PaleoBios, Volume 16, Number 3, July 4, 1994 AMELANCHIER HAWKINSAE SP. NOV. (ROSACEAE, MALOIDEAE) FROM THE MIDDLE MIOCENE OF STEWART VALLEY, NEVADA, AND A REVIEW OF THE GENUS IN THE NEVADA NEOGENE Howard E. Schorn and Nancy L. Gooch Museum of Paleontology, University of California, Berkeley, CA 94720 ABSTRACT Amelanchier hawkinsae sp. nov. grew in west-central Nevada from approximately 15-12 million years ago. The earliest records of the species at Purple Mountain and Stewart Valley coincide with a widespread global change in vegetation and floristics that took place at approximately 15.0 - 14.5 Ma, at what corresponds to the initiation of the Seldovian- Homerian transition. In the western conterminous United States this change brought on relatively drier conditions that were more pronounced earlier in the southwestern United States than to the north. During this period of change it is inferred that A. hawkinsae dispersed into the Nevada region, replacing the earlier A. magnifolia var. desatoyana (Axelrod) comb. nov. Prior to the 15.0-14.5 Ma period of change, A. magnifolia var. desatoyana in Nevada was contemporaneous with A. magnifolia Arnold var. magnifolia at Succor Creek to the north in Oregon. After the change, when A. hawkinsae occupied west- central Nevada, the A. magnifolia-comp\ex continued on in Oregon and Idaho where conditions were more mesic than in contemporary Nevada. INTRODUCTION Amelanchier Medicus is a shrub or small tree with simple, pinnately veined, toothed, deciduous leaves. It is related to such plants as mountain ash (Sorbus), apple (Malus), toyon (Heteromeles) and hawthorn (.Crataegus). Jones (1945) allies the genus most closely with Malocomeles Decaisne and Peraphyllum Nut tall. Amelanchier is placed with these and other genera in the Subfamily Maloideae (= Pyroideae or Pomoideae ) of the Family Rosaccae. Extant species of Amelanchier are distributed in Europe, northern Africa, eastern Asia and North America (Jones, 1946; Landry, 1975). In North America the various species are commonly referred to as serviceberry, juneberry, shadberry, shadbush, and saskatoon, among others. Jones (1946) gives an illuminating account of the derivation of these common English names. The Native American Shoshone name for the western serviceberry (A. alnifolia Nuttall) is duh-hee yemba (Train et al., 1941, p. 33). This serviceberry grows abundantly in the Wassuk Range just west of Hawthorne, Nevada, and we have noted it along the drainage courses in the pinyon-juniper belt of other nearby areas such as the Shoshone Range at Ichthyosaur State Park, Nevada. Partitioning of species in Amelanchier remains unsettled. Hybridization among species is common (Christopher et al., 1985). Jones (1946) recognized nineteen species and two varieties in North America. More recently, Landry (1975) reviewed the genus throughout its range and reduced the number of species to six, with five subspecies and seven varieties. Such divergent views indicate that the selection of characters for diagnosis is difficult because there are so few non-overlapping distinctive features (Landry, 1975) useful in analysis. The fossil record of Amelanchier in North America consists of leaf impressions (with two exceptions: MacGinitie, 1933, a queried flower; Axelrod and Ting, 1960, questionable pollen). The oldest record of the genus is from early middle Eocene (approximately 50 Ma) rocks at the One Mile Creek locality, Princeton, British Columbia (Wolfe and Wehr, 1988). With the numerous, large and diverse fossil collections made available in the last 50-75 years, and the ever increasing accuracy of radioisotopic dating methods (Swisher, 1992) and tephrochronologic methods (Michael E. Perkins, oral comm., 1993; Sarna-Wojcicki et al., 1992), it is now possible to look at the evolution and biogeographic distribution of taxa in considerable detail, and thereby hopefully derive more accurate inferences about the causes for the changes we recognize. The purposes of this report are 1) to describe a new species of Page 2 Amelanchier Schorn and Gooch Amelanchier from Nevada, 2) to assess the taxonomic status of the fossil record as it relates to the Nevada occurrences, and 3) to provide a general interpretation of the diversification, distribution and paleoecological history of Amelanchier during the Neogene in Nevada, Idaho and Oregon. Material. The material used to circumscribe the new Hawkins Serviceberry is from Stewart Valley, Mineral County, Nevada (Fig. 1). Forty-seven leaf impressions are available from three main sites in the valley. Preservation of the impressions ranges from those essentially equivalent in detail to cleared leaves of extant plants, to others that do not show venation below the third order. 120 IIS Figure 1. Location of Stewart Valley (*) and other floras discussed in text. The terminology we use in the leaf diagnoses and descriptions follows Hickey (1979). The first part of a diagnosis is presented as if the leaves were being scored for C.L.A.M.P. analysis (Wolfe, 1993). We have altered the character sequence presented by Wolfe (1993, figure 10, p. 26) to correspond with the sequence we are using to describe leaf features. The leaf impressions in Stewart Valley are closely associated with pollen, impressions of fish and an abundant, diverse and excellently preserved impression insect paleofauna. This diverse assemblage was initially discovered by John E. Mawby , Donald E. Savage and S. David Webb in 1959 during their field studies for the Department of Paleontology, University of California, Berkeley (Wolfe, 1964). The potential significance of this unique paleontological area was recognized by Wolfe (1964), and subsequently, largely under the impetus of Harvey I. Scudder (California Academy of Sciences, San Francisco), the valley is now recognized as a natural treasure. With the help of such local residents as Allen E. Conelly, Frances Hawkins and William M. O'Malia, and Bureau of Land Management staff members Tom Abbett, Linda Armentrout, Brian Hattoff, Norman Melvin and Thomas Owen, a portion of the valley has been set aside as an Area Of Critical Environmental Concern. A resulting Congressional Order withdrew this area from any further development that would alter or destroy this unique paleontological heritage. The continued interest and support of Bonnie Templeton (Director) and Pat Barker (State Archaeologist) of the Nevada State Bureau of Land Management, and District Archaeologists Patrica A. Hicks and Prill E. Mecham, continues to assure the future of this heritage. GEOLOGIC OCCURRENCE AND AGE OF THE TYPE MATERIAL The Stewart Spring flora (Wolfe, 1964) occurs in laminated to thin-bedded siliceous shales that formed as diatomaceous sediments in the deeper, anoxic part of a chemically stratified lake. The lake was approximately 6.5 km wide, east-west, and 16 km long at its maximum extent. The major drainages that brought leaves into the lake came from an easterly direction (Stewart, 1992). The plant-bearing shales are part of a sequence of lacustrine and fluvial sedimentary rocks 400-500 m thick (Schorn et al., 1989). Radiometrically dated volcanic rocks as well as fossil vertebrate horizons occur both below and above the plant horizon. A potassium-argon date from volcanic breccias that interfinger with the lowermost beds of the sedimentary sequence approximately 80-90 m below the Stewart Spring paleoflora is 15.1 Ma (biotite) or 15.4 Ma (plagioclase) (Morton et al., 1977). An 39Ar/40Ar age (Swisher, 1992) from approximately 60 m below the flora