used good topographic maps of riparian areas of the upper Rhine from 1838, 1852, 1872, and the present or from another that from a population of 1420 stands containing 1223 species made a selection of 328 riparian stands containing 331 species, then a volume by B. M. Mirkin et al. (1980) {Riparian vegetation of the Mongolian Peoples' Republic, Biological resources and natural conditions of the MPR, vol. 13. 284 pp. Nauka, Leningrad) should be of interest. It covers in considerable detail the vegetation of a dozen Mongolian rivers from their origins in the high mountains to their dis- appearance in the Gobi deserts or across the Soviet border. This group took data on 3500 stands and made 5000 plant collections, so we evidently still have some purely botanical work to do on California's riparian ecosystems!— Jack Major, Botany Department, University of California, Davis. Forest Succession and Stand Development Research in the Northwest. Proceedings of the Symposium held 26 March 1981, at CorvaUis, Oregon. Edited by Joseph E. Means. Forest Research Laboratory, Oregon State University, Corvallis 9733 1 . 1 982. $6.00. In creating his revised tolerance table for American forest trees. Baker (1949) made a significant observation. Noting that a number of the "outstanding men in the field 1985] REVIEWS 279 of silviculture" submitted a minority view on the place of a species in the tolerance hierarchy he commented: "We are at a loss to account for this, but the disturbing suggestion has been made that most of us accept the things we learned in school as the gospel truth; only a few of the best learn to observe and think for themselves." It is apparent from some of the papers of this symposium that many of the participants were among the outstanding men in the field. Although the study of succession is considered by many to be responsible for the birth of ecology in the United States and has been a major focus of an inordinate number of ecological studies and pub- lications, there is still a great deal to be learned about whatever we want to include under the heading of succession. The symposium proceedings cover a broad range of topics including studies of trees growing in computers, observations of real world situations where old dogmas were not verified, and a discussion of stratification of the sites as a means of under- standing the processes of succession. The papers are arranged by subject matter into two parts. Part I is devoted to the larger topic of forest succession. The introductory paper by D. M. Smith sets a fine stage for the papers that follow. He raises a number of interesting points about some of the differences that have developed in ecological thought and warns us against trying to force observed patterns into the few classes already described. A trap that he does not mention is letting our preconceived patterns influence the results of our observations. Hopefully none of us are subject to that mistake. The next four papers describe computer simulations. The first of these papers, by D. C. West and others, aptly introduces these, giving examples of how simulations may help us compress time into a comprehensible vector and make some predictions about long-term effects of management actions or natural variations of the environ- ment. Six papers are devoted to reporting observations of real world situations from northern California to southeast Alaska and inland to western Montana. For some reason Canada was not represented in the symposium. Papers by R. D. Pfister and S. F. Amo demonstrated the reason for using some form of habitat type classification for stratification of successional observations. Part II consists of papers on stand development. The introductory paper of this section by C. D. Oliver gives an overview of the growth, development, and importance of this newly-budded aspect of succession. Papers by B. C. Larsen and S. D. Viers give examples of why careful observations are better than conventional wisdom in trying to understand natural processes. Larsen demonstrated that much of a stand's growth may be concentrated in the dominant trees and that thinning does not always release the smaller trees. Viers demonstrated that redwoods would reproduce with or without fire in certain environments. In the summary paper, J. F. Franklin pointed out a lack of papers dealing with permanent plot studies. This results from a lack of research using that method. Also evident in most of the papers is our continued dependence on the eastern deciduous forest for successional concepts. These concepts are undoubtedly useful for a great number of areas in the west, but there is still a great deal to learn about succession where shade tolerance is not the driving force behind the patterns of succession we see, especially in many of the drier interior forests. The body of the text is reduced to the point where it is difficult to read. Some of the figures have lettering that is almost too small, but I didn't find any that were unreadable. Also, don't plan to read it more than a few times. My copy is now held together with a paper clip. All things considered, this is a valuable volume for those studying succession in the Pacific Northwest. The summary paper by Franklin is especially valuable. Many suggestions for further studies are made that, if followed, would certainly improve our understanding of succession. It was emphasized that more permanent plot studies should be started now to help validate the models and hypotheses that have evolved from comparing different-aged stands. — Don G. Despain, Yellowstone National Park, Wyoming. 280 MADRONO [Vol. 32 ANNOUNCEMENT A New Section for MADRONO Beginning with Volume 33, each issue of MADRONO will contain an ed- itorial page on which comments by the editor, invited contributions, unsoli- cited letters, and other remarks will be featured. This editorial page will serve as a vehicle for communication among our members and could include, for example, opinions of authorities on current trends in botany, rebuttals or comments on papers published in MADRONO, letters from the President of our society, and other noteworthy communications. The editors invite all members to participate in this forum; however, all editorials will be published at the discretion of the editors. ANNOUNCEMENT Attention Contributors of Floras New conventions Jbr the publication of floras have been adopted by the editors of MADRONO. All annotated catalogues and lengthy checklists will be printed in 8 point type and will appear in the body of papers after the discussion section and before acknowledgments. The catalogue should include latin names and authors, references to habitats and plant communities in which the plants are found, standardized estimates of the frequency and abundance of each taxon, and citation of a limited number of voucher specimens. The annotated catalogue also should include or be preceded by an explanation of categories and abbreviations. Nomenclature should be current and synonyms should follow names that do not appear in standard regional floras. Other sections of floras, including introduction, study area, methods, vegetation, analysis of the flora, etc., will be printed in 10 point type. We encourage the use of a limited number of high quality photographs of plant communities or topographic setting, but such photographs are not required. Standardization of the basic format of floras will improve the quality of such papers published in MADRONO and should reduce the amount of revision that may be required. We look forward to working with authors and welcome^ suggestions regarding conventions of format and style adopted for MADRONO. 1985] ANNOUNCEMENTS 281 ANNOUNCEMENT The Tenth Annual Graduate Student Meeting, sponsored by the California Botan- ical Society, was held at the University of California, Santa Barbara on 1 9 October 1985. More than 50 people attended the twenty presentations by students from 8 institutions. The following individuals received awards for their presentations: Completed Research Best Paper Second Place Third Place Research in Progress Best Paper Second Place Third Place Proposed Research Best Paper Second Place Jeffrey P. Hill Department of Botany and Plant Sciences University of California, Riverside William Bond Department of Biology University of California, Los Angeles Niall F. McCarten Department of Biological Sciences San Francisco State University Alan L. KoUer Department of Botany University of California, Davis Richard W. Kerrigan Department of Biological Sciences University of California, Santa Barbara Kirk E. Apt Department of Biological Sciences University of California, Santa Barbara Edith A. Reed Department of Ecology and Evolutionary Biology University of California, Irvine Christine Schierenbeck Department of Biological Sciences San Francisco State University The Society also would like to recognize and thank Kathy Rindlaub, Chairperson of the Organizing Committee, and the many committee members for providing a well organized and successful meeting; John Bleck for hosting a social hour and tour of the UCSB greenhouses; and Walter H. Muller for giving an entertaining and informative keynote address at the banquet. Next year, the meeting will be held at the University of California, Davis. 282 MADRONO [Vol. 32 REVIEWERS OF MANUSCRIPTS The editors thank all reviewers listed below for their assistance with papers pub- lished in volume 32. We are grateful for their generous contributions^of time and effort toward maintaining the quality of papers published in MADRONO. Wayne Armstrong Daniel I. Axelrod H. G. Baker Mary E. Barkworth James Bartel Derek Burch Judith M. Canne Kenton L. Chambers Susan G. Conard Lincoln Constance William B. Critchfield Robert W. Cruden Alva G. Day W. J. Ferlatte Amy Jean Gilmartin James R. Griffin Lawrence R. Heckard Douglass Henderson James Henrickson James Hickman John T. Howell Duncan Isely Dale Johnson Jon E. Keeley Sterling Keeley Robert W. Lichvar John Little Jack Maze Elizabeth McClintock Steven P. McLaughlin John McNeill Robert J. Meinke Richard A. Minnich Reid Moran Arthur M. Phillips Donald J. Pinkava Duncan M. Porter Barry A. Prigge J. Rzedowski Clark Schaack Rob Schlesing James R. Shevock James P. Smith Richard Spellenberg John L. Strother Ronald J. Taylor Robert F. Thome Frank C. Vasek William A. Weber Grady L. Webster Dieter H. Wilken 1985] EDITORS' REPORT FOR VOLUME 32 283 EDITORS' REPORT FOR VOLUME 32 This annual report provides an opportunity for the editors to communicate the status of manuscripts received for publication in MADRONO and to comment on other aspects of the journal. Between 1 Jul 1984 and 30 Jun 1985, 86 manuscripts were received (34 articles, 1 2 notes, and 40 individual noteworthy collections). This total includes manuscripts received by the previous editor and the new editors. The current status of all unpublished manuscripts, including those received after 30 Jun 1985, is 33 in review (7, 3, 23), 7 in revision (7, 0, 0), 14 awaiting decision by the editors (9, 3, 2), and 1 9 accepted (8, 1 , 10). There are three unpublished book reviews. Volume 32 included 83 published manuscripts (22, 7, 54) and 7 reviews, totaling 283 pages plus an index. The period between submittal and publication has averaged about one year. The past year has included the transition between editors, and hence also has been a period of readjustment of review and publication schedules. The new editors apol- ogize for the lateness of some communications with authors and for the slowness of the review process during the past few months. We expect the editorial process to improve with volume 33 and look forward to working with the present and future contributors. We thank the authors for their patience during this early period of our editorships and are grateful for the guidance provided by the Executive Council. We give special thanks to the former editor. Dr. Christopher Davidson, Director of the Idaho Botanical Garden, for his kind and helpful assistance during the transition period. On behalf of the Society, we express sincere appreciation for the professional guidance of the journal during his AVi years as editor. The many important^papers and notes published during his editorship illustrate the strength of MADRONO and the contribution Chris has made. We shall strive to continue this record of excellence and to expand the coverage and readership of MADRONO. We also take this opportunity to encourage members of the Society to submit well- written manuscripts for review and for potential publication in MADRONO. The strength of our journal depends on the quality and quantity of papers submitted by members. In addition to the various papers published traditionally in MADRONO, the editors welcome manuscripts on other topics such as marine phycology and bryology. We also welcome all suggestions from authors and other members to help us maintain or improve the status of MADRONO as an important botanical journal. W.R.F. and J.R.H. 30 Oct 1985 Dates of publication of MADRONO, volume 32 No. 1, pp. 1-63: 15 Feb 1985 No. 2, pp. 65-130: 26 Apr 1985 No. 3, pp. 131-195: 19 Aug 1985 No. 4, pp. 197-289: 20 Dec 1985 284 MADRONO [Vol. 32 INDEX TO VOLUME 32 Classified entries: major subjects, key words, and results; botanical names and plant families (new names are in boldface); geographical areas; reviews. Incidental references to taxa (including most lists and tables) are not indexed separately. Species appearing in Noteworthy Collections are indexed under name, family, state, or country. Authors and titles are listed alphabetically in the Table of Contents. Abies concolor, inhibition of radicle growth, 118; varieties in southern CA, 65. Adenostoma fasciculatum, post-fire seed- ling establishment, 148. Alpine flora: San Juan Mtns., CO, 253. Amargosa Range, CA: full-glacial vege- tation, 1 1 . Amsinckia carinata, rediscovery in OR, 124. Antennaria microphylla, new record for AZ, 121. Aquilegia chrysantha, new record for AZ, 121. Aristida oligantha, range extension for WY, 125. Arizona: floral nectar-sugar composition of species, 78; revised flora of Tuma- moc Hill, 225. New records: Antennaria microphylla, \25; Aquilegia chrysantha, 121; Car- ex haydeniana, 121; Corchorus hir- tus, 191; Ibervilla tenuisecta, 191; Pinus flexilis, 121; Potentilla nivea, 122. Range extension: Cowania subintegra, ni. Asarina procumbens, systematic rela- tionship to New World species, 168. Aster sibericus, new record for UT, 125. Asteraceae: Lasthenia maritima, new combination, 131. New records: Antennaria microphylla in AZ, 121; Aster sibericus in UT, 125; Gnaphalium viscoscum in UT, 125; Pedis angustifolia var. angus- tifolia in WY, 127. Astragalus robbinsii, new record for UT, 125. Baja California Norte: Salvia leucophylla complex, 273; 5". chionopeplica, 273. Berberidaceae: Vancouveria hexandra seed dispersal, 56. Boisduvalia glabella, new record for WY, 126. Boraginaceae: Cryptantha celosioides, new record for NV, 123. Rediscovery: Amsinckia carinata in OR, 124. Brassicaceae: New records: Draba api- culata in CA, 1 9 1 ; D. douglasii in UT, 1 25; Z). spectabilis var. oxyloba in WY, 192; Thellungiella salsuginea in WY, 128. Range extension: Rorippa truncata in WY, 128. Burroughsia: notes on, 186. Cactaceae: Opuntia pulchella, new record for CA, 123. Range extension: Opuntia macrorhiza var. macrorhiza in WY, 127. California: Adenostoma fasciculatum and Ceanothus greggii post-fire seedling establishment, 148; chaparral, 148; desert sand dune flora, 197; Salvia leu- cophylla complex, 273; Spartina dis- tribution and taxonomic notes, 158. New combinations: Chamaesyce ab- ramsiana, C. hooveri, C. ocellata subsp. rattanii, C. serpyllifolia subsp. hirtula, 188; Lasthenia maritima, 131. New records: Dalea ornata, 123; Ivesia baileyi, 123; Juncus cyperoides, 191; Opuntia pulchella, 123. New taxa: Linanthus nuttallii subsp. howellii, 102; Mimulus norrisii, 179; Paronychia ahartii, 87. Range extension: Dedeckera eureken- sis, 122. Carex haydeniana, new record for AZ, 121. Caryophyllaceae: LoeJIingia squarrosa subsp. texana, new record for WY, 1 27; Paronychia ahartii, new species from CA, 87. Ceanothus: C. greggii, post-fire seedling establishment, 148; C. velutinus, rad- icle growth inhibition by extract of, 118. Celtis occidentalis var. occidentalis, range extension for WY, 126. Cercocarpus: chromosome counts for C. montanus var. montanus, C ledifolius MadroSo, Vol. 32, No. 4, pp. 284-288, 20 December 1985 1985] INDEX 285 var. intercedens, C. ledifolius var. ledi- folius, 24. Chamaebatia: chromosome counts for C. australis and C. foliolosa, 24. Chamaebatiaria: chromosome count for C. millifolium, 24. Chamaesyce in the Galapagos Islands, 143. Chaparral: post-fire seedling establish- ment, 148. Chenopodiaceae: Grayia brandegei, new record for NM, 192. Chromosome numbers: Cercocarpus ledifolius var. intercedens, C. ledifolius var. ledifolius, C. montanus var. mon- tanus, Chamaebatia australis, C. foli- olosa, Chamaebatiaria millifolium, Coleogyne ramosissima, Holodiscus dumosus, Kelseya unijlora, Peraphyl- lum ramosissimum, Petrophytum caespitosum, 24; Mimulus spp., 91. Claytonia lanceolata var. flava, new re- cord for WY, 126. Coleogyne: chromosome count for C. ra- mosissima. Colorado: alpine flora, 253; Draba api- culata, new record for, 191. Compositae: see Asteraceae. Corchorus hirtus, new record for AZ, 191. Cowania subintegra, range extension for AZ, 122. Cracca: C. glabella, new combination and status, 98; identity in the U.S.A., 95. Crypt ant ha celosioides, new record for NV, 123. Cucurbitaceae: Ibervillea tenuisecta, new record for AZ, 191. Cyperaceae: Carex haydeniana, new rec- ord for AZ, 121; Scirpus heterochaetus, range extension for WY, 128. Cytotaxonomy: Rosaceae, 24. Dalea ornata, new record for CA, 123. Death Valley, CA: Neotoma (wood rat) middens and macrofossils, 1 1 ; Pleis- tocene vegetation, 1 1 . Dedeckera eurekensis, range extension for CA, 122. Dicentra pauciflora, range extension for CA, 57. Dispersal: Lasthenia maritima by sea- birds, 131; Vancouveria hexandra by yellow jackets, 56. Draba: New records: D. apiculata in CA, 191; D. douglasii in UT, 125; Z). spec- tabilis var. oxyloba in WY, 192. Ecuador: Chamaesyce in the Galapagos Islands, 143. Editors' report, 283; editors' announce- ments, 280. Eragrostis trichodes var. trichodes, new record for WY, 126. Eriogonum: New records for NV: E. crosbyae, 123; E. prociduum, 124. Erythronium elegans, new species from OR, 49. Euphorbia serpens, new record for WY, 126. Euphorbiaceae: Chamaesyce in the Ga- lapagos Islands, 143. New combinations: Chamaesyce ab- ramsiana, C. hooveri, C. rattanii, C. serpyllifolia subsp. hirtula, 187. New record: Euphorbia serpens in WY, 126. Fabaceae: Cracca in the U.S.A., 95; C. glabella, new combination and status, 98. New records: Astragalus robbinsii in UT, 125; Dalea ornata in CA, 123. Festuca: F. altaica complex: nomencla- ture and taxonomy, 1 ; F. altaica subsp. hallii, new combination, 9. Floras: Addenda to San Luis Obispo Co., CA, 214; alpine flora, San Juan Mtns., CO, 253; revised flora, Tumamoc Hill, AZ, 225; sand dune flora. Great Basin and Mojave deserts, CA, NV, OR, 1 97. Fremont: ecological solution to missing cannon, 106. Gnaphalium viscoscum, new record for UT, 125. Gramineae: see Poaceae. Grayia brandegei, new record for NM, 192. Great Basin Desert: sand dune flora, 197. Grossulariaceae: Ribes laxiflorum, new record for UT, 125. Growth inhibition: extracts ofCeanothus velutinus on Abies concolor, 118. Holodiscus dumosus, chromosome count, 24. Hydrophyllaceae: New records: Phacelia constancei in CO, 57; P. thermalis in NV, 124. Ibervillea tenuisecta, new record for AZ, 191. Insect damage: Pinus coulteri, 29. 286 MADRONO [Vol. 32 Ivesia: I. baileyi, new record for CA, 123; /. rhypara, range extension for NV, 1 24. Juncaceae: /. cyperoides, new record for CA and North America, 191. Kelseya uniflora, chromosome count, 24. Klamath Mtns., CA: Linanthus nuttallii subsp. howellii, new subspecies, 102. Lamiaceae: Salvia leucophylla complex, 273. New records: Monardella odoratissi- ma subsp. glauca in WY, 127; Scu- tellaria holmgreniorum in NV, 1 24. Larix occidentalis, new record for WY, 126. Lasthenia maritima, new combination, 139; dispersal by seabirds, 131. Leguminosae: see Fabaceae. Lemnaceae: Range extensions in CA: Spirodela punctata, 57; Wolffia bo- realis, 57. Leptodactylon watsonii, new record for WY, 126. Liliaceae: Erythronium elegans, new species from OR, 49; Veratrum te- nuipetalum, new record for WY, 128. Linanthus nuttallii subsp. howellii, new subspecies from CA, 102. Linaria canadensis var. texana, range ex- tension for WY, 127. Loasaceae: Mentzelia packardiae, new record for NV, 1 24. Loeflingia squarrosa subsp. texana, new record for WY, 127. Los Padres National Forest, CA: Pinus coulteri, serotiny and cone habit vari- ation, 29. Mentzelia packardiae, new record for NV, 124. Mexico: Burroughsia, notes on, 186; Pi- nus patula var. longepedunculata, range extension for Oaxaca, 191; Salvia leu- cophylla complex in Baja California Norte, 273. Mimulus: M. glabratus complex and chromosome numbers, 91; M. nor- risii, new species from CA, 179. Mojave Desert: Pleistocene vegetation, 11; sand dune flora, 197. Monardella odoratissima subsp. glauca, new record for WY, 127. Nectar-sugar composition of flowers, 78. Neotoma (wood rat): macrofossils in Death Valley, CA, 11. Nevada: desert sand dune flora, 197. New records: Cryptantha celosioides, 123; Eriogonum crosbyae, 123; E. prociduum, MA; Ivesia rhypara, 124; Mentzelia packardiae, 1 24; Phacelia thermalis, 124; Scutellaria holm- greniorum, 124. New Mexico: floral nectar-sugar com- position of species, 78. New records: Grayia brandegei, 192; Rhynchelytrum repens, 192. North America: Juncus cyperoides, new record, 191. Onagraceae: Boisduvalia glabella, new record for WY, 126. Opuntia: O. pulchella, new record for CA, 123; O. macrorhiza var. macrorhiza, range extension for WY, 127. Oregon: Amsinckia carinata, rediscovery of, 124; desert sand dune flora, 197; Erythronium elegans, new species, 49; Pleuropogon oregonus, rediscovery and reproductive biology, 189. Papaveraceae: Dicentra pauciflora, range extension for CA, 57. Paronychia ahartii, new species from CA, 87. Peraphyllum ramosissimum, chromo- some count, 24. Petrophytum caespitosum, chromosome count, 24. Phacelia: New records: P. constancei in CO, 57; P. thermalis in NV, 124. Physalis hederaefolia var. comata, range extension for WY, 128. Pinaceae: Abies concolor, inhibition of radicle growth, 118; A. concolor, va- rieties in southern CA, 65; Pinus coul- teri, serotiny and cone habit variation, 29. New records: Larix occidentalis in WY, 1 26; Pinus flexilis in AZ, 121. Range extension: Pinus patula var. longepedunculata in Oaxaca, Mexi- co, 191. Pinus: P. coulteri, serotiny and cone habit variation, 29; P. flexilis, new record for AZ, 121; P. patula var. longepedun- culata, range extension for Oaxaca, Mexico, 191. Pleistocene vegetation: Death Valley, CA, 11. 