BLM LIBRARY 88005847 THE PRESETTLEMENT VEGETATION OF THE CALIFORNIA DESERT June 1980 l_ Sys THE PRESETTLEMENT VEGETATION OF THE CALIFORNIA DESERT June 1980 W. Geoffrey Spauling College of Forest Resources and Quaternary Research Center University of Washington Seattle, Washington 98195 for Bureau of Land Management 1695 Spruce Street Riverside, California 92507 Contract CA-060-CT8-65 Bureau of Land Management Library Bldg. 50, Denver Federal Center Denver, CO 80225 THE PRE SETTLEMENT VEGETATION OF THE CALIFORNIA DESERT Submitted to the California Bureau of Land Management, Desert Planning Staff Under Contract // CA -DGe>' Ct8 - Co S BY W. Geoffrey Spaulding College of Forest Resources and Quaternary Research Center University of Washington Seattle, WA 98195 ACKNOWLEDGEMENTS The following people have offered their kind assistance and advice , ,ln8 the course of this research: Ken Cole Hyruin Johnson Estella Leopold Paul Martin Jim Mead Peter Rowlands Sue Spaulding Ray Turner Roger Twitchell Tom Van Devender ii TABLE OF CONTENTS ACKNOWLEDGEMENTS ii LIST OF TABLES v LIST OF FIGURES vi INTRODUCTION 1 METHODS 2 Field Techniques 5 Laboratory Techniques 6 Data Treatment 6 Age Determination 7 A Chronological Framework 7 THE STUDY AREA. 9 The Marble Mountains 10 Marble Mountains Locality A 10 Marble^ Mountains Locality B 12 < Marble Mountains Locality C 13 The Eureka Valley And Vicinity 13 Horse Thief Hills Locality A 15 Horse Thief Hills Locality B 16 The Eureka View Locality 17 THE FOSSIL RECORDS 18 The Marble Mountains 18 Marble Mountains Locality A *. 18 Marble Mountains Locality B And C 24 The Eureka Valley And Vicinity ; 35 Horse Thief Hills Locality A... 35 Horse Thief Hills Locality B 38 The Eureka View Locality 41 iii The Regional Fossil Record 49 The Latest Wisconsin At Low Elevations 49 The Latest Wisconsin At Intermediate Elevations .53 The Early Holocene at Low Elevations ,...55 Early Holocene Middens From Intermediate Elevations 60 Middle and Late Holocene Packrat Middens 63 DISCUSSION 69 Vegetation Dynamics 70 The Marble Mountains 70 The Eureka Valley 75 The Sheep Range 83 Precipitation Gradients '. 84 Creosote Bush and Glacial Refugia 86 LITERATURE CITED 93 iv LIST OF TABLES. Table No. Page No, 1. Packrat midden sites, samples, and radiocarbon dates 19 from the Marble Mountains, San Bernadino County, California. 2. Plants from Marble Mountains Locality A. 22 3. Plants from Marble Mountains Locality B. 25 AA. Plant s pecies from Marble Mountains Locality C, Sites 29 6 and 7. AB. Plants from Marble Mountains Locality C, site 9. 31 5. Packrat midden sites, samples, and radiocarbon dates 36 from the northeastern Eureka Valley, Inyo County, California. 6. Plants from the Horse Thief Hills 1 site. 39 7. Plants from the Horse Thief Hills 2 site. . AO 8. Plants from the Horse Thief Hills 3 site. A2 9. Plants from the Eureka View Locality. . A A 10. Plants from the Owl. Canyon 1 midden, Amargosa Desert, 52 Nye County, Nevada. 11. Plants from late Wisconsin and early Holocene packrat 56 midden sites in the Sheep Range, Clark County, Nevada. 12. Plants from middle and late Holocene packrat midden 6A sites in the Sheep Range, Clark County, Nevada. 13. Tossil records of Larrea tridentata from the Wisconsin 87 Glacial Age and early Holocene. LIST OF FIGURES Figure No. Page No, 1. The California Desert and adjacent areas. 3 2. The Marble Mountains study area. 11 3. The Eureka Valley Study area. 14 A. The estimated relative abundance of selected plants 71 in Neotoma midden samples from Marble Mountains Locality B and C. 5. Packrat midden samples from the CDCA and adjacent 73 regions plotted against radiocarbon age. 6. The total number of identified plant species (N) in 76 Neotoma midden samples from Marble Mountains Locality B and Locality C plotted against radiocarbon age. 7. The estimated relative abundance of selected plant 78 species in macrofossil assemblages from the Eureka View Locality plotted against radiocarbon age. 8. The total number of identified plant species (N) in 82 macrofossil assemblages from the Eureka View Locality '. plotted against radiocarbon age. 9. Fossil sites of Wisconsin and early Holocene age con- 90 taining creosote bush. vi INTRODUCTION The biotic provinces of North America, as we know them now, are relatively recent developments. The environment of the last Ice Age was drastically different than today's and, during that time, the dis- tribution and composition of natural communities differed correspondingly. For at least the last million and a half years, a period known as the Pleistocene, the global climate has alternated between long glacial and shorter interglacial intervals. At least six glacial to interglacial cycles have occured in the last half million years. The average length of the cold, glacial periods exceeds 50,000 years while that of the warmer, interglacials is less than 15,000 years. Superimposed on these long term oscillations are fluctuations of shorter frequency and lower amplitude. The climate of the present interglacial, the Holocene, has been subject to frequent changes of relatively low intensity (Imbrie and Imbrie 1980; Turekian 1971; Wright 1976; Davis 1976). The last shift in global environment from an Ice Age to the warmer climate that we enjoy today occured between 14,000 and 8,000 years ago. One result was the last major episode of plant species migration ' and the ultimate segregation of those species into plant communities of modern aspect. In the strict biogeographic sense, the Mojave Desert has existed for only the last 8,000 years or so. The regional vegetation and climate of the preceding Wisconsin glacial age was different. Hence, at the outset, we are presented with the fact that the ecosystems under scrutiny in the California Desert Conservation Area (CDCA) are subject to low frequency perturbations of sufficient amplitude to alter their aspect considerably. The desert-adapted -1- -2- plants and animals being studied are subject to, and are the product of, selective pressures exerted by a long series of major environmental changes. This study examines the impact of the last such change, the end of the Wisconsin Glacial Age, on vegetation of the present Mojave and Colorado (northwestern Sonoran) Deserts (Fig. 1). The development of postglacial desertscrub (sensu Brown and Lowe 1974) vegetation through the last 8,000 years is also discussed. METHODS Paleoecological reconstruction depends upon a reliable source of fossil material. In temperate regions depositional basins such as lakes and bogs yield well preserved organic remains. Sites such as these can provide stratigraphically continuous records of vegetation that stretch back tens of thousands of years. However, sedimentary deposits containing well preserved organic remains are scarce in the deserts of North America. In the early 1960fs a new paleoecological tool was introduced (Wells and Jorgensen 1964; Wells 1976) and its use has shed a great deal of light on the biotic history of arid North America. Packrat (Neotoma spp.) midden analysis (paleonidology) is the reconstruction of past vegetation using the mummified plant remains gathered and secreted in rock crevices and caves by packrats. The habit of col- lecting massive amounts of vegetation and constructing a dwelling with that debris is a generic trait of this cricetid rodent (Finley 1958; Van Devender 1973). Within a packrat stickhouse or den there are various functionally specific areas including nesting chambers, -3- Figure 1. The California desert and adjacent areas. Subdivisions of the CDCA defined by the California Bureau of Land Management, Desert Planning Staff. A dashed line delimits the Colorado (Sonoran) and Mojave Deserts. EV, Eureka Valley; MM, Marble Mountains. Other fossil sites are: 1, Funeral Range; 2, Kofa Mountains; 3, Lower Grand Canyon; 4, Lucerne Valley and vici- nity; 5, southern Nevada Test Site; 6, Newberry Mountains; 7, New Water Mountains; 8, Ord Mountain; 9, Owl Canyon; 10, Picacho Peak; 11, Scodie Mountains; 12, southeastern Sheep Range; 13, western Sheep Range; 14, Titus Canyon, Inyo Moun- tains; 15, Turtle Mountains; 16, Wellton Hills; 17, Whipple Mts, e e © © © »0 >© / . / < «* — ' z < © © »o^ ,tm *S»X o ' z * z „ J «w I > • •- < / *" *> -» o ...••• l^i— ^ X 3 i \ z ( < cr LJ < UJ cr K •1 • v © o 1 CO UJ Q < 2 2 -D t>4 * < 2 O < O w — c O o UJ "•= or .. ** _/^ o: > C >* o -J < cr u to o Z •> UJ .*! o .- < o IKm « O o UJ o < — X ii < -4- food caches, and trash middens. Vegetable debris from the latter areas is the focus of this study. In rocky habitats packrats often live in rock shelters or crevices, using the naturally occuring shelter to augment that provided by their den. Middens often serve as urination and defecation points and the vegetable debris in sheltered middens fre- quently becomes cemented in a matrix of crystalized urine (amberat) . These dessicated masses of plant debris, encased in rock-hard amberat, occur in cavities in the hills and mountains of the North American deserts, Middens that are protected in a rock shelter may persist as long as the shelter lasts. Cavities in slow weathering rock types have yielded middens and associated paleobotanical data for the last 50,000 radio- carbon years. Ancient middens normally have a thick weathering rind composed of degraded midden debris, adhering dust, and packrat urine. Amberat is hygroscopic, it absorbs water from the air during periods of high atmospheric humidity and, during such times, becomes partially | fluid again; Fresh breaks in a midden are quickly sealed by a new layer of amberat drawn by capillary action from the interior of the midden. With time, the amberat becomes convoluted and takes on the aspect of the original weathering, rind. This self-sealing contributes to the long-term viability of fossil middens. Several factors combine to make packrat midden analysis a power- ful paleoecological tool. The most striking is the n^ar perfect pre- servation of leaves, flowers, seeds, and other delicate plant parts that may be tens of thousands of years old. Given an adequate reference collection, more than 60% of the fossil taxa in a midden can be iden- tified to the species level. Packrats are comprehensive collectors of -5- vegetation within their limited (generally less than 30 m; Stones and Hayward 1968; Bleich and Schwartz 1975) home range. Modern middens invariably contain all the plants that are common at a site as well as most of the uncommon and rare species. Age determination of fossil middens is achieved through radiocarbon dating. Field Techniques. Field searches for midden sites were made along outcrops of shelter forming rocks, particularly along cliff bases. Carbonaceous sediments and certain igneous rocks yield ancient middens for this study. Shelters that contain fossil packrat middens occur only in rocky habitats and this study is necessarily restricted to rocky slope vegetation. Once a midden is located its position in the shelter and exposed stratigraphy is carefully examined. Ancient middens that appear com- plex or jumbled, without easily discernable layering, are disregarded. Any indurated midden with a heavy weathering rind and of uncomplicated aspect is considered suitable for sampling. The weathering rind is removed from selected middens and they are again examined for evidence of multiple layering. Samples from individual midden units are removed with a hammer and cold chisel or crowbar and carefully wrapped. A list of all the plants occuring within 30 m of the midden site is made. The distribution of plant communities, principal species, and their re- lationships to the mosaic of available habitats is noted for the entire 0 locality. The quantitative relationships between dominance of species in a plant community and importance of plant fragments in a midden are poorly understood. Quantitative data gathered on packrat midden site vegetation is restricted to qualitative ordinations of abundance. -6- Laboratory Techniques. Once in the lab midden samples are thor- oughly cleaned of weathering rind and other contaminants that adhere to their surfaces. They are then placed in a bucket of water to disolve the amberat and release the plant fragments. The resultant wet mass is poured from the bucket through a nested pair of six mesh (3.35 mm) and 20 mesh (0.8A mm) soil sieves and flushed with water. The captured material is dried and the identifiable plant .fragments, bones, etc.... are sorted into plastic vials. The plant fragments are examined under moderate magnification and compared with modern reference specimens. A list is compiled with estimates of the relative abundance of each plant taxon in the fossil assemblage. Data Treatment. Possible relative abundance values for the fossil plants in a midden sample range from 5 (very abundant), to 4 (abundant), to 3 (common), to 2 (occasional), to 1 (rare), to 0 (one or two fragments, a possible contaminant). The lowest number of fossil taxa (N) in , any macrofossil assemblage in this study is 8, the greatest number recorded is 33. The total number of woody perennials, excluding grasses, (N ) in each macrofossil assemblage is presented as well as P the percentage of woody perennials (N/N x 100). Similarity indices (IS) are used to express the percentage of shared taxa between the macrofossil assemblage and the plants at the site today. The similarity index chosen for this study is that proposed by Soren§en (1948; Mueller- Dombois and Ellenberg 1974) and takes the form : I IS ' tVb x 10° where A is the total number of species (N) noted within 30 m of the midden -7- site, B is the total number of species occuring in the midden sample, and c is the number of plants common to both the modern vegetation and the midden sample. Possible values range from 0 (no species in common) to 100 (all species shared) although test comparisons between recent middens and the surrounding plant community have failed to yield IS values greater than 85. Invariably there seem to be a few plants in the midden that were not noted on the field list, and a few noted on the list that are not found in the midden debris. Age Determination. I prefer to use a single extralocal plant species from a midden sample for radiocarbon ( C) dating. This method, suggested by Van Devender (1977a) , is desirable since the date is di- rectly associated with at least one plant species and there is no chance of contamination by recent plant debris. But desertscrub middens frequen- tly lack sufficient mass of any one species to radiocarbon date. Many of the age determinations used for this study are on undifferentiated twigs or packrat fecal pellets. Whatever the material dated, in the absence of discordant multiple dates and given a favorable stratigraphic setting, it is assumed that the C date reflects the age of the entire macrofossil assemblage. A C date is expressed as a mean date with a standard deviation reflecting various uncertainties including background radiation at the counter, sample size, counting time, number of counts, etc.... (Long and Rippeteau 1974; Sheppard 1975). A Chronological Framework. Packrat midden analysis is used to study late Quaternary climates but, for the purpose of this vegetation analysis, the chronology of climatic change, as it is known, is presented first. The chronological divisions of the latest Wisconsin and early -8- Holocene follow those presented by Van Devender and Spaulding (1979) and Phillips (1977). The latest Wisconsin (Stage III of Phillips 1977) extends from ca. 14,000 B.P. (radiocarbon years before present) to ca. 11,000 B.P. and is the waning phase of the last Ice Age. The start of the early Holocene at ca. 11,000 B.P. is marked by the retreat of many mesophytic species from the interior basins of the southwestern United States (Van Devender and Spaulding 1979; Martin and Mehringer 1965) . The extirpation of pinyon pines from the low elevation wood- lands of the present Chihuahuan, Sonoran, and Mojave Deserts was a near synchronous event at ca. 11,000 B.P. (Van Devender and Spaulding 1979). The early Holocene, persisting to ca. 7800 B.P. (Stage IV of Phillips 1977), was a time of transition characterized by an increasingly xeric climate. Desertscrub species assume gradual dominance throughout this period, culminating at ca. 7800 B.P. with the disappearance of the last of the woodland plants (Van Devender 1977b, Phillips 1977). The middle Holocene bioclimatic phase begins at this time and terminates ca. 3700 B.P At about this time in the White Mountains at the northern border of the CDCA upper tree line retreats dramatically (La Marche 1973, 1974). The White Mountains record of higher spring through autumn temperatures during the middle Holocene supports Antevs' (1948, 1955) concept of an Altithermal climatic phase in this region. Martin (1963) suggests that the middle Holocene may have been a period of increased summer preci- pitation in the monsoonal Southwest. This hypothesis is not inconsistent with a climate characterized by increased warm season temperatures and, perhaps, increased dryness in the Great Basin Desert. The late Holocene, . from ca. 3700 B.P. to present, appears in the White Mountains chronology -9- (La Marche 1973, 1974) to be a time of generally cooler temperatures. There is evidence for a more mesic late Holocene climatic regime in pollen and packrat midden data from the northern Mojave and southeastern Great Basin Deserts of Nevada (Spaulding 1977, unpub. ; Madsen 1973). THE STUDY AREA The fossil plant data from two areas, one at the southern boun- dary of the Mojave Desert and one in the northern Mojave Desert, are the basis for this study. The significance of other fossil sites in the vicinity is also discussed. Climatically induced changes in plant species distribution and abundance may be more pronounced in boundary region vegetation where more species are at the limits of their habitat tolerance. Hence, the two main study areas were selected to be near the bio tic boundaries of the Mojave Desert. The southern limit of the Mojave Desert is gradational and poorly defined. Describing the borders of the Colorado and Mojave subdivisions of the CDCA, the floristic boundary is defined by the decreasing number and importance of species adapted to a bimodal, winter-summer precipi- tation regime (Fig. 1; Shreve and Wiggins 1964; Hastings et al. 1972). Plants such as Dalea spinosa, Beloperone calif ornica, Cercidium f loridum, Fouquieria splendens, Hyptis emoryi, and Simmondsia chinensis occur in the Sonoran Desert (of which the "Colorado Desert" is but a subdivision) to the south and east. In this southerly region summer1 rainfall comprises not only a significant percentage of the annual total but is also pre- dictable. The staggered distribution limits of these monsoonal-adapted species form the biogeographic criteria for the southern Mojave boundary. -10- In contrast, the Mojave Desert's northern boundary is based on the limits of one species, Larrea tridentata (creosote bush), and is correspondingly well defined. A shrub that is ubiquitous in the warm deserts of North America, the northern limit of Larrea tridentata is controlled by minimum temperatures (Beatley 1974, 1975) and defines both the northern limit of the Mojave Desert and the southern boundary of the Intermountain Region (Cronquist et al. 1971). The Marble Mountains The Marble Mountains trend from northwest to southeast and lie at the northern end of the Cadiz Valley (ca. Lat. 34 40' N, ca. Lon. 115 35' W) . The mountain range is immediately south of the juncture of the southern limits of the Central and Eastern Mojave subdivisions of the CDCA (Fig. 1). Where the mountain slopes meet the bajada ele- vations range from ca. 300 m (1000 ft) in the south to ca. 920 m (3000 ft) at the northern terminus (Fig. 2). Castle Peak at 1170 m (3842 ft) is the highest point in the Marble Mountains and the majority of the range lies below 900 m (2950 ft) elevation. Marble Mountains Locality A. Near the Iron Hat Mine in the south- western Marble Mountains (Fig. 2) is an area of steep limestone slopes and ledges designated Locality A (Cadiz, CA, 15' quadrangle: T6N, R14E, NW*£, NW"J£, Sec. 19). These limestone outcrops provide anomalous habitats in a mountain range of Tertiary (?) rhyolite, rhyolitic breccia, and associated volcanic rocks. The sere south to southwest facing slopes, from 365 m to 490 m (1200 ft to 1600 ft) elevation, support widely scattered vegetation. The only common shrubs on the rocks are Encelia -11- Figure 2. The Marble Mountains study area. The Marble Mountains 1, 2 and 3 sites are at Locality A, the MM 4 and 5 sites at Locality B, and the MM 6, 7, and 9 sites at Locality C. GARBLE MOUNTAINS STUDY AREA 4 mi 1 I I i 2 3 4 5 km Contours at 400 feet . wi o td 4J c X 3 u c «;; a) c .o co U OJ (0 ^ s o- oj a) 4J o >4-l E a> O JO »-l u-i co u co to 0) 0) 4J ^ to •o c o 3 .O CO O O o JO CO y «h o T3 i-i rt CO U « •o a, c • to cq CO • 0) CO . iH «H < a cs § o ► nu ai ■H toO 0) CO «H 4J jO ■H * fi » S^ CO 4J CJ c s 4i D » *0 O CO •OU 0) «H -H E O U s o 4J -H J_> CO *o CO ^ CO U •x s o O U JO 03 0) CO 0* m j CO •H T3 P O) 0> 4J 4J CO CO ID s O 4J CO /—s •a • o c c o .o • l-l .£> CO CO CJ .H O «-' CO C*S >s 4-> W CO CO 3 o 4-> CO 0) e 0) bO CD C toO ■H "O. CJ OJ^ CO iH — "9 S c co to m < o cq c ■H 0) to PL4 • • Pm CO PQ o ^•s CO /-N m r*» to oo so r» ,r-» r-4 <-> + -" VO vO + 1 m i X vo x o cj sr cj CO 0) too cu 0) c o CO 4-> CU CO Cu 0) o i h •H CO »H CO bO 3 3 rH •H CO CJ 4J CO m-i *a c S CO m vO »3- tN CO s •H OJ CO CO toO ■H 4J CQ /-s — ' SO CO I o u Ph CO *3 •O 01 3 0) CO CO CO 0) co to too JJ o 01 •H CO 14-4 $1 CO u E • -H o o o 4-1 CO CO o ■H tO OJ S "*-• 5S Oh « CQ O 00 <""N CO CN i00 + 't; O I Os X o o CO 0) T3 c CO to OJ Cu oo o c rH •H CO CJ CO CO U-i 3 fH s CO CO 4-1 o o CO ' a o OJ c CO • OJ CO 4-1 CJ too c CO •H o •H > CJ s 4-1 ^^ CQ O m /-s to co ,°° + l7! vO m I CN x CO CJ CQ o - C J C wO M 3 J^ 3 »-i CO O CO O CO s: 2 z 2: s CM CO ^H 0-1 CM o»^ '**' v-»- N-^ w m «* •a- m 4 •n »n CQ 0} 01 03 CO 0) CO 3 K c 3 3 c c ■H ■H •H •H •H «H CO M 0) CO U td 0> to OJ to 0) to 4J J •H 4J iH 4J iH 4J "»H 4J iH 4J c < XJ c J3 3 wO 3 X) 8 ^3 3 3 CJ M 3 H 3 U 3 l-i 3 H 3 O O to O to O CO O CO O to O £ iJ X, S s £ £ £ £ £ £ £ -20- CO CO CO 3 3 CD o o a S to • OJ CO CO CO s M OJ CM •H c 3 *-3 C CO CO 60 to o cu c to iH to CO rH I E (1) CO o 0 CJ 60 4-1 4_) CD «H o o •H & OJ QJ S *J :s 55 PQ /-s m m 00 CO -H CM vD + |>< O o *— CO PQ *~> vO O 00 SD r-t cn vo + |>< o tN oo PQ /-> o oo <— i vD + |* CO i/~i ^ PQ /-v co m co vD •— < CN vO . I + X o in ■•— ' o CO Cm PQ O *-s rH CTl CN 00 + |VD m x m o in >— ' Pm PQ /-s PQ /-v O *H O CT\ O CT> VO rH cr\ •— i CN vD t—t vO + lx + IX o O ir\ v^ o ^ CO CN NO m r>. ir> CJ •H CO 4J OJ •H 60 CO "O QJ CD •O r-i c to t> e 60 CO c ■H CO a cu to a' vw O .rH CO CO CO V a o CJ ^-i •H CO 4-1 •H CO CO 3 QJ <-t "O CO C 4J to 60 C C tfl \ •H CJ CO tO QJ «4H 60 "O ExJ 0) OT rH 4J •H CO a) *o c CO to QJ a 60 o c rH •H CO CJ to CO "4-4 3 rH Sc CO w 4J E o m H H CJ O to c CN CN CU CO QJ to QJ to rH 4-» rH 4-t H 4J X> C X C X C In - M 3 U 3 C0 O tO O tO O on on co c cu to QJ (0 -Q C X C M 3 M 3 tO O tO O XX XX on co B «H QJ tO rH 4J XI C jh 3 fO O X 2C ON CO e ■H OJ CO fH 4J X) C U 3 to o £ X -21- ca. 475 m (1560 ft) elevation (Fig. 2). Marble Mountains 2 is the oldest midden, with a radiocarbon age of 10,465 + 330 B.P. (GX-6178; Table 1). The fossil assemblage differs considerably from the present sparse creosote bush-brittle bush vegetation (Table 2) . MM 2 contains higher elevation desertscrub species and is dominated by Ephedra cf . calif ornica and Lycium sp. (perhaps L, californicum) , neither of which presently occur near the site. Ephedra californica does occur within 6 km (3.6 mi) on bajadas as low as 490 m (1600 ft) elevation, but never on the low, bedrock slopes. Lycium californicum occurs in the Marble Mountains at least as low as 790 m (2600 ft) elevation on south facing rhyolite slopes. The leaves of Salvia cf. funerea and S. mohavensis (Mojave sage) occur in the MM 2 assemblage. The former persists at the site today but the latter does not. Salvia mohavensis occurs today in mixed desertscrub vegetation above ca. 1100 m (3700 ft) elevation in the Van Winkle Mountains, ca. 18 km (11 mi) to the north. Not only does' this macrofossil assemblage lack Larrea tridentata and Encelia f arinosa, the most important plants at the site today, but it also lacks the woodland species that are common in contemporaneous middens from low elevation sites further east. The Marble Mountains 1 and 3 middens contain late Holocene plant 14 macrofossils. The former midden has a C date of 3080 + 165 B.P. (GX-6177), and the latter, 1465 + 180 B.P. (GX-6179; Table 1). Both assemblages contain abundant Larrea tridentata and Encelia f arinosa, the current dominants at the sites (Table 2). Neither Suggests appreciable change in the local vegetation in the last 3100 years except, perhaps, in the distribution of Bebbia juncea and Peucephyllum schottii. Both -22- CO CO e 4J o c • c u to y 0 H CO •— \ — 1 u-l 2: 0- o lt-1 U bO 4J o U-l X c a en o O 0) «H 0) "O CO eo to - .l-i c 3 C M N 0) •H >— ' 5 CO XI liH •• £ CO >% 3 CJ C 5 4J y o c C -H y o> E a; c o v E .H CO 3 CO CO C E S CO CJ 4J cu e -H T> o l-i o u C - 4-1 o o C 3 <"> CO OJ .o #s 4J 0) cO •" ^ * c 0' *J CO CO HOC t-l i-l a > co t a ■H T> 4.) • »• CO 4J C c co C 01 CO 3 CO > c JO iH CO .O D •H o JO CO * oj < CO CO CO c bO 4J O CO >> c >% a •> i-l U CO M CO JO •H C OJ CO H E H i-l > CO CU « E •> -H CO u 0 » CO c co o "o m 4-1 C to J c 0) * • • a) J-l i-l M X OJ E QJ «H c u bO CX CO -H C CO co CO CO Q u >, o U 0) U-l •O U-l c >* c o 3 JD -H o O 0) 0 CO 3 3 J= X T> E 4J 4J a> cu U-l Vu M »-l o c «H CO o 0) JO C U p a) »- to c OJ OJ > <8 «H CO c o *E 0) oj a 3 JO §^h * c o >» U 0) i-t i-l 4J ^ »* CO ■•> tfl «H to CO OJ 4J t-l » o u O CO u niM CO 4-1 tH C 01 t-l «H Quo) ■ 6 ft. CO 4J m D--H i— 1 Z en I) « hi P-> CO c o CU -H CO i—l CO <— I JO M CO 2 Cu CN CQ CO O OJ c m H •H rn XI CO u 4J + CO c S 3 m O ^D £ (o-l C/j • l-^lltlllroiiii^niiimii CM I -4 I I — < | -4 | I I ~-l I •? ~4 I |^H| | | | fl | | l—l I I l«—i— i| | | | ftOI I X I CN r-4 I | fl I I I N I HH X CO ^^ /-N N-^ /-N T-l ■: /-s X ^H CO V-'^-N^-s X X to ^ o 3 XX ^ ^v w ^| 14-1 iH to CO V-/N-/ i-H 3 •H /— N CO CJ CU *-* JO CO >l C 4-1 X JO •H a, E CO ^-\ /— s 4-1 ■H ■H X l-i CO w D. ^. c •H o 3 QJ •H XI to X c •H C o« o 3 QJ X l-i -H JO r-l TO CO >-^ 1-1 w CO 4.) <""s o >> • o 00 CO u w o c u iH O CO iH O 4J X •H 4-1 a th C 4J >> U-l C •H 0) JO ^ «H CO /-s O CO CO O a CO CO > CO CO iH CO •H re -t-i JO 4-1 X CO 4J X u-i ■H >v d .