is 14.96 ± Schorn and Gooch Amelanchier Page 3 0.24 (biotite) and 14.86 ± 0.53 (plagioclase). Thin (2-3 cm thick) air-fall ash beds in the diatomite approximately 80 m above the Stewart Spring paleoflora are correlated with the Grouse Canyon tuff of the Nevada Test Site and are dated at 13.7 Ma (Michael E. Perkins, oral comm., 1993). An air-fall tuff approximately 330 m disconformably above the Stewart Valley plant beds is assigned to the 11.6 Ma Rainier Mesa Tuff event of the Timber Mountain Caldera complex in Nye County, Nevada (Michael E. Perkins, oral comm., 1993). This is the bed Evernden et al. (1964) dated at 10.7 Ma (11.0 Ma recalculated by the new constants provided by Dalrymple, 1979), and the bed that provided an 39Ar/40Ar date by Swisher (1992) of 11.743 ± 0.032 Ma. An early Barstovian palcofauna approximately 60 m below the plant-bearing shales (Donald E. Savage and Carl C. Swisher III, unpublished data), and an Early Clarendonian palcofauna (Mawby, 1965) disconformably above the plant beds give independent age control and support the radiometric dates and tephrochronology that indicate the Stewart Valley paleoflora is approximately 14.1 Ma. SYSTEMATIC PALEONTOLOGY Family ROSACEAE Genus AMELANCHIER Medicus Amelanchier haivkinsae Schorn et Gooch sp. nov. Plate 1, Figures 1-8,10 Amelanchier apiculata auct. non Brown. Axelrod, 1956, Univ. Calif. Publ. Geol. Sci., v. 33, p. 298 pro parte, pi. 8, fig. 9, and pi. 14, figs. 9, 11 only (not pi. 14, fig. 10 and homeotype no. 4138 which are indeterminate). Symphoricarpos wassukana Axelrod, 1956, ibid., v. 33, p. 312 pro parte, homeotype no. 4178 only (not pi. 9, fig. 2 which is Symphoricarpos, or pi. 9, fig. 3 which is Robinia). Amelanchier alvordensis auct. non Axelrod. Axelrod, 1962, ibid., v. 39, p. 234, pi. 49, figs. 1-4. Amelanchier cusicki auct. non Fernard. Wolfe, 1964, U. S. Geol. Survey Prof. Paper 454-N, p. 24, pi. 10, fig. 9. Diagnosis. - A species of Amelanchier with leaf shape elliptic to ovate; lamina length from 0.6- 3.1 cm, average length approximately 2.0 cm; length to width ratio 1:1 - 1:3; leaf size class ranges from leptophyll 1 to approximately microphyll 1; apex acute to obtuse; base acute to rounded to slightly cordate; teeth regularly spaced, acute, simple (very rarely doubly serrate), apically directed and typically restricted to the apical one half or less of the lamina; typically 6-8 secondary veins that depart the midrib at approximately 45° in the medial part of the lamina; vein which innervates the tooth passes through the apical (upper) 1/2 of the tooth lamina; 5-7 tertiary veins per intersecondary area in medial part of lamina, tertiaries typically oriented from 60-90° to the secondaries. Description. - Holotype, UCMP Paleobot. No. 1; lamina 3.1 cm long and 1.9 cm wide, shape elliptic; petiole 1.3 cm long; 8 simple teeth per side, restricted to the apical two thirds of the lamina; 9 or 10 arcuate, apically-directed secondaries, the basal 2 camptodromous, the remaining semicraspedodromous to craspedo- dromous, vein innervating tooth passes through the apical (upper) 1/2 of the tooth lamina; the secondary veins, especially in the medial part of the lamina are thicker at their point of origin with the midrib and distinctly taper distally; the tertiary veins are oriented from approximately 60 to 90° to the secondaries; 5 to 7 well developed tertiaries per intersecondary area in the medial part of the lamina; the fourth order venation forming fairly regular quadrangular mesh; fifth order veins a freely ending simple dendritic pattern (Plate 1, Figure 10). Etymology. - The species is named for Ms. Frances Hawkins of Hawthorne, Nevada. Frances has been a leader in the organization, implementation and growth of the Mineral County Museum. Her generous and continued public support for the community of Hawthorne and for the preservation of the Stewart Valley Fossil Area is deeply appreciated. Discussion. - The holotype specimen (Plate 1, Figure 1) was chosen because it one of the best preserved and most complete leaves available. It is somewhat atypical of the suite of leaves because of the tooth positioned in the basal one half of the lamina, but this expresses the known range of that character. Wolfe (1964) assigned the earlier collections from Stewart Valley to A. cusicki Fernard. Indeed, the shape of the lamina and teeth are quite similar to those of this extant species. However, with new and larger collections it can Page 4 Amelanchier Schorn and Gooch be shown that the restriction of the teeth to the apical part of the lamina is more nearly that of extant plants assigned to A. basalticola Piper and A. pallida Greene. These two species are closely related (Jones, 1946; Landry, 1975). All three extant species, however, have the tertiary veins oriented more nearly perpendicular to the midrib. Considering the controversial taxonomic status of the extant members of the genus and the distinct combination of characters present in the fossils, the fossils are recognized as a new species. Amelanchier hawkinsae is known only from west-central Nevada during the interval between approximately 15-12 Ma. This species is also present at the unpublished Purple Mountain flora located in the eastern reaches of the Truckee River Gorge (Daniel I. Axelrod, oral comm., 1990; Axelrod, in press). Occurrence. - Stewart Valley, (holotype UCMP Paleobot. No. 1, paratypes UCMP Paleobot. Nos. 2-9), Chloropagus, Purple Mountain, Aldrich Station and Chalk Hills, Nevada. DIVERSIFICATION AND PALEOECOLOGY OF THE NEOGENE AMELANCHIERS FROM NEVADA Ten characters (Table 1) were used to compare Amelanchier hawkinsae with other fossil species (discussed below). A detailed survey of the leaf architecture of the genus has not been done; in large part because of the controversial taxonomic status of extant Amelanchier. However, as a generalization, the majority of extant and fossil "species" have leaves that are typically 3-5 cm long, with tertiary veins that are relatively numerous, well organized and oriented essentially perpendicular to the midrib; conditions that appear to be typical of the less derived members of the Rosaceae. These character states are inferred to be ancestral within Amelanchier. If this suggestion is accepted, then A. magnifolia var. desatoyana (see below) is the least modified form of the fossil taxa under consideration (Table 1; Fig. 2). Amelanchier magnifolia var. desatoyana and A. magnifolia var. magnifolia differ primarily in size (Table 1); the lamina length of A. magnifolia var. desatoyana averages approximately 3 cm long, whereas A. magnifolia var. magnifolia averages approximately 5.5 cm long. Both forms are part of the A. magnifolia- complex (Fig. 2). Graham (1965) suggested that this larger leaf size indicates that more mesic conditions existed at Trout Creek than at other localities where leaves of A. magnifolia var. magnifolia have an average smaller size (approximately 3-5 cm). However the smaller- leaf forms at Thorn Creek, Succor Creek and Hidden Lake are associated with vegetation that indicates these areas were more mesic than either Trout Creek or Hog Creek (Univ. Idaho, loc. no. P15) where the larger Amelanchier leaves occur. We suggest an alternative explanation to that of Graham (1965). At Trout Creek, the type locality for A. magnifolia var. magnifolia, the species is represented by 215 (3.6%) of the 6000 specimens collected (Graham, 1965, p. 40). Although well below the four (more probably two) forms of coreaceous evergreen oak leaves that make up 65% of the