1985] INDEX 287 Pleuropogon oregonus, rediscovery and reproductive biology, 189. Poaceae: Festuca altaica complex, 1; F. altaica subsp. hallii, new combination, 9; Pleuropogon oregonus, rediscovery and reproductive biology, 189; Spar- tina in northern CA, distribution and taxonomic notes, 158; Ventenata du- bia, identification of, 1 20. New record: Rhynchelytrum repens in NM, 192. Range extensions: Aristida oligantha in WY, 125; Eragrostis var. tri- chodes in WY, 125. Polemoniaceae: Leptodactylon watsonii, new record for WY, 126; Linanthus nuttallii subsp. howellii, new subspe- cies in CA, 102. Pollination biology: floral nectar-sugar composition of species, 78. Polygonaceae: New records for NV: Er- iogonum crosbyae, 123; prociduum, 124. Range extension: Dedeckera eureken- sis in CA, 122. Portulacaceae: Claytonia lanceolata var. flava, new record for WY, 126. Potentilla: New records: P. hookeriana in WY, 128; P. nivea in AZ, 122. Ranunculaceae: Aquilegia chrysantha, new record for AZ, 121. Reproductive biology: Pleuropogon ore- gonus, 189. Reviews: F. G. Barth, Insects and flow- ers: the biology of a partnership, 276; R. D. Dom, Vascular plants of Mon- tana, 193; A. Cronquist, A. H. Holm- gren, N. H. Holmgren, J. L. Reveal, and P. K. Holmgren, Intermountain flora: vascular plants of the Inter- mountain West, U.S.A., vol. 4, 58; J. E. Keeley, Bibliographies on chaparral and fire ecology of other Mediterra- nean ecosystems, 276; J. E. Means, ed.. Forest succession and stand develop- ment research in the Northwest, 278; R. E. Warner and K. M. Hendrix, eds., California riparian systems: ecology, conservation, and productive manage- ment, 277; R. L. Williams, Aven Nel- son of Wyoming, 194. Rhamnaceae: Ceanothus greggii, post-fire seedUng establishment of, 148; C ve- lutinus, radicle growth inhibition by extract of, 118. Rhynchelytrum repens, new record for NM, 192. Ribes laxiflorum, new record for UT, 125. Rorippa truncata, range extension for WY, 128. Rosaceae: Adenostoma fasciculatum, post-fire seedling establishment, 148. Chromosome numbers: Cercocarpus ledifolius var. intercedens, C. ledi- folius var. ledifolius, C. montanus var. montanus, Chamaebatia aus- tralis, C. foliolosa, Chamaebatiaria millifolium, Coleogyne ramosissi- ma, Holodiscus dumosus, Kelseya uniflora, Peraphyllum ramosissi- mum, Petrophytum caespitosum, 24. New records: Ivesia baileyi in CA, 123; Potentilla hookeriana in WY, 128; P. nivea in AZ, 122. Range extensions: Cowania subintegra in AZ, 122; Ivesia rhypara in NV, 124. Salvia: S. leucophylla complex in CA and Baja California Norte, 273; S. chio- nopeplica in Mexico, 273. San Juan Mtns., CO: alpine vascular flora of, 253. San Luis Obispo Co., CA: addenda to the vascular flora of, 214. Sand dune flora: Great Basin and Mojave deserts, 197. Scirpus heterochaetus, range extension for WY, 128. Scrophulariaceae: Asarina procumbens, systematic relationship to New World species, 168; Mimulus norrisii, new species from CA, 179. Chromosome numbers in Mimulus: M. andicolus, M. andicolus x M. gla- bratus var. glabratus, M. glabratus var. glabratus, M. glabratus var. fre- montii, M. pilosiusculus, 92. Range extension: Linaria canadensis var. texana in WY, 127. Scutellaria holmgreniorum, new record for NV, 124. SeedUng morphology: Abies concolor, 65. Serotiny: Pinus coulteri, 29. Sierra Nevada: Mimulus norrisii, new species from CA, 179. Solanaceae: Phy sails hederaefolia var. comata, range extension for WY, 128. Southern Coast Ranges: Pinus coulteri, serotiny and cone habit variation, 29. 288 MADRONO [Vol. 32 Spartina, distribution in northern CA and taxonomic notes, 158. Spirodela punctata, range extension for CA, 57. Sweetwater Mtns., CA: ecological solu- tion to missing Fremont cannon, 106. Terpenoid chemistry: Abies concolor, 65. Thellungiella salsuginea, new record for WY, 128. Tiliaceae: Corchorus hirtus, new record for AZ, 191. Toiyabe National Forest, CA: ecological solution to missing Fremont cannon, 106. Tumamoc Hill, AZ: revised vascular flora, 225. Ulmaceae: Celtis occidentalis var. occi- dentalis, range extension for WY, 126. Utah: New records: Aster sibericus, 125; Astragalus robbinsii, 125; Draba doug- lasii, 125; Gnaphalium viscoscum, 125; Ribes laxiflorum, 125. Vancouveria hexandra, seed dispersal by yellow jackets, 56. Ventenata dubia, identification of, 1 20. Veratrum tenuipetalum, new record for WY, 128. Verbenaceae: Burroughsia in Mexico, 186. White fir {Abies concolor): inhibition of radicle growth, 118; varieties in south- em CA, 65. Wolffia borealis, range extension for CA, 57. Wyoming: New records: Aristida oligan- tha, 125; Boisduvalia glabella, 126; Celtis occidentalis var. occidentalis, 126; Claytonia lanceolata var. flava, 126; Draba spectabilis var. oxyloba, 192; Eragrostis trichodes var. tri- chodes, 126; Euphorbia serpens, 126; Larix occidentalis, 126; Leptodactylon watsonii, 126; Loejlingia squarrosa subsp. texana, 1 27; Linaria canadensis var. texana, 127; Monardella odora- tissima subsp. glauca, 127; Opuntia macrorhiza var. macrorhiza. 111; Pec- tis angustifolia var. angustifolia, 127; Physalis hederaefolia var. comata, 128; Potentilla hookeriana, 128; Rorippa truncata, 128; Scirpus heterochaetus, 128; Thellungiella salsuginea, 128; Ve- ratrum tenuipetalum, 128. Yellow jackets: dispersal of Vancouveria hexandra seeds, 56. CALIFORNIA BOTANICAL SOCIETY MEETING PROGRAM FOR 1985-1986 8:00 PM University of California, Berkeley LSB 4093 DATE SPEAKER & TOPIC SEP 19 Deborah LeToumeau, University of California, Santa Cruz Mutualism in a tropical rainforest. OCT 17 Don Santana, Gavilan College Studies in North American Fritillaria. NOV 21 Pam Matson, NASA, Ames Disturbance effects on nutrient cycling in forest ecosystems. JAN 16 Howard Wilshire, USGS, Menlo Park Residual impacts of World War II armored maneuvers in Cahfomia's deserts. ♦FEB 22 Prof. Joseph Ewan, Tulane University What happened beside the Golden Gate? Was it imitation or innovation? * ANNUAL BANQUET -Location to be announced. MAR 20 Steve Botti, NPS, Yosemite Rare and endangered flora of Yosemite. APR 17 Thomas Fuller, Sacramento, CA Plants and livestock poisoning. MAY 15 Paul Zinke, University of California, Berkeley Analytical variation of foliage in relation to soils. I MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME XXXII 1985 BOARD OF EDITORS Class of: 198 5 — Sterling C. Keeley, Whittier College, Whittier, CA Arthur C. Gibson, University of California, Los Angeles 1986— Amy Jean Gilmartin, Washington State University, Pullman Robert A. Schlising, California State University, Chico 1987— J. Rzedowski, Instituto de Ecologia, A.C., Mexico Dorothy Douglas, Boise State University, Boise, ID 1988 — Susan G. Conard, USDA Forest Service, Riverside, CA William B. Critchfield, USDA Forest Service, Berkeley, CA 1989 — Frank Vasek, University of California, Riverside Barbara Ertter, University of California, Berkeley Editor— Christopher Davidson [32(1-3)] Idaho Botanical Garden, P.O. Box 2140, Boise, ID 83701 -Wayne R. Ferren, Jr. [32(4)] Associate Editor— J. Robert Haller [32(4)] Department of Biological Sciences University of California, Santa Barbara, CA 93106 Published quarterly by the California Botanical Society, Inc. Life Sciences Building, University of California, Berkeley 94720 Printed by Allen Press, Inc., Lawrence, KS 66044 ii To Rimo Bacigalupi, "Ba[t]ch," first Curator of the Jepson Her- barium and Library (1950-1968) for his lifelong devotion to Nature and The Arts, for his extraordinary knowledge of the California flora that he generously shared with colleagues and successive generations of students, for his helpfulness as a linguist and botanical Latinist, for his warm personality, and for his civilizing influence, volume 32 is affectionately dedicated. Photo taken in 1959 by Marion Cave. iii TABLE OF CONTENTS Allen, Robert L. (see Keil, David J. et al.) Anderson, J. (see Schaack, C. G.) Atwood, Duane (see Goodrich, S.) Baker, H. G., Review of Insects and flowers: the biology of a partnership (by Friedrich G. Barth, translated by M. A. Biederman-Thorson) 276 Borchert, Mark, Serotiny and cone-habit variation in populations of Pinus coulteri (Pinaceae) in the Southern Coast Ranges of California 29 Bowers, Janice E. and Raymond M. Turner, A revised vascular flora of Tumamoc Hill, Tucson, Arizona 225 But, Paul P. H., Jimmy Kagan, Virginia L. Crosby, and J. Stephen Shelly, Rediscovery and reproductive biology of Pleuropogon oregonus (Poaceae) 189 Chambers, Kenton L., Pitfalls in identifying Ventenata dubia (Poaceae) 120 Chambers, Kenton L. (see also Hammond, Paul C.) CoNARD, Susan G., Inhibition of Abies concolor radicle growth by extracts of Ceanothus velutinus 1 1 8 CoNARD, Susan G., Review of Bibliographies on chaparral and the fire ecology of other Mediterranean systems (by Jon E. Keeley) 276 Crosby, Virginia L. (see But, Paul P. H.) Despain, Don G., Review of Forest succession and stand development research in the Northwest (edited by Joseph E. Means) 278 Dorn, Robert D., Review of Aven Nelson of Wyoming (by Roger L. Williams) 194 Elisens, Wayne J., The systematic relationship of Asarina procumbens to New World species in Tribe Antirrhineae (Scrophulariaceae) 168 Ellis, Barbara A. (see Kummerow, Jocken et al.) Ertter, Barbara, Paronychia ahartii (Caryophyllaceae), a new species from CaUfomia 87 Freeman, C. Edward and Richard D. Worthington, Some floral nectar- sugar compositions of species from southeastern Arizona and southwestern New Mexico 78 Goodrich, Sherel, Duane Atwood, and Elizabeth Neese, Noteworthy col- lections of Aster sibiricus, Gnaphalium viscoscum, Draba douglasii, As- tragalus robbinsii, and Ribes laxiflorum 125 Hammond, Paul C. and Kenton L. Chambers, A new species of Erythronium from the Coast Range of Oregon 49 Harms, Vernon L,, A reconsideration of the nomenclature and taxonomy of the Festuca altaica complex (Poaceae) in North America 1 Hartman, Emily L. and Mary Lou Rottman, The alpine vascular flora of three cirque basins in the San Juan Mountains, Colorado 253 Hartman, Emily L. (see also Price, Robert A. et al.) Hartman, Ronald L., B. E. Nelson, and Keith H. Dueholm, Noteworthy collections of Aristida oligantha, Boisduvalia glabella, Celt is occidentalis var. occidentalis, Claytonia lanceolata var. flava, Eragrostis trichodes var. tricodes. Euphorbia serpens, Larix occidentalis, Leptodactylon watsonii, Loeflingia squarrosa subsp. texana, Linaria canadensis var. texana, Mon- ardella odoratissima subsp. glauca, Opuntia macrorhiza var. macrorhiza, Pedis angustifolia var. angustifolia, Physalis hederaefolia var. comota, Potentilla hookeriana, Rorippa truncata, Scirpus heterochaetus, Thellun- giella salsuginea, and Veratrum tenuipetalum 125 Heckard, Lawrence R. and James R. Shevock, Mimulus norrisii (Scrophu- lariaceae), a new species from the southern Sierra Nevada 179 iv Henrickson, James, Review of Intermountain flora: vascular plants of the intermountain west, U.S.A. Volume 4 (by A. H. Holmgren, N. H. Holm- gren, J. L. Reveal, and P. K. Holmgren) 58 Henrickson, James, Notes on the genus Burroughsia (Verbenaceae) 186 HuFT, Michael J. and Henk van der Werff, Observations on Chamaesyce (Euphorbiaceae) in the Galapagos Islands 143 JoKERST, James D., Noteworthy collection of Juncus cyperoides 191 JossELYN, Michael (see Spicher, Douglas) JoYAL, Elaine, Noteworthy collection of Amsinckia carinata 124 Kagan, Jimmy (see But, Paul P. H.) Keil, David J., Robert L. Allen, Joy H. Nishida, and Eric A. Wise, Addenda to the vascular flora of San Luis Obispo County, California 214 Kelley, Walter and Dieter H. Wilken, Noteworthy collection of Phacelia constancei 57 KouTNiK, Daryl L., New combinations in California Chamaesyce (Euphorbiaceae) 1 8 7 Kummerow, Jocken, Barbara A. Ellis, and James N. Mills, Post-fire seedling establishment of Adenostoma fasciculatum and Ceanothus greggii in southern California chaparral 148 Lavin, Matt, The identity of Cracca Bentham (Fabaceae, Robinieae) in the United States 95 Levin, Geoffrey A., Noteworthy collection of Grayia brandegei 192 Major, Jack, Review of California riparian systems: ecology, conservation, and productive management (edited by Richard E. Warner and Kathleen M. Hendrix) 277 Major, Jack, Review of Vascular plants of Montana (by Robert D. Dom) 193 Marrs-Smith, Gayle, Noteworthy collections of Corchorus hirtus and Iber- villea tenuisecta 191 McArthur, E. Durant and Stewart C. Sanderson, A cytotaxonomic con- tribution to the western North American rosaceous flora 24 McIntosh, Laird, Noteworthy collection of Rhynchelytrum repens 1 92 Mills, James N. (see Kummerow, Jocken et al.) Morefield, James D., Noteworthy collections of Dedeckera eurekensis and Opuntia pulchella 1 23 Morefield, James D. (see C. G. Schaack) Neese, Elizabeth (see Goodrich, Sherel) Neisess, Kurt, Notes on the Salvia leucophylla complex (Lamiaceae) of Cal- ifornia and Baja California Norte 273 Nelson, T. W. and R. Patterson, A new subspecies of perennial Linanthus (Polemoniaceae) from the Klamath Mountains, California 102 Nishida, Joy H. (see Keil, David J. et al.) Odion, Dennis C, Noteworthy collection of Dicentra pauciflora 57 Pack, Steven R. (see Vickery, Robert K., Jr.) Patterson, R. (see Nelson, T. W.) Pavlik, Bruce M., Sand dune flora of the Great Basin and Mojave Deserts of Cahfomia, Nevada, and Oregon 197 Pellmyr, Olle, Yellow jackets disperse Vancouveria seeds (Berberidaceae) 56 Perry, J. P., Jr., Noteworthy collection of Pinus patula var. longepedun- culata 192 Phillips, Dennis R. (see Vickery, R. K., Jr.) Price, Robert A., Noteworthy collection of Draba spectabilis var. oxyloba 192 Price, Robert A., Mary Lou Rottman, and Emily Hartman, Noteworthy collection of Draba apiculata 1 9 1 Reveal, Jack L. and James L. Reveal, The missing Fremont Cannon— an ecological solution? 106 V Reveal, James L. (see Reveal, Jack L.) RoMiNGER, J. M. (see Schaack, C. G.) RoTTMAN, Mary Lou (see Hartman, Emily L.) RoTTMAN, Mary Lou (see Price, Robert A. et al.) Richards, Daniel V., Noteworthy collections of Spirodela punctata and Wolf- fia borealis 57 Sanderson, Stewart C. (see McArthur, E. Durant) Schaack, C. G. and J. D. Morefield, Noteworthy collections of Antennaria microphylla, Aquilegia chrysantha, Carex haydeniana, Pinus jlexilis, and Potentilla nivea 121 Schaack, C. G., J. D. Morefield, J. M. Rominger, and J. Anderson, Note- worthy collection of Cowania subintegra 122 Schoolcraft, Gary and Arnold Tiehm, Noteworthy collections of Dalea ornata, Ivesia baileyi, Cryptantha celosioides, Eriogonum crosbyae, E. pro- ciduum, Ivesia rhypara, Mentzelia packardiae, Phacelia thermalis, and Scutellaria holmgreniorum 123 Shelly, J. Stephen (see But, Paul P. H.) Shevock, James R. (see Heckard, Lawrence R.) Spicher, Douglas and Michael Josselyn, Spartina (Gramineae) in northern California: distribution and taxonomic notes 158 Tiehm, Arnold (see Schoolcraft, Gary) Turner, Raymond M. (see Bowers, Janice E.) van der Werff, Henk (see Huft, Michael J.) Vasek, Frank C, Southern California white fir (Pinaceae) 65 Vasey, Michael C, The specific status of Lasthenia maritima (Asteraceae), an endemic of seabird-breeding habitats 131 VicKERY, Robert K., Jr., Steven A. Werner, Dennis R. Phillips, and Steven R. Pack, Chromosome counts in Section Simiolus of the genus Mimulus (Scrophulariaceae). X. The M. glabratus complex 9 1 Wells, Philip V. and Deborah Woodcock, Full-glacial vegetation of Death Valley, California: juniper woodland opening to Yucca semidesert 1 1 Werner, Steven A. (see Vickery, Robert K., Jr.) Wilken, Dieter H. (see Kelley, Walter) Wise, Eric A. (see Keil, David J. et al.) Woodcock, Deborah (see Wells, Philip V.) Worthington, Richard D. (see Freeman, C. Edward) vi Subscriptions— Membership Membership in the California Botanical Society is open to individuals ($ 1 8 per year; students $ 1 0 per year for a maximum of seven years). Members of the Society receive Madrono free. Family memberships ($20) include one ten-page publishing allot- ment and one journal. Emeritus rates are available from the Corresponding Secretary. Institutional subscriptions to Madrono are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Madrono should be sent to the Corresponding Secretary. Information for Contributors Manuscripts submitted for publication in Madrono should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For ar- ticles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed ending with figures, and number each page. Do not use a separate cover page, "erasable" paper, or footnotes. Manuscripts prepared on dot matrix printers may not be considered. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including cap- tions) to the Madrono page, and designed for reduction to % original size. Scales should be included in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff card- board and sent flat. No illustrations larger than 22 x 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in Madrono 22:288. 1974. In- stitutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 7th edition. Abbreviations of serial titles should be those in Botanico-Periodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted ten pages in Madrono each year or 20 pages over a two-year period. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $60.00 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing Madrono on a reasonable schedule, with equity among all members for access to its pages. Printer's fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author's changes after typesetting @ $3.00 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. CALIFORNIA BOTANICAL SOCIETY ^ IE 33, NUMBER 1 JANUARY 1986 IB MADRONO A WEST AMERICAN JOURNAL OF BOTANY Contents Montane Meadow Plant Associations of Sequoia National Park, California Charles B. Halpern Vegetation of Torrey Lake Mire, Central Cascade Range, Oregon Robert E. Frenkel, William H. Moir, and John A. Christy Grasslands as Compared to Adjacent Quercus garryana Woodland Understories Exposed to Different Grazing Regimes Loretta Saenz and /. O. Sawyer, Jr. Function of Olfactory and Visual Stimuli in Pollination of Lysichiton americanum (Araceae) by a Staphylinid Beetle Olle Pellmyr and Joseph M. Patt An Apparent Case of Introgression between Pinyon Pines of the New York Mountains, Eastern Mojave Desert James des Lauriers and Mark Ikeda Systematics of Collinsia parviflora and C. grandiflora (Scrophulariaceae) Fred R. Ganders and Gerda R. Krause A New Species of Lomatium (Apiaceae) from Southwestern Oregon James S. Kagan 24 40 47 55 63 71 NOTES Spineless Petioles in Washingtonia filifera (Arecaceae) James W. Cornett NOTEWORTHY COLLECTIONS California Nevada Oregon Washington REVIEWS EDITORIAL ANNOUNCEMENTS 76 79 79 80 80 81 83 46, 62, 70, 75, 84 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Madrono (ISSN 0024-9637) is published quarterly by the Cahfomia Botanical So- ciety, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to James R. Shevock, Botany Dept., California Academy of Science, San Francisco, CA 94 11 8. Editor— Wayne R. Ferren, Jr. Associate Editor— J. Robert Haller Department of Biological Sciences University of California Santa Barbara, CA 93106 Board of Editors Class of: 1986— Amy Jean Gilmartin, Washington State University, Pullman Robert A. Schlising, California State University, Chico 1987— J. RzEDOWSKi, Instituto de Ecologia, A.C., Mexico Dorothy Douglas, Boise State University, Boise, ID 1988— Susan G. Conard, USDA Forest Service, Riverside, CA William B. Critchfield, USDA Forest Service, Berkeley, CA 1989 — Frank Vasek, University of California, Riverside Barbara Ertter, University of California, Berkeley 1990— Barry D. Tanowitz, University of California, Santa Barbara Thomas R. Van Devender, Arizona-Sonora Desert Museum, Tucson CALIFORNIA BOTANICAL SOCIETY, INC. Officers for 1985-86 President: Charles F. Quibell, Department of Biology, Sonoma State University, Rohnert Park, CA 94928 Eirst Vice President: Rodney G. Myatt, Department of Biological Sciences, San Jose State University, San Jose, CA 95 114 Second Vice President: David H. Wagner, Herbarium, Department of Biology, University of Oregon, Eugene, OR 97403 Recording Secretary: V. Thomas Parker, Department of Biological Sciences, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: James R. Shevock, Department of Botany, California Academy of Sciences, San Francisco, CA 94 1 1 8 Treasurer: Cherie L. R. Wetzel, Department of Biology, City College of San Fran- cisco, 50 Phelan Ave., San Francisco, CA 941 12 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, Rolf W. Benseler, Department of Biological Sciences, California State University, Hayward, CA 94542; the Editor of Madrono; three elected Council Members: James C. Hickman, Department of Botany, Uni- versity of California, Berkeley, CA 94720; Thomas Fuller, 171 Westcott Way, Sacramento, CA 95864; Annetta Carter, Department of Botany, University of California, Berkeley, CA 94720; and a Graduate Student Representative, Joseph M. DiToMASO, Department of Botany, University of California, Davis, CA 95616. MONTANE MEADOW PLANT ASSOCIATIONS OF SEQUOIA NATIONAL PARK, CALIFORNIA Charles B. Halpern Department of Botany and Plant Pathology, Oregon State University, Corvallis 97331 Abstract Twelve plant associations are recognized and described for montane meadows of Sequoia National Park based on 8 1 releves. Three major groups are defined by growth- form dominants: mixed forb and grass associations, Carex and Scirpus associations, and Eleocharis associations. Major environmental factors influencing vegetation dis- tribution include: 1) a complex moisture gradient incorporating water depth and movement, and 2) site exposure and shading. Monitoring of water wells indicates that seasonal fluctuations of the water table are important in structuring the vegetation. Montane meadows, a common feature of Sequoia National Park in the southern Sierra Nevada of California, punctuate a landscape dominated by mixed conifer forest. Scenic vistas, a rich and colorful flora, proximity to Giant Sequoia groves, and accessibility result in disproportionate public visitation to these sites. Nevertheless, mon- tane meadows in the Park and in the southern Sierra Nevada in general have been poorly described. Studies of montane meadows and their environmental controls were initiated as part of a comprehensive study of riparian ecosys- tems and the interactions between terrestrial (forest and meadow) and stream systems in Sequoia National Park. As hydric sites, these montane meadows may be grouped functionally with forest riparian systems. Stream channels, overland flows, and pooled water are common. Plant community physiognomy, composition, and distri- bution reflect strong seasonal and spatial hydrologic patterns. Disturbance history and microenvironmental characteristics also influence vegetation composition and structure. It is difficult to as- sess the degree to which montane meadows in Sequoia National Park are recovering from a history of human and livestock use. Although disturbance currently appears minimal, present-day mead- ow vegetation may reflect the burning activities of aboriginal man as well as the widespread grazing of sheep and cattle during the late 1800s and early 1900s. The sites studied, however, do not exhibit the characteristics of habitat deterioration (trampling, surface ero- sion, hummock formation, gullying, and obvious reduction in vege- tation cover) reported from many subalpine meadows in the south- em Sierra Nevada (Armstrong 1 942, Sumner 1 94 1 , 1 948, Sharsmith 1959, Hubbard et al. 1965, 1966, Harkin and Schultz 1967, Leonard MADRofvfo, Vol. 33, No. 1, pp. 1-23, 27 March 1986 2 MADRONO [Vol. 33 et al. 1967, 1968, Giffen et al. 1969). Whereas these studies are qualitative, subsequent studies by Bennett (1965) and Strand (1972) provide a more quantitative basis for the evaluation of disturbance and subsequent plant succession. DeBenedetti and Parsons (1979a) reviewed the history of human and domestic livestock use of mead- ows in the southern Sierra Nevada, providing examples of subse- quent resource problems and evaluating the effectiveness of man- agement actions. Natural disturbance in the form of lightning fire may play an infrequent yet important role in subalpine meadows of the southern Sierra Nevada, particularly along the forest-meadow ecotone (DeBenedetti and Parsons 1979b, 1984). Natural fire in montane meadows of Sequoia National Park has not been reported in the literature and its historical role is unknown. The focus of this paper is the composition and distribution of montane meadow plant communities and their relationship to major environmental features in Sequoia National Park. It provides basic information for managers as well as a baseline for future research. The classification presented complements studies of subalpine meadows in the southern Sierra Nevada and in Sequoia National Park in particular (Sumner 1941, Sharsmith 1959, Bennett 1965, Harkin and Schultz 1967, Strand 1972, Ratlifri979, 1982, Benedict and Major 1982, Benedict 1981, 1983). Study Area Meadows examined were located in the mixed conifer forest zone of Sequoia National Park (Rundel et al. 1977) between 1493 and 2390 m elevation (Fig. 1). Sample plots were concentrated in the Giant Forest area and included Log, Crescent, Circle, Huckleberry, and Round Meadows and Vasey's Paradise; Long, Cahoon, Cabin, and Halstead Meadows were sampled outside the Sequoiadendron groves. Ten unnamed meadows were sampled and two additional sites were included from Kings Canyon National Park. Forest composition surrounding meadow sites varies. Pinus pon- derosa, P. lambertiana, Abies concolor, A. magniflca var. shastensis, Calocedrus decurrens, and Sequoiadendron giganteum are the most common tree species within the closed canopy forests. Understory dominants include Chrysolepis sempervirens, Ceanothus cordulatus, and Pteridium aquilinum. A variety of herbs comprise only minimal cover in the ground layer. Forest-meadow ecotones are abrupt both in vegetation and environment; tree encroachment is minimal. Long-term climatic records are available for the Giant Forest (elevation 1966 m) (Parsons and DeBenedetti 1979). The regional climate is Mediterranean with warm, relatively dry summers and cool wet winters. Hydric montane meadows, however, are less in- 1986] HALPERN: MONTANE MEADOWS 3 Fig. 1 . Location of the study area in Sequoia National Park, California. fluenced by regional climate than are surrounding forests, as they receive surface as well as sub- surface water throughout the growing season. Although average annual precipitation is 113 cm, June through September averages less than 3 cm (Rundel 1972); most precipitation occurs from December to March as snow. Mean annual snowfall at the Giant Forest exceeds 500 cm and depths of greater than 2 m are common in mid- winter. The average date when moun- tain basins are free of snow is May 20 (Wood 1975). Average min- imum temperatures range from — 6.7°C in February to 11.8°C in August. Average maximum temperatures range from 3.4°C in De- cember and January to 27.4°C in August (Parsons and DeBenedetti 1979). Methods Vegetation sampling. Field sampling was conducted during Sep- tember 1982 using a modification of the reconnaissance method of 4 MADRONO [Vol. 33 Franklin et al. (1970). A total of 81 plots in 20 meadows was sam- pled. Each plot was located subjectively in an area of visually ho- mogeneous vegetation and habitat. Although the shape varied to accommodate vegetation patterns, plots were most often circular and located within larger areas of similar vegetation to minimize edge effects. Sample plot areas were 250-500 m^; homogeneous units smaller than this were not sampled. Areas of recent natural or man- caused disturbance as well as areas that lacked visually uniform topographic or hydrologic features also were avoided. For each plot visual estimates of projected crown cover were recorded for each vascular plant species. Cover estimates also were made of the various substrate types (bedrock, loose rock, mineral soil, coarse and fine litter, and moss). Environmental features such as elevation, slope, aspect, landform, topography, and hydrologic characteristics also were recorded. Field notes included descriptions of the following: 1) sample plot species composition and physiognomy, 2) hydrologic regime, 3) neighboring vegetation, 4) surrounding forest vegetation, and 5) apparent forest-meadow ecotone changes (seedling and sap- ling encroachment, meadow expansion, or forest to meadow tree- fall). Voucher specimens of unidentified species were collected for identification and incorporation into the Oregon State University Herbarium (OSC). Nomenclature of vascular plants follows Munz (1959, 1968). Nomenclature of mosses follows Lawton (1971). Vegetation analysis. Vegetation data were analyzed using two complementary approaches: cluster analysis and ordination analysis. Cluster analysis utilized indicator species analysis (Hill et al. 1975) using the computer program TWINSPAN (Hill 1979a) and manual table sorting techniques (Mueller-Dombois and Ellenberg 1974, Westhoff" and van der Maarel 1978). Ordination analysis utilized correspondence analysis (Hill 1973, 1974) as implemented by the program DECORANA (Hill 1979b, Hill and Gauch 1980). Both TWINSPAN and DECORANA are part of the Cornell Ecology Pro- gram Series; other programs were developed at Oregon State Uni- versity (B. G. Smith, unpublished programs). TWINSPAN is a hierarchical, polythetic, divisive classification technique that uses reciprocal averaging (RA) to produce a classi- fication of samples and species based on differential species. DE- CORANA is an eigenvector ordination technique derived from re- ciprocal averaging that attempts to correct two problems of RA— an arch distortion effect and a compression of the axis ends relative to the axis middle (Gauch 1982). An octave transformation of species cover values was performed to compress the range of abundance. The octave scale is logarithmic (base 2) and the transformation prevents the few very abundant species from dominating the anal- ysis. Montane meadow plant associations were delineated based upon i 1986] HALPERN: MONTANE MEADOWS 5 the correspondence of TWINSPAN clusters with manual table sort- ing results. Ten of the initial 81 samples were ecotonal or outlier stands and could not be assigned successfully to a recognizable as- sociation. Because only one sample was available for each, desig- nation at the association level was not justified. Subsequently, the ecotonal and outlier samples were excluded from DECORANA or- dination analysis. Associations were plotted on ordination axes and a final classification was developed based upon subjective consid- eration of group homogeneity with field observations. Water table sampling. To assess seasonal water table fluctuations in a variety of vegetation types, 1 6 permanent perforated PVC pipe water wells (15 cm diameter) were established along a transect line perpendicular to the long axes of Log and Crescent Meadows, Giant Forest. Wells were placed subjectively in homogeneous vegetation representing selected plant associations. A meter stick was lowered to the water surface to establish depth from ground level. Biweekly measurements of water table depth were taken from 6 July through 8 November 1983. Results and Discussion Floristic Analysis A total of 1 1 6 vascular plant species and 6 bryophyte genera were identified within the montane meadow sample plots of Sequoia Na- tional Park. The vascular flora included 38 families and 77 genera. The 10 families with the greatest number of species are presented in Table 1. The Gramineae had the largest number of genera (14) and species (18). The Cyperaceae was represented by 3 genera and 1 7 species, and the Compositae by 9 genera and 1 2 species. Canopy cover of the Cyperaceae, however, dominates these meadows due to the prominence of Carex, Scirpus, and Eleocharis species. Species with the greatest frequency of occurrence in the samples (constancy) are listed in Table 2. Oxypolis Occident alis (Umbelliferae) is nearly ubiquitous, with 78% constancy and 25% characteristic cover (av- erage cover for only those plots in which the species occurs) (Pak- arinen 1984). Other important species include Scirpus microcarpus, Glyceria elata, Eleocharis montevidensis, and Carex rostrata, with constancies of 47 to 60% and characteristic covers of 13 to 20%. Species such as Athyrium filix-femina, Carex amplifolia, and Vac- cinium occidentale are relatively uncommon, but are diagnostic of particular plant associations, and often dominate cover therein. Vegetation Analysis Twelve plant associations and one phase are recognized from the montane meadows of Sequoia National Park. These are grouped 6 MADRONO [Vol. 33 Table 1 . Ten Vascular Plant Families with the Greatest Number of Species. Family Genera Species Gramineae 14 18 Cyperaceae 3 17 Compositae 9 12 Scrophulariaceae 4 6 Juncaceae 2 5 Liliaceae 4 4 Salicaceae 1 4 Orchidaceae 3 3 Umbelliferae 3 3 Polygonaceae 2 3 into three broad types based on growth-form dominants (Table 3): mixed forb and grass associations, Carex and Scirpus associations, and Eleocharis associations. The association concept used herein refers to a recurring assemblage of plant species with visually ho- mogeneous composition and physiognomy representing a modal position in the pattern of vegetation, and, possibly, environment. Association names reflect the diagnostic and often dominant species. Phase names represent recognizable variation in an association at- tributed to the presence of one or more species. Species constancy and characteristic cover are compared between plant associations in Tables 4-6. Only species exceeding 49% con- stancy in at least one association have been included for ease in interpretation. Stand tables containing constancy and characteristic cover statistics for all sample plots within an association are available from the author. Within the following descriptions of associations, "channeled flows" refers to perennial stream courses, "overland flows" refers to unrestricted, generally seasonal runoff* across mead- ow surfaces, "pooled and standing water" refers to relatively still water above the soil surface, and "stagnant water" refers to water not subject to movement at or above the soil surface. A. Mixed forb and grass types. Six plant associations comprise the Mixed Forb and Grass Types. A mixture of herbaceous peren- nials or grass species, or both, dominate these sites, although Scirpus microcarpus is occasionally abundant (Table 4). Typically, the Mixed Forb and Grass Types occur in the drier portions of montane mead- ows. 1 . Glyceria elata-Lotus oblongifolius Association. This is an herb- rich association with a mosaic appearance. Local dominance of in- dividual species within the mosaic is not accompanied by observable differences in microenvironment; the patterning is likely the result 1986] HALPERN: MONTANE MEADOWS 7 Table 2. Twenty Most Common Montane Meadow Species, Ranked by Constancy. 'Growth-form key: K = herb, G = grass, and S = sedge or rush. 2 Characteristic cover represents the average cover for only those samples in which the species occurs. Growth- Characteristic Species form' Constancy (%) cover (%y Oxypolis occidentalis H 77.8 25.2 Glyceria elata G 60.5 7.3 Scirpus microcarpus S 55.6 20.0 Lotus oblongifolius H 55.6 3.5 Eleocharis montevidensis S 53.1 16.5 Veratrum califomicum H 50.6 21.7 Carex rostrata S 46.9 13.7 Dodecatheon jeffreyi H 45.7 5.2 Epilobium exaltatum H 44.4 0.5 Stachys albens H 44.4 4.7 Polygonum bistort oides H 42.0 1.3 Carex nebrascensis S 38.3 7.9 Juncus oxymeris S 34.6 2.9 Senecio triangularis H 32.1 3.3 Habenaria dilatata H 32.1 0.2 Deschampsia caespitosa G 29.6 2.6 Perideridia parishii H 29.6 1.7 Cinna latifolia G 29.6 0.8 Agrostis scabra G 27.2 1.9 Castilleja miniata H 27.2 0.7 Table 3. Montane Meadow Plant Associations of Sequoia National Park. Association acronyms are indicated in parentheses. Mixed Forb and Grass Types 1 . Glyceria elata-Lotus oblongifolius Association (GLEl^LOOB) 2. Elymus glaucus-Heracleum lanatum Association (ELGl^HELA) 3. Agrostis scabra Association (AGSC) 4. Glyceria elata-Scirpus microcarpus Association (GLEl^SCMI) 5. Calamagrostis canadensis-Scirpus microcarpus Association (CACA-SCMI) 6. Athyrium filix-femina Association (ATFI) Carex and Scirpus Types 7. Scirpus microcarpus-Oxypolis occidentalis Association (SCMI-OXOC) 8. Carex amplifolia-Oxypolis occidentalis Association (CAAM-OXOC) 9. Carex nebrascensis-Oxypolis occidentalis Association (CANE-OXOC) 10. Carex rostrata Association (CAR02) Eleocharis Types 1 la. Eleocharis montevidensis-Oxypolis occidentalis Association, Eleocharis montevidensis Phase (ELMO-OXOC-ELMO Phase) 1 lb. Eleocharis montevidensis-Oxypolis occidentalis Association, Carex rostrata Phase (ELM0-0X0C-CAR02 Phase) 1 2. Eleocharis montevidensis-Moss Association (ELMO-MOSS) MADRONO [Vol. 33 E < u P-l O u o o c YX ^ ^ C £ CO A)inO|0>/^'/^0>/^i I I I I - <^ I CM .r^ I H I 5^ "3 C ^ Q S s: ^ ^ o S.I 1^ -si s I 5 ^ >; -C) S (3 -0,3^:5 CI, Qq Co Co O ^ Co Co Co ^ ^ o ti; I I I I I I o o TfO'*rs|(NO"<^Ooo^ "Tl- I i(N^ o ^ I o I o o o I ^ oo I H ■y^ O (N O ^ I CnI (N I .to ^ to o 2= ^ 2 a ^3 U Co U 10 MADRONO [Vol. 33 of initial establishment and vegetative spread of rhizomatous pe- rennial forbs. Species of greatest constancy and cover include Lotus oblongifolius, Senecio clarkianus, Castilleja miniata, Solidago can- adensis, and Glyceria elata. Species of Cyperaceae occur occasion- ally, but are more common on wetter sites. This association is com- mon along meadow edges and on elevated flats in the driest portions of montane meadows. 2. Elymus glaucus-Heracleum lanatum Association. This asso- ciation is characterized by a dominance of Elymus glaucus and a host of perennial forbs, including Heracleum lanatum, Solidago can- adensis, Stachys albens, and Senecio triangularis. Floristic compo- sition is similar to the Glyceria elata-Lotus oblongifolius Associa- tion, but dominance has shifted to grasses. This association occurs along montane meadow edges and on elevated flats where the water table falls well below the soil surface throughout the growing season. 3. Agrostis scabra Association. This association is limited to small areas in several montane meadows. It resembles the two mixed forb and grass associations described previously, but also has abundant Agrostis scabra. Phleum pratense is a common grass associate, whereas Stachys albens, Solidago canadensis, and Lotus oblongifo- lius are common herbs. This association is restricted to drier, convex landforms that have a relatively deep water table. The abundance of Agrostis scabra and Phleum pratense suggest previous disturbance on these sites. 4. Glyceria elata-Scirpus microcarpus Association. Glyceria elata reaches its peak abundance within this association, and Scirpus mi- crocarpus is a codominant. Stachys albens is a consistent associate and Solidago canadensis and Veratrum californicum are locally common. These sites typically exhibit channeled or overland water flows. 5. Calamagrostis canadensis-Scirpus microcarpus Association. Calamagrostis canadensis and Scirpus microcarpus exceed 100% total canopy cover in this species-poor association. Glyceria elata is the only other species with 100% constancy. This association commonly occurs near meadow edges on moist to saturated sites adjacent to channeled or overland flows. 6. Athyrium filix-femina Association. A nearly complete upper canopy of the fern Athyrium filix-femina characterizes this associ- ation. Senecio triangularis and Oxypolis occidentalis are common, as are the grasses Glyceria elata and Cinna latifolia. This association is restricted to narrow, cool, and shaded meadows that are commonly the smaller basin swales in which forest canopy shading is important. It also can occur in modified form as streamside vegetation where forest canopies are relatively dense. These sites typically have sat- urated soils, and seeps are common. 1986] HALPERN: MONTANE MEADOWS 11 B. Carex AND SciRPUS TYPES. Four plant associations comprise the Carex and Scirpus Types. Sites are typically dominated by Scir- pus microcarpus, or coarse-leaved species of Carex, or both (Table 5). Species richness is generally low. These associations occupy areas with pooled, channeled, or overland flows. 7. Scirpus microcarpus-Oxypolis occidentalis Association. Height and density of the vegetation suggest that this type has the highest standing crop of any montane meadow type. Scirpus microcarpus reaches greatest abundance in this association; Oxypolis occidentalis is an important codominant; and Athyrium filix-femina, Stachys albens, Mimulus guttatus, and Equisetum arvense are frequent as- sociates. This vegetation type is widespread, usually occurring along stream channels and within areas of overland flow where the water table remains at or above the soil surface throughout the growing season. 8. Carex amplifolia-Oxypolis occidentalis Association. This as- sociation is restricted to small, shaded, swale meadows similar to those supporting the Athyrium filix-femina Association. Carex am- plifolia dominates these sites and Oxypolis occidentalis is of sec- ondary importance. Common associates include Glyceria elata, Cin- na latifolia, Athyrium filix-femina, Mimulus guttatus, and Cardamine breweri. Soils typically are saturated throughout the growing season. 