-( 01 iH O CO ^H Eg 4-1 CJ c w O re >-• -H h i-l CJ •H 0) a CO l-i • O a 0 CO c rH CO 4J 4-1 c re o CO >> o CO -H JO /-». Uj OJ C o QJ • 3 u /-s -H CU c • iH iH c- I-l • 0) CO QJ c X cj u CO -H OJ u-i a CO X >. CO OJ a. H — E o o- *o •H O o OJ w C 3 >-l • JO u >^ l-i -a CO CO • 3 E CO to 1-1 c CO J= CO co 4J CO U-l c CO O CO •H CO re U-l i-l cu QJ 3 CO 4-1 CO JO ^H o U-l o E E o 3, . • E •H U Cu CO — i JO CJ i-l CU tfl CO i-) •H c -o 4-1 CO 3 3 I-l 4> a. cu CO 4-t CO i-l l-i >J OJ c TJ 0 CO •H c CO CJ CO CO C C 3 CJ CO -H rH O CO CO — — i o i-l -C CO CO N QO CO -H O -H >-i O O CU CO CO o re E QJ E i-t •H Bi >•> 4-1 o •H c_> •H •H 4-1 X c i-4 -a Cp bO c CJ CO — 1 l—l QJ O N C u 4-1 0) 4-1 QJ CO •H JO I-l M Cu CO •H QJ QJ o O o o •H 4-1 bo M ■H 4-1 cx C c CJ ■H bO •H 1-1 JO • o t^ 4-1 JO U JO ■H •H •H I-l — 1 CU C u CJ c •H 9 o 3 t-l M U JJ a' U-l J= • l-i •H u c a U I-l l-i CU -H >> CO re >> CU — 1 a D. 01 QJ < < Q cc J-l aJ CO C S =5 O S CO OJ C rH «H .O CO CO C S 3 O 0 (J | —< en I .-< I X CO CO ^^ u /— » •u X OJ X CO td S*> CO H •H •H 3 r-l CO S-i CO C O ■H 3 0) m «4-l l-i H ^ p •H CO a. <■ O — i O •-h <— < r» so vo oo — * I cm w m i o o a aw Z Z 2 M -24- are presently restricted to the mesic upper bajada in the vicinity of Locality A (Fig. 2) , but Bebbia juncea occurs in both samples and Peuce- phyllum schottii in MM 3 . Marble Mountains Localities B And C. The Marble Mountains 4 and 5 sites lie at 885 m and 900 m (2900 ft and 2950 ft) elevation, respec- tively. Samples from these two middens provide early, middle, and late Holocene-age macrofossil assemblages. The oldest, MM 5(1), is radio- carbon dated twice. Miscellaneous twigs and Prunus fasciculata (desert- 14 almond) seeds yield a C date of 10,090 + 380 B.P. (GX-6182) while a second date on Neotoma feces is contemporaneous within one standard deviation at 10,325 + 350 B.P. (GX-6183; Table 1). A single age, aver- aged from these two values (Long and Rippeteau 1974), is 10,210 + 260 B.P, Like MM 2 in the southern Marble Mountains, the MM 5(1) assemblage provides evidence for early Holocene mixed desertscrub vegetation at a site currently dominated by brittle bush and creosote bush (Table 3) . The MM 5(1) midden sample lacks woodland plants, with the exception of the rare occurence of Cowania mexicana (Table 3). Higher elevation desert shrubs, such as cf. Haplopappus laricifolius (turpentine brush), H. cuneatus, Prunus fasciculata, Coleogyne ramosissima (blackbrush) , Salvia mohavensis, and Yucca schidigera (Mojave yucca), are most common in the MM 5(1) assemblage. The diversity of this macrofossil assemblage is striking compared to the younger midden records from the same locality or the modern vegetation (Table 3). It is, however, no? unusual relative to the present mixed desertscrub communities on the north slopes of the Van Winkle Mountains at ca. 1220 m (4000 ft) elevation, 10 km (6 mi) to -25- ei c o x» « c a iH a x o C a o CN a H 0) to IT) /— » CM oj e m 4360 + 210 B.P. ,£) CO p j~»* CO C /- S J~ 260 B.P. 10,210 + CO en >> CO ^-' x» 0) c <* ^ iH •H •H .O ctj a M XJ u CO C ^s 0 S 3 CM •J a c ^ « xi c 3 0 r V •H JO u « r 1680 + 150 B.P. 9515 + 185 B.P. CO xi o C jo b. CO JO I I CM I I «-l I I NHHH | | I — i I | <■ I O I I I CM c o o H |rt |HHCSrtN-l |Nh I | ! —I I O CN I I I I w I I I I I ~H | rH CM I I I l l I I I en l I I I I l -< I I — * I I i—iim—il l l I l I cm en l i-h| I l 3 ^ ^"s c X ^-s X M *■" x /-* >-^ O >— ' X CJ CO -s 6 w 03 ■H -a CO X 3 0) •H > >H E W r-s. XJ 0) «»-s CO e a) 60 •H Xl CO *o X ■»-» o u •H CO /— \ CO >-* rH o w CO •H /^<^s ,fl M CO X ■•-» 3 s: •H to XX "H CO w CO . CO CJ /-N XJ CO 0) c s- ' ^ • • CO C iH Cl CO •H ^-n X C •H . to to 0) . . u-i U-i o CO O CO O CJ X x to •H Cl O CO u • Cl Cfl cl u cj e o c * o> c CO •H >-" u 60 to £ QJ CO • Cl CO CJ CO CO •H CO Cl U CO •H CO •H CO CO 60 3 XJ TD o. CO u aj OJ rJ X CO C 3 M • Uh c E 01 0) CO •o CO CO CO c CO J= ,C a> CO CO 0) CO a c 3 Cl CO M 3 cfl CO 3 3 •H XJ XJ QJ E A tO rH t-i U-l co e CO XJ O 3 60 JC CO •H CO CO CO rH t| H C C c XJ XJ OJ 3 E CO s: XJ CO XJ •H ^ *o "O ■H CO 1—1 CO CO >! CO C •H rco o CO CO 3 p rH u o 0) fO C CO o •H •H D. to CO QJ N N 60 •H cfl X) •H o •H •H O OJ U-I •H CO •H •H CO o c XJ XJ 01 CO ■H ^ •H •H O c XJ U X c rH iH 60| 01 c l-i o CJ U M M •H CO CO rH J- -O CJ J-i H 0) Cfl a 3 rC •H O) OJ O J= •H XJ o 0) CO CO J3 CO •H •H O XJ ,£ •H O O rH 5 >* U XJ j3 CJ CJ •H ' >-l Cm y 1 1 1 U i-j CO CO a) U J= JS O o u 3 •H CJ c c U •, •. «l 01 Vj < < < < < ea PQ CJ u U O U U a U w UJ ux w W U b -26- m /— - CNI (0 s^ CD c in H •H X> CD M J-J CO c s-~ >. S 3 O N— - a m OJ CO c 2 3 CM O *-* a <* c o CJ CI JQ 5 03 CD •H O 0) o\ CO I I I t i«-i| icni I 111— «tcsi i i i i i i i H | NN I I I ^h I l co i n h i r-> i m i n -' J-l o • U-J a- i-i 03 iH CO X u CO • • »H 03 03 03 CO CO JO u J= X v^ CO 4-1 CO .H 3 c OJ 1-1 u 03 CO a u o 4J o 01 X 3 03 U a, c 03 O o u o Mk3 D. CO . 03 J-i a. oj CO • a> o. o 03 H CO CD CO 03 03 3 to c/3 to u CO CJ o c CO 03 >J O 3 >- -3- r^ o Cl •-h mm tr> /— s ^-n B^S c 4J »a- to c O %«• • g •-N r* • 0) Cm •o rH • e X) « (0 t-l •"N CO -< o \D S CO 0) 4J § + ■H IT) CO CM cr> ■h co CJ ^ 4J •H H CO o o J CO CO 4J c 3 o s • CO QJ rH H O .O -Q CO R S CO E QJ 0 -C u u u_i U-I CO 0 QJ •H C U o CJ •H a. 4-1 CO CO C 4_i to c H CO a. H X m O V CO /-s *J QJ CU CO •H X /-n «H C •o • o 0) U-l «H E X CO T3 0) 3 t-l Xj"3 V> QJ o •H Cu u M •H •H CO fH W rH O o 4-1 CO •H u ^-s en «H .£> 4J 4-1 4-1 CD 3 . J3 o CO rH CO ^■v CO H X > to c CO 00 CO o a 4_i r-l QJ 4-1 CO X c w CO 01 • 3 o 1-< c CO •H co c CJ C CO CO CO QJ >w> QJ • « «H V4 U-4 60 E O o> o CJ CO •H 3 3 4-1 o 4J CJ o • Cu CO Cu C X) O C Cu • 01 o> c CO CO o E U a • C E C CO CO • QJ Cu CO QJ CO «H CO in a u o CO •H CO -H CO Cu a a) 3 •H u •a a. co co U V4 0) 0) CO «4-l c 3 M • U-l N CO CO CO CO T3 •0 CO CO CO CO c CO -a JS JS CO CO CO cu CO. »Cu 4-1 CO . «H 3 Cu •H CO rH CO CO 3 3 •H e 4-1 J-l X JC CO CO iH S-J U-4 co E — 1 3 Cu N CU O CO !u CO ■H QJ CO CO •H CO rH •n rH CO c c 4-1 4J 4-1 01 0) 3 o 4-1 CO cu cu rH 3 4-J •H .* CO "3 •a CO •H CO rH rH CO CO c 3 -H QJ o JC O CO •H rH •H U4 -O CO •H •H 4-1 a 4-1 .o •H • o O >> ?^ O . 4-1 JC o JS -H o M <-t 4J Cu • c J-l 1 1 *r U t-l u CO CO QJ t-i uj J3 JS l-i u 3 14-1 t-l o c Cu )-i CO QJ •H 3 CO U-4 3 CO < < < < < < CQ PQ u CJ cj u u U (J o UJ u w w Id U, bui O •x. CJ •-J hJ -30- 03 S C « c 3 o £ iH Xl M ^> CO i-H £ ^ I | i _< i _ j I | HCM | - < | I —4 | O I ~-4 I I I I — I — < | I -H | I «-" — I I I I |^_H| | IIO^H-H^r-H^CM sD CO cu c H ■H X) to J-l J-l to c £ 3 O £ 4J c 0 < *9 X) 03 o ■H O 0) I I I — H I I N | I -4 I | Icnl^l I HH(OHH | l-H| CO • 3 CD 4J c x a x o 0) • 60 CO X3 CM 0) ►J « « a •H u rH CO >, o JC o a J= o J-l c c o to 60 o . -H to a. i-i CO J-J • • X CO . c U-l O o £ CO O- «4-l co o X CO u E 3 •H U X H o E a. • »-i co a. o co CO o E CD >, >> CM o u at Pm X CO CO 1-1 x: j-1 0 X3 o •H DEI rc X CO CO a> u i—i to >j cu CO CM en o. to CO u CD CO 60 3 •H a. "3 >-i •H CO J= u CJ o CO 01 H cn st n*> 00 st fO riHinN N N <■ 00 CN -H ST ST o> cy> r^. -H ^ ST O O a. ojco Z 2 Z M -31- c 0 CN P-. N^ • o PO n c 0 _J CO O Q] C cr> iH •H •— ( p. K 0) JO J-4 to w -H to c ss 3 O c (0 O CN 53 ^ m u v_^ 0 cn x> CM P-. • to •">. • PL, H .—1 pa. • V-^ CQ 0) o> O CI 0) -H O c/5 .h to tNl VD JO C ,N >-: -H + . • to to +1 o> S JJ m C tn. m 0) 3 m m «J O « \D »4 2 O r- ^ «H i—i • Cm e *H • u ^s pa S O) a. in iH en \D en JO c CN c H ■H -* to to -H « 2 4J 4J c to c 3 o 3 O o> 0 2 00 2 ^ 4) H X) u i R5 0} £ O E JO o u I IM n CD 01 4-1 £ c 4J « Cl, o m JO i i ■ •— * i ^^ •— • •— t est i •— t i •— • i i i i f-H i I tn •-« l •-* —t lllC o •—i •— t i | NHH I | H I CM 'I H | | I —I I I »H i T-l <& «*5 *H I |CM| I NHHH | | | —4 — * | CN i | | » — I I «— I — t ■ I CM C"> ST /-n cn /-n X <0 /~s 3 X N-/ -H /-N tO X C >-• iH X *j ^ u •H O /■"N W tO o to ■H (4-1 X 4J 3 CJ CO •H *->. C E •H •H E O >-^ tO ^v 4J en •H ^N 10 C •o o OJ PC •H U-l e X| cn CO 3 M Xl-u o •H CJ u •H CO •H CO ^-1 •^ «H iH 4-J CO w to •H J-i /-* •H JO 1-4 en J-i 4-1 OJ 13 3 CO tH »H en H X > •H en CO PC CO •H CJ OJ to 0) c ^ O • • en 3 .-1 c cn 1-4 •H c cn • tn en O) • • • • u twin O PC • o OJ o >H U 3 3 • a o en CJ o a a a Cu J3 u cj E c a . (V c > en CJ Ou • a w e cu en • oj en oj en en CO CO CO CX CJ CO •H CO Cu Cu en 3 4-1 -o CX en cj a) OJ OJ M en c 3 M • . U-l • cn CO en m *o CO en en C CO en JO JO „c CO CO CO OJ CO u-i CO. a • 3 Cu C 3 tO CO 3 3 •H •H 4J 4-1 4-1 OJ JO JO CO rH 1-1 VH a cn e cn a a o cn o JO CO •H CO CO ■H i—l T~\ rH 4J c c c a 4-1 4-1 4-1 s 3 cn cu .H 3 03 M •u ■H j*: T3 TO CO tO rH o CO CO CO t» c c •H cn CO CO CO C to CO a 1-4 , C en o •H •H •H 60 CO OJ CO M N N to CO CO JO -H o •H Li J- o CJ o Cu to CO ■H a CO O c 4-1 4-1 e tO •H J*! c ■H •H •H o 4-1 4J M X c ^H TO •a PO 3 MJ O 30 Cu CJ o u M •H en en OJ l-i 40 CJ 0) M u J-i 01 a. a 3 CO •H OJ CU 0) o 4-1 CO rH •H OJ u CO JO w •H ■H 4-1 4-> JO •H to o o o iH s^ ^ CJ 4-1 JO o JO JO •H cn rH CU • c ex PD 1 P 1 M M 1-1 en O) l-i JO JO JO JO o 1-1 i-i 3 •H CJ c Cu Cu l-i 0) •H CO IM 3 in < ■5 < < < < PQ BQ u u u u u o u O CO u w w M w fn In 30 CJ -> -32- cs o\ ai to tH C •° TJ ■p co re j-j * § o ^ I • • I I I I «-H I «-• I I HP) I I —i I I I I CSJ | — i — ( | — I 4J I C O o CSI 0) fH 03 .o c (0 CO 3 O s 0 r-1 03 .£> C td to o 2 03 0) •H u 01 CO f « — ' — ■ I I I I I «-< *- • — <—<»-« i i . — i i ., — i i i —i —i i in —i —i I —i — i CM I I— ICMI | ICMI I I 1^1 I I I^OI I | HNHHol | H(0 I X X to AJ to e 01 •o •H • u 03 to CU J-i to 03 • 3 ax 03 *J c X! a x o 03 W iH 01 • 60f d. -H 03 X CO to a iH VJ fH CO >. cj X o a X o AJ c c o CO 00 o •H CO i-i AJ • (4-1 CO Cu cfl u aj CO CO o > >^ !-i 4-1 3 CO CJ iH 0) Cu 1-1 CO . e 4-1 0) CJ S 03 a to en u a 03 to X >i *s U o B H 0) X o U-l •H 03 03 • CO O. I-i 03 CJ CO en U* X >J CO Cu X a c 03 o u 03 CO B cu > ra j3 . o ^e CO 03 CJ J3 as J= • a to CO . 03 -H E 03 03 O CO 01 3 IM u AJ CO a ■H 01 • 03 0) u > to a o u • CO 0) •H 03 a H a. u »J 13 u ra 03 O £* •H CO CO Vj C •C •H in CU CO CO CO O > CO Dk 03 CJ 03 i-H . J= •H ^ u CO V4-I Q. AJ J= 3 • C/j O c/3 to H >i H -33- CN C> 0) CT; r-i c .c ■rl U CO a 4-1 X c o o 53 CN as 0) r-{ CO jo C CO CO o s rH CO .o C ^ -H co co O X en to ■>» U -0 a a •H u u 9 •a- cj> oo o cm en CN m r^ ej\ m en «-i . •— i •—I N-^ ^J- o o Q. O.CO Z 2 Z M -34- occassional in MM 7(1) and abundant enough in MM 9(1). to provide suf- ficient mass for radiocarbon dating (Table 1). Ribes cf. velutinum ' in MM 9(1)1 represents an intriguing early Holocene occurence (Table AB) . The- leaves of a sagebrush (Artemisia sec. Tridentatae) occur in both MM 7(1) and MM 9(l)j and, while not strictly a woodland species, it is an occurence not replicated by the older Locality B samples. The xeric desertscrub species that characterize the Locality C vegetation today are rare or lacking in MM 7(1) and MM 9(1)., with the exception of Encelia farinosa (Table AA, AB) . Brittle bush is the first of the warm desert dominants to appear in the Locality B fossil record. Such appears to be the case at Locality C as well. Marble Mountains 6, at 7930 +285 B.P. (GX-6185; Tbale 1), is ca. 1000 years younger than MM 7(1) and MM 9(1). and is a mix of some extralocal desertscrub species and the xeric shrubs that occur at the site today. Larrea tridentata is the most abundant fossil species in MM 6. Salvia mohavensis and cf . Haplopappus laricifolius are common and Opuntia acanthocarpa occassional extralocals (Table AA) . Juniperus sp. , Artemisia sec. Tridentatae, Coleogyne ramosissima, Haplopappus cuneatus, Yucca brevifolia, and Y_. yhipplei are lacking. Accompanying the dramatic increase of Larrea in the fossil record, and the demise of most mesophytic species, is the appearance of Peucephyllum schottii and an apparent increase of Ambrosia dumosa. Two annuals occur in relatively large amounts in MM 6, Eschscholtzia sp. (California poppy) and Filago sp. (Ta*ble AA) . Marble Mountains 7(2), at AA75 + 170 B.P. (GX-6187; Table 1), is the youngest reliably dated Locality C midden sample.