collection, Amelanchier is the fifth (more probably the third) most abundant species in the Trout Creek paleoflora. The relative abundance at Trout Creek suggests that these plants occupied habitats immediately along the northward facing drainage courses leading into the lake at the southern rim of the White Horse Caldera (Rytuba and McKee, 1984). The large leaf size further suggests these habitats were deeply shaded. If this interpretation is accepted, the large-leafed forms of A. magnifolia var. magnifolia can be viewed as a segregate of the A. magnifolia complex adapted for high shade tolerance. Leaves of both the Nevada serviceberrics average smaller in size than the northern A. magnifolia var. magnifolia or A. alvordensis (see below). Leaves of A. hawkinsae average approximately 2.0 cm long. The reduction in leaf size, restriction of the teeth to the apical half of the lamina and changes in venation of A. hawkinsae are inferred to be adaptations to the less mesic and more open habitats that became available in Nevada after approximately 15- 14.5 Ma. The morphology of A. hawkinsae appears to be so highly derived that it is difficult to infer its ancestry; was it derived in place from the A. magnifolia var. desatoyana, or was it an immigrant from the north or east probably related to the A. alvordensis lineage? From what is known of the record it appears the ancestral stock of A. hawkinsae could have been either the A. alvordensis or A. magnifolia var. desatoyana groups. The features that distinguish A. hawkinsae from other species/varieties are 1) thicker Schorn and Gooch Amelanchier Page 5 Table 1. Comparison of Amelanchier kawkinsae, A. alvordensis, A. magnifolia var. desaloyana and __________A. magnifolia var. magnifolia.____________________________________________________________ Character/Character State A. hawk. .......Species/Variety................................ A. alvord. A. magni. v. desatoy. v. magni. (character: + = present, o = absent) A. Orientation of tertiary veins 1. = perpendicular to midrib 0 + (apical) + + 2. = perpendicular to secondaries + + (basal) 0 0 B. Number of tertiary veins per inter-secondary in medial part of lamina 3. typically 12-14 0 0 0 + 4. typically 8-10 0 + + 0 5. typically 5-7 + 0 0 0 c. Organization of tertiary veins 6. relatively regular 0 0 + + 7. relatively irregular + + 0 0 D. Thickness of secondary veins in medial portion of lamina 8. relatively thin 0 f + + 9. relatively thick + 0 0 0 E. Angle of secondary departure from midrib in medial portion of lamina 10. angle typically < 45° + 0 + + 11. angle typically > 45° 0 + 0 0 F. Tooth-vein relationship 12. vein enters in basal to medial half of tooth 13. vein enters in medial 0 c + + to apical half of tooth + t 0 c G. Position of teeth (typically) 14. near base to apex of lamina 0 - + + 15. apical 1/2 or less of lamina + 1 ' + H. Tooth shape (after Hickey, 1979) 16. convex-concave (A 3) 0 0 0 + 17. concave-convex (C 1) 0 • 0 0 18. straight-convex (B 1) + + - + 19. straight-acuminate (B 4) ( + - o 20. convex-convex (A 1) o + 0 0 1. 21. acuminate-acuminate (D 4) Base + 0 0 0 22. obtuse to rounded + + - + J. 23. cordate Length of lamina + + 0 0 24. > 4 cm 0 + + + 25. 4-3 cm 0 + + 0 26. < 3 cm + + + 0 K. Average length of lamina in cm = 2.0 = 3.5 = 3.0 = 5.5 Page 6 Amelanchier Schom and Gooch secondary veins with somewhat more irregular courses, 2) a reduced number of functional tertiary veins, 3) more irregularly organized tertiaries, 4) tertiaries oriented more nearly perpendicular to the secondaries, and 5) the virtual elimination of the sixth order venation. These features are correlated with each other and with a reduction in leaf size, and are thought to represent adaptations to a more xeric environment. Although it appears that A. hawkinsae could have been derived from either the A. magnifolia var. desatoyana or A. alvordensis stock, it is (phylogenetically) more parsimonious to conclude that A. hawkinsae is derived from the A. alvordensis lineage (Fig. 2). Both A. hawkinsae and A. alvordensis share the presumably derived character states of basal tertiaries perpendicular to the secondaries and more irregular tertiaries. However, in accepting this conclusion, we are assuming that A. alvordensis or its descendants have records between approximately 23-15 Ma that are not yet discovered or have not been preserved (Fig. 2). Features such as size and shape of the lamina, number of secondaries or tooth shape are more variable and overlap among species (Table NEVADA Chalk Hills Aldrich Station Chloropagus Stewart Valley Purple Mtn. Kurpie Mtn. mm r- Buffalo Canyon Eastgate A. hawkinsae Ma OR EGON-IDAHO 1 2-- 1 3-- X ,«tf ,\e* T [31 A. mag. desatoyana 1 5-- 1 6-- 1 7-- A. mag. magnifolia .J . — --7 / / .--"7 Hog Creek, ID Trout Creek, OR Thorn Creek, ID Hidden Lake, OR Succor Creek, OR 22 23 -- A. i i alvordensis .--7 Alvord Creek, OR 31 -- 32-- I c~"~\ 1 - i ¦ _ —— 7 / / __.- _-"""""I ¦-n i A. gray! Gray Ranch, OR Figure 2. Inferred relationships and distribution of Amelanchier from Nevada, Oregon and Idaho. Schorn and Gooch Amclanchier Page 7 1). The organization of the tertiary venation is more consistently distributed within the genus and is, therefore, more important for indicating relationship. The tertiary venation of Amelanchier alvordensis is both perpendicular and oblique to the midrib; typically perpendicular in the apical 1/2 - 2/3 of the lamina and oblique (i.e., more perpendicular to the secondaries) in the basal 1/3 - 1/2 of the lamina. Also, the vein innervates the tooth in the apical 1/2 of the tooth in both A. hawkinsae and A. alvordensis. Amelanchier hawkinsae could have been derived from the A. alvordensis- type leaf by: 1) a more robust development of the secondaries, 2) change in orientation of all tertiary venation perpendicular to the secondaries, 3) a reduction of the number of tertiary veins per intersecondary area, 4) the loss of, or very rare occurrence of, doubly-serrate teeth, 5) a general overall reduction in leaf size and 6) a change in shape to a more narrow, elliptic-ovate shape. The close relationship between A. hawkinsae and A. alvordensis is most strongly supported by the organization of the tertiary veins and the tooth-vein relationship (characters C and F in Table 1). With the change toward drier climates that occurred in western North America at approximately 15-14.5 Ma, (concurrent with the initiation of the Seldovian-Homerian transition) (Wolfe and Tanai 1980; Schorn, 1984, 1986; Axelrod, 1985, 1991, 1992; Axelrod and Schorn, 1994), we infer that A. hawkinsae dispersed into the west-central Nevada area and replaced the earlier, more mesic, forms of A. magnifolia var. desatoyana (Fig. 2). During the same time that A. hawkinsae occupied west-central Nevada, the A. magnifolia-comp\vx continued on in the paleofloras of Oregon and Idaho where conditions were relatively more mesic than in contemporary Nevada (Axelrod, 1992; Axelrod and Schorn, 1994). In drier areas such as Trout Creek, Oregon (Graham, 1965) and Hog Creek, Idaho (this report), A. magnifolia var. magnifolia is represented by its large-leaved, shade adapted forms (Fig. 2). ADDITIONAL TAXONOMIC NOTATIONS AND REVISIONS OF FOSSIL AMELANCHIER The following taxonomic revisions are necessary for the preceding discussion of the inferred relationships and Ncogene history of Amelanchier in Nevada. The published record and some new and undescribed species are reviewed, but they are included in these revisions only as they have direct bearing on the Nevada