9. Carex nebrascensis-Oxypolis occidentalis Association. Carex nebrascensis reaches peak abundance in this association, whereas Oxypolis occidentalis, Carex rostrata, and Eleocharis montevidensis are codominant species. This vegetation type is related composi- tionally and structurally to the Eleocharis montevidensis-Oxypolis occidentalis type. Habitats typically are flat to gently sloping; pooled to slightly-flowing water is present throughout the growing season. 10. Carex rostrata Association. This association occurs under a variety of topographic and hydrologic regimes. Under deeply-pooled water, pure stands of Carex rostrata develop. On sloping sites or under conditions of decreased water tables, C rostrata abundance and vigor decrease and species diversity increases. The common associated herbs include Dodecatheon jeffreyi, Polygonum bistor- toides, Oxypolis occidentalis, and Perideridia parishii. Moss can be locally abundant. C. Eleocharis types. Two plant associations and one phase com- prise the Eleocharis Types. Typically, sites are dominated by the fine-stemmed spike-rush, Eleocharis montevidensis (Table 6). Standing or nearly stagnant water at or above the soil surface is characteristic of these communities. 1 la. Eleocharis montevidensis-Oxypolis occidentalis Association, Eleocharis montevidensis Phase. This association resembles the Car- ex nebrascensis-Oxypolis occidentalis Association; however, Carex 12 MADRONO [Vol. 33 X5 H ^ i uj o s a < .s 2 O :S z C i u ° O -Ct c/) CO O ^ a: II z > > O UJ ^ - Si I 2 z Z < o ^ (J ^ 3 O -3 (U CX E Z S I I I I I I "^H I I I I I I o o o o o I I I I fN fN C3 - S S s; V S Z S C 00 14 MADRONO [Vol. 33 Table 6. Constancy and Average Cover Synthesis Table for the Associa- tions OF THE Eleocharis Types. 'See Table 3 for association acronyms. -CON = constancy (%). ^COV = average cover (%) based only on those samples in which species occurs. Plant association': ELMO-OXOC ELMO-OXOC ELMO-MOSS CAR02 PHASE ELMO PHASE Number of plots per type: 2 5 13 Mean number of species per plot (s.d.): 16.0(11.3) 16.6 (6.8) 19.1 (4.8) CON^ COV^ CON cov CON COV Shrub species Vcicciniufyi occidcntcilc 54 15 Herb species Habenaria di I at at a 1 1 1 s T 1 l eratrum californicum 50 1 60 1 15 T I^UIU^ (JUlUll^lJUllU^ 100 13 60 5 69 3 i^.KypUllo ULLlU.crllU.llJ 100 41 100 62 100 17 Dodecatheon jeffreyi 100 1 5 60 7 100 1 5 Camassia leichtlinii 50 3 60 T 69 Perideridia parishii 50 T 40 3 92 8 Polygonum bistortoides - - 60 T 69 3 Hypericum anagalloides — — 20 7 54 9 Spiranthes romanzoffiana 20 T 62 1 Mimulus primuloides 69 21 Aster alpigenus 85 5 Grass species Glyceria elata 50 1 60 2 8 T Deschampsia caespitosa 50 1 62 8 ivi unienuergici jiiijurmis 69 8 Sedge and rush species Scirpus microcarpus 100 8 60 19 15 23 Carex rostrata 100 45 60 2 69 15 Juncus oxymeris 100 5 80 2 77 17 Eleocharis montevidensis 100 63 100 66 100 46 Carex nebrascensis 60 5 62 15 Carex ormantha 69 9 Bryophyte species Sphagnum/ Philonotis/ Aulacomnium 50 4 60 13 92 64 is less important and Eleocharis assumes dominance. Although species richness can be high, few species have constancies greater than 60%. The physiognomy is two-layered, having a tall Oxypolis overstory and an open Eleocharis understory. Water remains at or slightly above the soil surface throughout the growing season. 1 lb. Eleocharis montevidensis-Oxypolis occidentalis Association, 1986] HALPERN: MONTANE MEADOWS 15 Carex rostrata Phase. This uncommon phase occurs on habitats with shghtly higher standing or flowing water regimes than the typical community. The physiognomy is similarly two-layered, but the understory is denser due to the abundance of Carex rostrata. Ele- ocharis montevidensis and Oxypolis occidentalis are dominant, and Dodecatheon jeffreyi and Lotus oblongifolius are common herbs. 12. Eleocharis montevidensis-Moss Association. This association is characterized by 1) a moss mat composed primarily of Sphagnum, Philonotis, and Aulacomnium, occurring singly or in combination; 2) an abundance of Eleocharis; and 3) a characteristic mosaic of mat-forming vascular species such as Aster alpi genus, Hypericum anagalloides, Mimulus primuloides, and Muhlenbergia filiformis. Juncus oxymeris and Perideridia parishii are taller diagnostic as- sociates. The average cover of moss is 60%. Standing to stagnant surface water typifies level sites whereas surface seeps typify sloping sites. Relation to Other Sierra Nevada and Cascade Meadows Several meadow associations of Sequoia National Park are similar structurally and, in certain instances, floristically to montane mead- ows found elsewhere in the Sierra Nevada and in the Cascade Range of Oregon. The physiognomy and floristic character of the Eleocharis mon- tevidensis-Moss Association tie it to many montane mire systems throughout the Sierra Nevada and the Oregon Cascade Range. The complex of matted-boggy species with a taller, open Eleocharis layer is characteristic. Eleocharis pauciflora is the diagnostic counterpart in the montane zone of the Cascade Range and in the subalpine zone of the southern Sierra Nevada. Within the Sierra, Benedict (1981, 1983) describes an Eleocharis pauciflora Association and an Eleocharis pauciflora-Mimulus primuloides variant from the Rock Creek and Whitney Creek drainages of Sequoia National Park. Sim- ilarly, Ratliff" (1979, 1982) defines an Eleocharis pauciflora type (few- flowered spike-rush/Site Class H) within the subalpine zone of Yosemite, Sequoia, and Kings Canyon National Parks, and the Stan- islaus, Sierra, and Sequoia National Forests. In the Western Cascades of Oregon, Hickman (1976) alludes to a phase of his Bog Association that may have a similar assemblage of low-growing herbs. Halpem et al. (1984) describe similar vegetation, defined as the Eleocharis pauciflora community type, within the Three Sisters Wilderness Area, Oregon. An Eleocharis pauciflora/hryophyiQ community at Sphagnum Bog, Crater Lake National Park, Oregon (very similar in composition and physiognomy to that in Sequoia National Park), is described by Seyer (1979). At Multorpor Fen, Mt. Hood National 16 MADRONO [Vol. 33 Forest, Oregon, Seyer (1983) also describes an Eleocharis /herbs/ Aulacomnium-Sphagnum community, which is a similar low stat- ure, moss-mat community with permanently saturated soils. Camp- bell (1973) describes an Eleocharis-Aulacomnium community at Hunts Cove, Mt. Jefferson, Oregon, within a larger Carex scopulo- rum meadow complex. The Carex nebrascensis-Oxypolis occidentalis and the Carex ros- trata Associations, typical of standing to slightly flowing water re- gimes in montane meadows of Sequoia National Park, have ana- logues elsewhere. Ratliff' (1979, 1982) describes a Nebraska sedge class (Site Class G) common on nearly level, imperfectly to mod- erately well-drained, subalpine sites in the southern Sierra Nevada. Carex nebrascensis- and Carex ro^/rara -dominated vegetation is described from Grass Lake, California, by Beguin and Major (1975). Benedict (1981, 1983) describes a subalpine Carex rostrata-Mimu- lus primuloides Association from the Whitney and Rock Creek drainages of Sequoia National Park. It appears similar to the herb- rich variant of the montane Carex rostrata association of the Park, occurring on sites with depressed water tables. Ratliff' (1979, 1982) describes a Carex rostrata type (beaked sedge/Site Class A) occu- pying poorly and imperfectly drained sites. Carex rostrata assem- blages are also important in many hydric montane meadows throughout the Oregon Cascade Range. Campbell (1973) describes a Carex rostrata-Sphagnum community at Hunts Cove, Mt. Jeffer- son, Oregon. Carex rostrata-dommdiXtd. reeds wamps at Sphagnum Bog, Crater Lake National Park, and Gold Lake Bog near Willamette Pass, Oregon, have been described by Seyer (1 979). A Carex rostrata community with C sitchensis has been reported for Big Springs near Nash Crater, Oregon (Roach 1958). Frenkel (pers. comm.) identifies similar reedswamp vegetation at Torrey Lake Mire, Oregon. Com- parable assemblages are scattered throughout the Three Sisters Wil- derness Area, Oregon, under standing to slightly flowing water con- ditions. Several associations of the Mixed Forb and Grass Types within the montane meadows of Sequoia National Park contain herb species common to meadows of the Sierra Nevada and Oregon Cascade Range. The particular compositions and physiognomies of these assemblages, however, may be specific to the Park. This uncertainty reflects the paucity of reports of similar associations in the montane and subalpine meadow literature. Similarly, basin swale commu- nities dominated by Athyrium filix-femina may represent a rather unique aspect of montane meadows in Sequoia National Park. Al- though the fern is common in coastal forested swamps in Oregon and Washington (Franklin and Dyrness 1973) and along mountain streams in the Cascade Range and Sierra Nevada, extensive meadow swards have not been described outside of Sequoia National Park. 1986] HALPERN: MONTANE MEADOWS 17 The prominence of Oxypolis occidentalis is perhaps the most unique floristic aspect of the montane meadows of Sequoia National Park. A tall, leafy umbel of marshy meadows and shallow water, Oxypolis ranges from Tulare Co. to Eldorado Co. in the Sierra Ne- vada, and north to Crater Lake in Oregon (Jepson 1936). It has been reported as a fairly common component of only two geographically limited hydric communities in the subalpine of the southern Sierra Nevada and at Crater Lake National Park (Ratliff 1979, Seyer 1979). In contrast, in montane meadows of Sequoia National Park, Oxypo- lis occurs in 11 of 1 2 associations and dominates in nearly half of these. In most cordilleran wet meadows, graminoids are the sole dominants, but in similar communities in Sequoia National Park, Oxypolis occidentalis plays a significant role as a codominant. Detrended Correspondence Analysis The results of the detrended correspondence analysis are useful in describing vegetation patterns and inferring complex environ- mental gradients (Fig. 2). The overlay of TWINSPAN groups on the ordination reveals the spatial relationship between associations within the two dimensional compositional space portrayed. Interpretation of these axes as environmental gradients is possible if we consider stand compositions, species autecology, and environmental infor- mation. DCA ordination yielded the four eigenvalues of 0.63, 0.31, 0.19, and 0.16, which suggest that only the first two axes are important. The first eigenvector, or axis of the ordination, is 4.6 standard de- viation units long, representing a moderate turnover in species com- position within samples along that gradient (Gauch 1982). Field observations suggest this axis represents a complex moisture gradient that incorporates water table depth and water movement. The driest representatives of this gradient (toward the left end of Axis 1) are the Elymus glaucus-Heracleum lanatum, Glyceria elata-Lotus ob- longifolius, and Agrostis scabra Associations. They typify sites with seasonal lowering of the sub-surface water table. Located more cen- trally along Axis 1 is vegetation that typifies channeled or flowing water sites— the Scirpus microcarpus-Oxypolis occidentalis Asso- ciation is a representative. To the right of these are associations that exhibit shallow to deep and standing to slightly flowing water throughout the growing season. Included are both phases of the Eleocharis montevidensis-Oxypolis occidentalis Association, the Carex rostrata Association, and the Carex nebrascensis-Oxypolis occidentalis Association. To the extreme right lies the Eleocharis montevidensis-Moss Association typical of sites with persistent, standing to stagnant, surface water. The second DCA axis is not as easily interpreted as the first. The 18 MADRONO [Vol. 33 DCA AXIS I Fig. 2. Detrended correspondence analysis ordination of samples. Letters indicate samples representing the same association, as defined in Table 3. axis is 2.8 standard deviation units long, representing a significantly smaller turnover in species composition than along Axis 1 . The forb- grass associations in the left portion of the ordination space have a better separation than the Carex and Eleocharis Associations to the right. Field observations suggest that at its ends, the axis seems to describe extremes in meadow exposure and shading, reflecting site location in one of two broad landform classes: 1) within small con- cave openings or swales in mixed conifer forest or in Sequoiadendron groves, where bordering trees provide much shade; or 2) in large, broad basins of greater area and minimal shading. Typical swale vegetation occurs highest on Axis 2 and is repre- sented by the Carex amplifolia-Oxypolis occidentalis and Athyrium filix-femina Associations. At the other extreme, open basin vege- tation (represented by the Agrostis scabra and Glyceria elata-Lotus oblongifolius Associations) occurs lowest on Axis 2. Those types with intermediate positions along Axis 2 (the Glyceria elata-Scirpus microcarpus and Scirpus microcarpus-Oxypolis occidentalis Asso- ciations) occur over a wider range of habitats. They are found within more shaded sites in swale, stringer, or streamside meadows, as well as within large, open meadow basins. The Calamagrostis canaden- sis-Scirpus microcarpus Association has affinities to the Glyceria elata-Lotus oblongifolius Association as it is restricted to open ba- sins. The former usually occurs farther from drier edges and closer to drainage channels and areas of overland flow than the latter. 1986] HALPERN: MONTANE MEADOWS 19 LOG MEADOW WELL NO. VEGETATION 1,4,5 Glyceria - Lotus 2 Scirpus 3 Agrostis 6 Carex rostrata - Oxy poli s 7 C. nebrascensis - Oxypolis 8 Sci rpus - Oxypoli s CRESCENT MEADOW WELL NO. VEGETATION 1,5,6,7 Glyceria - Lotus 2,4 Eleoctioris- Oxypolis 3 C. nebrascensis-Oxypolis 8 Scirpus - Oxypol is JULY AUG SEPT OCT NOV JULY AUG SEPT OCT NOV DATE DATE Fig. 3. Water table depths for Log and Crescent Meadows, Giant Forest, for 6 Jul to 8 Nov 1983. Water Table Dynamics Species distributions and vegetation composition largely reflect meadow hydrology. Detrended correspondence analysis ordination indicates that plant associations segregate along a complex moisture gradient reflecting water table depth and movement. Results from seasonal water table measurements reinforce this interpretation (Fig. 3). Although the water table patterns represent only two meadow sites for a single growing season during a year with an unusually high snow-pack and greater than average summer rains, the distri- ► bution of vegetation nevertheless reflects spatial differences in mead- ow hydrology. Several general trends are evident from the water well transects in Log and Crescent Meadows. Water was progressively drawn down from 6 July through 12 September in those communities that ex- perienced a fluctuating water regime (see Fig. 3, Glyceria elata-Lotus oblongifolius Association, Wells 1, 4, 5, Log Meadow; Glyceria elata- Lotus oblongifolius Association, Wells 5, 6, 7, Crescent Meadow; and Agrostis scabra Association, Well 3, Log Meadow). In several instances, the water table fell below well bottoms (see arrows. Fig. 20 MADRONO [Vol. 33 3). Small but distinct increases in the water table during the period of decline may correspond to summer rains (8-10 August, 1.5 cm; 15-19 August, 4.0 cm). From 12 September through final sampling on 8 November, the water table progressively increased to levels that were near initial sampling heights in most areas even though fall rains had been minimal (approximately 7.1 cm from 22 Sep- tember to 8 November). Apparently enough water remains within side-slopes to restore the water table to early summer levels when evapotranspiration is reduced during senescence of late summer vegetation. Wood (1975) described similar patterns in a subalpine meadow at the Central Sierra Snow Laboratory. Rates of decline and increase in the water table were variable across meadow transects; maximum rates of change appeared in the Agrostis scabra community of Log Meadow. Here, levels dropped an average of 0.21 cm per day through mid-September and rose an average of 0.25 cm per day through early November. Wood (1975) reported a stronger water table rise, 1 .2 cm per day, in his subalpine meadow. Water wells located more centrally in the meadows (in vegetation dominated by Scirpus micwcarpus-Oxypolis occidentalis, Carex rostrata-Oxypolis occidentalis, Carex nebrascensis-Oxypolis occi- dentalis, and Eleocharis montevidensis-Oxypolis occidentalis) showed minor fluctuations in growing season water depths. These were com- munities with water essentially at or above the soil surface through- out the sampling period. Eleocharis montevidensis-Oxypolis occi- dentalis sites (Wells 2, 4, Crescent Meadow) showed minor water table fluctuations of 1 to 2 cm. The Carex nebrascens is -dominsited sites (Well 7, Log Meadow, and Well 3, Crescent Meadow) main- tained water levels between -0.5 and +2.0 cm. Scirpus microcar- /?W5-dominated sites exhibited standing water tables as great as 9 cm, but levels slowly declined to 0.5 and 5.0 cm by September (Well 8, Log Meadow, and Well 8, Crescent Meadow, respectively). The Carex rostrata site maintained a stable, standing water table at 2.0- 3.0 cm (Well 6, Log Meadow). A Scirpus microcarpus site near a stream channel in Log Meadow (Well 2) appeared anomalous in that water table fluctuations resem- bled those more characteristic of Glyceria elata-Lotus oblongifolius meadow edge communities. Its occurrence on elevated coarse sand deposits may explain the rather large 1 7 cm depression of the water table from 6 July through 1 2 September. Patterns within a Glyceria elata-Lotus oblongifolius site in Crescent Meadow (Well 1) also appeared anomalous as the water table remained stable beneath the soil surface through the growing season. Unusually high winter snow- pack and August rainfall may be masking more distinct water table patterns in these meadows, particularly in communities with per- manently saturated soils. Typically, the water table in these sites 1986] HALPERN: MONTANE MEADOWS 21 may drop both differentially, and more quickly to lower mid-sum- mer depths. Nevertheless, seasonal measurements, field observa- tions, and ordination results suggest the importance of water depth and movement in montane meadow vegetation patterns. Conclusion The analysis of montane meadow vegetation in Sequoia National Park complements similar research within subalpine meadows of the southern Sierra Nevada and provides baseline data for future research and management. Twelve plant associations and one phase are segregated floristically, reflecting environmental variation in 1) water table depth and water movement, and 2) site exposure and shading. Vegetation similar floristically and physiognomically to that described herein exists elsewhere in the Sierra Nevada and in the western Cascade Range of Oregon. Although spatial and seasonal patterns of water depth and movement influence the composition and distribution of plant communities, future research is necessary to address the relative importance of microenvironmental param- eters and disturbance. Acknowledgments I thank Teresa Magee for assistance with vegetation sampling and Annie Esperanza, Patti Haggerty, Sara Molden, and Tom Stohlgren for help with water table measure- ments. Larry Norris aided in location of many of the sites and Bradley Smith provided much guidance with computer analyses. Joseph Antos, Jerry Franklin, Robert Frenkel, Arthur McKee, Annette Olson, and Dean Taylor provided useful criticism of the manuscript. Finally, I thank Jerry Franklin and David Parsons for the opportunity to pursue this research as part of the PULSE studies of 1982 and 1983. The study was supported in part by the National Park Service, under an Interagency Agreement Grant IA-9088-82-01 with the Pacific Northwest Forest and Range Experiment Sta- tion, USDA, Forest Service. Literature Cited Armstrong, J. E. 1 942. A study of grazing conditions in the Roaring River District, Kings Canyon National Park, with recommendations. National Park Service Report. Beguin, C. and J. Major. 1975. Contribution a I'etude phytosociologique et eco- logique des marais de la Sierra Nevada (Califomie). Phytocoenologia 2:349-367. Benedict, N. B. 1981. The vegetation and ecology of subalpine meadows of the southern Sierra Nevada, California. Ph.D. dissertation, Univ. California, Davis. . 1983. Plant associations of subalpine meadows. Sequoia National Park, California. Arctic and Alpine Research 15:383-396. and J. Major. 1982. A physiographic classification of subalpine meadows of the Sierra Nevada, California. Madroiio 29:1-12. Bennett, P. 1965. An investigation of grazing impact of 10 meadows in Sequoia- Kings Canyon National Park. M.A. thesis, San Jose State College, California. Campbell, A. G. 1973. Vegetative ecology of Hunts Cove, Mt. Jefferson, Oregon. M.S. thesis, Oregon State Univ., Corvallis. DeBenedetti, S. and D. J. Parsons. 1979a. Mountain meadow management and 22 MADRONO [Vol. 33 research in Sequoia and Kings Canyon National Parks: a review and update. In R. L. Linn, ed., Proceedings of the First Conference on Scientific Research in the National Parks. U.S. D.I. Nat. Park Serv. Trans, and Proc. Series No. 5. Washington, DC 2:1305-131 1. and . 1979b. Natural fire in subalpine meadows: a case description from the Sierra Nevada. J. Forest. (Washington) 77:477-479. and . 1984. Post-fire succession in a Sierran subalpine meadow. Amer. Midi. Naturalist 111:118-125. Franklin, J. F. and C. T. Dyrness. 