- There is little discernable difference between this assemblage and the present vegetation -35- at the site. Both are dominated by Larrea tridentata and Encelia farinosa, with smaller amounts of Ambrosia dumosa, Bebbia juncea, and Peucephyllum frhottii (Table 4A) . A trace amount of Salvia mohavensis occurs, but this may .well be a contaminant. MM 7(2), as well as the older MM 7(1) assem- blage, contains the seeds and spines of Ferocactus acanthodes , which does not occur at Locality C today. The Eureka Valley And Vicinity The Horse Thief Hills localities (Fig. 3) are unproductive areas for midden sites, the rocks are fractured and metamorphosed with few cav- ities. This is unsurprising since these localities are within half a kilometer of the Furnace Creek Faultzone. The activity of the fault pro- bably produces an unstable geomorphic setting. However, the cliffs on the northeast border of the Eureka Valley, several kilometers further south (Fig. 3), are a more productive midden area. Horse Thief Hills Locality A. Horse Thief Hills 1 is a small solu- tion cavity ca. 30 m above the bed of an arroyo tributary to Willow Wash (Fig. 3). The east sloping talus below the cliff containing the cavity, at 1615 m (5300 ft) elevation, is derived from dolomite of the Bonanza King Formation. The local plant community is dominated by the Great Basin shrubs Atriplex conf ertifolia and Chrysothamnus viscidiflorus. In contrast to the present Great Basin desertscrub community, the Horse Thief Hills 1 midden contains abundant woodland species. The twigs and seeds of the most common fossil, Juniperus osteosperma (Utah juniper), 14 are C dated at 10,690 + 280 B.P. (GX-6217; Table 5). Symphoricarpos c^« longiflorus is also abundant while extralocal woodland taxa include, -36- cO •H -O U OJ CU 4-> 4J fl SO TD 33 0) u CO /—> ♦o . o e c o XI • J-i JQ CO CO V tH O »-' CO OS >S 4-> M CO CO 4-1 E •H •H J2 J-i CO CU x. o *« c O 01 4J tH •h a CO B CO CO CO OJ ex CO o 0) 4J CO o CO >-l 4-1 CJ CO cu 3 CO O a> T3 c C CO CO H tH CO 0) 60 CJ •H CO o 00 /-~\ CM r^ -H + CM vo o 1 o> X vD o Pm • /-s PQ 00 i— I O CM O vD CM I v u CQ CO U-4 a> VM 60 •H •a iH a> CJ iH CU a) 4-> CO 4J CO •H OJ •H CU 0 a. 6 a o o o o tH .H tH tH O CD o CO t) *o CO CO 60 3 60 3 C H c iH •H CO •H CO CJ 4J CJ 4J a) CO cw -o U-l T3 C C W CO W CO CO 60 •H * 4-1 H CO CO OJ 3 CU CJ O CJ QJ CU CU IN c CO IH CO tH CO E tH E o 0) o 4_) CJ 4-1 o CO O OJ •H CU 25 s SB Pm cu o > 5 60 c «H CJ CO <4-l Cm- CQ CTN PQ o PQ — ■ i .—1 CM CM O CM O CM O CM o VD O vd O vD CM 1 CM 1 CM | V V X! V X CO 60 •H CO s 4-1 4J ' to * •u CO 2 3 CU o •o CU •H c M CO - 1 4-1 tH (0 . CU CU CO 4-1 U u 60 c CO u •H o •H to > CJ 23 -J 4J O CM fcJ to -d" CM H CM — 1 CM < CM i*<* O vD + 1 1 X o X m o 2= Q CI CO CO CO U-i 3 vw T3 tH c S CO CO CO CO 4-1 S S S S E m m m o »o i— i CO r^ * » t— i vD vD vD m m •a- •— 1 t— l —4 ■H — i CO CO tA •J *~s /-N <^N .-9 •J •— i CN CN M M •>— ' >— ' \— ' < EC «-h «N S PO cn m •-H CN Pk CO CO fo CO CO CQ , S ? £ w < tH tH W PQ tH tH tH W CU CU M tH tH M iH tH iH M «H «H S3 >h •H 3= •H as B •H as •H as ■H as > e > > M CO *o • o c c o X) CO CO CJ o •H *o CO U CO CO 4-1 E «H ■H jO H CO en •H co O c CO t-l r-H 0) CJ CO •H S DP O o 00 m CN I x: o m >-> £> 01 "O o •H c CO 00 o DP o m 00 CN — I CM CO 4-> CO 4-J 05 c •H CJ fc "O & •H OJ »-i •o J-l CJ CO •H 0) CO c v-4 00 CO Iw •H bO cO 5 J-i ■J 4-1 o + 1 o CO o en Pm PQ O v£> m cm -M CM +Vi in cj> co ^-^ m E Pm • DP O f- —I CM CM CM ♦'A m o ON "> — m in 03 03 0) 3 00 iH •o CO a> 4J o f-t •H o 4J V •H •H 4J 4-1 0} E •h cd •H Q> o E o- E oo «H o o o *o O t-l iH tH 0) •O OJ = O 03 O -H a iH U iH OJ OJ «-l CJ C 03 o •H CJ £ Ph • DP © 00 ON CM -* CM + 'i m cj on ^— ' r>» P-. • DP O ON CM CM CN CM + l^ m o co w vr m CN co 0) CX 03 o OJ 4-1 03 O CO >-i OJ C 03 00 ■H 4J Oh DP o CO ^n in o + |CN vO O I CM X 0) 3 « 4-1 CJ •H 4_> 03 ■H 0} E CO o -a >-{ 0) O iH •o 60 c C CO •H CJ 0) co a u_i o CO 03 S o CO -a- CO 00 •H 4J 03 3 O O) C CO 0) CJ CO ■H a DP o — < m co CM CM o o CO »—' co CO • 4J n U N-X w v^ < CQ c CN «N co sr <• «* m m 0 u » > > » 5 5 5 > s^ • OJ 0) OJ a) 0) a> a) 0) O •H •H •H ■H •H •H •H •H «o •* c > > > > > > > > 0 Q) CU CO CO cd cd CO CO a) CO "< 4J fH co b ^ J2 ^ ^ i2 ,2 J2 ^ A 0) OJ 0) 01 Ol 0) 0) Ol «: h u h h h h u h k CO 9 3 3 3 3 3 3 3 CO H W H U W W W W -38- besides Utah juniper, Amelanchier utahensis (service-berry), Chamae- batiaria millefolium (fern-bush), Purshia cf. tridentata (bitterbrush) , Ephedra cf. viridis, and Cercocarpus intricatus (Table 6). The last two species occur in restricted mesic habitats within half a kilometer of HTH 1. Other plants that do not occur at the site today, such as Artemisia sec. Tridentatae, Prunus cf. andersonii (desert-peach), and Salvia dorrii, have affinities with mesic desertscrub communities although many are also common in xeric woodlands. Atriplex conferti- folia is the only plant common at the site today that is also abundant in the macrofossil assemblage (Table 6). The Horse Thief Hills 2 site, at 1635 m (5360 ft) elevation, is near a crest overlooking Willow Wash. The site is not as protected as HTH 1 and the plant community is correspondingly more xeric. Atriplex confertifolia dominates the local community with Eurotia lanata, Ephedra nevadensis, and Hecastocleis shockleyi being common. The HTH 2 packrat midden is a partly indurated mass with a radiocarbon date of <200 B.P., confirming its young appearance (GX-6218; Table 5). The most common plants in the midden appear to be the most common in the surrounding vegetation (Table 7). Horse Thief Hills Locality B. The single midden site at Locality B, HTH 3, is isolated in the middle of Willow Wash canyon with a less rugged topography than the preceding sites. A west sloping alluvial fan leads away from the rock outcrop, coalescing with the Willows/ash fan ca. 35 m from the midden site. Atriplex confertifolia is the dominant shrub with Da lea fremontii and Hymenoclea salsola common associates. The latter is Table 6. Plants from the Horse Thief Hills 1 site, explanation of the symbols. See Table 2 for an Sp_e< Horse Thief Hills 1 es (10,690 + 280 B.P.) Anelanchier titahensis Amsinckia intermedia Aristida sp. (X) Artemisia sec. Tridentatae Astragalus sp. Atriplex confertifolia (X) Erickellia sp. Cercocarpus intricatus Chamaebat iar ia millefolium Chenopodium sp. Chrysothamnus viscidif lorus ( X_ ) cf. Chrysothamnus sp. Crypt ant ha sp. Dalea f remontii (X) Ephedra nevadensis (X) Ephedra cf. viridis Eriogonum heermannii (X) Eriogonum sp. Eurotia lanata (X) Fendlerella utahensis (X) Forsellesia nevadensis (X) Forsellesia sp. Galium sp. (X) Hecastocleis shockleyi (X) Hilaria jamesii (X) Juniperus osteosperma Mirabilis multiflora (X) Mortonia utahensis *l Opuntia basilaris (X) Opuntia cf . polyacantha Oryzopsis hymenoides (X) Penstemon sp. cf . Prunus andersonii Purshia cf . tridentata Salvia dorrii Scopulophila rixfordii Sphaeralcea sp. _Stanleya cf. pinnata (X) Symphoricarpos cf. longiflorus (X) Tetradymia axillaris (X) Tetradymia cf. canescens (X) Tetradymia sp. 2 1 3 1 A 1 2 1 1 1 1 1 2 5 1 3 1 1 3 1 1 1 2 N (19) NP (14) VN (100) (74%) IS 27 19 70% 35 -40- Table 7. Plants from the Horse Thief Hills 2 site, explanation of the symbols. See Table 2 for an S£ ecies Apropyron sp. Artemisia spinescens (X) Atriplex conf ertifolia (X_) Brickellia sp. Chrysothamnus viscidif lorus (X) Ephedra nevadensis (X) Eriogonum heermannii (X) Eurotia lanata (X) Fendlerella utahensis (X) Forsellesia nevadensis Gilia cf. latifolia (X) Hecastocleis shockleyi (X) Hilaria jamesii (X) Lepidium cf. fremontii Menodora spinescens (X) Mirabilis sp. Opuntia basilaris Oryzopsis hymenoides (X) Prunus f asciculata (X) Scopulophila rixfordii (X) Sphaeralcea sp. Stanleya cf. pinnata Symphoricarpos cf. longiflorus (X) Tetradymia axillaris (X) Horse Thief Hills 2 (<200 B.P.) N Np Nn/N (100) IS N(16) (13) (81%) 1 2 5 1 2 3 3 2 1 2 2 1 2 1 2 1 2 1 1 3 3 21 16 76% 70 -41- a disturbance plant normally restricted to washes. Both midden samples are well indurated but both yield subrecent radiocarbon dates (<200 B.P.; Table 5). The date on Horse Thief Hills 3(1) is on Neotoma feces (GX-6219) while the two subrecent dates on HTH 3(2) are on Neotoma feces and undifferentiated twigs (GX-6221 and GX-6220, respectively; Table 5). Both midden assemblages agree with the present plant community in that they are dominated by Atriplex conf ertifolia with common Dalea fremontii and Ephedra sp. (probably E_. nevadensis ; Table 8). However, both samples diverge from the modern vegetation in other aspects. Opuntia echinocarpa, present at the locality but not within 30 m of the midden site, is abun- dant in HTH 3(1) and probably represented by abundant Opuntia sp. spines in HTH 3(2) (Table 8). HTH 3(1) contains no Hymenoclea salsola but does contain the frequent remains of Lycium sp. and Menodora spinescens, both absent at the locality today. Hymenoclea salsola and Menodora spinescens are occassional in HTH 3(2) and Lycium sp. represented by rare fragments (Table 8). Larrea tridentata and Yucca brevifolia are scattered at Locality B, although none occur at the HTH 3 site. The remains of these two plants are found only in the HTH 3(2) assemblage. The Eureka View Locality. Eureka View is a network of midden sites on rugged exposures of limestone and dolomite rising above Eureka Valley (Fig. 3). The desertscrub communities on the dark, south to west facing calcareous slopes are diverse. All the sites occur within 100 m of each other and offer a time transgressive serfes of fossil oiddens from one locality and one community type, dominated by Atriplex confertifolia and Larrea tridentata. -42- Table 8. Plants from the Horse Thief Hills 3 site, explanation of the symbols. See Table 2 for an Species Horse Thief Hills 3 HTH 3(1) HTH 3(2) (<200 B.P.) (<200 B.P.) Acamptopappus sphaerocephalus (X) Amsinckia tesselata Aristida sp. (X) Artemisia spinescens Atriplex conf ertifolia (X) Chaenactis sp. Chrysothamnus viscidif lorus (X) Dalea fremontii (X) Ephedra nevadensis (X) Ephedra sp. Eriogonum fasciculatum (X) E. heermannii (X) E. inflatum (X) Eriogonum sp. Erioneuron pulchellum (X) Eurotia lanata (X) Grayia spinosa (X) Hapiopappus laricifolius (X) cf . Hapiopappus laricifolius Hilaria jamesii (X) Hymenoclea salsola (X) Larrea tridentata Leucelene ericoides (X) Lycium sp. Menodora spinescens Opuntia basilaris (X) (h echinocarpa (X) Opuntia sp. Oryzopsis hymenoides (X) Prunus fasciculata (X) Scopulophila rixfordii Sphaeralcea sp. (X) Stanleya cf. pinnata (X) Tetradymia axillaris (X) T. canescens (X) Yucca brevifolia (X) 2 3 4 1 1 1 2 2 M (25) 13 1? Np (19) 10 13 Np/N (100) (76%) 77Z 762 I? •» A7 48 -43- Eureka View 5, at ca. 1430 m (4690 ft), is the lowest elevation site and provides the two oldest macrofossil assemblages. It is about 10 m above the head of the alluvial fan that slopes away to the Eureka Valley bajada. EV 5A contains abundant Utah juniper twigs and seeds, 1/+C dated at 14,720 + 530 B.P. (GX-6230; Table 5). Other extralocals in this late Wisconsin woodland assemblage include Ephedra cf . viridis,_ Artemisia sec. Tridentatae, Forsellesia nevadensis, Salvia dorrii, Hecastocleis shockleyi and, remarkably, Pinus cf. f lexilis (limber pine; Table 9). Atriplex confertifolia is the only common desertscrub species in the EV 5A sample and the only fossil plant that is well- represented in the modern vegetation. Eureka View 5B, at 8330 + 250 B.P. (GX-6231), is the only early Holocene macrofossil assemblage and documents an apparent decline in species diversity at this site (Table 9). The woodland and Great Basin desertscrub species, common in the late Wisconsin EV 5A sample, are lacking in EV 5B along with most of the species present at the site today. Shadscale is the abundant fossil in EV 5B and Ephedra sp. is common with the extralocal Lepidium cf. fremontii (Mojave peppergrass) occassional (Table 9). The Eureka View 4 site, located ca. 20 m west of the EV 5 site, contains the next oldest suite of middens (Table 5). EV 4(2), at 6795 +190 B.P. (GX-6228 on cf. Atriplex confertifolia wood), is domi- nated by shadscale and contains common Haplopappus brickellioides and Encella virginensis (Table 9). The latter does not occur at the site tQday. Opuntia basilaris, Lycium sp., and Dalea fremontii are other frequent extralocals in the EV 4(2) assemblage. Ambrosia dumosa first •PPears in this sample. -44- * 8330 + > 250 B.P, £ 14,720 > + 530 B.P, 5 5435 + * 220 B.P. u 5 6795 ± f 190 B.P. 5595 + 210 B.P. i- « 150 B.P. W > £ 3930 + 180 B.P. « 1580 + 140 B.P. NO DATE | a 2635 + ^ * 140 B.P. w > I I — • I I l I m i i i i - I Ihh i Nnrt | | i n i N-— * C X l-i >--■ O O CO •H T3 > •H 0) cu: t- -w -O u CJ i) o ,c uj;_> a. u u U 3 .-I CO X. D. OJ LI > — ( O a. n 4J U n u o c c u ui|lO| X X u *^ .