material. Amelanchier alvordensis Axelrod Plate 1, Figure 11 Amelanchier alvordensis Axelrod, 1944, Carnegie Inst. Wash. Publ. 553, p. 255, pi. 44, fig. 3. Rosa alvordensis Axelrod, 1944, ibid., 553, p. 259, pi. 44, fig. 5. Ceanothus precuneatus auct. non Axelrod. Axelrod, 1944, ibid., 553, p. 261 pro parte, pi. 45, fig. 3 only (not pi. 45, fig. 2 which is Robinia sp.). Arbutus idahoensis auct. non (Knowlton) Brown. Axelrod, 1944, ibid., 553, p. 262, pi. 45, fig. 5. Discussion. - Neogene material from Nevada that was formerly referred to A. alvordensis (Axelrod, 1962) is reassigned to A. hawkinsae. The figured holotype of A. alvordensis is not typical of the large suite of leaf impressions now available from Alvord Creek, Oregon (H. W. Meyer and HES, unpublished data); the leaves range from orbicular with teeth from the base to apex, to those that are typically obovate. Doubly serrate teeth are also common in the suite. Wolfe (1964) and Axelrod (1991) discussed the apparent close relationship of A. alvordensis and A. grayi Chaney (1927) from the Oligocene Bridge Creek paleoflora (Gray Ranch locality). We concur with these authors that the two specimens from Gray Ranch are similar to A. alvordensis. Recent collecting and renewed research in the area by Manchester and Meyer (1987, and oral comm., 1992) recovered additional specimens that support the suggestion of the close morphological similarities of the two species. However, A. grayi has a greater number of tertiaries that arc more regularly organized, and thus A. grayi and A. alvordensis are considered (Fig. 2) distinct species (H.W. Meyer, oral comm., 1992). Further preparation of the holotype of Rosa alvordensis Axelrod from Alvord Creek (Axelrod, 1944) exposed a long petiole (J. A. Wolfe, oral comm., 1964). This specimen, the additional unfigured paratype of Rosa alvordensis, the leaf referred to Arbutus idahoensis (Axelrod, 1944), and one specimen (Axelrod, 1944, pi. 45, fig. 3) referred to Ceanothus, all have the venation and tooth type of Amelanchier alvordensis and are transferred to that species. Page 8 Amelanchier Schom and Gooch The record of A. alvordensis from the late Eocene at Copper Basin (Axelrod, 1966) is recommended for rejection. Both specimens are poorly preserved and incomplete. The margin and the venation below the secondaries of the leaf figured on plate 16, figure 11 (Axelrod, 1966) cannot be determined. The other leaf (pi. 16, fig. 12) is also difficult to evaluate; the preservation is very poor and it appears to lack both base and apex. The Alvord Creek paleoflora was originally considered early Pliocene (Axelrod, 1944) [= late Miocene as that Epoch is presently recognized]. Evernden and James (1964) dated an andesite directly above the plant-bearing shales at 21.3 Ma. The contact between the dated andesite and underlying shales appears conformable but Minor et al. (1987) and field work by H.W. Meyer and HES show the dated andesite to be part of the Steens Mountain Volcanics that is unconformable with the plant-bearing shales of the Alvord Creek Formation. Dates from the plant beds (23.8 ± 1 Ma; Barnett and Fisk, 1984) and from the slightly younger, or in part contemporaneous, Pike Creek Formation (21.8 ± 0.7 Ma, approximately 22.4 Ma, 23.6 ± 0.7 Ma; Minor et al., 1987) indicate Amelanchier from Alvord Creek should be considered transitional Oligo- Miocene. Occurrence. - Alvord Creek, Oregon. Amelanchier apiculata Brown Amelanchier apiculata Brown, 1949, Wash. Acad. Sci., Jour., v. 36, p. 226, figs 6, 7. Discussion. - This species is known from only two leaf impressions from the late Miocene at Cache Valley, Utah. Brown (1949) stated they closely resemble the leaves produced by A. utahensis Koehne and A. alnifolia Nuttall in the drier parts of their ranges. The tertiary venation of A. apiculata is perpendicular to the midrib (see Brown, 1949, fig. 7). The material from Chloropagus and Aldrich Station that was referred to this species (Axelrod, 1956) has tertiaries perpendicular to the secondaries and is transferred to A. hawkinsae. Occurrence. - Cache Valley, Utah. Amelanchier magnifolia Arnold var. magnifolia Plate 1, Figure 12 Amelanchier magnifolia Arnold, 1937, Univ. Michigan, Contr. Mus. Paleont., v. 5, p. 89, pi. 4, figs. 1, 4. Amelanchier grayi auct. non Chaney. MacGinitie, 1933, Carnegie Inst. Wash. Publ. 416, p. 58. Amelanchier dignatus auct. non (Knowlton) Brown. H. V. Smith, 1938, Michigan Acad. Sci., Papers, v. 23, p. 228. Amelanchier couleeana auct. non (Berry) Brown. Chaney and Axelrod, 1959, Carnegie Inst. Wash. Publ. 617, p. 183 pro parte, pi. 36, fig. 1 only (not homeotype 5121 which is Ulmaceae, cf. Zelkova). Graham, 1963, Amer. Jour. Bot, v. 50, p. 930. Graham, 1965, Kent State Univ. Bull., Res. Ser., No. 9, p. 89, pi. 15, figs. 1, 2, 4, 5. Diagnosis. - A species of Amelanchier with leaf shape elliptic to obovate; lamina length from 4.5 - 10.0 cm, average length approximately 5.5 cm; length to width ratio approximately 1.5:1 - 0.0:1; leaf size class ranges from microphyll 1 to microphyll 3; apex acute; base rounded; teeth regularly spaced, acute, simple (more rarely doubly serrate), apically directed and extending from near the base to the apex; typically 7-9 secondary veins that depart the midrib in the medial part of the lamina at approximately 45°; vein which innervates the tooth passes through the basal (lower) 1/2 of the tooth lamina; 10-14 tertiary veins per intersecondary area in medial part of lamina, tertiaries typically oriented from 60-90 degrees to the midrib. Discussion. - The type material of A . magnifolia from Trout Creek (Arnold, 1937) was synonymized with A. couleeana by Brown (1946). The different tertiary venation in these two taxa (Fig. 3) negates this synonymy. With our transfer of "Amelanchier" couleeana back to Phyllites couleeanus Berry (see below) it is necessary to reinstate the species epithet, magnifolia. All leaves listed in the above synonymy share a similar lamina and tooth shape, relatively thin secondaries and tertiary veins oriented nearly perpendicular to the midrib. They range in size from lamina approximately 6- 10 cm long at Trout Creek, Oregon (Arnold, 1937) down to smaller forms approximately 4.5 cm long at Hidden Lake, Oregon (Wolfe, oral, comm., 1991). Other than these differences in size, which we infer to reflect responses to local ecological conditions, this assemblage of fossils is considered a single species. In their discussion of A. couleeana (= A. magnifolia), Chaney and Axelrod (1959) referred two additional specimens to that species: Betula Schom and Gooch Amelanchier Page 9 bendirei Knowlton of Smith (1938b) and Phyllites dadrastrifolia Arnold (1937). We do not believe that either specimen should be assigned to Amelanchier. The specimen of Smith (1938b) is Betula by nature of the compound teeth. The poor preservation of Phyllites cladrastifolia (Arnold, 1937, p.101) suggests it is better left Incertae Sedis. A B Figure 3. Comparison of venation in (A) Phyllites couleeanus and (B) Amelanchier magnifolia var. magnifolia, both approximately 2X.