1973. Natural vegetation of Oregon and Wash- ington. U.S.D.A. Forest Serv. Gen. Techn. Rept. PNW-8. , , and W. H. Moir. 1970. A reconnaissance method for forest site classification. Shinrin Richi (Tokyo) 12:1-14. Gauch, H. G. 1 982. Multivariate analysis in community ecology. Cambridge Univ. Press, Cambridge. Giffen, a., C. M. Johnson, and P. Zinke. 1969. Ecological study of meadows in Lower Rock Creek, Sequoia National Park (unpubl.). Halpern, C. B., B. G. Smith, and J. F. Franklin. 1984. Composition, structure, and distribution of the ecosystems of the Three Sisters Biosphere Reserve/Wil- derness Area. Report to the U.S.D.A. Harkin, D. W. and A. M. Schultz. 1967. Ecological study of meadows in Lower Rock Creek, Sequoia National Park. Progress report for 1966 (unpubl.). Hickman, J. C. 1976. Non-forest vegetation of the central western Cascade Moun- tains of Oregon. Northw. Sci. 50:145-155. Hill, M. O. 1973. Reciprocal averaging: an eigenvector method of ordination. J. Ecol. 61:237-249. . 1974. Correspondence analysis: a neglected multivariate method. Applied Statistics 23:340-354. . 1979a. TWINSPAN— a FORTRAN program for arranging multivariate data in an ordered two-way table by classification of the individuals and attri- butes. Ecology and Systematics, Cornell Univ., Ithaca, NY. . 1979b. DECORANA-a FORTRAN program for detrended correspon- dence analysis and reciprocal averaging. Ecology and Systematics, Cornell Univ., Ithaca, NY. , R. G. H. Bunce, and M. W. Shaw. 1975. Indicator species analysis, a divisive polythetic method of classification and its application to a survey of native pinewoods in Scotland. J. Ecol. 63:597-613. and H. G. Gauch. 1980. Detrended correspondence analysis: an improved ordination technique. Vegetatio 42:47-58. Hubbard, R. L., C. E. Conrad, A. W. Magill, and D. L. Neal. 1966. The Sequoia National Park and Pacific Southwest Forest Range Exp. Sta. cooperative moun- tain meadow study. Part II: An ecological study of Lower Funston Meadow (unpubl.). , R. RiEGELHUTH, H. R. SANDERSON, A. W. MaGILL, D. L. NeAL, R. H. TwISS, and C. E. Conrad. 1965. A cooperative study of mountain meadows. Part I: Extensive survey and recommendations for further research (unpubl.). Jepson, W. L. 1936. A flora of California. Vol. 2. California School Book Depository, San Francisco. Lawton, E. 1971. Moss flora of the Pacific Northwest. The Hattori Botanical Laboratory, Nichinan, Japan. Leonard, R., D. Harkin, and P. Zinke. 1967. Ecological study of meadows in Lower Rock Creek, Sequoia National Park. Progress report for 1967 (unpubl.). , C. M. Johnson, P. Zinke, and A. Schultz. 1968. Ecological study of meadows in Lower Rock Creek, Sequoia National Park. Progress report for 1968 (unpubl.). MuELLER-DoMBOis, D. and H. Ellenberg. 1974. Aims and methods of vegetation ecology. Wiley, NY. 1986] HALPERN: MONTANE MEADOWS 23 MuNZ, p. A. 1959. A California flora. Univ. California Press, Berkeley. . 1968. Supplement to a California flora. Univ. California Press, Berkeley. Pakarinen, p. 1984. Cover estimation and sampling of boreal vegetation in north- em Europe. In R. Knapp, ed., Handbook of vegetation sampling: methods and taxon analysis in vegetation science, p. 35-44. W. Junk, The Hague, Netherlands. Parsons, D. J. and S. H. DeBenedetti. 1 979. Impact of fire suppression on a mixed conifer forest. Forest Ecol. Manage. 2:21-33. Ratliff, R. D. 1979. Meadow sites of the Sierra Nevada, California. Classification and species relationships. Ph.D. dissertation. New Mexico State Univ., Las Cruces. . 1 982. A meadow site classification for the Sierra Nevada, California. U.S.D.A. Forest Serv. Pacific Southw. Forest Range Exp. Sta., Gen. Techn. Rept. PSW-60. Roach, A. W. 1958. Phytosociology of the Nash Crater lava flows, Linn County, Oregon. Ecol. Monogr. 22:169-193. RuNDEL, P. W. 1972. Habitat restriction in Giant Sequoia: the environmental control of grove boundaries. Amer. Midi. NaturaHst 87:81-95. , D. J. Parsons, and D. T. Gordon. 1 977. Montane and subalpine vegetation of the Sierra Nevada and Cascade Ranges. In M. G. Barbour and J. Major, eds.. Terrestrial vegetation of California, p. 559-599. Wiley-Interscience, NY. Sever, S. C. 1979. Vegetation ecology of a mountain mire. Crater Lake National Park, Oregon. M.S. thesis, Oregon State Univ., Corvallis. . 1983. Ecological analysis of Multorpor Fen Preserve, Oregon. Report to the The Nature Conservancy. Sharsmith, C. W. 1959. A report on the status, changes and ecology of back country meadows in Sequoia and Kings Canyon National Parks (unpubl.). Strand, S. 1 972. An investigation of the relationship of pack stock to some aspects of meadow ecology for seven meadows in Kings Canyon National Park. M.S. thesis, California State Univ., San Jose. Sumner, E. L. 1941. Special report on range management and wildlife protection in Kings Canyon National Park (unpubl.). . 1948. Tourist damage to mountain meadows in Sequoia-Kings Canyon National Park 1935-1948. A review with recommendations (unpubl.). Westhoff, V. and E. van der Maarel. 1978. The Braun-Blanquet approach. W. Junk, The Hague, Netherlands. Wood, S. H. 1975. Holocene stratigraphy and chronology of mountain meadows. Sierra Nevada, California. Ph.D. dissertation, California Institute of Technology, Pasadena. (Received 7 Jan 1985; revision accepted 17 Oct 1985.) VEGETATION OF TORREY LAKE MIRE, CENTRAL CASCADE RANGE, OREGON Robert E. Frenkel Department of Geography, Oregon State University, Corvallis 97331 William H. Moir U.S. Forest Service Region 3, 517 Gold Avenue SW, Albuquerque, NM 87102 John A. Christy Botany Section, Milwaukee Public Museum, 800 West Wells Street, Milwaukee, WI 53233 Abstract Torrey Lake Mire is typical of many small isolated wetlands in the predominantly forested central Oregon Cascade Range. Four mire communities, identified by floristic similarity analysis, exhibit distinct zonation with respect to a complex moisture gradient observed in the field and shown by detrended correspondence analysis or- dination. In order of increasing moisture and soil saturation they are: Kalmia mi- cwphylla/ Sphagnum Bog, Vaccinium occidentale/Trifolium longipes Thicket, Carex sitchensis Fen, and Carex rostrata Reedswamp. Montane mires are often regarded as having higher species diversity than surrounding forest because mire communities are packed into small areas in response to a fairly sharp environmental gradient. Forest communities, occupying more uniform environments, extend over broad areas. Because of a sharp hydrological gradient, Torrey Lake Mire embraces several closely packed plant communities. Based on the jackknife procedure, individual mire com- munities are shown to have similar species diversity to that of the surrounding Tsuga mertensiana forest community. Mires and wet meadows are small scattered features within the predominantly forested Cascade Range. These dispersed wetlands are seldom studied although they are important faunal habitat, at- tractive recreational features, often rich in plant taxa, and valuable sites for research on vegetational history. The term 'mire' is used in this paper in preference to 'bog.' Until recently 'bog' has been applied loosely, and often improperly in North America, to wetland types ranging from convex ombrotrophic wetlands with peat (true bogs) to seasonally flooded minerotrophic wetlands without peat. International wetland classifications now em- ploy the neutral term 'mire' to encompass bogs, fens, and a number of other wetland types (Gore 1983). Proper classification of a mire requires detailed knowledge of its topographical, hydrological, chem- Madrono, Vol. 33, No. 1, pp. 24-39, 27 March 1986 1986] FRENKEL ET AL.: TORRE Y LAKE MIRE 25 ical, and biological characteristics, information generally unavailable for wetlands in the Cascade Range. 'Mire' is, therefore, broadly applied here to denote a minerotrophic, peat-based wetland in which sedges and mosses, including Sphagnum, dominate the ground layer. There is relatively little published research on mires and wet mon- tane meadows in the Oregon Cascade Range; most studies are reports or theses and consider only the vascular flora. Hansen (1947) studied pollen profiles from several mires in the Western and High Cascades, but provided incomplete descriptions of contemporary vegetation. Aller (1956) identified a ''bog-marsh" community at 1220 m ele- vation on Monument Peak in the Western Cascades and furnished a brief description and species list. Hickman (1968, 1976) surveyed the vascular flora of a number of non-forested communities in the Western Cascades among which were three very generalized wetland types including a "bog association" for which he provided a partial floristic list. In the Three Sisters Wilderness, in the central Cascades, Halpem et al. (1984) surveyed montane and subalpine meadows including 11 hydric and 8 mesic community types. Seyer (1983) assessed the vegetation of Multorpor Fen, near Mt. Hood, recog- nizing aquatic, low sedge, moss mound, Carex sitchensis, low shrub, and carr (shrubby mire) communities. A brief description of the Big Spring Mire assemblage appears in a study by Roach (1952) of the Nash Crater lava flow vegetation near Santiam Pass. The most complete study, which considered both ecological re- lationships and the bryophyte flora of an Oregon Cascade mire, is by Seyer (1979) who identified 1 1 communities at Sphagnum Bog, Crater Lake National Park, and related these to nutrient status, hydrology, and microtopography. In another detailed investigation in Hunts Cove, Mt. Jefferson Wilderness, Campbell (1 973) described 1 1 meadow associations embracing four hydric communities. These communities correlated strongly with snow persistence and, to a lesser extent, with soil moisture and nutrient status. Because of the lack of published studies, the primary purpose of our paper is to describe the main floristic and structural features of Torrey Lake Mire, a typical wetland of the central Cascades in Or- egon, and to relate the communities to those identified in published and unpublished regional research. Because there are only fragmen- tary accounts of the bryophyte composition of Cascade mires, a secondary objective is to describe the bryophyte flora of the mire and associated mire communities. A third objective is to relate mire communities to microtopography, associated hydrological condi- tions, and surrounding forest vegetation. There is also a need to document the resources of Torrey Lake Mire because it is a dis- tinctive element in the proposed Torrey-Charlton Research Natural Area, a 1075 ha tract in the Willamette and Deschutes National Forests. 26 MADRONO [Vol. 33 Location and Site Torrey Lake Mire lies at 1650 m elevation, 4 km north of Waldo Lake in the Willamette National Forest and approximately 200 m southeast of Torrey Lake into which it drains (Fig. 1). The 0.9 ha mire is the largest of more than 10 small wetlands in the 380 ha Torrey Lake Unit of the proposed research natural area (RNA). The RNA is being established to protect old-growth Tsuga mertensiana forest and associated wet meadows, ponds, lakes, and rock outcrops (McKee and Franklin 1977). Located within the High Cascades geological province, the mire is situated on a gently undulating plateau formed by a series of composite volcanoes that deposited scoriaceous materials, andesites, and basalts during the Pliocene and Pleistocene Epochs (Baldwin 1981). The area was subsequently glaciated and more recently cov- ered by Mount Mazama ash and pumice. A soil pit shows that Torrey Lake Mire developed directly on Mount Mazama ash (ca. 40 cm below the present surface); the present mire is, therefore, at least 6700 years old (Sarna-Wojcicki et al. 1983). The area has a cool, wet climate. Annual precipitation ranges from 1 600 to 2000 mm, more than three-fourths of which occurs as snow. 1986] FRENKEL ET AL.: TORREY LAKE MIRE 27 Snow depths often exceed 5 m and snow packs may persist from six to eight months. Mean January and July temperatures are — 3.5°C and 13.5°C respectively, whereas the mean January minimum is — 8.5°C and the mean July maximum is 21.5°C. Frosts can occur in any month (McKee and Franklin 1977). Torrey Lake Mire is an inclusion within a broad expanse of Tsuga mertensiana forest that mantles the crest of the central Cascades. The forest is included within the Tsuga mertensiana Zone (Franklin and Dymess 1973, Schuller 1978, Hemstrom etal. 1982). The major vegetation community surrounding the mire community is the Tsu- ga mertensiana/ Vaccinium scoparium association. Methods A single 60 m transect was established at right angles to the long axis of the mire between permanently marked endposts in the fring- ing upland forest, 2-3 m in elevation above the mire. The transect extends into the forest about 7 m on both sides of the mire (Fig. 1). Sixty-four 20 x 50 cm microplots were placed at 50 to 100 cm intervals along the transect and species canopy cover was estimated by Daubenmire (1959) cover classes. Substrate pH at 5 cm depth was measured colorimetrically at several locations within a com- munity. Adjacent forest was sampled with two 500 m^ circular re- connaissance plots centered 20 m from the mire edge in line with the transect. Additionally, the regional Tsuga mertensiana forest was sampled with 46 reconnaissance plots, each 500 m^. Forest plot placement followed landscape stratification based on aerial photog- raphy and field survey and was aimed at typifying the general vege- tation of the proposed Torrey-Charlton RNA. Reconnaissance plot methods followed FrankUn et al. (1970). A nearly complete species list was compiled by observations beyond plots for the mire and adjacent forest. Vascular plant nomenclature follows Hitchcock and Cronquist (1973) and bryophyte nomenclature follows Crum et al. (1973) and Crum (1984). Voucher specimens are deposited at OSC. Mire microplot and forest macroplot data were analyzed using the program TABORD, which structures phytosociological data in the form of an association table based on sample similarity (Maarel et al. 1978). Microplot data were also ordinated by detrended corre- spondence analysis, an eigenvector ordination technique derived from reciprocal averaging using the Cornell University ecology pro- gram DECORANA (Gauch 1982, Hill 1979). From the latter anal- ysis, major environmental controls were inferred, floristic changes across environmental gradients were quantified, and plant associa- tions based on TABORD were displayed with respect to environ- mental gradients. 28 MADRONO [Vol. 33 Mire and forest diversity were analyzed by PRHILL, an unpub- lished computer program developed by B. G. Smith that was de- signed to calculate jackknife estimates of M. O. Hill's diversity num- bers (Hill 1973, Heltsche and Forrester 1983) for each community type. By simulation, Heltsche and Forrester (1983) have shown that jackknife estimates of diversity permit community comparisons un- der a wide variation in sample size. Details of the forest segment of the study are to be published separately. Results and Discussion The flora of Torrey Lake Mire and the adjacent forest include 6 trees, 1 5 shrubs, 58 herbs, and 28 bryophytes (Table 1). The vascular plant families represented by the greatest number of species include the Cyperaceae (10), Gramineae (8), Rosaceae, Ranunculaceae, and Pinaceae (6), and Asteraceae and Ericaceae (5). A species list for the forest plots is available from the senior author. Plant communities. Two terrestrial and four mire communities are identified by TABORD similarity analysis and are summarized in Table 2 with respect to mean percent cover and percent constancy of dominant vascular plant species in microplots. Communities also are depicted by profile of dominant species cover as it occurs across the mire (Fig. 2). Table 1. Flora of Torrey Lake Mire and Adjacent Forest. Forest species occur in the Tsuga mertensiana/ Vaccinium scoparium Association and Xerophyllum tenax mire fringe; mire species occur in four mire communities. Mire Forest Trees Picea engelmannii Abies amabilis Pinus contorta var. latifolia A. lasiocarpa Pinus monticola Tsuga mertensiana Shrubs Kalmia microphylla var. microphylla Arctostaphylos nevadensis Lonicera caerulea Vaccinium caespitosum Ribes lacustre V. membranaceum Salix geyeriana V. ovalifolium S. myrtillifolia V. scoparium Spiraea densiflora S. douglasii var. menziesii Sorbus sitchensis Vaccinium caespitosum V. occidentale Herbs Agrostis idahoensis Achlys triphylla A. thurberiana Anemone deltoidea Apargidium boreale A. lyallii Arnica mollis Berberis nervosa Aster occidentalis Carex pensylvanica 1986] FRENKEL ET AL.: TORREY LAKE MIRE 29 Table 1. Continued. Mire Forest Calamagrostis canadensis var. scabra C. rossii Caltha biflom var. rotundifolia Chimaphila umbellata var. Carex buxbaumii occidentalis C. limosa Clintonia uniflora C. muricata Cornus canadensis C. rostrata Fragaria virginiana var. C. simulata platypetala C. sitchensis Gaultheria humifusa Danthonia intermedia Hypopitys monotropa Deschampsia atropurpurea Juncus parryi D. cespitosa Pyrola secunda Dodecatheon jeffreyi Rubus lasiococcus Eleocharis pauciflora Smilacina stellata Elymus glaucus Viola glabella Epilobium alpinum var. gracillimum V. orbiculata Hypericum anagalloides Xerophyllum tenax Juncus balticus var. balticus Juncus bolanderi Mimulus primuloides Muhlenbergia filiformis Nuphar polysepalum Pedicularis attolens Potentilla drummondii Ranunculus alismaefolius var. davisii R. flammula R. gormanii Scirpus congdonii Senecio cymbalarioides S. triangularis Tofieldia glutinosa Trifolium longipes var. shastense Utricularia vulgaris Veronica wormskjoldii Viola macloskeyi Bryophytes Aulacomnium palustre Andreaea blyttii Bryum pseudotriquetrum Brachythecium leibergii Calliergon stramineum B. collinum Cephalozia sp. Ceratodon purpureus Cephaloziella sp. Claopodium bolanderi Drepanocladus exannulatus Dicranoweisia crispula var. Philonotis fontana contermina Pohlia nutans Dicranum fuscescens Rhizomnium magnifolium D. pallidisetum Scapania irrigua Dryptodon patens Sphagnum capillifolium Kiaeria starkei S. squarrosum Marsupella sp. S. subsecundum Pohlia cruda P. nutans Polytrichum piliferum Rhacomitrium heterostichum var. sudeticum 30 MADRONO [Vol. 33 > a S OA a O 00 o o o o o o o -H m ir> in o \\\\ O 00 o o r- oo m O O O 00 — m 00 O m o O c<-) 00 !C ^ C H H ^ Q:^ ^ c ^ ^ ^ ^ k II III '2 5 g §1 "I 2"§ V jT* i.. s -2 1986] FRENKEL ET AL.: TORREY LAKE MIRE 31 00 O 00 C2 X) 00 C > ^ o o o o ^tt ^t3 d d ^ T3 TD o o o o "3 "3 "3 3 o o o o >s ^ ^ ^ U interpreted as an indication of wind-pollination. It is equally pos- sible, however, that this ratio is the result of the function of pollen as the pollinator reward, together with the relative inefficiency of beetles as pollinators compared, for example, to bees (Pellmyr 1985). Camazine and Niklas (1984) suggested that the closely related Symplocarpus foetidus may be wind-pollinated, rather than insect-pollinated. This conclusion was based on the scarcity of in- sects, documented by the authors as well as earlier investigators (e.g., Trelease 1879), aerodynamic properties of the spathe, poor syn- chronization of flowering within populations (Niklas, pers. comm.). 