-< • 3 U) a. a c K 0) c u E o 3 3 t-l C D -M | ~* | *-| I i | loll I I iocmi-^i II l-a-l I I t l«-«l I I i-^i-h llllr-illlll-Hlcsillllvflllll I | | i — l i i o I —i I m I —* I I n l n I I I »H — i i I i —• I I I x W > I MINI I I I lO |H I I I I I I I N I N I I I I I I -* I — ' I U > | | | 1 — . I | | I 04 | | I I I I H(1 I PIHrt I I ~ I I I — « I •— • I I i -* -« I i I I im-H| I I n I NwN I rtH I I I I I N I < CM > Ul cni- Il<-t|||~4||srr4ltllll~4H t-t ft a «H K ■ ■ O •n) «-l s ^H ^>> i-l •H c M-t J= ~-( >, (9 y-v «-l a < *4 ^-* u •H w • -c o ro tM _c o u o — H u -w 3 tH • o ^-t u 3 a. u o i_i I— 1 i-l u ro re a 9 V X U u a. a c ^ E -O —i 10 *J B c • ^ i— i a. • c 0) •H o to ^-v o X ~-l n CD o . . re Q. ^. u -r-t u a. E ai • t H M-l ro re X CU| re *-* re W 0 n — < o -o u a t-i 0) X J-> to ro • 10 3 u Cm M-l U-l u o — i re **" u — < ~- 3 *— 3 'J ■«H U 01 . u y-' n u Cm i— I u cr. u v. o M^ 09 re ■ • N C — 1 a. -H to M D. r: in w X) o (0 c — H • o kM £ (J • U Cy . C- O Cm u Cj o 3 W E E IS -C •H • c c c >, re ro to UmI "C u Ch — 1 Q. "H —1 (A U ro ■^ ro c h 3 => u t— i a. r, re --M re E X — c -mH O ^i O ro 10 l-i — M hi C i- Cm i-l CJ P! -h •^ n ro «H w »< ■M ^1 M-l Oi Cm —i — M — * re o -H u O a> rat ai o ro G <-* ro — " u •M< •#H ■#■4 ro rt c c a c tn o U O n z> > Cm ro D. c u •H u r_ a. •J c u Cm C1 u • t- c CJ 3 s C D ro re > c -H • o X •H K a ••"M 3 (S • ro MM 1 ro -1 D. • % a> ey x: JZ £ ■H re U-l (J C. -J u. o|o 3= = X — — 1 -J -J -J — u X z z o o o a. i. c- Cl, a. Cm -o o trt \n mO to -A 6- i i i i i i i i i i cm vo o r»- Jx» N j rs CM <-l VD —I f> O Is* -h H rt N >J en o r> — t rt ^ n ri sD -h Os 00 •— I •-! VO C"> u 2 | t* < V 1 CM 1 1 1 1 1 I O 1 1 CM 1 »<« •-H 1 1 CO —i — i 1 1 —l *— t vo m 1 1 1 1 1 ■ 1 - oa 1 ► (N 1 1 1 1 CM ro CA 00 e > 1 1 cs — i m m 1 1 3 1 ij I < ' u 1 ! ~H «» sC * — i 1 « i 11 3 -i — m i i Ol CM f> CM i — « *—* vC so >• > i i i i J i t i i 1 ^•N ^-x «S ^v i SO m 00 1 «H X i CM •_« *n I •-1 x^ i x^» X_^ N^ o i •— (5 i — *-i i ec M 1 c »-4 0 3 i ^* u i .£ ■H i O a i i «3 u i /-% >»4 i o e E ffl i o 4 1 * i ■w 1 V. 1 "O 3 i i z 1 H < i CL O.CA > i 2 S5 z •-» -45- Is 0 f> u; > « — W > . — - t < — . I — . I I I I I CM I I i— I | | i i c> m I -< I I — • I I t 1 1 I I t t - — • I I t —< I t _4| | I I I -4 I I | .-« i-l -H -< | t-l | »H I I I toil I I IC=>c^|^^l I I 1-3-1 t I I I— 'I I I l«— «l— ^1 lll!cnll»ll~«ioiiiii w ^N Wl X , CO ^-s — O. — •^ X o Ti o AQ o. iH 01 e Xl iJ 00 >> «H ^ u- . T3 ■H c 01 —- ( »-s .H w c X 4J •r4 O. Wi 0 01 u 61 X J* u o CO o. iJ CO en 0) co a •o c .* w {J Oh CO £ hi 0 M O IJ en -r-< IM IM CO m U O H u ai 0> c i-< X CO CO H a X 61 > o -H in • .e o (0 u JS 0 u u — i U -r^ 3 •H • 01 M i~ 3 a. to Oi IJ IM u 00 n CO CO 3 u X u 11 Q. CO c -» E J3 ■^i 09 *-> B c • "T* fH Cl c V ■w u 10 ^-O X •—I in 10 u . . CO Q. «-n M ■H U & c 01 • • «— I iH re re X Cv • IB ^ CO to 0 m —I o -a u-< a. i-* en x *■> W CO « (0 3 t-l ta. >w im t-l u .-H cfl *-» u O. — ' CO u Q. t-i U cr. u u o •H a CO » V. • N A -H Cj — h 10 hi Q. C n CO X) o m C —1 • O u -C u • u * . ** .a. v C. U a U 3 4J E E IS SZ -H • c B o >. CO re w vui -o o c -^ a. «h 1. « — < w u n ■H re o l-i 3 3 (J i—l o. a CO •^ CO E £ ■H — H CJ IM o re 10 t-i u -^ m a U c_ •u O CO -H ■^ E f- •^ in •»" — jZ T-l Cv Bt -H — ^ *— « CO u <—l M o tj 3 01 col 01 o CO G B! Cm 01 -a •o *a X) u ■j-j o i-1 4-1 CJ OJ Oi re in -rJ to 3 0) re XJ v 4J V) -J ■»-( •— 1 — -1 re — « u ■M *M •r« CO ct 0 P "■J c r/1 u CJ o in 3 > Ch re D< a L. 10 i~ — ( u r- a u e u a O. U • 1* c u 3 3 C 3 n CO ^ C — H • o -C ••-* t IV, 0 >u •»< ro •H Cn • "i. 0) 0» JS j; £ -tH re U-l CJ c ~> u. U u X SB X — -1 J _1 -J -J u S ■z. z O o O o- a- u. a, — C- cy5 o to io W to -46- i i i i i i i i (N \o o r- cn sj- O r^ — < —t -* r~ oo ~h .-t vO r"> C> ST O CO *h — »■% I so irv oo I X I cm -h m i >-/ | v^ v»* s-^ I id I H> o o a o.cA 2 2 z m -47- Eureka View 4(1) and Eureka View 4(3) have radiocarbon dates that are contemporaneous within one standard deviation and are ca. 1300 years K, younger than EV 4(2). A date of 5595 + 210 B.P. (GX-6227) for EV 4(1) is slightly older than a date of 5435 + 220 B.P. (GX-6229; Table 5) for EV 4(3). Atriplex confertifolia is the primary constituent of both assemblages with Encelia virginensis and Opuntia basilaris being abun- dant extralocals in EV 4(3) and common in EV 4(1) (Table 9). EV 4(1) is the oldest midden sample to contain the spines of Opuntia sp. (perhaps 0. echinocarpa) . Erioneuron pulchellum, which made its earliest appearance in EV 4(2), is occassional in both EV 4(1) and EV 4(3). Dalea fremontii, common in EV 4(2), is represented by only rare occur- ences in these two younger samples. A few fragments of Larrea triden- tata occur in the youngest assemblage, EV 4(3) (Table 9). A single twig of Juniperus osteosperma in EV 4(1) is a presumed contaminant. The Eureka View 2C macrofossil assemblage, dated at 3930 + 180 B.P. (GX-6225; Table 5), is ca. 1500 radiocarbon years younger than EV 4(3) and the youngest middle Holocene sample. Like the preceding assemblages, EV 2C is dominated by Atriplex confertifolia and contains common Opuntia basilaris. It contains two other cacti, Opuntia sp. and Echinocactus polycephalus (Table 9) . This is the oldest midden sample to contain cotton-top barrel cactus as well as appreciable quantities of Larrea trldentata and Peucephyllum schottii. Desert spruce was not found in the census of the modern vegetation. The nearest population known to me occur* on south facing slopes in lower Hanging Rock Canyon at ca. 1370 m (ca. A500 ft) elevation, 12 km (7 mi) to the southwest in the Last Chance Range. -48- Eureka View 1, at 1510 m (4950 ft), is the highest elevation site. At 2635 + 140 B.P. (GX-6222; Table 5) it is the oldest assemblage with an array of species closely resembling the modern community (Table 9). The twigs and leaves of creosote bush and shadscale are abundant. Viguiera reticulata and Ambrosia dumosa are both common in this sample, the former appearing in the fossil record for the first time. Echino- cactus polycephalus, Lepidium cf . fremontii, and Opuntia echinocarpa are occassional in EV 1, but they do not occur at the site today. A radiocarbon date of 1580 + 140 B.P. (GX-6224) for EV 2B places it ca. 2400 radiocarbon years younger than EV 2C, its companion midden from the same site (Table 5). The most important plants in EV 2B are the dominant species in the present vegetation, although it contains common beavertail prickly pear, Opuntia sp. (probably Ch echinocarpa) , Peucephyllum schottii, and Dalea fremontii, desertscrub species not pre- sent at the site today. (table 9). Amphipappus fremontii, a common plant at the Eureka View Locality today, makes a belated appearance in the fossil record in EV 2B. s' Eureka View 3 is the youngest radiocarbon dated midden from this locality. A C date of 535 + 150 B.P. (GX-6226; Table 5) is associated with a macrofossil assemblage containing abundant Larrea tridentata and common Atriplex confertifolia (Table 9) . It contains only two species that cannot be found at the site today, Opuntia sp. and Peucephyllum schottii. The last midden from this series, EV 2A, ha* no radiocarbon date. This sample of Larrea twigs was lost at the radiocarbon lab. Its similarity to other late Holocene macrofossil assemblages suggests that *t is less than 3500 years old.. -49- The Regional Fossil Record Late Quaternary age packrat middens have been found throughout- the Southwest, from Mexico as far south as Lat. 26 N (Van Devender 1978), to Trans-Pecos Texas and southern New Mexico (Van Devender et al. 1978, 1979; Van Devender and Everitt 1977; Wells 1966), west to the inland flanks of the Sierra Nevada (Van Devender and Spaulding 1979) , north to east-central Nevada (Thompson 1978). Fossil packrat middens from within the California Desert Conservation Area and adjacent Nevada and Arizona provide a biogeographic and temporal framework from which to assess the records presented up to this point. The latest Wisconsin (14,000 B.P. to 11,000 B.P.), early Holocene (11,000 B.P. to 7,800 B.P.), and then middle and late Holocene (7,800 B.P. to present) midden records will be dealt with in that order. The presentation is further separated into low ( < 1000 m; < 3280 ft) and intermediate (1000 m to 1800 m; 3280 ft to 5900 ft) elevation sites. Ancient packrat middens have been recovered from elevations as great as 2400 m (7870 ft) in Mojave Desert mountain ranges (Van Devender and Spaulding 1979). The Latest Wisconsin At Low Elevations. Five midden sites from the Kofa (Lat. 33°20' N, Lon. 114°00' W) and New Water Mountains (Lat. 33 36' N; Lon. 113 55' W; Fig. 1) of Yuma County, Arizona were studied by Van Devender (1973). The sites range in elevation from 550 m to 860 m (1800 ft to 2820 ft) in paloverde-saguaro (Cercidium micro phyllum- Cereus giganteus) vegetation more characteristic of the Arizona Uplands •ection of the eastern Sonoran Desert (Shreve and Wiggins 1964). Palo- verde-saguaro communities are zoned at higher elevations in the mountains t0 the east of the Colorado River Valley (Brown and Lowe 1977). The -50- fossil middens provide evidence for a xeric juniper woodland with lesser amounts of shrub live oak (Quercus turbinella) . Only the oldest mid^deh from the highest elevation site (Burro Canyon #1(1)), C dated at 14,400 + 330 B.P. (A-1315), contains abundant Pinus monophylla. Other woodland plants such as Rhus trilobata (skunk-bush), Berberis sp. (barberry), and northern desert species like Thamnosma montana (turpentine- broom) , Yucca brevifolia, and Atriplex conf ertifolia are present in the middens. The data also suggests the absence of typical, frost-sensitive Sonoran Desert species including Cercidium microphyllum and Cereus giganteus (Van Devender 1973) . Packrat middens from the Newberry Mountains in extreme southern Nevada (Lat. 35°15' N, Lon. 114°37' W; Fig. 1) were studied by Leskinen (1975) . A midden sample from 850 m (2790 ft) elevation in modern creo- sote bush vegetation documents an ancient pinyon-juniper-live oak wood- land. The oak species are identified as Quercus chrysolepis and G> dunnii and are associated with a radiocarbon date of 13,380 + 300 B.P. (GaK-1988) . The most comprehensive study of Mojave Desert paleocommunities to date is that by Phillips (1977). His Lower Grand Canyon study area (Lat. 36°06' N; Lon. 113°55' W; Tig. 1) encompasses 25 midden sites ranging from 440 m to 675 m (1440 ft to 2215 ft) elevation on calcareous slopes of the Inner Gorge. The present open slope desertscrub communi- ties are often dominated by Larrea tridentata and Encelia farinosa. Other important species include Fouquieria splendens (ocotillo) , Echino- cactus polycephalus, and Viguiera deltoidea (desert marigold) . The The latest Wisconsin of the Lower Grand Canyon, corresponding to Phillips' v«getational Stage III, was characterized by a woodland dominated by -51- juniper (as Juniperus sp.) and single-leaf ash (Fraxinus anomala) . Important associates include Atriplex confertifolia, Symphoricarpos sp. , Mortonia scabrella (tick-weed) , Nolina microcarpa (bear-grass) , Prunus fasciculata, and Yucca baccata (banana yucca) . Various Mojave Desert plants that do not occur in the Inner Gorge today were widespread in- cluding Coleogyne ramosissima, Yucca brevifolia, and Salvia dorrii (Phillips 1977). Only one high elevation midden from a north slope records a pinyon-juniper woodland. Desert Almond //10A at 635 m (2080 ft) contains both Pinus monophylla and P_. edulis (Rocky Mountain pinyon) associated with a radiocarbon date of 12,650 + 380 B.P. (A-1720; Phillips 1977). The Owl Canyon midden site is on the southeastern edge of the Amargosa Desert (Lat. 36 25' N, Lon. 116°15f W; Fig. 1), immediately outside the CDCA in the low hills east of Ash Meadows , Nevada (Mehringer and Warren 1976). Creosote bush and white bursage are widely scattered on west facing calcareous slopes at ca. 790 m (2600 ft) elevation. The macrofossil assemblage contains abundant Juniperus osteosperma and Agave utahensis (Utah agave), the former yielding a "" C date of 13,150 + 500 B.P. (1-4237; Mehringer and Warren 1976). Other extralocal perennials that are common in this fossil midden include Ribes cf . yelutinum and Petrophytum caespitosum (Table 10) . Few macrofossil records date from 14,000 B.P. to 11,000 B.P. at low elevation sites within the CDCA. A midden from 730 m (2400 ft) •levation in the Turtle Mountains of the Eastern Colorado Subdivision (Ut. 34°15f N, Lon. 114°50' W; Fig. 1) contains both pinyon and juniper. C date of 13,900 + 200 B.P. is reported for this assemblage but -52- Table 10. Plants from the Owl Canyon 1 midden, Amargosa Desert, Nye County, Nevada, ca. 790 m (ca. 2600 ft) elevation. Radiocarbon age: 13,150 + 500 B.P. (1-4237) on Juniperus osteosperma '' (Mehringer and Warren 1976) . See Table 2 for an explanation of the symbols. Relative Species abundance Agave utahensis 4 Ambrosia dump s a 0 Artemisia sec. Tridentatae 1 