________________ The record of Amelanchier from Mascall, Oregon (Chaney and Axelrod, 1959) is recommended for rejection. Wolfe (1964, p. 24) transferred the leaf referred to A. coveus (Chaney and Axelrod 1959, pi. 36, fig. 2, UCMP 2992) to the category of indeterminate specimen. The leaf from that site that Chaney and Axelrod assigned to A. couleeana (UCMP homeotype 5121) is also recommended for rejection; the straight secondaries and urticoid tooth type (Hickey and Wolfe, 1975) place this leaf in the Ulmaceae, cf. Zelkova. Amelanchier is present at Thorn Creek, Idaho (Smith, 1941; UCMP locality P4600), but the homeotype material from that locality that Chaney and Axelrod (1959) referred to A. coveus is rejected; their unfigured homeotype 3261 has numerous small teeth and camptodromous venation (= IPrunus), homeotype 3262 has glandular salicoid teeth (= IPopulus), and homeotype 3263 is missing. The two available specimens are so battered and incomplete that the queried generic assignments should be considered only possible affinities. As presently understood, A. magnifolia var. magnifolia is restricted to Idaho and Oregon where it occurs in what was a more elevated U- shaped paleofloristic province that bordered the lower elevation "Columbia River Basalt" province to the east, west and north within the 'U'. At its southern area the U-shaped "Paleocascade-Idaho" province passed southward into the still higher elevation "Paleonevada" province (Wolfe, 1969; Axelrod, 1985, 1991). Occurrence. - Hidden Lake, Trout Creek, Oregon; Succor Creek, Oregon and Idaho; Hog Creek and Thorn Creek, Idaho. Amelanchier magnifolia Arnold var. desatoyana (Axelrod) Schorn et Gooch comb. nov. Plate 1, Figure 9 Amelanchier desatoyana Axelrod, 1991, Univ. Calif. Publ. Geol. Sci., v. 135, p. 56, pi. 16, figs. 1-7. Amelanchier grayi auct. non Chaney. Axelrod, 1985, Univ. Calif. Publ. Geol. Sci., v. 129, p. 159, pi. 28, fig. 2; pi. 30, figs. 2, 3. Diagnosis. - A species of Amelanchier with leaf shape elliptic to obovate; lamina length from 1.5 - 4.3 cm, average length approximately 3.0 cm; length to width ratio 1.3:1 - 1.8:1; leaf size class ranges from leptophyll 2 - mierophyll 1+; apex acute to obtuse to rounded; base acute to rounded; teeth regularly spaced, acute, simple, apically directed and extending from the basal 1/3 of the lamina to the apex; typically 7-9 secondary veins that depart the midrib in the medial part of the lamina at < 45°; vein that innervates the tooth passes through the basal (lower) 1/2 of the tooth lamina; 8-14 (tipically 8-10) tertiary veins per intersecondary area in medial part of lamina, tertiaries typically oriented from 60-90 degrees to the midrib. Discussion. - This form is considered a variety of A. magnifolia, rather than a distinct species. The two forms share so many overlapping character states (Table 1) that it seems best to express their close relationship by assigning the Nevada material to varietal status. The variety differs from A. magnifolia var. magnifolia primarly by its smaller size, a feature noted by Axelrod (1985, p. 160), and because it is typically more ovate. This group of closely related forms can be considered the A. magnifolia-complex (Fig. 2). Amelanchier magnifolia var. desatoyana overlaps with the overall leaf shape of A. hawkinsae, but the two are differentiated by the nature of their secondary and higher order venation (Plate 1, Figures 8 and 9). In the A. magnifolia-complex the secondaries are Page 10 Amelanchier Schom and Gooch relatively thin, the tertiary venation is oriented essentially perpendicular to the midrib and there are typically 8-14 tertiary veins per intersecondary area in the medial portion of the lamina; in A. hawkinsae the secondaries are relatively thick, the tertiary veins are oriented primarily perpendicular to the secondaries (even in the apical portion of the leaves) and there are typically only 5-7 tertiary veins per intersecondary area in the medial portion of leaf. The age of A. magnifolia var. desatoyana requires discussion. This variety occurs in the Middlegate Formation at the Eastgate locality (Axelrod, 1985) and in the Buffalo Canyon Formation (Barrows, 1971) at the Buffalo Canyon locality (Axelrod, 1991). The most current published age assignments given for the Eastgate and Buffalo Canyon sites are by Axelrod (1985 and 1991 respectively). He assigned an age of 18.5 (+ 2.2) Ma for the Eastgate site (1985, p. 7), and either a 19.0 Ma (1985, p. 94) or 18.0 Ma age (1991, p. ix; which is an approximate average of three dates given in the appendix, p. 69) for the Buffalo Canyon paleoflora. The original age assigned to the Middlegate paleoflora, which is the lateral equivelant of the Eastgate site (Axelrod, 1985), is 15.9 Ma (Evcrnden and James, 1964), or 16.3 Ma recalculated. Also, the age assignments for the Buffalo Canyon site given above (Axelrod, 1985, 1991) are considerably older than his original age (= 16 Ma) earlier communicated to Smedman (1969, p. 3). The ages of these three paleofloras need to be resolved because unequivocal age assignments are critically important for any discussion concerning the timing of biological and physical events in this region of the Great Basin. The Middlegate flora was originally assigned a radiometric age of 15.9 Ma (Evcrnden and James, 1964). The Middlegate and Eastgate localities occur in the upper few meters of the Middlegate Formation and are considered lateral equivalents (Axelrod, 1985). Swisher (1992) redated this tuff at Middlegate and obtained an age of 15.479 ± 0.200 Ma. In Buffalo Canyon an air-fall ash approximately 15 m above the plant-bearing beds, dated by Swisher (1992) at 15.595 ± 0.018 Ma, more closely dates the paleoflora than the samples taken from 80-90 m below the plant-bearing beds (Axelrod, 1991) from rocks that are most likely older debris reworked into the basal beds of the Buffalo Canyon Formation (Michael E. Perkins, oral. comm. 1994). A preliminary megaflora zonation based on the Neogene plants of this local region of west- central Nevada (Schorn, 1986 and unpublished data) also suggests the dates for the Middlegate- Eastgate and Buffalo Canyon sites as given by Axelrod (1985, 1991) are excessive. Taxonomically, the oldest well-documented paleoflora in this region of west-central Nevada, is that at Fingerrock (Wolfe, 1964; Daniel I. Axelrod, oral comm., 1987). The paleoflora is from shales conformably below a 50-60 m thick andesite breccia unit, the upper flows of which are dated at 15.1 or 15.4 Ma (Morton et al., 1977). Dates from the base of the breccia unit are given as 16.0 Ma (Axelrod, 1985, p. 93). An 40Ar/39Ar date by Swisher (1992) from the plant-bearing beds of the Fingerrock flora is 15.419 ± 0.290. The overlying Stewart Spring paleoflora (Wolfe, 1964) is approximately 14.1 Ma (see earlier discussion in the section under geologic setting). These five floral sites from the Stewart Valley and Middlegate-Buffalo Canyon basins are within 85 km of each other along a north-south transect. We consider the community of species from Middlegate-Eastgate to be intermediate between the Fingerrock and Buffalo Canyon assemblage. Likewise, Buffalo Canyon shares some species with Middlegate-Eastgate, and some with the younger Stewart Valley paleofloras and is therefore considered intermediate in age between those paleofloras. Based on the known dates associated with the plants, the overlapping associations of species, especially those that are inferred to be derived species, suggest the ages of the Middlegate- Eastgate and Buffalo Canyon localities should be placed at approximately 16 - 15.5 Ma, not the 18.5 and 19.0 Ma given by Axelrod (1985) or the 18.0 Ma date for Buffalo Canyon assigned by Axelrod (1991). If the age relationships proposed here are accepted, it is significant to note that the floras in this area of west-central Nevada appear to initially cluster around a fairly short interval of time (approximately 16 - 15 Ma) that apparently corresponds to a distinct period