The Altithermal Revisited: Pollen Evidence from the Leonard Rockshelter

Journal of California and Great Basin Anthropology Vol. 1, No. 2, pp, 280-294 (1979),

ROGER BYRNE COLIN BUSBY R. F. HEIZER

URING the past thirty years, there has sistence and settlement patterns? In this paper, Dbeen a continuing debate among we present evidence that relates directly to the students of Great Basin prehistory over the first question and indirectly to the second. nature and significance of postglacial climatic More specifically, we report on the fossil change (Aikens 1970; Antevs 1948; Aschmann pollen content of two series of sediment 1958; Baumhoff and Heizer 1965; Bryan and samples from the Leonard Rockshelter, Gruhn 1964; Harper and Alder 1970; Jennings Pershing County, Nevada. 1957; Martin 1963; Swanson 1966). Several Leonard Rockshelter (NV-Pe-14) is one of symposia have been devoted to the subject a limited number of sites in the Great Basin (Fowler 1972, 1977; Elston 1976), but there with a clear record of human occupance that are still no signs of a consensus. One basic dates back to the early Hoiocene. The site was issue is the validity of the Antevs three- excavated in 1937 and 1950 by University of part model. Mehringer (1977), for example, California archaeological field parties and has cautioned against the acceptance of paleo- produced a varied assemblage of artifacts. climatic models which claim relevance for the Wooden atlatl foreshafts have been radio Great Basin as a whole. The question remains, carbon dated at 7038 + 350 B.P. (5088 B.C.), however, as to whether or not the Antevs' and indirect evidence in the form of obsidian reconstruction is valid. Was the Altithermal, flakes from a basal bat guano layer radio for example, a period of warmer and drier carbon dated at 11,199 ± 570 B.P. indicates climate? And, if so, was it warm and dry that the site may have been occupied as early as enough to account for changes in human sub- 9000 B.C. (Heizer 1951). Roger Byrne, Dept. of Geography and Museum of Paleon The site is also of interest in that it provides tology, Univ. of California, Berkeley, CA 94720. Colin Busby, clear evidence of environmental change during Dept, of Anthropology, Univ, of California, Berkeley, CA the Hoiocene. The rockshelter was formed by 94720. R, F. Heizer, Dept. of Anthropology, Univ, of Cali wave action along the southern shore of an arm fornia, Berkeley, CA 94720 (deceased). of Pleistocene Lake Lahontan, and has since

[280] THE ALTITHERMAL REVISITED 281 been partially filled by a distinctive sequence of Humboldt Sink (Fig. 3). The Humboldt River sediments, including beach gravels, bat guano, drains through this area into Humboldt Lake. aeolian silt, and rockfall. In 1950, Antevs inves The lake is now shallow (<20 ft. deep) and tigated the stratigraphy of the site and during the historic period has been highly interpreted it as supporting evidence for his variable in area. In 1883, for example, it reconstruction of postglacial climatic change covered an area of 20 mi.^ and the eastern in the Great Basin (Antevs 1955). In 1955, in shoreline was less than a mile from the rock the hope of obtaining further evidence on the shelter. In 1951, however, the lake had an environmental history of the site, Heizer sent area of only 11 mi.-, and the eastern shoreline H. P. Hansen (Oregon State University) was more than seven miles from Leonard (Fig. samples of guano for pollen analysis. Unfor 2). During the period of European settlement, tunately, the results proved to be negative and irrigation has drastically reduced the flow of no further pollen work was attempted at that the Humboldt River and during drought years time. the lake has dried up completely (Antevs In this paper, we report on a second 1938). This recent variability in the area of attempt to recover pollen from the Leonard Humboldt Lake underlines the fact that even Rockshelter sediments. In 1976, we directed small changes in hydrological conditions can our attention to two series of samples that have significant effects on a shallow lake of included the whole of the stratigraphic this kind. We shall reemphasize this impor sequence. The samples were collected during tant point later in the paper. the 1950 excavation, and had since been stored The present climate of this part of Nevada in the Lowie Museum of Anthropology is semi-arid. Mean annual precipitation totals (Berkeley). Like Hansen, we found it impos vary from the less than 5 in. at 3800 ft. to sible to extract pollen from the lower bat 30-35 in. at 8000-9000 ft. Most of the precipi guano layer, but the overlying silt and rock- tation is received in winter from mid-latitude fall layers yielded reasonably well-preserved cyclones, although summer thunderstorms pollen in countable quantities. The results of can also produce locally significant totals. the analysis in large part support the Antevs The annual temperature regime is character- model and also throw new light on the "sig nificance" of climatic change in the Great / Basin. l. c^^' ^'1- _ ^ ^ \ THE REGIONAL SETTING ; •• / X ' (-study Location Leonard Rockshelter is located 17 miles ^"'"'^ M \ ''' -x 1 south of Lovelock, Nevada, on a north-facing ' ''} } slope of the West Humboldt Range (Figs. 1 1 Jfiml r^' and 2). This area of the Great Basin has numerous cave and open sites, several of which V i ( \ \ ^ ' have been excavated or surface collected VJ • I'M (Loud and Harrington 1929; Heizer and ( 1 1 1 \ \ f^ ' ~ T Krieger 1956; Roust 1966; Baumhoff 1958; ) , Elsasser 1958; Heizer and Clewlow 1968; [. 1 1 ' 1 Heizer and Napton 1970; cf. Ranere 1970 for an overview). Less than a half-mile to the west Fig, 1. Location of the study area within the Great of the site is the broad and flat expanse of the Basin, 282 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY istically continental with hot summers and semi-arid nature of the climate and is domi cold winters; Lovelock, Nevada, for example, nated by several species of xerophytic shrubs. has a mean January temperature of 27°F. and On a regional scale, the site is close to the a July average of 75°F. (Brown I960). boundary between the Sagebrush-Grass Zone The vegetation of the area reflects the of the northern Great Basin and the more