Atriplex confertifolia 2 Chrysothamnus cf . viscidiflorus 2 cf. Eriastrum sp. 1 Eriogonum heermannii 1 Gutierrezia cf . microcephala 0 cf . Haplopappus nanus 1 Juniperus osteosperma 5 Larrea tridentata 1 Opuntia sp. 1 Petrophytum caespitosum 3 Ribes cf . velutinum 3 Salvia dorrii 1 Scopulophila rixfordii 1 Sphaeralcea sp. 1 Symphoricarpos cf . longiflorus 1 N - 18; N =18 P -53- few species are recorded as associates (Wells and Berger 1967). The o *■ Whipple Mountains of eastern San Bernadino County (Lat. 34 16 f N, Lon. 114°25I W; Fig. 1) have yielded a number of ancient packrat middens (Van Devender 1977b, King and Van Devender 1977; Mead et al. 1978) but the bulk of the data from these sites remains unpublished. The Redtail Peak #5 midden, radiocarbon dated at 12,960 + 210 B.P. (A-1666) , is the* only midden to contain both juniper and single-needle pinyon. At an elevation of only 510 m (1670 ft) it is the lowest altitude record of Ice Age pinyon- juniper woodland (Van Devender and Spaulding 1979) . Other associates of the latest Wisconsin woodland in the Whipple Moun- tains include Yucca brevifolia, Y_. whipplei, Nolina bigelovii, and Cercocarpus intricatus (Mead et al. 1978; Van Devender and Spaulding 1979). The Latest Wisconsin At Intermediate Elevations. Two packrat midden sites in the South Central Mojave Subdivision of the CDCA document xeric juniper woodlands in the Lucerne Valley and on Ord Mountain, north of the San Bernadino Mountains (Fig. 1). The fossil midden from the north side of Ord Mountain at an elevation of 1220 m (4000 ft; Lat. 34°40' N, Lon. 115°50' W) contains abundant Juniperus 14 psteosperma and Purshia glandulosa (cliff-rose) associated with a C date of 11,850 + 550 (UCR-149; King 1976 a,b). Pinus monophylla occurs in the midden. Larrea tridentata and Haplopappus cooperi are the domi- nant shrubs at the site today while Yucca schidigera and Y_. brevifolia are common (King 1976 a,b) . None of these species were noted in the oacrofossil assemblage. Southwest of Ord Mountain, the Lucerne Valley bidden at 1010 m (3300 ft; Lat. 34°35* N; Lon. 115°50' W) elevation -54- provides a complex sequence of macrofossil assemblages (King 1976 a,b). The two oldest samples are radiocarbon' dated at 12,100 + 400 B.P. (UCR-181; Samp. #13) and 11,100 + 420 B.P. (UCR-187; Samp. #10) and contain abundant juniper but no pinyon. The present xeric desertscrub vegetation is dominated by Larrea tridentata and Ephedra californica. The Scodie Mountains lie on the western edge of the South Western Mojave Subdivision of the CDCA (Lat. 35°36' N, Lon. 117°57' W; Fig. 1), forming part of the southern terminus of the Sierra Nevada. Three midden samples from the Robber's Roost area at 1130 ra (3710 ft) elevation are radiocarbon dated from 13,800 + 400 B.P. to 12,820 + 400 B.P. (A-1763 and A-1762, respectively; Van Devender and Spaulding 1979). They docu- ment the occurence of a pinyon- juniper woodland in what is now creosote bush desert. A detailed analysis has yet to be published but the middens contain a diverse fossil flora including Cercocarpus ledifolius, Yucca brevifolia, Quercus turbinella, Purshia glandulosa, and Artemisia tridentata (Van Devender and Spaulding 1979). Numerous ancient packrat middens have been collected in the Sheep Range of southern Nevada (Fig. 1; Spaulding 1977; Van Devender and Spaul- ding 1979). However, only one midden from intermediate elevations is of latest Wisconsin age. The Penthouse 2 site, at 1580 m (5180 ft) elevation, is at the crest of a limestone ridge in the southeastern Sheep Range (Lat. 36°28' N, Lon. 115°15' W) . The xeric east slope about the site supports a mixed Mojave desertscrub community dominated by Gutierrezia l_arothr ae (snakeweed) , Ephedra torreyana, Coldenia canescens, and Stipa arida» The Penthouse 2(2) macrofossil assemblage is radiocarbon dated -55- at 11,550 + 150 B.P. (A-1774) and contains abundant Utah juniper and Forsellesia nevadensis (greasebush) . The remains of Pinus monophylla and Ephedra viridis are. common while Ceanothus greggii (Mojave buckbrush) and Prunus fasciculata occur occasionally (Table 11). Wells and Berger (1967) report a midden site from 1280 m (4200 ft) elevation in the Funeral Range in the Eastern Mojave Subdivision of the CDCA (ca. Lat. 36°50' N, ca. Lon. 117°05' W; Fig. 1). A woodland • assemblage with abundant Juniperus osteosperma and Prunus fasciculata 14 and common Symphoricarpos longif lorus is associated with a C date of 11,600 + 160 B.P. . No pinyon was found although the presence of other mesophytes, such as Chamaebatiaria millefolium, Fraxinus anomala, and Cercocarpus ledifolius, is extraordinary in this arid range in the Death Valiey region. Another ancient midden from 1250 m (4100 ft) elevation on Mercury Ridge in the Nevada Test Site of south-central Nevada (ca. Lat. 36 40' N, ca. Lon. 115 50' W; Fig. 1) contains abun- dant juniper but no pinyon. The radiocarbon date for this low diversity midden sample is 12,700 +200 B.P. (Wells and Berger 1967). The Early Holocene At Low Elevations. Fossil records of desert- scrub vegetation from the Wellton Hills in Yuma County, Arizona are reported by Van Devender (1973; 1977b). These low hills at the north end of the Lechuguilla Desert (Lat. 32°36' N; Lon. 114°11* W; Fig. 1) are surrounded by sparse creosote bush-white bursage vegetation. The Wellton Hills //l midden at 160 m (530 ft) elevation is dated by two statistically equivalent radiocarbon ages run on Larrea tridentata (A-1407) and extralocal Ephedra nevadensis (A-1406; Van Devender 1973; -56- V jj II c • .e cu 03 to M "iH 3-1 O 4) 3 JO • jj O g. < Pi + 1 • 03 CO CN pa c JJ u-i ■n c u u vO O a) 03 pa co cn to iH JC cy» cn 01 u P. jj o c o ■H •t >s m 5 iH a O iH CO + ' c 0) c CO u pq P-, o • ■o • * »H c CN O P3 •o at O •H CN H 03 CO C_3 •O E 3 C CO PQ CT\ 00 O o « *h m u J= •H CO JJ 4J u C co ^ 03 c 3 a P* co iH O c a * a X X 41 Put 0) U 4) • M e U C C 0 O H Sn M • 0 c O •— s ,P- CO U-l •—4 + • cj V^ pa c CM r-H o r» o u •H 0) JS O CN (3 CO iH P-. oo -< 41 co JJ T3 n 03 03 C m 3 83 o 0) O + • n 03 •C x->^ P-, CO JJ CN o • c ^-^ -? pa C • 03 CN m H cd • P-4 • o gg •a 03 X -* m c co 4J ex. — * — * 0 > •H u 0) CQ n z C 3 3 * 03 o IS •o c JJ "O c •H 0 3 s w O 5 u 03 .* JJ H u 0 (Q JJ ca 60 ■ * c tJ 01 •H c 00 5 «: c o •-* q S-i c OS 00 CO «H a 03 CO -hi l -h I i^icni i i l l l l I -4 | | | l l l^l l I l -* i i cn I I l I l l CN I l I l XI I l X X I IX I I I IXI I I I I IXIXXIX cn | | I | | I I I I I -h | I I cn l I I I -< I I ' — 1 I - — • I 1 I • — l I t f I ' — » ' — tllC-J, — lltllllllll XI I IXI IXI IXI IXI I IXIXI IX I X I CO c 03 JC CO u 3 03 > CO e 03 CO JJ cd 3 - 3-1 JJ J-I 03 3 O CO c 03 3 co CO tH CJ 0) ■H jj o B «4-l •H •H T3 CO •H CO cO 0J •H •H U •H •H C CO •H CO •H cn CJ CO CO CO J-I CO > U 03 C c •H h o •H •H 9 • C CO 3-i > O H > 03 CO 0 03 03 e T-t r-l ra O. 0) •H CO 03 •H rH CO a ■H CO c X. o to JJ c CO CJ CO 3 U_l •H • •3 03 CO • J= 01 rH JJ • 03 rH >-i to c " CO o C3 C CO • -h a. • 00 C o O 0) O fl Q. X. CO J3 03 •H CO CO 0) E CJ o jj C- JJ CO a B — 1 CO QJ s c U-I 3 CO (0 3 U >-i 3 3 c CO •H o c^ CO c CO 3 XI •H CO CO -H jj OO CO C c CO )-i >; rH o CO • H O CO a JJ (0 03 3 CO 03 3 6 E u CJ CO 3 CO E H C a. jo CO •H ITS 3 3-i —J •H T-> in co D. ra CO 0) E J3 CJ .c 0) c o CO o •H > •H i— 1 X 03 iH r-H CO c CJ 3 3-i ^; JZ CO C jj o JJ M H u -C 01 o U3 CO 03 u_> •— 1 ^^ •<- ) o r-i jC CO _i JJ •H >> CO C > c (M CO >> 03 u •H T3 «H 00 -— C 03 QJ 03 03 •H jj CJ o o c 00 •H CO CO C3 U cu •H 3 e 3 E CO a. o Jii ^rj r— t CO •H 0 O CO CO 03 o B JJ — 1 JJ CO 3 o JJ 03 CD tH 9 M ■H u U CJ •a H JJ c CJ >. >. TD 03 CO a. IH a 03 CJ u CO O JJ JJ jj 3-i •H iH "3 E n ^I >-i u. u — i rH 5 >> =N i-H CO 00 U J-i 3-i . >-i CO jj • u U 3 CO CO O) 03 £ JZ O o o W • 3-i CO 03 < < < < < < < < < S3 pq pa u u U U CJ u U U C_J C_3 u CJ a a -57- c o >\ c to o -0 | | I I I I I -* I I I I I I I »-* I I | I I I I I I ! V"l I | I I | C O u I I I I I ! I I I I I I I I -< I I I I I , I I I I -< CM iO !l I I I I I X I X X I > I I X I I I I I X ' X I I I X I x I > I I X I I I l i-*i i i n I I l l I I l l^l I Nrt I n I i I i-*-h| I i ^ -hi i i i n i i i i i-^i i -^ <— « i i >a- -h i i^i i I sr o 1 »4 I i I I I IXI I IXXXIXXI IXXI IXXI I I I I I I X I J3 a o BB 3 a> • u a. co ai CJ •H CO 13 CO l-l CU •a CO ■H • CO o 4-1 CO a 4-1 c CO CO CO • VJ •H Q. C C CO M cO o 9 > c • CX J2 C E CO CU •H c CO 4-1 CO 01 CD ca CO a. co c 3 , u CO a. CO a •H M c >n •H co s CO j-J E 3 CO iH 3 o CO CU cu TD c E CO 3 •r-l CO 4-1 CU a CO u •H CO o c )-i H c CU X> cfl •H CJ CO 1-4. o S-i u M u u o CU 0-1 o -a M •H 00 ca a i—( c ■a c •H T3 0) 001 CU C 00 •H o 4-) 3 CJ 3 CU ■H aj 4J > CU 60 o JS ■H o C r; o i-H o 4-1 CO JS -c 'JS •H •H '•H U a. u iH >-i CO u u Q. . . a. U u • V U 3 3 3 CO cu CU o (d H w w u w w w W Ed ua W i-U fa u* fa fa CO M 3 • co fa CJ Q. -H CO CO CU CD OJ CO co o OJ 4-1 CO o 3 M CU cu, cu u E • 3 a u CO _Q CO CO 4 -58- < CM PQ C o >> c co u 03 CM O 3 O c J3 P- cu CQ 3 O 4-1 c 0) p- CM CM J= (X, o c en 0) •H o 0) a I I —* I t-< | -h i «-* | i i co I i^hi^i i i n i i i i lll.l-i|llllllll-H|||||||CMIO • I I I IXI IXI I X X X I IXI I I IXI IXI \o co -* co h h co n CM o en o H CO H vO 0> CO cm — i r-» I J I I I <-* I I N I I N I | |- | -i | | | | | |CM| I rH I I I I I I n I N | | r-l rH | I I 1 rH | I rH I XXI I IX X I I I X I I I X XXX I X X x CO CO 3 ■H -V: (4 H CO e VJ o O fH cd 3 O rH 4-1 iH r-l CO rH U-l •H >> cn CO CO o U-l ■H V Xi CO a) 3 JS 4-1 6 CO •H 60 u a Xi T3 4J a ■H CO 3 3 CJ c o o j-> •H CO i CO CJ 4J r-l ■H CO cn co O a c O 0) H >s ai Cu 00 • X •H i-l CO OJ ■H •H CU cn o <4-l -C 4-1 r-l • Xi a CO 0) £ Cu CU: l-i CJ X> > U cfl CO u CO CO U E •H u O a s 3 CO o CU CO O t-t CU OJ cu -a cn CO CO > c CO o< cn en c c 4J a H nj rH • o CJ CJ E ■H o a •H CO CU CO c •H o o 01 ^ CO o TJ 14-1 -H U-4 TJ rH r-l o u •H ■H E E. u t-> CO CO CO w £ E i— I j= •H e C u a CO CO c CO U H >, to JO cu •H ■H •H a OJ 01 -N a r-l 0) cn 4-1 co -J -4 CO cu o ■a o CO 4-1 4J 4-1 o 4J 4J 4J o 03 CO U-l 3 CO ■H CU CU sz CO CU Xi CO c CO JZ o c c N CO cn •H u co 3 c cn cu > CO CO Cu a cn Cu H E u a a a 3 >< c (3 -J U >M C CO a 3 JO H Xi — CU •H . H 4J CO cj 03 ■H CM a r-l 1) a cu OJ _c ■H 0 -4 JC t-» CO Cu c 4J u CJ Xi 3 2 z O o c Pu p- cu PU p- Oh P- P-. Pi PS 00 W oo C/2 00 cn cn H H >- CO M 60 o r*. in o CM rH co 2 CM — i so | CO CM \0 I o o a cucn Z z z H -59- Mead et al. 1978). They yield an average date (Long and Rippeteau 1974) of 10,665 + 325 B.P. Ambrosia dumosa is also present while other extra- local species, besides Ephedra, are Encelia frutescens (common) , Brickellia atractyloides, and Holacantha emoryi (both rare) . Extralocal shrubs are the most important component of the Wellton Hills //l midden although it contains the plants that dominate the vegetation today. Other dates on extralocal Ephedra nevadensis from the Wellton Hills are 8750 + 320 B.P. (A-1399, WH #2) and 8150 + 320 (A-1364; WH #5) while creosote bush fossils yield dates of 7950 + 370 B.P. (A-1400, WH #2) and 6600 + 370 B.P. (A-1365, WH //5; Van Devender 1973, 1977b; Mead et al. 1978) An early Holocene midden from 245 m (800 ft) elevation on Picacho Peak, Imperial County, California (Fig. 1) contains a desertscrub assem- blage radiocarbon dated at 8650 + 280 B.P. (Van Devender and Spaulding 1979). Of the common species in the midden, only Ephedra nevadensis is lacking in the modern flora (K.L. Cole, pers. comm.). This contrasts with the early Holocene Marble Mountains desertscrub assemblages which are dominated by extralocal plants. Other packrat midden sites that date from 11,000 B.P. to 7800 B.P. document the occurence of juniper woodland (Van Devender 1977b). Ancient oiddens from the Lower Grand Canyon in Arizona (Fig. 1) provide an ela- borate record of this period (Phillips 1977). They contain abundant woodland species although Phillips (ibid.) notes a progressive decrease *a mesophytic plants and a corresponding increase in xecic species through the early Holocene. In this area, many of the desert shrubs destined to dominate the middle and late Holocene vegetation were apparently present <^ring the preceding woodland phase. Phillips (1977, 87-88) summarizes -60- this phenomenon in the Lower Grand Canyon thusly ( parentheses mine) : "The Stage III and Stage IV (latest Wisconsin and early ., Holocene, respectively) changes in vegetation involved not a replacement of woodland by desert from other areas, but a gradual decrease in abundance and number of wood- land species and a relative increase in importance of desert species, most of which were already there." Fouquieria splendens is one of the desert plants that apparently did not enter the area until after the close of the early Holocene. Fossil sites from 365 m (1200 ft) to 520 m (1700 ft) elevation in the Whipple Mountains of the Eastern Colorado Subdivision of the CDCA (Fig.