of extensional tectonics and basin development in this part of the present Great Basin. The temporal relationships expressed in Figure 2 are based on available radioisotopic dates and the preliminary megaplant zonation suggested for the area (HES, unpublished data). Occurrence. - Eastgate and Buffalo Canyon, Nevada. Schorn and Gooch Amelanchier Page 11 Amelanchier sp. Amelanchier couleeana auct. non (Berry) Brown. Axelrod, 1964, Univ. Calif. Publ. Geol. Sci., v. 51, p. 123, pi. 13, figs. 9,10. Quercus mccanni auct. non Berry. Axelrod, 1964, ibid., v. 51, p. 118 pro parte, pi. 11, fig. 11 only (not pi. 11, fig. 10, which is Zelkova). Discussion. - All specimens from Trapper Creek, Idaho (Axelrod, 1964) are incomplete and difficult to evaluate. Although the leaves appear to be distinct from any previously described forms, it is not possible to describe a new species on the basis of this material. The (?)obovate shape, presence of doubly serrate teeth and the nature of the tertiary and higher order venation suggest the relationship of this material is with the A. grayi-alvordensis complex. Occurrence. - Trapper Creek, Idaho. Phyllites couleeanus Berry Plate 1, Figure 13 Phyllites couleeanus Berry, 1931, U. S. Geol. Survey Prof. Paper 170-C, p. 42, pi. 13, fig. 12. Discussion. - The single specimen of Phyllites couleeanus Berry (1931) was reassigned to Amelanchier couleeana by Brown (1946). It is a leaf impression preserved in ferruginous siltstone that is interbedded between basalts dated below at 17.2 Ma and above at 16.4 Ma (recalculated from Gray and Kittleman, 1967). The specimen is incomplete and poorly preserved (Berry, 1931, plate 1, figure 13), but where visible, the tertiaries are essentially perpendicular to the secondaries (Fig. 3A). The nature of the tertiary venation, the poor preservation and incomplete nature of this single specimen are considered sufficient criteria to relegate this leaf/leaflet to Insertae Sedis and reinstate it to Berry's original Phyllites couleeanus. Additional material from the vicinity of Spokane, Washington that Berry (1929, pi. 55, fig. 4) referred to Amelanchier scudderi Cockerell is considered a new species of Amelanchier (Wolfe, 1960), but will not be described here. Another specimen from the nearby Vera site that Brown (1937a, pi. 53, fig. 11; 1946, p. 349) referred to Amelanchier dignatus (Knowlton) Brown is transferred to Alnus sp. This leaf has two well- developed teeth between each primary tooth. Both the primary and secondary teeth produce an occasional subsidiary tooth, and there are numerous, evenly spaced tertiary veins oriented perpendicular to the secondaries. The leaf from the Beaverhead basins area of Montana that Becker (1969, p. 97, pi. 30, fig. 25) referred to this species appears to be entire-margined. It seems doubtful that this is an Amelanchier but until the specimen has been examined we prefer to consider it ?'Amelanchier sp. All other material referred to A. couleeana (Chaney and Axelrod, 1959; Graham, 1963, 1965; Axelrod, 1964) subsequent to Brown (1946) is discussed elsewhere under the appropriate species. Occurrence. - Grand Coulee, Washington. Ulmus dignatus (Knowlton) Schorn et Gooch comb. nov. Celastrus dignatus Knowlton, 1902, U.S. Geol. Survey Bull., 204, p. 71, pi. 11, fig. 5. Myrica oregoniana Knowlton, 1902, U.S. Geol. Survey Bull., 204, p. 33, pi. 3, fig. 4. Ulmus knowltoni Tanai and Wolfe, 1977, U.S. Geol. Survey Prof. Paper 1026, p. 5, pi. 1C, F, and G; pi. 2A, C, H, I and J. Discussion. - The leaf from Van Horn's Ranch that was originally described by Knowlton (1902) as Celastrus was transferred to Amelanchier dignatus by Brown (1935). He made one further reference (Brown, 1937b) to the species. Later, Brown (1946) synonymized the original Celastrus dignatus with the new combination of Amelanchier couleeana (Berry) Brown. LaMotte (1952, p. 361), and Chaney and Axelrod (1959), recognized that the specimen of "Celastrus" belongs to the Ulmaceae and transferred it to Zelkova oregoniana (Knowlton) Brown. This specimen is an elm of the type Tanai and Wolfe (1977, p. 5) assigned to Ulmus knowltoni nomina nov. (Wolfe, oral comm., 1982). They could not use the epithet oregoniana because it is preoccupied by an Eocene elm from Ashland, Oregon (Knowlton, 1900), so they set up the epithet knowltoni to hold the holotype of Myrica oregoniana Knowlton (1902) which is an elm. In setting forth this new name Tanai and Wolfe overlooked the epithet dignatus as the next available name for the Knowlton elm; that transfer is made here. Occurrence. - Van Horn's Ranch, Oregon. Page 12 Amelanchier Schorn and Gooch ACKNOWLEDGEMENTS A number of people contributed to this discussion. We thank Patrick F. Fields for his many discussions with HES concerning the taxonomic and floristic comparisons between the ancient Neogene floras of the paleo-Nevada and paleo-Oregon-Idaho provinces; much of what is presented here is an outgrowth of those, generally friendly, discussions. We especially thank Wes Wehr and Jack A. Wolfe for reading early drafts of the manuscript and adding a number of helpful suggestions. Discussion with them and Daniel I. Axelrod, Anthony D. Barnosky, William M. Krebs, Herbert W. Meyer, James J. Rytuba, Donald E. Savage, Geraldine E. Swartz, Charles J. Smiley and John H. Stewart have added materially to various aspects of the paper. Dr. Smiley also generously made available his collections of Amelanchier from Hog Creek, Idaho. A special thanks is extended to Michael E. Perkins and Carl C. Swisher III for the radioisotopic age determinations that are so critical for timing the physical and biological events of this region. Comparative leaf material was studied at the U.S. Geological Survey Cleared Leaf Collection with Jack A. Wolfe, and the Herbarium of the University of California, Berkeley, through the courtesy of Thomas O. Duncan and Barbara Ertter. James P. Ferrigno , Francis M. Hueber and Scott L. Wing kindly made loans of fossil material available from the Smithsonian Institution. We thank Patrick F. Fields and Wes Wehr for their constructive reviews that helped provide a better presentation. This is contribution number 1624 of the University of California Museum of Paleontology. LITERATURE CITED Arnold, C.A. 1937. Observations on the fossil flora of eastern and southeastern Oregon. Part 1. University of Michigan, Contributions from the Museum of Paleontology 5(8):79-102. Axelrod, D.I. 1944. The Alvord Creek flora. In, R.W. Chaney (ed.), Pliocene Floras of California and Oregon. Carnegie Institution of Washington Publication 553(9):225-262. ----- 1956. Mio-Pliocene floras from west central Nevada. University of California Publications in Geological Sciences 33:1-322. ----- 1962. A Pliocene Sequoiadendron forest from western Nevada. University of California Publications in Geological Sciences 39(3):195-268. ----- 1964. The Miocene Trapper Creek flora of southern Idaho. University of California Publications in Geological Sciences 51:1-181. ----- 1966. The Eocene Copper Basin flora of northeastern Nevada. University of California Publications in Geological Sciences 59:1-124. ----- 198 5. Miocene floras from the Middlegate basin. University of California Publications in Geological Sciences 129:1-279. ----- 1991. The Early Miocene Buffalo Canyon flora. University of California Publications in Geological Sciences 135:1-76. ----- 1992. Miocene floristic change at 15 Ma, Nevada to Washington, U.S.A. In, B.S. Venkatchala, D.L. Dilcher and H.K. Maheshwari (eds.), Essays in Evolutionary Plant Biology. The Palaeobotanist 41:234 -239. ----- in press. The Purple Mountain flora, western Nevada. University of California Publications in Geological Sciences. and Schorn, H.E. 1994. The 15 Ma floristic crisis at Gillam Spring, Washoe County, northwestern Nevada. PaleoBios 16(2):1-10. ----- and Ting, W.S. 1960. Late Pliocene floras east of the Sierra Nevada. University of California Publications in Geological Sciences 39:1-118. Barnett, S.F. and Fisk, L.H. 1984. Palynology of the ?Miocene Alvord Creek Formation, southwestern Oregon. Palynology 8:253. Barrows, K.J. 1971. Geology of the southern Desatoya Mountains, Churchill and Lander Counties, Nevada. Unpublished. Ph.D. Dissertation, University of California, Los Angeles, California. 