52 MADRONO [Vol. 33 Table 2. Results from Experiments on Function of Color and Fragrance IN Attraction of Pelecomalius testaceum. Numbers represent individual beetles observed during 30 minute intervals. No. of ap- No. of proaches alight- Experiment only ings G-test 1. Spathe in sealed dish 4 1 G = 1.386 Leaf in sealed dish 0 0 df = 1; n.s. 2. Leaf in sealed dish 1 1 G = 14 699 Spathe hidden between leaves ,p i ' ^nnm _r.jju 1 ic di=l;p< 0.001 m perforated dish 2 15 3. Spathe in sealed dish 1 3 Leaf in sealed dish 0 0 Spathe in vase 0 38 4. Inflorescence with female- stage spadix 1 13 G= 1.202 Inflorescence with male- df = 1; n.s. stage spadix 2 8 G = 68.002 df= 2; p < 0.001 and considerable flexibility in flowering time between years. Al- though the aerodynamic properties of the open spathe of L. ameri- canum are unknown, potential pollinators are present and flowering time is relatively constant between years (Meeuse, pers. comm.) and highly synchronized within populations. These characteristics to- gether with the absence of Lysichiton pollen on the pollen traps suggest insect pollination rather than wind-pollination in L. ameri- canum. Successful attraction to female-phase flowers, where rewards are lacking, may be a case of "pollination by deceit" (sensu Baker 1976). The inflorescence, however, still serves the function of a mating site, a crucial resource to the beetles, and thus the system cannot be considered entirely deceptive. The behavior experiments indicate a sequential function of stimuli in attraction of beetles. Odor alone induced search behavior, whereas non-odorous yellow dishes did not. When the entire spathe was exposed, the number of alightings increased with a factor of about 2.5. This suggests that the fragrance elicits search behavior in P. testaceum for yellow objects. A few similar cases are known from nocturnal as well as diurnal moths— the fragrance of Melandrium album (Caryophyllaceae) induces search for white objects in Hadena (Noctuidae) (Brantjes 1 976), and several plusiine moths behave sim- ilarly when exposed to fragrance of Platanthera spp. (Orchidaceae) 1986] PELLMYR AND PATT: POLLINATION OF LYSICHITON 53 (Nilsson 1978) and other flowers (Schremmer 1941). Although strong fragrance is typical of most beetle-pollinated flowers (Faegri and van der Pijl 1979), the sensory steps in pollinator attraction are usually not known. It is important to point out that these experiments did not allow the conclusion that the beetles possess color vision. Be- cause we could not match green leaves and yellow spathes for bright- ness, it is possible that the insects only detected differences in bright- ness. Knoll (1926) used purple and white glass models to study the effects of odor on the pollinators of the purple-spathed Arum nigrum. Presence of the carrion-like odor resulted in as many visits to each of the models as to the natural ffowers. The results demonstrate that the odor could attract the visitors, but his experiments are incon- clusive as to whether visual stimuli are important in pollinator at- traction. Pelecomalius testaceum is apparently heavily specialized on Ly- sichiton americanum during its adult life. The inflorescence is a focal spot for mating activities, it provides some degree of shelter during mating, and it is a cornucopia of food. The life-cycle of the genus Pelecomalius may be closely tied to Lysichiton. A number of species were described by Casey (1886, 1893) from small numbers of in- dividuals collected on or around L. americanum in California. Sev- eral were not recognized by Schwarz (1892), and Hatch (1957) listed only two species, both with distributions well matching that of Ly- sichiton americanum. In Japan and the northern Pacific parts of Asia, Lysichiton cam- tschatkense grows in much the same habitats as its North American vicariant (Ohwi 1984). The only obvious diflerence from L. ameri- canum is its bright white spathe. Studies of the pollination biology of L. camtschatkense could provide data for a better understanding of the evolution of interaction between Pelecomalius and Lysichiton. Acknowledgments We thank Bastiaan J. D. Meeuse for sharing his knowledge of the Araceae, for showing us several populations of L. americanum, and for having the 1 984 pollination class perform thin-layer chromatography of spathe pigment extractions for us; Douglas Ewing for access to the greenhouse cage; the Remote Sensing Lab at the University of Washington for help with reflectance analyses; and B. J. D. Meeuse, Angela M. Brown, Robert A. Schlising, and two anonymous reviewers for comments on the manuscript. Literature Cited Baker, H. G. 1976. "Mistake" pollination as a reproductive system with special reference to the Caricaceae. In J. Burley and B. T. Styles, eds.. Tropical trees: variation, breeding, and conservation, p. 161-169. Academic Press, New York. Beach, J. H. 1982. Beetle pollination of Cyclanthus bipartitus (Cyclanthaceae). Amer. J. Bot. 69:1074-1081. Brantjes, N. B. M. 1976. Riddles around the pollination of Melandrium album 54 MADRONO [Vol. 33 (Mill.) Garcke (Caryophyllaceae) during the oviposition by Hadena bicruris Hufn. (Noctuidae, Lepidoptera), I + II. Proc. Konikl. Ned. Akad. Wetensch., Ser. C, Biol. Med. Sci. 79:1-12, 127-141. Camazine, S. and K. J. Niklas. 1984. Aerobiology of Symplocarpus foetidus: in- teractions between the spathe and spadix. Amer. J. Bot. 71:843-850. Casey, T. L. 1 886. Descriptive notices of North American Coleoptera. I. Bull. Cahf. Acad. Sci. 2:157-264. . 1893. Coleopterological notices. V. Ann. New York Acad. Sci. 7:281-606. Crowson, R. a. 1981. The biology of the Coleoptera. Academic Press, London. Cruden, R. W. 1977. Pollen-ovule ratios: a conservative indicator of breeding systems in flowering plants. Evolution 31:32-46. Faegri, K. and L. van der Pijl. 1979. The principles of pollination ecology, 3rd ed. Pergamon Press, Oxford. Hatch, M. H. 1957. The beetles of the Pacific Northwest, part II. Univ. Washington Press, Seattle. Knoll, F. 1926. Insekten und Blumen. IV. Die Arum-Bliitenstande und Ihre Be- sucher. Abh. K. K. Zool.-Bot. Ges. Wien 12:383-481 + 1 plate. Knuth, p. 1898-1904. Handbuch der Bliitenbiologie I-III. Engelmann, Leipzig. Meeuse, B. and S. Morris. 1984. The sex life of flowers. Facts on File Publ., New York. NiLSSON, L. A. 1978. Pollination ecology and adaptation in Platanthera chlorantha (Orchidaceae). Bot. Not. 131:35-51. Ohwi, J. 1984. Flora of Japan, 2nd ed. Smithsonian Press, Washington, DC. Pellmyr, O. 1985. Pollination biology of Cimicifuga arizonica (Ranunculaceae). Bot. Gaz. (Chicago) 146:404-412. . In press. Cyclocephala: visitor and probable pollinator of Caladium bicolor (Araceae). Acta Amaz. Raven, P. H., H. Evert, and H. Curtis. 1981. The biology of plants, 3rd ed. Worth Publishers, New York. ScHREMMER, F. 1941. Sinncsphysiologic und Blumenbesuch des Falters von Plusia gamma L. Zool. Jahrb. 74:375-434. Schrottky, C. 1910. Die Befruchtung von Philodendron und Caladium durch einen Kafer (Erioscelis emarginata Mann.). Zeitschr. Wiss. Insektenbiol. 6:67-68. ScHWARz, E. A. 1892. In protocol from meetings. Proc. Entomol. Soc. Wash. 2: 396. Trelease, W. 1879. On the fertilization of Symplocarpus foetidus. Amer. Naturalist 13:580-581. VAN DER PijL, L. 1937. Biological and physiological observations on the inflorescence of Amorphophallus. Rev. Trav. Bot. Neerl. 34:157-167. VoGEL, S. 1978. Pilzmiickenblumen als Pilzmimeten. Flora 167:367-398. Whigham, D. 1974. An ecological life history study of Uvularia perfoliata L. Am. Midi. Naturalist 91:343-359. Williams, N. H. and R. L. Dressler. 1976. Euglossine pollination of Spathiphyllum (Araceae). Selbyana 1:349-356. (Received 28 Jan 1985; revision accepted 17 Oct 1985.) AN APPARENT CASE OF INTROGRESSION BETWEEN PINYON PINES OF THE NEW YORK MOUNTAINS, EASTERN MOJAVE DESERT James des Lauriers Department of Biology, Chaffey Community College, Alta Loma, CA 91701 Mark Ikeda Department of Biological Sciences, California State Polytechnic University, Pomona 91768 Abstract The piny on pines in the New York Mountains of the Mojave Desert are poly- morphic along an altitudinal gradient. The possibility that this gradient is the result of hybridization between Pinus monophylla and P. edulis is the subject of this study. Counts of needle number per fascicle, resin canal number per needle, and cone width were made over the entire altitudinal range of distribution of the pine community. We conclude that there are two species present and that an extensive altitudinal zone of introgression exists at intermediate elevations. The characterization of the pinyon pines in the New York Moun- tains, San Bernardino County, California, has been addressed by several authors (Lanner 1974, Vasek and Thome 1977, Trombulak and Cody 1980, Thome et al. 1981). The slopes of the New York Mountains are described as containing the westemmost sympatric distribution of Pinus edulis and P. monophylla (Griffin and Critch- field 1972). Lanner (1974), however, concluded that P. monophylla is the only species present and that the two-needle form represents a mutant to the ancestral two-needle condition. His argument was based on the resemblance of other morphological characters to those of P. monophylla. Trombulak and Cody (1980) found that the pro- portion of two-needle fascicles on individual trees showed a nearly disjunct distribution with practically no trees displaying interme- diate proportions, and concluded that both P. monophylla and P. edulis were present with the latter species limited to higher eleva- tions. Their transects were near Prospect Canyon on the north side of New York Peak. Reported here are the results of a new study carried out in a south- facing drainage on the same mountain. We describe an altitudinal zone of intergradation between the one- and two-needle forms and speculate on its significance. Madro^^o, Vol. 33, No. 1, pp. 55-62, 27 March 1986 56 MADRONO [Vol. 33 Fig. 1 . Location of the transect, a. Map of California. Dark square shows location of New York Mountains, b. New York Mountain range showing both Prospect and Caruthers Canyons, c. Topographic map with 100 m contour intervals showing tran- sect location (dashed line) in Caruthers Canyon. The 5 isolated stations (0) are on outcrops at the lower limit of pinyon distribution. Study Site and Methods Data were collected during the month of May in the years 1981- 1985. A transect was placed in the Caruthers Canyon drainage (Fig. 1). The transect followed the jeep road north from the lower limit of pinyons at an elevation of 1620 m to the Giant Ledge Mine, then up an east-facing slope to the peak. The sampling stations along the transect tended to occupy the more mesic situations in the drain- age. Elevations and positions were determined using a Thommen temperature compensated altimeter and by triangulation using a Brunton Pocket Transit and USGS topographic maps (Ivanpah, CA, and Midhills, CA, quadrangles). At 5-60 m increments in elevation, 8-25 trees were sampled. The terminal 1 5-20 cm segment of a twig was taken at about chest height from the side of the tree most exposed to illumination. From the cut end of the twig the first 20 fascicles were classified as bearing one or two needles. Three-needle fascicles were reported as two- 1986] DES LAURIERS AND IKEDA: PINYON PINES 57 needle ones. The mean number of two-needle fascicles was computed for each collecting station. Resin canal numbers were determined from an additional 11-15 twigs (one twig per tree), obtained at each station in a manner similar to that described above. Trees were selected for sampling without regard to which had been sampled previously for fascicle ratio. Five needles were selected at random from near the cut end of each twig. Counts of resin canals in razor-sectioned needles were made under a 10 X hand lens. When two-needle fascicles were sampled only one of the two needles was counted. Sampling was repeated at all but the highest two stations, where measurements of fascicle ratio and resin canal numbers were taken from one twig from each of ten trees at each station. Each twig yielded 20 fascicle counts and five resin canal counts. Thirty intact seed cones were collected haphazardly from the ground at each station. Vernier calipers were used to obtain the greatest diameter of each cone, to the nearest 0. 1 mm. Results The frequency of one- and two-needle fascicles on individual trees is shown in Fig. 2B. Of the 388 trees sampled, 122 (31%) showed a range between 20-80 percent two-needle fascicles. A strong altitu- dinal shift occurred in the proportion of two-needle fascicles. The proportion of two-needle fascicles (Y) plotted against eleva- tion (X) is illustrated in Fig. 3. The proportion of two-needle fascicles rises sharply between 1700-1900 m and then changes more slowly to the peak. The data are best described by the sigmoid equation The elevation at which 50 percent of the sampled fascicles contain two needles (Xq.s, the turnover elevation) is computed from the regression equation to be 1940 m. The average number of resin canals per needle (Y) is plotted against elevation (X) (Fig. 3). The average number of resin canals per needle declines sharply between 1620-1860 m. From that ele- vation to the peak, the number of resin canals changes only slightly (Fig. 3). The data are described best by the sigmoid equation ! ^=l + 7.50x lO-^^X^3s --^ = 0.87, p«0.01. (eq. 2) The mean number of resin canals for all needles sampled below 1 900 m equals 5.5 ± 0.9, n = 950 (X ± 95% confidence interval). This 58 MADRONO [Vol. 33 # RESIN CANALS 2-NEEDLE FASCICLES per needle per sample Fig. 2. A. Frequency distribution of the proportion of the sample of needles (n) against resin canal number per needle in four elevation zones. B. The proportion of individual tree samples (n) is plotted against the number of two-needle fascicles per 20 needle sample in four elevation zones. value does not differ significantly from that reported for P. mono- phylla (4.9 ± 1.1) by Lanner (1974). For the trees above 1900 m the number of resin canals equals 2.7 ± 0.2, n = 445. Lanner's value for P. edulis (2.0 ± 0. 1) is smaller than that we obtained. The mean resin canal number above 1900 m is significantly lower (t = 4.12; p < 0.001) than it is at lower elevations. There is a highly significant correlation (t = 8.23, p^ 0.01) be- tween the mean number of two-needle fascicles and the mean num- ber of resin canals per needle (Fig. 4). The values near the x-axis are 1986] DES LAURIERS AND IKEDA: PINYON PINES 59 © 20 10 ® 0 ® 0© 0 0. +0 0 © 0 0 ® 0 0 0 0 10 - 5 CD Z 2 < ' Ul 1 1 1 1 1 1 1700 1900 2100 ELEVATION (m) Fig. 3. Change in the mean number of two-needle fascicles per 20 fascicle sample over elevation (circles) and the change in mean number of resin canals per needle over elevation (crosses). Needle number is plotted to the left and resin canal number is at the right. from sites at the highest elevations and are typical of P. edulis. The values near the y-axis are from low elevation sites and resemble P. monophylla. The intermediate points are from stations at inter- mediate elevations. Seed cone diameter varies independently of elevation. The mean maximum cone diameter for the entire sample equals 54.8 ± 2.5, n = 660. This value is not significantly different from Lanner's (1974) value for widely dispersed hybrid populations in the western U.S. (53.2 ± 5.9, range 30-87 mm). Lanner also reports cone diameters for P. edulis and P. monophylla as 42.5 ± 4.6, range 20-68 mm and 64.6 ± 4.8, range 39-80 mm, respectively. Discussion Lanner (1974) concluded that the ratio of one-needled to two- needled fascicles is heritable. Experimental P. monophylla x P. ed- ulis hybrids exhibited a 0.5 ratio. In comparing mesic and xeric sites, he also concluded that for P. monophylla the ratio is stable and insensitive to site differences. The transition of trees with one-needle fascicles and several resin canals at lower elevations and on arid slopes toward two-needle, two-resin-canal plants in higher, presumably moister locations is compatible with expectations about P. monophylla and P. edulis. Trombulak and Cody (1980) present a scenario of progressive re- placement of P. edulis by P. monophylla upward on the slopes in- 60 MADRONO [Vol. 33 CO 7 n 6 - 4 - -I 20 0 5 10 15 2-NEEDLE FASCICLES per SAMPLE Fig. 4. Relationship between the mean number of two-needle fascicles per 20 fascicle sample and the mean number of resin canals per needle per 50 needle sample. Each point represents one sample station containing 10 trees. duced by generally more arid conditions since pluvial times. The Xo.5 needle ratio values they obtained (1680 m and 1834 m) are lower on north-facing slopes than is the value we obtained (1940 m, computed from equation 1) on the south side of the range (Fig. 3). We would add that frost hardiness, especially in seedlings, may be greater in P. edulis (R. M. Beeks, pers. comm.), thus making it a superior competitor only at greater elevation and on north-facing slopes. Together these results are consistent with the Trombulak and Cody interpretation of an interaction between the two taxa that is mediated by moisture and temperature (see also Wells 1979, 1983). The decline in resin canal number that occurs over the same altitudinal range (Figs. 2A, 3) supports the evidence that a zone of intergradation exists between the two taxa. This trait, however, is highly sensitive to site differences. Lanner (1974) showed a 40 per- cent increase in the number of resin canals in P. monophylla when the samples from arid sites (X = 5.06 canals) were compared to mesic ones (X = 3.62 canals) in the Raft River Mountains, Utah. The impact of this moisture response in our study is minimized by our deliberate choice of mesic sites to sample. This shifting rela- tionship of needle number to resin canal number (Fig. 4) is a pattern 1986] DES LAURIERS AND IKEDA: PINYON PINES 61 found elsewhere in hybrid swarms of P. monophylla x p, edulis (Lanner 1974). Alternatively, but less parsimoniously, the pattern of intergradation might be produced by primary differentiation of altitudinal races or species. Finally, seed cone size variation is chaotic. The population mean, however, is the same as that obtained for hybrid populations else- where in the southwestern U.S. (Lanner 1974). Because this trait overlaps broadly in the two species, it is of little value here in attempting to characterize hybrids. Endler (1982) defined the width of contact zones between hybrid- izing species as the region having hybrid index values between 0.2 and 0.8. A substantial number (31%) of the individuals in our sample have those intermediate proportions of two-needle fascicles at mid- dle elevations (Fig. 2B). The distribution shown in Fig. 2B differs significantly (x^ = 202; p < 0.00 1) from that obtained by Trombulak and Cody (1980, p. 64). They found practically no trees having intermediate proportions of two-needle fascicles in Prospect Canyon on the north side of New York Peak. They interpret their result as distinguishing clearly between P. monophylla and P. edulis in the New York Mountains. The distinction fails in Caruthers Canyon, on the south side. In fact, the pattern demonstrated (Fig. 3) shows a zone of intergradation and suggests introgression between about 1700 and 1900 m elevation. In conclusion, the trees near the peak appear to be a remnant stand of P. edulis whose distinctiveness is reduced by hybridization with the regionally abundant P. monophylla. The fact that populations in Caruthers and Prospect Canyons show different patterns is par- ticularly interesting. Any explanation for the difference awaits further study. Acknowledgments The students in the Chaffey College Population Biology course during 1981-1985 sweated up hot mountainsides and braved grumpy rattlesnakes. Their eager coop- eration is greatly appreciated. Various drafts of this paper were commented on by Drs. Richard M. Beeks, Curtis Clark, Ronald M. Lanner, and David Moriarty. Dr. Moriarty also provided statistical advice. Literature Cited Endler, J. 1982. Problems in distinguishing historical from ecological factors in biogeography. Amer. Zool. 22:441-452. Griffin, J. R. and W. B. Critchfield. 1972. The distribution of forest trees in California. U.S.D.A. Forest Serv. Res. Pap. PSW-82. Lanner, R. M. 1 974. Natural hybridization between Pinus edulis and Pinus mono- phylla in the American southwest. Silvae Genet. 23:108-1 16. Thorne, R. F., B. a. Prigge, and J. Hendrickson. 1981. A flora of the higher ranges and the Kelso dunes of the eastern Mojave desert in California. Aliso 10: 71-186. 62 MADRONO [Vol. 33 Trombulak, S. C. and M. L. Cody. 1980. Elevational distribution of Pinus edulis and P. monophylla (Pinaceae) in the New York Mountains, eastern Mojave Desert. Madrono 27:61-67. Vasek, F. C. and R. F. Thorne. 1977. Transmontane coniferous vegetation. In M. G. Barbour and J. Major, eds.. Terrestrial vegetation of California, p. 797-834. J. Wiley-Interscience, New York. Wells, P. V. 1979. An equable glaciopluvial in the West: pleniglacial evidence of increased precipitation on a gradient from the Great Basin to the Sonoran and Chihuahuan Deserts. Quaternary Research 12:31-325. . 1983. Paleobiogeography of the montane islands in the Great Basin since the last glaciopluvial. Ecol. Monogr. 