* 1) document early Holocene juniper woodland (Van Devender 1977b; King and Van Devender 1977; Mead et al. 1978). Associated extralocal species include Salvia mohavensis, Nolina bigelovii, Yucca brevifolia, and Y_. whipplei. A complete analysis of these sites has yet to be pub- lished. Extensive early Holocene juniper woodland in the Whipple Moun- tains contrasts with the fossil record of desertscrub vegetation from the higher elevation Marble Mountains to the northwest (Fig. 1). Early Holocene Midden's From Intermediate Elevations. A packrat midden from modern creosote bush desert at 850 m (2790 ft) elevation in the Newberry Mountains of extreme southern Nevada contains the fossils of Juniperus sp. , Quercus dunnii, and perhaps (£. chrysolepis (Leskinen 1975). A radiocarbon date on oak acorns from this assemblage is 9500 + 240 B.P. (A-1077). Further north in Nevada the Sheep Range and several small ranges in the southern Nevada Test Site have early Holocene midden records (Fig. 1). The Basin Canyon 2 site, in mixed Mojave desertscrub vegetation on -61- the west side of the Sheep Range (Lat. 36°43' N, Lon. 115°15* W) , is at 1630 m (5340 ft) elevation, Gutierrezia sarothrae, Artemisia bigelovii, Fallugia paradoxa, and Yucca brevifolia are common today on this north- west- facing limestone slope. The Basin Canyon 2B midden, radiocarbon dated at 19,200 + 580 B.P. (A-1741), provides a full glacial perspective with an assemblage containing abundant Juniperus osteosperma, common Atriplex conf ertifolia, and some montane species (Table 11). The early Holocene BC 2A midden still contains abundant extralocal juniper, dated at 9365 + 320 B.P. (UCR-727). However, the other common fossil plants, Prunus fasciculata, Atriplex canescens (four -wing saltbush) , and Symphori- v carpos longiflorus, persist at the site today. Atriplex canescens appears more abundant in BC 2A than in the present community and Gutierrezia saro- thrae, the most common plant at the site today, is lacking in BC 2A (Table 11). Penthouse 1(1), at an elevation of 1600 m (5240 ft) on a xeric lime- 14 stone ridge in the southeastern Sheep Range, has a C date of 8100 + 120 B.P. (A-1771). Gutierrezia sarothrae is the most abundant species in the present community and the midden sample. Ph 1(1) contains extralocal species that can be grouped into. three types. The occasional twigs of Juniperus osteosperma and Forsellesia nevadensis, among others (Table 11), represent plants that were important in the preceding late Wisconsin but •re now restricted to woodland communities some distance away. A second group is represented by Coleogyne ramosissima and Buddie ja utahensis , desert shrubs that occur nearby but only in more mesic habitats. The third and most unusual group is represented by the remains of Larrea illdentata and cf. Lycium andersonii (Table 11). Neither species occurs n*«r the site today and their presence in the Ph 1(1) assemblage represents -62- a displacement upward relative to their current position on the lower, warmer talus slopes in this part of the Sheep Range. Wells and Berger (1967) discuss a suite of six early Holocene fossil assemblages from the mountain ranges in the southern Nevada Test Site from Lat. 36°37' N to Lat. 36°55' N (ca. Lon. 115°50' W; Fig. 1). The midden sites range from 1100 m to 1830 m (3610 ft to 6000 ft) elevation and record juniper woodland in what is now desertscrub vegetation. An exception is the Spotted Range //2 midden (9450 + 90 B.P.; 1550 m, 5080 ft) which con- tains common Pinus monophylla. Other early Holocene variations include the Ranger Mountains //2 midden (10,100 + 160 B.P.; 1100 m,v3610 ft eleva- tion) in which Atriplex confertifolia shares dominance with juniper and Mercury Ridge #1 (9000 + 250 B.P.; 1390 m, 4560 ft elevation) in which Cowania mexicana exceeds juniper in abundance (Wells and Berger 1967). The Spotted Range #2 occurence of Pinus monophylla is anomalous. The ma- crofossil record from throughout the Southwest indicates that pinyon was eliminated from low elevations by 11,000 B.P. (Van Devender and Spaulding 1979). The Lucerne Valley midden in the South Central Mojave Subdivision of the CDCA at 1010 m (3300 ft; Fig. 1) elevation records a juniper woodland 14 in current creosote bush desert (King 1976 a,b). Two samples with " C dates of 8300 + 780 B.P. (UCR-186) and 7800 + 350 B.P. (UCR-249) contain common Juniper and Encelia frutescens but no Larrea tridentata* is reported. One •ample also contains common Purshia glandulosa (King 1976 a,b) . The Lucerne v*lley midden substantiates the early Holocene record from Negro Butte, ca. 5 km northwest of King's site at 1070 m (3510 ft) elevation (Wells and ^tger 1967). The Negro Butte midden contains abundant Juniperus osteosperma *nd Purshia glandulosa as well as some Yucca brevifolia (ibid.) . The midden i -63- samples discussed by King (1976 a,b) contain a yucca identified only to the generic level. Middle and Late Holocene Packrat Middens. Other than the Marble Mountains and Eureka Valley fossil sites, only middens from the Lucerne Valley (King 1976 a,b) and the Sheep Range provide vegetational data on this time period. The Sunset Cove (5880 +250 B.P.; UCR-134) and Lucerne Peak (5800 + 250 B.P.; UCR-135) middens from above Lucerne Valley at 970 m (3080 ft) and 1100 m (3610 ft) elevation, respectively, suggest some vege- tation change over the last 5800 years (King 1976 a,b). The Lucerne Peak midden contains little Larrea tridentata, in contrast to its present abun- dance at the site, and Encelia frutescens, common in the midden, is not noted in the modern species list. The Sunset Cove midden contains Opuntia basilaris and Opuntia sp. which are both lacking in the modern flora (King 1976 a,b). Ambrosia dumosa and Opuntia sp. are not noted for the present vegetation of the Lucerne Valley midden site but it is common in midden samples ranging in age from ca. 4300 B.P. to ca. 3600 B.P. Middle to late Holocene records in the Sheep Range span the last 5200 radiocarbon years. The Desert View midden site, on the range Ts west flank at 1810 m (5940 ft) elevation, lies near the upper limit of blackbrush de- fiertscrub. A single pinyon tree grows in a crack above the midden site today but the lower limit of woodland lies at about 1950 m (6400 ft) ele- vation, ca. h km away. The Desert View macrofossil assemblage is radiocar- bon dated at 5210 + 95 B.P. (WSU-2044) on extralocal Juniperus osteosperma. Prickly pears (Opuntia phaeacantha and 0_. polyacantha) in the Desert View •ample do not presently occur on the bedrock slopes near the site (Table 12). The Willow Wash 2 packrat midden site (not to be confused with the Ulow Wash at the northern Eureka Valley) is in the bottom of a canyon at -64- 2 3040 + u 75 B.P. «0 O c o c re u c oL c TO o - 3530 + u 90 B.P. re S 1990 + h 70 B.P. 3 33l° ± h 100 B.P. 3520 + 100 B.P, recent debris 4125 + 85 B.P. 5210 + 95 B.P. u 41 D. I | ■ — • t . — • ~ * t_roi | I I I I — I I I I CM I I i J — | -. | | | _ cs, , , , , , , , , , , , , _ , XI I I x X I X I IXt I I I I I x i I x I I X X I (N I t I I — --llllll-.lliiiirs.-OMl llllll— t > — o * • t « • — » > t ■ fl ■ I — . CM CM I XIIIXIIXIIIIIIIIXIIIXIIXI iti— it— ; . — ■ f » — i i i i . — . i i i j i i t i i I I I I I ••« •« 1 <9 I I I I I ^ — I I I I N I I Ul 01 c *-t IM "H ■— T-« •H 3 • to b > C H > 0) ■U —J 10 «4 u •H u n a. ■»< ■— m iH — « re u c CO e ec 1.1 O • re C to to > •h T3 O re • J= 01 o C • 41 • 00 C 41 a. oi O •— 05 . tc c O O 0) E o a. .c a. a* «»H u_ to c to Bl 1 > c O. E •— n •r^ r— t X X (fl .— i . — i re 'H a u 4-j -a >— -: JS c 2 -* *~ (/) 0 in K3 o o -— ■ — i i — i "H ■ — i sz re c o in u ij E ^i n U 0] ij ■— ■^J •*-< CUJ ■—1 U 0) OJ o o Ul ul -— o CX > — J3 a> o —i a> u -— a^ (J u ■a i-i c u •H c 1 — 1 >> >> Ul u >. i i i i i i i i i -« i i i — xiiiiiiiiixiiiixiixttiiiiiixiix: ~,|||||1-.||llllllr">lllllllllCNCNI- CN|||lll«H|lllllll4»-4|ICM XXI I I IXI I X I I I X I X I I IXXI I X X I 1 X X I I . i Itllllttttt — « ft* - — J • I 1 I 3 1111 — « <— , S I I — il-.|lllllll-t l a 3 c — * 3 o o> n 01 re T3 re o c <0 •— oc U • 0) en U j: Xi ■H c • •»-( o c C3 a. o c •t* a u E in . o re o > "3 as • >H •^ Q. —l CL, 0) ••-» U) U) CI u •-^ 3 -»-< to -3 4-1 1—1 a u .— i a. u o. m o • -5 a» -o n •H • 3 C u JJ as 3 M-4 c *H v< U> c W-l a. co B) -r-l en u A3 re • re — ( u_ CL (0 c (0 U O X o « 05 -^ B • D. u-i c E O X -H c a re (J re II T! re E -^ -r* HI AJ 01 u 01 1j 01 re u D. 9) c 3 c a 10 re o-l -H u •^ E J3 JC oi c E c l* o u o u IM c > — * Ifl B BJ w B o u nj 4J pH -4 o in u u u -r* — 1 3 jQ •w tr 01 u H c LJ *J 19 3 3 (3 a c c U c r^ "O o *w4 -3 V cc CJ C CC c •H O 0 i-i 3 ^-i ■j 01 K CL, a* oi a O .0 •»< — * u -l-« SI OJ i-> > 01 co o JS ■H c c •u JZ u o ~h ~-> AJ in 3 >l >> •— ' in IA IB • £. j3 JB u J '.c — < **4 •n -^H o o. u ipH c in ki C w L. "i CD 01 u u* (J u • u c C • • D. L* u • • U u u 3 • 3 m 01 Ol o u u u a a O a u U4 w w Id u Lu Ul UJ w UJ Ui ua UJ UJ UJ UJ UJ IU UJ u. u- u- u. -66- <0 o u I I I I I m I i i i i i i i i i i i i CM I I I I I I — I I I X I I I Xl I J I I t 1 t I I ■ | « x I I X I I I I X I ■ I '«»'•»■» I I I I I I I I -< I I I I — I I -, I t I iiiiiiMliiiiiiiiitn«^i i i i i t — , ■ ( ^_ •XXIIIIIIXIIIXtXtlllliliixilt llllll-*llll»^l||MI-*ll — •litlitit III— I i o I I I I I I i i — ioii I I I — — i cn i i I I X I I I X I X I t I I I I x *l I I I I I I I I I I I I — II— — IO-lll-.lllll-H-.ll I I I «HJ >« 'I «H I I IXXIIXIIIIIIIXIIXIII I I I I I I I I I I I I - I •» I I I I I — i i i i i i i I I X t I I I I X I I I I X I I X I I ■M — I I I I I — I I I I I I I I I 0) tj •> n o 9 c Ti «W V •H 1 t-H 60 «-l II O 01 c 4) oo «0 u-< E 3 TJ ■ to O n w. — . Obi C en 01 3 4J c X 1* 0) i-l v-4 c TJ u •r4 81 u (0 C. eg ■^ C •H u 01 CB T* TO a. 60 p o 01 m IJ u o T-. CO u «rt 0 — ( in C 9 u • c 0 TO c — • C n U c 0 01 3 i-> to a -h cfll 41 4J o p X a. o CO 0) CO CO CO TO 01 •^^ CO a a w w — i "3 u C H 3 4-1 to w ^ c • — < a JC JS E u • o — — ■SI u 01 C • u u o T-( Q. -H u 4J t-1 >y «l o. • m — CO ■a t/1 Ci O -3 u u D. 01 Cu 3 1-1 0 c A-i o a "3 to ■H C TO CO n c- c o >• '/I H m U eg CO u —> ^ C- u u u — CO -*H C 01 >- o. CO ■J) fc' E £ —i — B - u a CJ (5 ■--< a. — c ^ a. o. • U u u u 3 3 >x C c l-l >• r • 0" u-i 01 ^ ig ty -~* D- • • • , •. •. a l- ai SI 11 JS u. U u U = = -) ~J — J -1 u -1 -J z: -*- 2: o o| o Ol C| O o o 0- c- cu &- -67- 0 H I U c _| 0 ^-^ >> H c U es V » O c I en i — — Illtt — illlt — •! in I n | h I I I -« i I l l — i I X| I X X X X X I • IXI I I X X I I N I N I I — I CN I I — I Mini I I l—l I I I— I I— I I — -" en I I IXIXXXI I IXI IXIXI I N I - I «- I - I I I I — — I I — i — i _ i « «■> i — i i n i I cm i en I XXIXllXXXI I IXI 1X1X1 I I imi i -< i <^ i — i I i i in I I I I X| I X X X I I IXI I IXI I N I N I I I I— I I I I I — • — CN I IXXXIXIXIXI I I IXXXI 01 3 u O T. i-t o vw c «0 ■»-t 0) ■** ■H to u oa c •< 0] Id 3 a c 01 tn w 01 OO c o -( U) c 01 <0 C •— * -— » JO • *H a (0 J3 -hI cc r c (0 e a. a. in c • AJ £ — •** u 3 w > u — H —1 n ■H » CU U U) o ui -H JZ JS 0) u £> 1- o u *. V u rrj to nj CJ 0 c 0J rrj ■z 1 — 1 0 u cj • U ■H -*H rr: CJ u-i B n 09 o -j w. •— t -r-» T3 < — 1 ~-. « a. -»* {j. H E ro •w Q. E c — u (0 ta >N to u. >, tn .O > 1) HI V) u i-i to 1_ u 01 a -J T3 o 01 u cy c- •0 ~ 1 c ■»* -J 01 #~t n r n rg c <0 U m tn 3 3 on | > ~i ig C Q. a. u u t£ U J3 u >> u: »0 3 c 3 >— 1 -- g <0 -,-, £ u 1-1 <0 CJ u JB ■»-• O U o j= to CJ ^X, tJ *-j 01 CJ _c 3 . 