349 pp. Becker, H.F. 1969. Fossil plants of the Tertiary Beaverhead basins in southwestern Montana. Palaeontographica 127(B):1-142. Berry, W.E. 1929. A revision of the flora of the Latah formation. U. S. Geological Survey Professional Paper 154-H:225-264. ----- 1931. A Miocene flora from Grand Coulee, Washington. U. S. Geological Survey Professional Paper 170-G31-42. Brown, R.W. 1935. Miocene leaves, fruits and seeds from Idaho, Oregon and Washington. Journal of Paleontology 9(7):572-587. Schorn and Gooch Amelanchier Page 13 ----- 1937a. Additions to some fossil floras of the Western United States. U. S. Geological Survey Professsional Paper 186-J: 163-206. ----- 1937b. Further additions to some fossil floras of the western United States. Journal of the Washington Academy of Science 27(12):506-517. ----- 1946. Alterations in some fossil and living floras. Journal of the Washington Academy of Science 36(10):344-355. ----- 1949. Pliocene plants from Cache Valley, Utah. Journal of the Washington Academy of Science 39(7):224-229. Chaney, R.W. 1927. Geology and palaeontology of the Crooked River Basin, with special reference to the Bridge Creek flora. With a chapter on the Tertiary insects by T.D.A. Cockerell. In, R. Kellogg, J.C. Merriam, C. Stock, R.W. Chaney and H.L. Mason (authors), Additions to the Palaeontology of the Pacific Coast and Great Basin Regions of North America. Carnegie Institution of Washington Publication 346(4):45-138. ----- and Axelrod, D.I. 1959. Miocene floras of the Columbia Plateau. Part II. Systematic considerations. Carnegie Institution of Washington Publication 617(2):135-237. Christopher, S.C., Greene, C.W., Neubauer,B.F. and Higgins, J.M. 1985. Apomixis in Amelanchier laevis, shadbush (Rosaccae, Maloideae). American Journal of Botany 72(9):1397-1403. Dalrymple, G.B. 1979. Critical tables for conversion of K-Ar ages from old to new constants. Geology 7(11)558-560. Evernden, J.F., and James, G.T. 1964. Potassium argon dates and the Tertiary floras of North America. American Journal of Science 262(8):945-974. —, Savage, D.E. , Curtis, G.H. and James, G.T. 1964. Potassium-argon dates and the Cenozoic mammalian chronology of North America. American Journal of Science 262(2):145-198. Graham, A. 1963. Systematic revision of the Sucker Creek and Trout Creek Miocene floras of southeastern Oregon. American Journal of Botany 50(9):921-936. ----- 1965. The Sucker Creek and Trout Creek Miocene floras of southeastern Oregon. Kent State University Bulletin, Research Series 9:1-147. Gray, J. and Kittleman, L.R. 1967. Geochronomctry of the Columbia River Basalt and associated floras of eastern Washington and western Idaho. American Journal of Science 265(4)257-291. Hickey, L.J. 1979. A revised classification of the architecture of dicotyledonous leaves. In, C.R. Metcalfe and L. Chalk, Anatomy Of The Dicotyledons. Volume I. Systematic Anatomy Of Leaf And Stem, With A Brief History Of The Subject. Clarendon Press, Oxford, pp. 25-39. and Wolfe, J.A. 1975. The bases of angiosperm phylogcny: vegetative morphology. Annals of the Missouri Botanical Garden 62(3)538-589. Jones, G.N. 1945. Maiacomeles, a genus of Mexican and Guatemalan shrubs. Madrono 8(2):33-39. ----- 1946. American species of Amelanchier. University of Illinois Press, Illinois Biological Monographs 20(2):1-126. Knowlton, F.H. 1900. Fossil plants associated with the lavas of the Cascade Range. U.S. Geological Survey, 20th Annual Report, Part 3:37-64. ----- 1902. Fossil flora of the John Day Basin, Oregon. U.S. Geological Survey Bulletin 204:1-153. LaMotte, R.S. 1952. Catalogue of the Cenozoic plants of North America through 1950. Geological Society of America Memoir 51:1- 381. Landry, P. 1975. Le concept d'espece et la taxinomie du genre Amelanchier (Rosacees). Bulletin, Societe botanique de France 122:243- 252. MacGinitie, H.D. 1933. The Trout Creek flora of southeastern Oregon. In, Fossil Floras of Yellowstone National Park and Southeastern Oregon. Carnegie Institution of Washington Publication 416(2):21-68. Manchester, S.R. and Meyer, H.W. 1987. Oligocene fossil plants of the John Day Formation, Fossil, Oregon. Oregon Geology 49(10):115-127. Mawby, J.E. 1965. Pliocene vertebrates and stratigraphy in Stewart and lone Valleys, Nevada. Unpublished. Ph.D. Dissertation, University of California, Berkeley, California. 210 pp. Minor, S.A., Rytuba, J.J., Goeldner, C.A. and Tegtmeyer, K.J. 1987. Geologic Map of the Alvord Hot Springs Quadrangle, Harney County, Oregon. U.S. Geological Survey Miscellaneous Field Studies Map MF-1916. Page 14 Amelanchier Schorn and Gooch Morton, J.L., Silberman,M.L., Bonham, H.F., Jr., Garside, J.L. and Noble, D.C. 1977. K-Ar ages of volcanic rocks, plutonic rocks, and ore deposits in Nevada and eastern California - determinations run under the USGS-NBMG cooperative program. Isochron/West 20:19- 29. Rytuba, J.J. and McKee, E.H. 1984. Peralkaline ash flow tuffs and calderas of the McDermitt volcanic field, southeast Oregon and north central Nevada. Journal of Geophysical Research 89(B10):8616-8628. Sarna-Wojcicki, A.M., Caxr, M.D., Fleck, R.J., King, B.S., Meyer, C.E., Miller, D.M., Nakata, J.K., Pringle, M.S., Jr., Wan, E., Brown, F.H., Nash, W.P., Perkins, M.E., Reheis, M.C. and Yount, J.C. 1992. Developing a regional Neogene chrono- stratigraphic framework for the Great Basin. Geological Society of America Abstracts with Programs 24(6):60. Schorn, H.E. 1984. Palynology of the Late Middle Miocene sequence, Stewart Valley, Nevada. Palynology 8:259-260. ----- 1986. Vegetation and climate ca. 17-12 Ma in the Great Basin of western Nevada: data from Stewart Valley. Geological Society of America Abstracts with Programs 18(5):410. —, Scudder, H.I., Savage, D.E. and Firby, J.R. 1989. General stratigraphy and paleontology of the Stewart Valley area, Mineral County, Nevada. In G. Liu, R. Tsuchi and L. Qibin (eds.), Proceedings on the International Symposium of Pacific Neogene Continental and Marine Events, National Working Group of China for IGCP-246. Nanjing University Press, Nanjing, June, 1989, pp. 157-173. Smedman, G. 1969. An investigation of the diatoms from four Tertiary lake bed deposits in western Nevada. PaleoBios 9:1-16. Smith, H.V. 1938a. Notes on fossil plants from Hog Creek in southwestern Idaho. Papers of the Michigan Academy of Science, Arts and Letters 23:223-231. ----- 1938b. Some new and interesting late Tertiary plants from Sucker Creek, Idaho- Oregon boundary. Bulletin of the Torrey Botanical Club 65:557-564. ----- 1941. A Miocene flora from Thorn Creek, Idaho. The American Midland Naturalist 25:473-522. Stewart, J.H. 1992. Paleogcography and tectonic setting of Miocene continental strata in the northern part of the Walker Lane Belt. In, S.P. Craig (ed.), Structure, tectonics and mineralization of the Walker Lane. Geological Society of Nevada, Walker Lane Symposium, Proceedings Volume, pp. 53-62. Swisher, C.C., III. 1992. 