53:341-382. (Received 9 Jan 1985; revision accepted 18 Oct 1985.) ANNOUNCEMENT Symposium to Honor G. Ledyard Stebbins An International Symposium will be held in Davis, California, on 12-14 September 1986, to honor Professor G. Ledyard Stebbins in the year of his 80th birthday. Invited talks by leading plant biologists will include topics in population and ecological genetics, organelle and nu- clear molecular genetics, morphogenesis and plant development, and evolution and systematics. For further information please contact: Dr. L. D. Gottlieb, Department of Genetics, or Dr. S. K. Jain, Department of Agronomy and Range Science, University of California, Davis, CA 95616. SYSTEMATICS OF COLLINS lA PARVI FLORA AND C GRANDIFLORA (SCROPHULARIACEAE) Fred R. Ganders Department of Botany, University of British Columbia, Vancouver, B.C. V6T 2B1, Canada Gerda R. Krause Department of Biology, Langara Campus Vancouver Community College, Vancouver, B.C. V5Y 2Z6, Canada Abstract Collinsia parviflora and C. grandiflora are distinguished primarily on the basis of flower size, but intermediate populations are more numerous than either small- or large-flowered populations in coastal British Columbia. Experimental hybridizations demonstrated that populations of all flower sizes are completely interfertile, and all are tetraploid with In = 28. Flower size in F, hybrids was intermediate between the parents, and the F2 generation showed a greater range than the F,, indicating segre- gation at several loci. We consider the two taxa to be intergrading varieties of one species, C. parviflora Dougl. ex Lindl. var. parviflora and C. parviflora var. grandiflora (Dougl. ex Lindl.) Ganders & Krause. Collinsia parviflora Dougl. ex Lindl. and C grandiflora Dougl. ex Lindl. are the only species of this predominantly Califomian genus that are found in western Washington and British Columbia. They are very similar, being distinguished primarily on the basis of flower size. In their extreme forms, the flowers are very different in ap- pearance. Collinsia parviflora has corollas 4-5 mm in length and the flowers are highly to completely autogamous. Collinsia grandiflora has corollas 10-17 mm in length and sets few seeds if pollinators are excluded, although it is completely self-compatible. West of the Cascade Mountains in British Columbia and Washington State most populations of Collinsia have flowers intermediate in size (Fig. 1). These intermediate populations have been named as varieties of one species or the other or even treated as a separate species, although recent floristic treatments do not accept the intermediate taxa (Hitch- cock et al. 1959, Taylor and MacBryde 1977). In her monograph of the genus, Newsom (1929) noted, 'Tn studying these two closely allied species, one finds a continuous series of intergradation from the largest-flowered grandiflora var. typica through grandiflora var. pusilla to the smallest-flowered parviflora, and yet the two extremes differ greatly." Nevertheless, all authors have maintained the two as specifically distinct. The dividing line between C parviflora and C. grandiflora, however, has been drawn arbitrarily at different ffow- Madrono, Vol. 33, No. 1, pp. 63-70, 27 March 1986 64 MADRONO [Vol. 33 er sizes by different authors. Only St. John (1956) and Peck (1961) formally recognized any overlap in flower size between the two species. Some authors, notably Gilkey and Dennis (1980), make the two species appear to be quite distinct. Garber (1956, 1958, 1974) reported both C. parviflora and C. grandiflom to be diploids with n = 1 . Unfortunately Garber cited neither voucher specimens nor localities for the plants on which the counts were based, although the plants were probably from Cali- fornia because the seeds from which he grew the plants were sent to him by people living in California. Taylor and Mulligan (1968) reported C. parviflora from Haida Point, Graham Island, in the Queen Charlotte Islands, British Columbia, to be tetraploid with n = 14. The distribution of flower sizes among populations in British Co- lumbia suggests that the two species recognized traditionally may just represent the extremes of a single variable species. Rower sizes are not correlated with other morphological characters (Krause 1978). The reports of both diploid and tetraploid chromosome numbers suggest, however, the possibility that ploidy level may be correlated with flower size. We analyzed experimental intra- and interpopu- lational hybridizations and counted chromosome numbers to de- termine the taxonomic relationship of the two species and their intermediate populations. Materials and Methods Seeds were collected from the localities shown in Fig. 1, all in British Columbia except one population from Anacortes, WA, USA. Plants were grown from seed in growth chambers under a long-day regime of 16 hr light at 20°C and 8 hr dark at 10°C. Total corolla length was measured on these plants. Interpopulational hybridizations were made in growth chambers. Seed parents were emasculated in bud by removing anthers with fine forceps, and were pollinated several days later. Crosses involving small-flowered plants were difficult to make because emasculation of small buds was difficult and small flowers produced few pollen grains. Hybrids involving the Lindeman Lake population were ob- tained using this population as the male parent only, because the small buds were destroyed by our attempts to emasculate them. Inheritance of flower size was investigated in five interpopula- tional crosses. In the genetic studies of corolla size, corolla length was measured from the tip of the lips to the bend in the corolla tube, in order to compensate for differences in the total corolla length caused solely by differences in the angle of the bend in the corolla tube. Therefore, the lengths reported in the inheritance studies are less than the total corolla lengths shown in Fig. 1 . 1986] GANDERS AND KRAUSE: COLLINSIA 65 Fig. 1 . Range of corolla lengths in mm in populations of Collinsia in southwestern British Columbia and northwestern Washington State and the localities of the pop- ulations studied. The numbers following the locality are the range of corolla lengths in each population. 66 MADRONO [Vol. 33 Table 1. Numbers of Hybrids Obtained in Inter- and Intrapopulational Crosses in Collinsia. All crosses attempted were successful and all hybrids completely fertile. I^d.ICIll pupUlallUlla Fi progeny (including reciprocals) F2 progeny Carlos Island x Carlos Island 40 1395 Carlos Island x Jack Point 27 2529 blk rails x hlk rails 14 13 Elk Falls X Botanic Valley 47 337 Elk Falls X Mt. Douglas Park 165 749 Jack Point x Jack Point 70 1401 Jack Point x Botanic Valley 67 410 Jack Point x Lindeman Lake 2 31 Jack Point x Mt. Douglas Park 42 94 Jack Point x Nanoose Hill 58 1161 Nanoose Hill x Botanic Valley 83 984 Nanoose Hill x Lindeman Lake 1 0 Nanoose Hill x Mt. Douglas Park 16 30 For chromosome counts buds were fixed in a 6:3:2 solution of ethanol, chloroform, and propionic acid and stained in alcoholic hydrochloric acid-carmine (Snow 1963). Vouchers are at UBC. Results Chromosome counts of2;i = 28(2«=14 pairs) were obtained from six populations: Botanic Valley, Carlos Island, Elk Falls, Jack Point, Lindeman Lake, and Nanoose Hill. These populations cover a range of flower sizes, including the populations with the largest and smallest flowers studied. The counts confirm the tetraploid sta- tus of Collinsia in British Columbia. All intra- and interpopulational crosses produced vigorous, fertile hybrids; F2 generation plants were also vigorous and fully fertile. More than 600 Fj hybrid plants and more than 6000 F2 hybrids were grown (Table 1). There are no intrinsic isolating mechanisms between C. parviflora and C. grandiflora or intermediate popula- tions. Interpopulation Fi hybrids were confirmed by intermediacy in flower size or by the use of single gene markers for flower color or leaf pigmentation (Griffiths et al. 1977, Krause 1978). A few progeny from accidental self-pollinations were detected in this way in crosses involving small-flowered plants as seed parents. Plants from self-poUinations and intrapopulational crosses showed the same narrow range of flower sizes found in the parental popu- lations (Krause 1978). This suggests that plants within natural pop- ulations are homozygous at most loci determining flower size. Av- 1986] GANDERS AND KRAUSE: COLL/A^5/^ 67 Table 2. Summary of Crosses to Investigate Inheritance of Flower Sizes. Flower size (mm) Sample size F2 populations Range Mean s.d. Flowers Plants Botanic Valley 3-6 4.5 U.Oh 27 9 Nanoose Hill 5-7 6.1 U.JO 15 5 4-6 4.6 U.oo 53 18 ^ 2 4-7 5.2 0.73 380 160 Lindeman Lake 3-4 3.7 n AQ 15 5 Jack Point 6-8 6.7 U. J / 18 6 (F, not measured) F, 4-8 5.8 0.98 21 10 Botanie Valley 3-6 4.5 U.0'+ 27 9 Jack Point 6-8 6.7 U.J / 18 6 4-7 5.4 U. /o 49 18 F2 4-9 6.5 1.01 290 120 Mt. Douglas Park 3-5 4.4 0.53 47 18 Elk Falls Park 8-13 10.6 1.19 53 18 F, 6-10 8.6 0.90 70 20 F2 4-11 7.5 1.22 388 154 Botanie Valley 3-6 4.5 0.64 27 9 Elk Falls Park 8-13 10.6 1.19 53 18 F, 6-9 7.5 0.82 33 11 F2 4-11 7.3 1.39 187 107 erage flower size in interpopulation hybrids was intermediate between the parental averages, except in crosses between the Botanie Valley and Nanoose Hill populations. Both of these populations have small flowers and flower size of F, hybrids was not significantly different from the Botanie Valley parental population. Mean flower sizes of the F2 generations from these interpopulation crosses also were in- termediate between the parental populations, except that the F2 from the cross between the Botanie Valley and Jack Point populations was not significantly different from the Jack Point parental popu- lation. The ranges and standard deviations of flower size in the F2 generations were all greater than in the Fi hybrids (Table 2). This indicates segregation at several loci controlling flower size. It was impractical to attempt to determine the number of loci because flower size is also affected by environmental and developmental factors (even in growth chambers) that tend to obscure discrete size classes and smooth out the size frequency distributions. In particular, flower size on a given plant tends to decrease with age of the plant, so that the first formed flowers are largest. Developmental diflerences in flower size were most prominent in plants with relatively large flowers; those with very small flowers were quite uniform in flower size throughout their life cycle. 68 MADRONO [Vol. 33 Discussion Populations of Collinsia in British Columbia are uniformly tet- raploid and are completely interfertile. Different populations exhibit a range of flower sizes from the smallest reported for C parviflora to the largest C. grandiflora, but most populations in coastal British Columbia are intermediate in flower size. Although it is possible that diploid populations may exist in California (Garber 1 974), chro- mosome number is not correlated with flower size. We do not believe that there is any biological basis for regarding C parviflora and C. grandiflora as separate species. Most genera of Pacific Coast annuals that have been studied biosystematically have well developed prezygotic or post-zygotic interspecific isolating mechanisms. Most recognized species of Collinsia will not form fertile interspecific hybrids (Garber 1974). All reported attempts to cross C. parviflora and C. grandiflora with other species of Collinsia have failed (Ahloowalia and Garber 1961, Garber 1974). We think that the species of Collinsia that form fertile hybrids should be regarded as conspecific varieties (or subspecies) rather than as sep- arate species. The extremes in flower size are conspicuously different and prob- ably represent adaptations for outcrossing in large flowers and self pollination in small flowers. Seed set in unpollinated large flowers isolated from insects was low, but small flowers set full capsules and often self-pollinated before the corolla opened. Populations with flowers of intermediate size probably experience some outcrossing in nature but also are capable of full seed set by automatic selfing. Individual populations, especially those with smaller flowers, are quite uniform in flower size, indicating fixation of alleles at most loci controlling flower size. But, diflferent populations show every degree of intermediacy in the range of flower size in the species. These populations are probably fixed for alleles for large flowers or small flowers at diflerent numbers of loci. No other morphological character is correlated with flower size, and it is clear that all pop- ulations belong to a single species that is variable in flower size. Large-flowered populations are restricted to the Pacific Coast from British Columbia to California west of the Cascades and Sierra Ne- vada (Newsom 1929). Small-flowered populations extend eastward to Manitoba, Michigan, Colorado, and Arizona, as well as occurring along the Pacific Coast. The small-flowered plants have a geograph- ical range much larger than that of the large-flowered plants, and the ranges are largely but not completely allopatric. Therefore, it is useful taxonomically to recognize the widespread small-flowered plants as distinct from the large-flowered plants. A new combination for the large-flowered variety is herein proposed. 1986] GANDERS AND KRAUSE: COLLINSIA 69 Collinsia parviflora Dougl. ex. Lindl. var. grandiflora (Dougl. ex Lindl.) Ganders & Krause, comb, nov. — Collinsia grandiflora Dougl. ex Lindl. Bot. Reg. 13:pl. 1107. 1827. -Type: Garden specimens from seed collected by Douglas on the banks of the Columbia a hundred miles or more from the ocean; the plate fixes application of the name. Collinsia grandiflora Dougl. ex Lindl. var. nana A. Gray, Proc. Amer. Acad. Arts 8:394. 1872. -Type: Oregon, Hall 366 (Ho- lotype: GH!). These rather weak varieties intergrade completely on the west coast, however, and the division between them is arbitrary. Inter- mediate populations are common west of the Cascade-Sierra axis, and some intermediates also occur in eastern Washington and north- ern Idaho. The two varieties can be separated as follows: 1 A. Corollas 4-7 mm long; corolla tube saccate or gibbous, usually declined less than 45° or even almost erect C parviflora var. parviflora IB. Corollas 7-19 mm long; corolla tube usually saccate, usually declined about 45-90° C. parviflora var. grandiflora Acknowledgments This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada and the University of British Columbia. We thank Dr. Ken Carey, Dr. A. J. F. Griffiths, Dr. Katherine Beamish, and Dr. Wilfred Schofield for some of the seed samples. Literature Cited Ahloowalia, B. S. and E. D. Garber. 1961. The genus Collinsia. XIII. Cytogenetic studies of interspecific hybrids involving species with pedicelled flowers. Bot. Gaz. (Chicago) 122:219-228. Garber, E. D. 1956. The genus Collinsia I. Chromosome number and chiasma frequency of species in the two sections. Bot. Gaz. (Chicago) 1 18:71-73. . 1958. The genus Collinsia VII. Additional chromosome numbers and chias- mata frequencies. Bot. Gaz. (Chicago) 120:55-56. . 1974. Collinsia. In R. C. King, ed., Handbook of genetics, vol. 2, p. 333- 361. Plenum Press, New York. Gilkey, H. M. and L. R. J. Dennis. 1 980. Handbook of northwestern plants. Oregon * State Univ. Bookstores, Inc., Corvallis, OR. Griffiths, A. J. P., G. Krause, and F. R. Ganders. 1977. A leaf spot polymorphism in Collinsia grandiflora (Scrophulariaceae). Canad. J. Bot. 55:645-661. Hitchcock, C. L., A. Cronquist, M. Ownbey, and J. W. Thompson. 1 959. Vascular plants of the Pacific Northwest. Part 4. Ericaceae to Campanulaceae. Univ. Washington Press, Seattle. Krause, G. R. 1978. Genetic studies in populations of Collinsia parviflora Dougl. ex Lindl. (Scrophulariaceae). M.S. thesis, Univ. British Columbia, Vancouver. Newsom, V. M. 1929. A revision of the genus Collinsia. Bot. Gaz. (Chicago) 87: 260-301. Peck, M. E. 1961. A manual of the higher plants of Oregon. Binfords and Mort, Portland, OR. 70 MADRONO [Vol. 33 Snow, R. 1963. Alcoholic hydrochloric acid-carmine as a stain for chromosomes in squash preparations. Stain Technol. 38:9-13. St. John, H. 1956. Flora of southeastern Washington and adjacent Idaho. State Coll. Washington Press, Pullman. Taylor, R. L. and B. MacBryde. 1977. Vascular plants of British Columbia, a descriptive resource inventory. Techn. Bull. No. 4. Univ. British Columbia Bot. Card., Univ. British Columbia Press, Vancouver. and G. A. Mulligan. 1968. Flora of the Queen Charlotte Islands. Part 2. Research Branch, Canada Dept. Agriculture Monograph No. 4, Ottawa. (Received 8 Aug 1983; revision accepted 10 Sep 1985.) ANNOUNCEMENT Ash Valley Research Natural Area The Susan ville District of the Bureau of Land Management has re- cently designated 1 1 20 acres of public land in Lassen County, California, as a Research Natural Area. The notification designating the area as the Ash Valley RNA was published in the Federal Register, Vol. 49, No. 236, Thursday, 6 December 1 984. The designation was made to provide protection for a unique assemblage of sensitive plants. The area is to be preserved for research and education purposes and the continued existence of the sensitive plants and their habitat. Present are three plants listed by the U.S. Fish and Wildlife Service as "candidates for listing as threatened or endangered" {Astragalus tegetarioides, Eriogo- num prociduum, and Ivesia paniculata) and three plants listed by the California Native Plant Society (Dimeresia howellii, Draba douglasii var. douglasii, and Erigeron elegantulus). A Habitat Management Plan (HMP) was recently completed by the Bureau of Land Management to establish guidelines for the management and protection of this unique botanical area. The objectives of this plan were to ensure that the sensitive plants and their habitats are protected, to ensure that the area is preserved in its natural state for research and educational purposes, and to generate interest into the scientific research of the area. The Bureau hopes that any knowledge gained from scientific research will aid in the understanding and successful management of this area. Anyone desiring a copy of the Ash Valley Research Natural Area HMP or wishing to use the area for research purposes should contact the District Manager, Bureau of Land Management, 705 Hall Street, Susanville, California 96 1 30. A NEW SPECIES OF LOMATIUM (APIACEAE) FROM SOUTHWESTERN OREGON James S. Kagan The Nature Conservancy, 1234 NW 25th Ave., Portland, OR 97210 Abstract Lomatium cookii, a new species from vernal pools in the Agate Desert area near Medford, Oregon, is described and illustrated. The species appears to be local and rare. It is morphologically most similar to another rare Oregon endemic, Lomatium brads hawii, of the southern Willamette Valley, and to a northern California endemic, Lomatium humile. The Agate Desert region, an approximately 10,000 ha area of the Rogue River Valley north of Medford, Oregon, has long been known for its botanical diversity (Detling 1968). In 1981, during a search for populations of Limnanthes floccosa subsp. grandiflora in the Agate Desert, a new species of Lomatium was discovered. Lomatium cookii J. S. Kagan, sp. nov. Planta perennis, glabris, acaulis vel subacaulis, 1.5-5 dm alta; radice elongata, angusta, 1.5-3 dm longa aut incrassata ob caudicem 2-8 ramosum plantam multicaulem formantem. Foliis radicalibus oblongis 8-17 cm longis petiolo excluso, 2.5-10 cm latis, tematis deinde tripinnatisectis, segmentis ultimis linearibus, acutis, non- nunquam apiculatis, distinctis, 6-12 mm longis, minus quam 1 mm latis; petiolis vaginatis, 5-22 cm longis. Pedunculis foliis longioribus, 1.5-4 dm longis; umbellis 6-12 radiatis, radiis fertilibus 2-9 cm longis, inaequaliter elongatis, radiis sterilibus 1-2 cm longis; invo- lucro nullo; bracteis involucelliorum 8-12, 6-10 mm longis, lineari- bus, viridibus, marginibus scariosis; pedicellis fruitificantibus 1-3 mm longis; floribus flavis. Fructibus oblongis, 8-13 mm longis, 4-6 mm latis, alis lateralibus suberosis, crassis, alis dorsalibus temis, filiformibus, elevatis; vittis plerumque obsoletis (Fig. 1). Plant perennial, glabrous, acaulescent or subacaulescent, 1 .5-5 dm tall; root elongate, narrow, 1.5-3 dm long, simple, occasionally sur- mounted by a thickened 2-8-branched caudex. Leaves oblong, 8-1 7 cm long excluding petiole, 2.5-10 cm broad, ternate then tripin- natisect, ultimate segments linear, acute, or sometimes apiculate, 6- 1 2 mm long, less than 1 mm broad; petioles vaginate, 5-22 cm long. Peduncles exceeding leaves, 1.5-4 dm long; umbels 6-12-radiate, fertile rays 2-9 cm long, unequally elongate, sterile rays 1-2 cm long; Madrono, Vol. 33, No. 1, pp. 71-75, 27 March 1986 72 MADRONO [Vol. 33 Fig. 1 . Lomatium cookii. A. Habit. B. Bractlets of involucel. C. Fruits. D. Fruit cross section. involucre none; involucel bracklets 8-12, 6-10 mm long, linear, green, margins scarious; fruiting pedicels 1-3 mm long; flowers yel- low. Fruits oblong, 8-13 mm long, 4-6 mm wide, lateral wings corky, thick, dorsal ribs three, filiform, elevated; oil tubes obsolete. Type: USA, Oregon, Jackson Co., Agate Desert vernal pools, just nw. of junction of Antelope and Table Rock Roads (T36S R2W S24 nw. 1/4); 380 m elev.; 1 Jun 1983, J. S. Kagan 6018301 (Holotype: ORE; isotypes: UC, OSC). Paratypes. USA, Oregon, Jackson Co., Agate Desert vernal pool margins, off' road to Jackson Co. Sports Park (T36S RIW S22 se. •A), 12 May 1983, /. S. Kagan 5128303, 04, 05 (ORE, OSC); Agate Desert vernal pool margins, just south of Hwy. 140, 4 mi e. of Hwy. 1986] KAGAN: NEW LOMATIUM 73 64 (T36S RIW S23 sw. V4), 13 Apr 1983, /. S, Kagan 4138301 (ORE). Ecology and phenology. Lomatium cookii is found on margins and bottoms of vernal ponds in an area having "patterned ground" topography. The flat landscape is composed of a continuous series of shallow vernal pools surrounded by low mounds. The pools have stony, pebbly bottoms, whereas the mounds are composed of rela- tively rock-free light clay of volcanic origin. The area is an ancient alluvial outwash plain adjacent to the Rogue River. The vernal pools have standing water only in the winter and early spring, from De- cember to April or May, and appear somewhat dry and barren from July to October. Common species (nomenclature follows Munz 1959) of the vernal pools include Deschampsia danthonioides, Alopecurus geniculatus, Plagiobothrys austinae, P. bracteatus, P. nothofulvus, Juncus uncialis, Navarretia leucocephala, N. tagetina, and Lim- nanthes floccosa. The mounds are usually dominated by non-native grasses including Bromus mollis, B. commutatus, Cynosurus echina- tus, Elymus caput- medusae, and Poa bulbosa, but also can have a significant component of native forbs such as Saxifraga integrifolia, Viola douglasii, Brodiaea pulchella, B. hyacinthina, and Lupinus bicolor. This species blooms from mid-March through mid-May, depend- ing on the season. Many plants bloom while in standing water. The plants usually have 2-5 flowering umbels. The earliest inflorescences are largely male, with most of the flowers having undeveloped ova- ries. The later umbels have perfect flowers and produce almost all of the fruits. The only floral visitor observed was a small black moth, as yet unidentified. Lomatium cookii is protogynous, having stigmas that are exposed and receptive to pollen prior to anther dehiscence in the same flower. Mature fruit is present from late May through July. Relationship to other species. There are 2 other species of Lomati- um occurring in the Agate Desert area. Lomatium utriculatum (Nutt.) Coult. and Rose commonly occurs on the mounds, and is easily distinguishable from L. cookii by its obovate involucel bracklets, cauline habit, and thin winged fruits. Lomatium macrocarpum (Nutt.) Coult. and Rose occurs on vernal pool margins, as well as on hillsides at the edges of mounded prairie areas. In the Rogue Valley, it is distinguishable from L. cookii by its pubescent herbage, pale white to tan flowers, and narrow, thin winged fruits. Lomatium cookii is very similar to L. bradshawii (Rose) Math, and Const., a species indigenous to wet prairies of the southern Willamette Valley, Oregon. It is most easily diflerentiated by the narrow, linear involucel bracklets, as opposed to the wider, temately or bitemately divided bractlets of L. bradshawii. In cross section. 