3 V/} ( ft. fi- cu cc oc i/} (fl CO en l/> 00 f- H H >- >- ^ m r4 n «> n n « >a K r>» \o •* r« rt « (Jl * 14 ON %o O I «M CM ON I M «n o o vr (N cn oo m n o> n ^ N H B «I M -* cn «* l en cm r*» I «» en vo en cs CM ON NO M o «o en o en «n oo r^ r^ tn en I CM .CM ON I K vO O r^ no cm cm r* •» M vO VO —> I CM -H NO | no «M m tt> cm cm oo m on -* en i CM cm co I o o (X Z 2 o. -68- 1540 m (5040 ft) elevation in the southeastern Sheep Range. The site is surrounded by diverse ripparian desertscrub vegetation. The Willow Wash 2 macrofossil assemblage is radiocarbon dated at 4125 + 85 B.P. (WSU-2039) and is dominated by the species that presently occur at the site (Table 12). Four taxa that are in the midden but not at the site now are Chilop- sis linearis (desert willow), Brickellia sp., Opuntia basilaris, and Agave utahensis (Table 12) . Two frost damaged desert willows grow at the mouth of the canyon, ca. 1 km downstream at 1470 m (4880 ft) elevation and are the only known representatives of this tree from the main mass of the Sheep Range. ^ The Canyon Two midden at 1800 m (5920 ft) elevation, again in the southeastern Sheep Range, provides assemblages that are radiocarbon dated at 3310 + 100 B.P. (A-1531, top layer) and 1990 + 70 B.P. (A-1532, bottom layer; Spaulding 1977). Like Desert View, the site is near the present lower limit of woodland which extends down to 1980 m (6500 ft) elevation. The midden samples contain the desertscrub species that dominate the modern vegetation but Juniperus osteosperma occurs in both and is one of the most Important speices in the younger, bottom layer (CT(2)), which also contains Pinus monophylla (Table 12). The presence of these conifers at the Canyon Two site suggests a slight downward displacement of the lower limit of wood- land at ca. 3300 B.P. with a larger displacement occur ing by ca. 2000 B.P. (Spaulding 1977). The Basin Wash midden site, from the west flank of the Sheep Range »t 1650 m (5420 ft) elevation, is near the bottom of an arroyo in a rip- parian desertscrub community. It contains the shrubs currently important *t the site as well as abundant Utah Juniper and common single-needle pin- 14 yon (Table 12). Juniper twigs and seeds yield a " C date of 3520 +100 -69- B.P. (WSU-1854). Both conifers occur sporadically in the vicinity of the site. Modern midden debris from the Basin Wash site reflects their rarity (Table 12) . The present lower limit of woodland in this part of the Sheep Range lies at ca. 1790 m (5860 ft) elevation, about 1.5 km west of the Basin Wash site. Like the Desert View and Canyon Two middens, the data from Basin Wash suggests a depression of the lower woodland boundary by at least 100 m. The Sawmill Canyon middens on the east flank of the Sheep Range at 1755 m (5760 ft) elevation provide late Holocene macrofossil records from within the current pinyon-juniper woodland. Located immediately east of the highest part of the Sheep Range, the site is im an area where forest and woodland vegetation extends to unusually low elevations. The Sawmill Canyon 1A and ID macrofossil assemblages are radiocarbon dated at 3530 + 90 B.P. (WSU-2045) and 3040 + 75 B.P. (WSU-2041), re- spectively. They show no appreciable departure from the current vegetation (Table 12). t DISCUSSION The last 14,000 years have seen the decline of Ice Age plant asso- ciations and the ultimate demise' of many mesophytic species in the interior basins of western North America. The concommitent expansion of xerophytic plants led to the development of the present interglacial plant communities. The fossil packrat midden record is complex and reflects changes that are, to some degree, unique to each area. The tlaing of the retreat of mesic plants, as well as the advent of xeric ■Pecies, varies from place to place and from plant to plant. Some -70- desert shrubs apparently migrated considerable distances in post-Wisconsin times, while others were present in the preceding late Wisconsin vegetation. v Vegetation Dynamics The Marble Mountains. The early Holocene vegetation of this area (Fig. 1,2) was dominated by mesic desertscrub plants. The fossil record at Marble Mountains 2 (465 m elevation) implies vegetation in which Ephedra cf . californica and Lycium sp. were important. Larrea tridentata and Encelia farinosa, the dominant taxa today, are not in the fossil sample. At higher elevations, from 840 m to 900 m elevation, contempora- neous early Holocene middens record diverse, mixed desertscrub vegeta- tion in what is now creosote bush-brittle bush desert. Salvia mohavensis, Yucca schidigera, cf . Kaplopappus laricifolius, and H. cuneatus are among the most common species in these shrub assemblages (Fig. 4). None of these early Holocene plants occur near the fossil sites today, although most may be found in higher elevation (ca. 1400 m) desertscrub communities with similar aspect. An exception to the early Holocene record of desertscrub vegetation is found in two Marble Mountains middens, both dated at ca. 8900 B.P. They are the only samples to contain Juniperus sp., Yucca brevifolia, and Y. whipplei (Fig. 4) and represent either a reversal in the early Holocene trend toward a xeric climate, or a stochastic, developmental phase of vegetation not directly related to any forcing climatic-factor. The Juniper represented in these two assemblages may be either the mesic Juniperus osteosperma, or the xerophytic £. californica (California juni- per). The latter is found In higher elevation desertscrub vegetation -71- Figure 4. The estimated relative abundance of selected plants in Neotoma midden samples from Marble Mountains Locality B and C v plotted against radiocarbon age. 5, very abundant; 4, abun- < dant; 3, common; 2, occasional; 1, rare; closed circle, one or two fragments, a possible contaminant. % I, Ji ll • • J L I il ■ jL_l I Ii ■ i j^l_L 1 Oi if) t 1 1 r GO o IO I m .i i E IT) in ID i- U1) lf> 0) o c a c < > m o D E (0 m m rviv ^#a SJD9JL D -72- within the CDCA today, associated with many of the plants found in these early Holocene middens. Reversals in the early Holocene drying-warming trend have yet to be consistently defined in the Southwest. Mehringer (1967) suggests that such a mesic phase occured between ca. 8500 B.P. and ca. 8,000 B.P. based on pollen stratigraphic data from Tule Springs in the Las Vegas Valley of southern Nevada. The minor role played by woodland species in the early Holocene vegetation of the Marble Mountains is surprising. The three oldest macrofossil assemblages, dating from ca. 10,500 B.P. to ca. 9500 B.P., are virtually devoid of any woodland indicators (Table 2,3). However, contemporaneous early Holocene woodland is well documented vin the eastern Colorado Subdivision of the CDCA as well as the Mojave Desert of Arizona (Fig. 5). Woodland records range upward from elevations as low as 365 m (1200 ft) on igneous substrate of the Whipple Mountains (King and Van Devender 1977) and 460 m (1520 ft) on limestone of the Lower Grand Canyon (Phillips 1977). Only middens from the low eleva- tion Wellton Hills sites in Yuma County, Arizona and Picacho Peak in Imperial County, California contain additional evidence of early Holo- cene desertscrub communities (Van Devender 1973, 1977b; Van Devender and Spaulding 1979; Fig. 1,5). The conclusion appears inescapable that, during the early Holocene, the vegetation of the Marble Mountains was desertscrub while sites of •lmilar age and aspect to the east supported xeric woodland. However, • contributing cause to these discrepancies may lie in a bias introduced ty midden sampling technique. Extralocal woodland species have been •aployed as a field index of antiquity. In the Whipple Mountains and — JC c o o en o 1 -} x x J, \ <>J> *4 ■ « 1 •♦ •>i 1 •■ 6= I O o o 1^ o o — i — i — i — i — i — i — r o o o o o to (UJ) UOIj 0A9| 3 w 2 ,r c c *> — J5 a> c a> o O >. o u -0D -K a> c a> o o E 4 -73- Figure 5. Packrat midden samples from the CDCA and adjacent regions plotted against radiocarbon age and elevation. Open circles indicate desert scrub assemblages, closed circles juniper woodland, and closed squares- pinyon-juniper woodland. The broad dashed line indicates the present lower limit of woodland in the Sheep Range of southern Nevada. The short dashed line, the approximate lower limit of woodland in the Granite Mountains, ca. 5 km north of the Marble Mountains. -74- Lower Grand Canyon only middens containing obvious extralocal woodland taxa were chosen for analysis. The consequence is sampling biased, against ancient middens that may contain records of desertscrub vegetation. Sampling bias cannot account for differences between the Marble Mountains record and contemporaneous woodland samples from the systema- tically excavated Lucerne Valley midden (King 1976 a,b; Fig. 5). The Lucerne Valley site is at ca. 1010 m (3310 ft) elevation, ca. 110 m (400 ft) higher than the Marble Mountains sites, and apparently lay above the lower limit of woodland during the early Holocene. Questions of sampling aside, it is plausible that an early Holocene east-west gradient of increasingly xeric vegetation existed in the southern Mojave and eastern Colorado Deserts. Low elevations may have supported xeric juniper woodland in the Colorado River Valley but only desertscrub vegetation further west. Today perennial vegetation becomes increasingly sparse and of more xeric aspect on a similar east-west gradient. Excepting an apparent reversal about 8900 B.P. (Fig. 4,5), the early Holocene, fossil record from the Marble Mountains reflects the progressive development of xeric desertscrub vegetation Of the peren- nials that occur at the sites today, only Bebbia juncea is reliably documented in the oldest fossil sample. Most of the current dominant •Pedes were probably absent during the latest Wisconsin and immigrated to the area during the early Holocene. Encelia farinosa was the first to appear, in trace quantities at ca. 10,200 B.P., and then commonly « the next youngest fossil sample at ca. 9500 B.P. (Fig. 4). It is *o important element in all younger midden samples. The apparent im- | ■igration of Larrea tridentata, Ambrosia dumosa, and Peucephyllum -75- schottii occured later (Fig. 4). Trace amounts of creosote bush first appear at ca. 8900 B.P. but it apparently did not become important until sometime between ca. 8900 B.P. and ca. 8000 B.P. Figure 6 presents the total number of fossil taxa in midden samples from Marble Mountains Locality B and C. It is assumed that the sample diversity is directly proportional to the diversity of the vegetation at the sites. There is no dramatic early Holocene species diversity decrease since the local extinction of mesophytes was accompanied by the near synchronous immigration and expansion of xeric plants (Fig. 4,6). There is a notable diversity decrease in the middle and late Holocene, but more samples are needed to establish the validity of this trend. The little evidence for vegetation change at the Marble Mountains during the middle and late Holocene is equivocal. More samples of this age are needed for analyses of vegetation changes which, at this area, appear to have been quite subtle. The Eureka Valley. Plant macrofossils from the northern boundary of the California Mojave Desert (Fig. 1,3) document changes in high elevation (1430 m to 1635 m) desertscrub vegetation. Ancient middens of latest Wisconsin and early Holocene age are scarce, but the record of the middle and late Holocene is good. The two oldest midden samples are dated at ca. 14,700 B.P. (Eureka View 5A) and ca. 10,700 B.P. (Horse Thief Hills 1) and suggest the presence of a juniper-shadscale woodland. Juniperus osteosperma and Atriplex confertifolia are abundant in both samples while other shared taxa include desertscrub (Salvia dorrii, Artemisia sec. Tridentatae) and woodland (Symphoricarpos cf. longiflorus) species. The HTH 1 site is -76- Figure 6. The total number of identified plant species (N) in Neotoma midden samples from Marble Mountains Locality B and Locality C plotted against their radiocarbon age. S 01 J - z- 5- f - 5 - ) 6J - 8- 9 - /O- i /WHtfs) *r — I- 10 \ MflSCZ) ♦ MM 7(2) 1 MM nniQ 1 I MM?0) 4 mi (2) \ nnsa) /£ 20 25 30 -77- more mesic and the midden contains records of Amelanchier utahensis , Chamaebatiaria millefolium, and Prunus cf . andersonii. Hecastocleis shockleyi is present at the Horse Thief Hills 1 site today but absent from the midden. It is common, however, in the Eureka View 5A assemblage implying that, during the Ice Age, the distribution of Mojave prickleleaf had shifted from north facing to south facing slopes. The remains of Pinus cf. flexilis in EV 5A suggest a nearly continuous extension of this subalpine species from the Inyo Mountains to the arid Last Chance Range. The abundance of Atriplex confertifolia in the two oldest Eureka Valley middens points to an important difference between the record of changing vegetation here and that from the Marble Mountains. In the latter area there was a near complete turnover in the composition of plant communities from the early Holocene to present. Plants that are important in the current Marble Mountains vegetation are almost totally lacking in the earliest fossil assemblages (Fig. A). In contrast, for the last ca. 15,000 years shadscale has played an important role in the plant communities bordering the northern Eureka Valley. Terminal Wis- consin and Holocene vegetation change took place within a persistent matrix of Atriplex confertifolia (Fig. 7). The only early Holocene sample from Eureka View lacks Juniperus osteosperma. It is uncertain when, between ca. 1A,700 B.P. and the EV 5B date of ca. 8300 B.P., juniper-shadscale woodland gave way to •hadscale desertscrub (Fig. 7). It appears that desertscrub vegetation ttay have developed earlier here than further to the southeast in the highlands of the Nevada Test Site (Wells and Berger 1967), Sheep Range, -78- Figure 7. The estimated relative abundance of selected plant species in macrofossil assemblages from the Eureka View Locality plotted against radiocarbon age. Note the change in time scale. 5, very abundant; A, abundant; 3, common; 2, occasional; 1, rare; Closed circle, one or two fragments, a possible contaminant. I_^_i l_L_L J L l_L_L CD N J L J I I I 1L J III J L III 1 I I H 1 I o CM s § 5 S t r T CO IT) O m o ID ID U o o m o ID O m C o c < > o o Ji O " s . J lJ U A^TTTTT