40Ar/39Ar dating and its application to the calibration of the North American Land Mammals Ages. Unpublished Ph.D. Dissertation, University of California, Berkeley, California. 239 pp. Tanai, T. and Wolfe, J.A. 1977. Revisions of Ulmus and Zelkova in the middle and late Tertiary of western North America. U.S. Geological Survey Professional Paper 1026: 1-14. Train, P., Henrichs, J.R. and Archer, W.A. 1941. Contributions toward a flora of Nevada. No. 33. Medicinal uses of plants by Indian Tribes of Nevada. Part I, pp. 1-61. Division of Plant Exploration and Introduction, Bureau of Plant Industry, U.S. Department of Agriculture, Beltsville, Maryland. Wolfe, J.A. 1960. Early Miocene floras of northwest Oregon. Unpublished Ph.D. Dissertation, University of California, Berkeley, California. 253 pp. ----- 1964. Miocene floras from Fingerrock Wash, southwestern Nevada. U. S. Geological Survey Professional Paper 454-N:l-36. ----- 1969. Neogene floristic and vegetational history of the Pacific Northwest. Madrono 20(3):83-110. — 1993. A method of obtaining climatic parameters from leaf assemblages. U.S. Geological Survey Bulletin 2040:1-71. ----- and Tanai, T. 1980. The Miocene Seldovia Point Flora from the Kenai Group, Alaska. U.S. Geological Survey Professional Paper 1105:1-52. and Wehr, W. 1988. Rosaceous Chamaebatiaria-like foliage from the Paleogene of western North America. Aliso 12(1 ):177-200. Schom and Gooch Amelanchier Page 15 APPENDIX A: LIST OF TAXONOMIC REVISIONS The following list summarizes the taxonomic revisions (and their occurrences) recognized in this report. A few specimens are recommended for rejection and are so noted. Amelanchier alvordensis auct. non Axelrod. Axelrod, 1962, p. 234. = Amelanchier hawkinsae Schorn et Gooch (Chalk Hills, Nevada). Amelanchier alvordensis auct. non Axelrod. Axelrod, 1966, p. 71, pi. 16, figs. 11,12 = Insertae Sedis, indeterminate ?leaf (Copper Basin, Nevada). Recommend rejection. Amelanchier apiculata auct. non Brown. Axelrod, 1956, p. 298 pro parte, pi. 8, fig. 9; pi. 14, figs. 9 and 11 only = Amelanchier hawkinsae Schorn et Gooch (Aldrich Station and Chloropagus, Nevada). Amelanchier apiculata auct. non Brown. Axelrod, 1956, p. 298 pro parte, pi. 14, fig. 10 = Incertae Sedis, indeterminate ?leaf (Chloropagus, Nevada). Recommend rejection. Amelanchier apiculata auct. non Brown. Axelrod, 1956, p. 298 pro parte, homeotype no. 4138 only = Incertae Sedis, indeterminate ?leaf (Aldrich Station, Horsethief Canyon, Nevada). Recommend rejection. Amelanchier couleeana (Berry) Brown, 1946, p. 349, = Incertae Sedis, Phyllites couieeanus Berry (Grand Coulee, Washington). Recommend rejection. Amelanchier couleeana auct. non (Berry) Brown. Chaney and Axelrod, 1959, p. 183 pro parte, pi. 36, fig. 1 only = Amelanchier magnifolia Arnold var. magnifolia (Hog Creek, Idaho). Amelanchier couleeana auct. non (Berry) Brown. Chaney and Axelrod, 1959, p. 183 pro parte, homeotype no. 5121 only = Uimaceae, cf. Zelkova (White Hills [Mascall], Oregon). Amelanchier couleeana auct. non (Berry) Brown. Graham, 1963, p. 930 = Amelanchier magnifolia Arnold var. magnifolia (Trout Creek and Succor Creek, Oregon). Amelanchier couleeana auct. non (Berry) Brown. Axelrod, 1964, p. 123, pi. 13, figs. 9, 10 and homeotypes nos. 8454,8455, 8569 = Amelanchier sp. (Trapper Creek, Idaho). Amelanchier couleeana auct. non (Berry) Brown. Graham, 1965, p. 89, pi. 15, figs. 1, 2,4,5 = Amelanchier magnifolia Arnold var. magnifolia (Trout Creek and Succor Creek, Oregon). Amelanchier couleeana auct. non (Berry) Brown. Becker, 1969, p. 97, pi. 30, fig. 25 = ?Amelanchier sp. (Beaverhead Basin, Montana). Amelanchier coveus (Chaney) Chaney et Axelrod, Chaney and Axelrod, 1959, p. 183 pro parte, homeotype no. 3261 = IPrunus (Thorn Creek, Idaho). Recommend rejection. Amelanchier coveus (Chaney) Chaney et Axelrod, Chaney and Axelrod, 1959, p. 183 pro parte, homeotype no. 3262 = tPopulus (Thorn Creek, Idaho). Recommend rejection. Amelanchier coveus (Chaney) Chaney et Axelrod, Chaney and Axelrod, 1959, p. 183 pro parte, homeotype no. 3263 = missing (Thorn Creek, Idaho). Amelanchier cusicki auct. non Fernard. Wolfe, 1964, p. 24, pi. 19, fig. 11 = Amelanchier hawkinsae Schorn et Gooch (Stewart Valley, Nevada). Amelanchier desatoyana Axelrod, 1991, p. 56, pi. 16, figs. 1-7 = Amelanchier magnifolia var. desatoyana (Axelrod) Schorn et Gooch comb. nov. (Buffalo Canyon, Nevada). Amelanchier dignatus auct. non (Knowlton) Brown. Brown, 1937a, p. 176, pi. 53, fig. 11 = Alnus sp. (Vera, Washington). Amelanchier dignatus auct. non (Knowlton) Brown. Smith, 1938a, p. 228 = Amelanchier magnifolia Arnold var. magnifolia (Hog Creek, Idaho). Amelanchier grayi auct. non Chaney. MacGinitie, 1933, p. 58 = Amelanchier magnifolia Arnold var. magnifolia (Trout Creek, Oregon). Amelanchier grayi auct. non Chaney. Axelrod, 1985, p. 159, pi. 28, fig. 2; pi. 30, figs. 2, 3 and homeotypes nos. 7014,7015 = Amelanchier magnifolia var. desatoyana (Axelrod) Schorn et Gooch comb. nov. (Eastgate, Nevada). Amelanchier scudderi auct. non Cockerell. Berry, 1929, p. 252, pi. 55, fig. 4 = Amelanchier sp. nov., fide Wolfe, 1960, not Page 16 Amelanchier Schom and Gooch described here (Latah - Spokane Brickyard, Washington). Arbutus idahoensis auct. non (Knowlton) Brown. Axelrod, 1944, p. 262, pi. 45, fig. 5 = Amelanchier alvordensis Axelrod (Alvord Creek, Oregon). Ceanothus precuneatus auct. non Axelrod. Axelrod, 1944, p. 261 pro parte, pi. 45, fig. 3 only = Amelanchier alvordensis Axelrod (Alvord Creek, Oregon). Ceanothus precuneatus auct. non Axelrod. Axelrod, 1944, p. 261 pro parte, pi. 45, fig. 2 only = Robinia sp. (Alvord Creek, Oregon). Celastrus dignatus Knowlton, 1902, p. 71, pi. 11, fig. 5 = Ulmus dignatus (Knowlton) Schorn et Gooch comb. nov. (Van Horn's Ranch [Mascall], Oregon). Myrica oregoniana Knowlton, 1902, p. 33, pi. 3, fig. 4 = Ulmus dignatus (Knowlton) Schorn et Gooch comb. nov. (White Hills [Mascall], Oregon). Quercus mccanni auct. non Berry. Axelrod, 1964, p. 118 pro parte, pi. 11, fig. 11 only = Amelanchier sp. (Trapper Creek, Idaho). Rosa alvordensis Axelrod, 1944, p. 259, pi. 44, fig. 5, and homeotype no. 2125 = Amelanchier alvordensis Axelrod (Alvord Creek, Oregon). Symphorocarpos wassukana Axelrod, 1956, p. 312 pro parte, homeotype no. 4178 only = Amelanchier hawkinsae Schorn et Gooch (Aldrich Station, Nevada). Symphorocarpos wassukana Axelrod, 1956, p. 312 pro parte, pi. 9, fig. 3 = Robinia sp. (Aldrich Station, Nevada). Ulmus knowltoni Tanai et Wolfe, Tanai and Wolfe, 1977, p. 5, pi. 1, figs. 1C, F, G; pi. 2 figs. 2A, C, H-J = Ulmus dignatus (Knowlton) Schorn et Gooch comb. nov. (White Hills [Mascall], Oregon). Plate 1. Figure 1, Amelanchier hawkinsae, IX, holotype, UCMP Paleobot. No. 1; Figures 2-7, A. hawkinsae, IX, paratypes, UCMP Paleobot. Nos. 2-7; Figure 8, A. hawkinsae, approximately 3X enlargement of figure 1; Figure 9, A. magnifolia var. desatoyana, approximately 3X enlargement, UCMP Paleobot. No. 9615b; Figure 10, A. hawkinsae, approximately 5X enlargement of right half of figure 3; Figure 11, A. alvordensis Axelrod, approximately 5X enlargement, holotype, UCMP Paleobot. No. 2110; Figure 12, A. magnifolia var. magnifolia approximately 5X enlargement of topotype material from Trout Creek, Oregon, UCMP Paleobot. Loc. No. 275; Figure 13, Phyllites couleeana Berry, approximately 2.5X enlargement, holotype, USNM 38155. Schom and Gooch Amelanchier Page 17