74 MADRONO [Vol. 33 the fruit of L. cookii has dorsal ribs that are clearly raised and are 80-120 ixm in diameter, whereas L. bradshawii has ribs that are less distinct, 35-50 iim in diameter, and are imbedded in the thick, stratified cortex. The endocarp of the fruit of L. cookii is three layers thick, whereas in L. bradshawii it is 5-6 layers thick. Plants of L. cookii tend to be shorter but more robust, having more umbels and stems per plant than plants of L. bradshawii. In ungrazed popula- tions, L. cookii has a mean pedicel height of 24 cm in fruit, and averages 4 fruiting umbels per plant (n = 34); whereas L. bradshawii has a mean pedicel height of 3 1 cm, and averages less than 2 fruiting umbels per plant (n = 100). Both species grow from a long, slender taproot. Plants of L. cookii appear more robust because, in about half of the plants observed, the taproot is surmounted by a multiple caudex (2, 4, and 6 are the most common stem numbers). Of the more than 500 plants of L. bradshawii observed, only 6 plants had multiple caudices (Kagan 1980). Lomatium cookii also is similar to Lomatium humile (Coult. & Rose) Hoover. This taxon also is described as Lomatium caruifolium (H. & A.) C. & R. var. denticulatum (Jepson) Jepson (Jepson 1936). Lomatium cookii and L. humile both occur in plains with vernal ponds and have similar fruits and bractlets. Lomatium humile has 4 to 6 dorsal ribs on the fruit, whereas L. cookii has only 2. Also, the fruit of L. humile is inflated in the commissure, whereas L. cookii has fruit that is not inflated. Although the bractlets of these taxa are fairly similar in appearance, those of L. cookii are linear and 6-10 mm long, whereas those of L. humile are orbicular to lanceolate, and usually shorter, 3-5 mm long. The two taxa are most easily distinguished by the leaves, which in L. humile have ultimate segments ranging from 2-60 mm long (Abrams 1951), and averaging about 1 5-20 mm. The ultimate segments of the leaves of L. cookii are much smaller, usually between 6-10 mm long. Distribution and status. Lomatium cookii is both rare and threat- ened (Oregon Natural Heritage Data Base 1985). In thorough searches of the Agate Desert area in 1982 and 1983, only 3 small populations of the species were identified. Fewer than 2000 plants have been observed, and 90% of these were at the type locality. Plants are readily eaten by cattle. Most available habitat has been very heavily grazed, and in these areas plants of this species are not present. In addition, large portions of the Agate Desert, including the type locality, are in the process of being developed for heavy industry, causing additional habitat destruction. Thus, it appears that L. cookii may be a good candidate for the Endangered Species List of the U.S. Fish and WildUfe Service. 1986] KAGAN: NEW LOMATIUM 75 Acknowledgments I have the honor to name the species after Stanton A. Cook, Professor of Plant Ecology at the University of Oregon. I would like to thank Dr. Kenton Chambers, Dr. David Wagner, Dr. Lincoln Constance, and Julie Kierstead as well as the Ma- drono reviewers for their help with this work, and Gaylee Goodrich for the illustration. Literature Cited Abrams, L. 1951. Illustrated flora of the Pacific states. Vol. IIL Stanford Univ. Press, Stanford, CA. Detling, L. E. 1968. Historical background of the flora of the Pacific Northwest. Museum Nat. Hist., Univ. Oregon, Eugene, Bull. No. 13. Jepson, W. L. 1936. A flora of California, Vol. IL California School Book Deposito- ry, San Francisco, CA. Kagan, J. S. 1980. The biology of Lomatium bradshawii (Apiaceae), a rare plant of Oregon. M.S. thesis, Univ. Oregon, Eugene. MuNZ, P. A. 1959. A California flora. Univ. California Press, Berkeley. Oregon Natural Heritage Data Base. 1985. Rare, threatened, and endangered plants and animals of Oregon. Oregon Natural Heritage Data Base, Portland. (Received 23 Nov 1984; revision accepted 10 Sep 1985.) ANNOUNCEMENT Phytologia Memoir VI, A Preliminary, Verified List of Plant Collec- tors in Mexico, is a 179 page booklet containing the names of over 4000 plant collectors of Mexico with some 800 references to their collecting activities. The work should prove of interest to all who have or intend to collect in Mexico. It can be ordered from the compiler. Dr. Irving W. Knobloch, 438 Tuhp Tree, E. Lansing, MI 48823, for $16.00 do- mestic, $17.00 foreign surface mail, and $19.00 foreign air mail (iden- tification number 479-22-0904 A). NOTES Spineless Petioles in Washingtonia filifera (Arecaceae).— The flattened petioles of the desert fan palm, Washingtonia filifera (Lindl.) Wendl. are described as being more or less spiniferous (Munz, A flora of Southern California, Univ. California Press, 1974) or as having spines along either edge (Shreve and Wiggins, Vegetation and flora of the Sonoran Desert, Stanford Univ. Press, 1964). Spines are usually assumed to aid in the protection of plant tissue from herbivore attacks (Louw and Seely, Ecology of desert organisms, Longman Group, 1982). In W. filifera, spine density is greatest around the apical meristem where the emergent, spine-covered petioles are in closest proximity to each other. I have examined over 300 specimens of W. filifera and found that petiole spines are absent or nearly so in individuals exceeding 14 m in height (Fig. 1). The spines also are absent on distal ends of petioles in trees exceeding 8 m in height and cover a decreasing percentage of the petiole in taller trees (Table 1). These tendencies persist regardless of whether the trees are growing under cultivated or natural conditions. If the spines function to deter herbivores from consuming vital plant tissues, then it can be assumed that petiole spines in small to intermediate individuals of W. filifera have evolved as a result of the protection afforded the apical meristem. (Extensive damage to this portion results in death of the tree because Washingtonia does not produce vegetative offshoots as do some palms.) It appears, however, that the utility of this protection decreases or is lost as W. filifera exceeds 1 4 m in height. Palms that develop spines only when small to intermediate may be at a selective advantage if energy for spine development is utilized for a more beneficial trait as the trees become taller and the utility of armament decreases. Other than the infrequent nibbling of seedlings by desert cottontails (Sylvilagus audubonii) and the browsing of leaf tips by domestic cattle (in the three locations where livestock and native palms occur together), no known herbivores consume the foliage of W. filifera today. During past geological epochs, numerous mammals existed within the present range of W. filifera, and some probably browsed upon palms. No known species, however, could have browsed upon palms exceeding 1 2 m in height. The giant camel, Titanotylopus, occurred throughout southern California during the Pliocene and Pleistocene epochs, but had a vertical reach of just 5 m (D. Whistler, pers. comm.). Ground sloths were widespread in western North America during the Pliocene and Pleistocene (Anderson, In Martin and Klein, eds.. Quaternary extinc- tions, Univ. Arizona Press, 1984). The largest sloth known from the southern half of California is the big-tongued sloth, Glossotherium (Stock, Rancho La Brea, Los Angeles Co. Mus. Nat. Hist., 1956). When erect, this sloth could not reach more than 5 m (G. Jefferson, pers. comm.). Mastodons (Mammut) appeared in the Miocene and persisted into the early Holocene in southern California. Some stood 3 m at the shoulder and had a vertical reach of about 6 m (Anderson 1984). Mammoths (Mam- muthus) were the largest terrestrial herbivores to occur in western North America from the Pliocene forward. Their fossils are abundant and widespread throughout the West (Agenbroad, In Martin and Klein 1984). They are believed to have fed upon a tremendous variety of plant material, including palms, but did not have a vertical reach of more than 9 m (G. Jefferson, pers. comm.). Whether or not these herbivores were sympatric with Washingtonia, and thus able to provide the selective pressure behind the development of spined petioles in young palms, is conjecture. No fossils attributable to Washingtonia have been identified. Natural populations of W. filifera are presently confined to the Sonoran Desert of southeastern California, Baja California, and western Arizona (Vogl and McHargue, Madrono, Vol. 33, No. 1, pp. 76-78, 27 March 1986 1986] NOTES 77 Fig. 1 . Leaves of Washingtonia filifera: left, spines on a petiole from a 4 m palm; right, a spineless petiole from a 1 2 m palm. Ecology 47:532-540, 1966; Brown et al., J. Ariz. Acad. Sci. 1 1:37-41, 1976). Axelrod {In Barbour and Major, eds.. Terrestrial vegetation of California, Wiley and Sons, 1977) implies that Washingtonia had a broader range in past epochs than at present. In addition, palm fossils of Pliocene age, assigned to the genus Sabal, occur over much of the southern half of California including what is now the Mojave Desert (Axelrod, Evolution 2:127-144, 1948). These fossils may be misidentified. The taxo- nomic affinity of fossil palm leaves from California is based (in part) upon the presence or absence of spines on the petioles (Axelrod, pers. comm.). As I have shown, however, the petioles of Washingtonia may lack spines and therefore could be assigned erro- neously to the genus Sabal. 78 MADRONO [Vol. 33 Table 1. Percent of Sampled Trees of Washingtonia filifera (by Height Category) with Spines on All or Part of Petioles, n = number of trees in each catergory. Palm ^^^^^ Percent of petiole length covered with spines categones 1 Z 1 (meters) n 100 90 80 70 60 50 40 30 20 10 0 0-4 20 100 4-6 18 72 11 6 6 6 6-8 33 55 6 6 3 3 12 9 3 3 8-10 24 4 8 8 13 8 4 29 25 10-12 43 2 2 2 4 4 2 12 58 12 12-14 34 3 21 50 26 14-16 45 4 42 53 16-18 59 25 75 18-20 29 10 90 Financial support for this research was provided by the Richard King Mellon Foundation of Pittsburgh, Pennsylvania.— James W. Cornett, Natural Science De- partment, Palm Springs Desert Museum, Palm Springs, CA 92263. (Received 18 Feb 1985; revision accepted 16 Oct 1985.) NOTEWORTHY COLLECTIONS California CoRYPHANTHA viviPARA (Nutt.) Britt. & Rosc var. DESERTii (Engclm.) W. T. Marsh. (Cactaceae). — Inyo Co., Funeral Mts., eastern end of Echo Canyon, calcareous flats and slopes, 1300-1600 m, T27N R3E, 18 Jan 1984, Peterson and Annable 1054, 1056 (UNLV). Significance. First record for Inyo Co. and Death Valley National Monument. A range extension of 56 km sw. of Shoshone Mt., Nye Co., NV (Beatley, Vascular plants of the Nevada Test Site and Central Southern Nevada, Technical Information Center, VA, 1976). Stipa scribneri Vasey (Poaceae). — Inyo Co., Cottonwood Mts., near White Top Mt. and w. of Tin Mt., calcareous talus slopes and rock outcrops, 1 900-2 1 00 m, T 1 2- 13S R41-42E, 20 Jun 1981, Peterson 345 (UNLV, UTC), 27 Jun 1982, Peterson 629 (UNLV, UTC), 21 Jul 1982, Peterson 659, 662 (UNLV, UTC). (Verified by M. E. Barkworth.) Significance. First record for CA and a range extension of ca. 200 km w. of the Sheep Range, Clark Co., NV. LiNANTHUS NUDATUS Greene (Polemoniaceae). — Inyo Co., Cottonwood Mts., Hunter Mt., granitic sandy clay loam slopes, 2050 m, T16S R41E, 2 Jun 1981, Peterson 228 (UCSB, UNLV), 25 May 1982, Dunn, Annable and Peterson 530 (UCSB, UNLV). (Verified by D. M. Smith.) Significance. First record for Inyo Co. and Death Valley National Monument. A range extension of 86 km ne. of the southern Sierra Nevada, Tulare Co., CA. — Paul M. Peterson and Carol R. Annable, see notes below. Nevada Carex occidentalis Bailey (Cyperaceae). — Clark Co., Spring Mts., 6 km s. of Charleston Peak, sandy clay loam ridges, 3292 m, T20S R56E SI sw.'A, 19 May 1984, Charlton, Wisura, Annable and Peterson 2460 (DES, UNLV). (Determined by J. H. Lehr.) Significance. First record for Clark Co. extending the range approximately 122 km s. of Pahute Mesa, Nye Co., NV, and 214 km w. of Shivwits Plateau, Mojave Co., AZ. Phacelia barnebyana J. T. Howell (Hydrophylaceae). — Clark Co., Spring Mts., near Lucky Strike Tunnels, calcareous rocky slopes, 1950 m, T19S R57E SI ne.'A, 14 May 1983, Lathrop and Peterson 952 (UNLV). Significance. First record for the Spring Mts. between known locations 54 km s. in the Clark Mts., San Bernardino Co., CA (Thome, Prigge and Henrickson, Ahso 10: 71-186, 1981) and 35 km ne. in the Sheep Range, Clark Co., NV (Ackerman, pers. comm.). MiMULUs montioides Gray (Scrophulariaceae).— Clark Co., Spring Mts., ne. of Grassy Spring, calcareous rock outcrops, 1585 m, T18S R58E S32 se.V4, 14 May 1983, Lathrop and Peterson 960 (UNLV). Significance. First record for the Spring Mts. between known locations 140 km nw. in the Panamint Range, Inyo Co., CA, and 51 km n. in the Spotted Range, Clark Co., NV.— Paul M. Peterson and Carol R. Annable, Dept. Biol. Sci., Univ. Nevada, Las Vegas 89154. MADROf^o, Vol. 33, No. 1, pp. 79-80, 27 March 1986 80 MADRONO [Vol. 33 Oregon DowNiNGiA iNSiGNis Greene (Campanulaceae). — Malheur Co., Barren Valley, s. of Dickinson Ranch, T27S R38-39E, black adobe soil along edge of playa, 1030 m, 17 Jun 1984, E. Joyal 517 (CAS, NY, OSC). Significance. Second report for OR, the first by Cusick in 1885, also from "Barren Valley." This central CA taxon was thought to be extirpated in OR.— Elaine Joyal, see note below. Washington Astragalus geyeri Gray (Fabaceae). — Grant Co., Lower Crab Creek n. of Saddle Mt., Columbia River Natl. Wildlife Refuge, T16N R24E S32, 175 m, stabilized sand dunes, 13 Jun 1984, E. Joyal 488 (BLM-Spokane, NY, OSC, WSU). (Verified by R. Bameby, NY.) Significance. Second report for WA, an extension of ca. 110 km nw. from Wallula Gap, WA-OR, otherwise a Great Basin taxon. Claytonia exigua T. & G. (=Montia spathulata (Dougl.) Howell) (Portula- caceae). -Grant Co., Saddle Mt., above lower Crab Creek, T15N R24E SV2, 300- 400 m, steep basalt talus on n. slope, 15 May 1984, E. Joyal 473 (BLM— Spokane, OSC). (Verified by K. Chambers, OSC.) Significance. Extension ne. of ca. 100 km from the Columbia River Gorge, David Douglas' type collection is from "the vallies of the Rocky Mtns.," an ambiguous locale, possibly in WA or BC. All subsequent reports from WA e. of the Cascades have been in the immediate vicinity of the C.R.G. — Elaine Joyal, Dept. Botany, Oregon State Univ., Corvallis, OR 97331. Utricularia inflata Walt. (Lentibulariaceae). — Kitsap Co. Horseshoe Lk., 15 km s. from Port Orchard, in water near the public fishing area, w. side of the lake, 90 m, 26 Oct 1980, A. and O. Ceska 4913 (K, V, WTU). Sterile specimens, but with typical coiled winter buds. Verified by P. Taylor (K). Significance. The first report from Washington and from w. USA. Previously known from NJ and PL, westward to e. TX. Most probably introduced, but the mechanism of introduction is not known.— Adolf Ceska and Oldriska Ceska, P.O. Box 1761, Victoria, B.C. V8W 2Y1, Canada. REVIEWS California Serpentines: Flora, Vegetation, Geology, Soils, and Management Prob- lems. By Arthur R. Kruckeberg. University of California Publications in Botany, Vol. 78. 1985. $10.95. ISBN 0-520-09701-7 (pbk). Was there field botany in California before "serpentine"? Yes— but as this volume eloquently explains, attempts to understand serpentine-plant relationships have prompted a great deal of important fieldwork and have enriched our systematic, ecological, and evolutionary literature. Arthur Kruckeberg worked on serpentine prob- lems in California during the height of serpentine enthusiasm in the 1950s. His professional interest in serpentine continues, and he is well versed in the international literature on ultramafic research. I can think of no one better qualified to assemble this sort of comprehensive review. Kruckeberg first discusses the slowness of California botanists to appreciate the nature of serpentine habitats. These collectors were attracted to the sterile, rocky slopes that harbored so many endemics, disjuncts, and puzzling distributions. But they didn't say much about the rocks and soils involved. Discussion of calcium- magnesium imbalances and the host of related physiological problems in serpentine soils did not surface in California botanical or agricultural literature until the 1930s. The sections on geology, soil and mineral nutrition, and physiological and mor- phological responses are well developed with more than enough detail for most readers. After discussing ultramafic, serpentinite, peridotite, and other rock and min- eral terms, Kruckeberg reverts to the common and convenient usage of "serpentine" for rock, soil, or vegetation of ultramafic affinity. One uncited reference that is relevant to the physiological response section is J. L. Jenkinson's work on Pinus ponderosa seedlings from serpentine and non-serpentine sources grown on serpentine and non- serpentine soils (U.S. Forest Service Resarch Paper PSW- 127/ 1977). Some botanical readers may pass over the above sections quickly and settle down to see if their favorite serpentine taxa are dealt with properly in the serpentine vege- tation and flora sections (with related appendices). A great deal of information is summarized in helpful ways in this material. One value of such summaries is to prod constructive critics into new fieldwork to check on apparent inconsistencies and omissions. The highlight of this volume may be the Streptanthus case history in the evolu- tionary ecology section. After review of established principles of ecotypic adaptation to serpentine, the principles are applied to Streptanthus, a genus on which Kruckeberg speaks with great authority. This is a most interesting discussion. The volume concludes with sections on exploitation threats and conservation needs. Although California serpentines have been mined for well over a century for mercury and chromium, massive geothermal developments and open-pit mining operations for trace amounts of gold and nickel pose new threats. Not mentioned is a serpentine mineral use that is declining— asbestos. Only two asbestos mines currently operate in the United States, and one is within the New Idria serpentine mass. A defunct asbestos mine nearby is considered one of the greatest "toxic waste" problems in California. —James R. Griffin, Hastings Natural History Reservation, Carmel Valley, CA 93924. Vascular Plants of the Channel Islands of Southern California and Guadalupe Island, Baja California, Mexico. By Gary D. Wallace. Contributions in Science, No. 365, Natural History Museum of Los Angeles County. 1985. 136 pp., $18.00 + $1.80 shipping. (Available from the bookstore. Museum of Natural History, 900 Exposition Blvd., Los Angeles, CA 90007.) Madrono, Vol. 33, No. 1, pp. 81-82, 27 March 1986 82 MADRONO [Vol. 33 A century ago E. L. Greene published his catalog of the flowering plants and ferns of Guadalupe Island, providing the first in a series of studies by California botanists of the interesting floras of the Channel Islands of southern California and of Guadalupe Island, Baja California. Despite the relative closeness of some of these islands to the mainland, they support distinctive floras (and faunas) that exhibit many features of truly oceanic islands. Woodiness occurs in groups that are otherwise commonly herbaceous (as in Eriogonum). Endemics include relic genera (such as Lyonothamnus and Baeriopsis) as well as taxa that appear to have evolved in situ from ancestral immigrants (as in some taxa of Malacothrix, Camissonia). The total vascular flora consists of 758 native taxa (one-fifth of them endemic) and 227 introduced taxa; the abstract is worded to indicate only 621 native taxa. In this volume, Wallace's objectives are to "provide a current guide to the floras . . . which clearly distinguishes between those records based on available herbarium specimens and those based on reports from the literature," to "present a picture of floristic relationships," and to "draw attention ... to the insular distributions of native taxa and to the numerous misidentifications and synonymous treatments of those taxa." The flora is presented in tabular fashion, with occurrences of each taxon (exotics included) noted on individual islands. No descriptions of the taxa are given. Where a literature report is in some question, reference to a literature citation is given (as, for example, Eastwood's 1 94 1 undocumented list of species). However, the insular occurrences of rather few taxa (such as Calyptridium monandrum, Centaurium muh- lenbergii) appear to be based on such questionable reports; rather it is the insular distribution of these taxa that is involved. Appendices list "selected" exsiccatae and their location and present an index to the disposition of synonyms, misidentifications, etc. In the latter index, the nature of the disposition is not clear. Baeria chrysostoma F. & M. = Lasthenia californica DC. ex Lindl., but Baeria macrantha (Gray) Gray, listed in identical format does not. This latter species does not occur on these islands, and its record there represents a misidentification of L. californica. Wallace's phytogeographic analysis is of particular interest to insular biogeographers (and should be read in conjunction with Johnson et al., American Nat. 102:297-306, 1968). The islands of the northern group (