8 -

LEONARD ROCKSHELTER

-^ '-"' ,.-...AOOO

-HUMBOLlir ^:'^OVELOCK'CAVE ;>r I.".-'- *;•,.• ••••'.'J..* "v-;.'"•-'•;. V ^•;f'- " ;'.'.-!.''.• 'SINK\:.:>ft' A • . • i:.-.. -"/- 4684 •••'.-,••'"'.•'•,' ^- ^ ' ~ - ' -•,'^'i'^'i**'f-'r-'-'-.''.V.^ A. •/•..• ."'.I ^ .'-; -';•: ..'*-^.,'- --'"•'•":,•':•• >V--

OCAJ_A* -CAVE

r J 1882 lake level (3929') + Surface site

-W^:: Alkali flat O 5 Miles •'I---':;-' •4 Contour interval 500' O 8 Km Fig, 2, Map of the study area indicating major topographic and cultural features. THE ALTITHERMAL REVISITED 283

Fig. 3. The Humboldt Sink from the Leonard Rockshelter. centrally-located Shadscale Zone (Billings marsh covered the area to the north of the sink. 1951:104). Local variation in species com Most of this area has since been reclaimed for position is largely a function of edaphic con agriculture by deep drainage canals, and only ditions. The following summary is based on a few relic areas remain. The dominant species Billings (1951) and Cronquist et al. (1972). here are cattail {Typha dotningensis) and tules On the higher ground, sagebrush {Arte- {Scirpus americanus, S. acutus). mesia tridentata) is the dominant species. THE SITE AND ITS STRATIGRAPHY On the lower slopes and on the well-drained Lahontan sediments, greasewood {Sarcobatus In 1936, the Leonard Rockshelter was baileyi) and shadscale {Atriplex confertifolia) mined for bat guano by Thomas Derby. are more important. These spiny shrubs are During the mining operation, Derby dis generally evenly spaced and provide a rather covered and saved several artifacts which thin cover. Other species less commonly en were later described by Heizer (1938). In 1937, countered are bud sagebrush {Artemisia a University of California field party made spinescens). Mormon tea {Ephedra nevaden- some preliminary archaeological investi sis) hop sage {Grayia spinosa), and winter fat gations at the site, and several more artifacts {Eurotia lanata). Perennial herbs and annuals were recovered from a basal guano layer. It may have been more common prior to grazing was recognized at the time that these artifacts by cattle and sheep, but are now rare. On might represent an early period of occupation, clayey soil and in poorly drained depressions although no means of absolute dating were at the edge of the sink greasewood {Sarcobatus then available. In 1949, samples of the guano vermiculatus) is dominant, but gives way to and three greasewood atlatl foreshafts were iodine bush {Allenrolfea occidentalis), salt- sent to W. F. Libby for radiocarbon dating. grass {Distichlis spicata var. strict a), and The dates proved to be unexpectedly old, Salicornia europaea subs, rubra on the more 8600 ±300 B.P. (6710 B.C.) and 7038 ±350 xeric sites. Cottonwoods {Populus fremontii) (5088 B.C.), respectively (Arnold and Libby form gallery forests along the river, and sage 1950), and it was therefore decided to exca brush {Artemisia tridentata) is also encoun vate the site more thoroughly. A University tered in the valley bottoms. During the nine of California field party spent five weeks at teenth century, extensive areas of freshwater the site in 1950 and established the natural 284 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY stratigraphy and history of human occupance. guano layer in Areas B and C. The guano The results of this excavation were presented accumulated over a period of 4000 years as in a preliminary report by Heizer (1951). indicated by a date of 7038 B.P. for the atlatl During the 1950 excavation, four areas of the foreshafts near the top of the deposit. Both site were excavated (Fig. 4). The sediment Heizer (1951) and Antevs (1955) interpreted samples used in the present study were taken the guano layer to be evidence of a humid from Areas B and C. Area B proved to be the climate and high lake levels; chronologically, most productive archaeologically and also it corresponds with the Anathermal 12,000 to provides a useful basis for a discussion of the 7500 B.P. Above the basal guano layer is a stratigraphy of the site (Fig. 5). layer of fine sand intermixed with angular The rockshelter, which has a basal eleva rock fragments (Fig. 5, unit C). The rock tion of 4175 ft. above sea level, was formed fragments account for some 20-30% of the by wave action along the shoreline of Lake total deposit. No radiocarbon dates were Lahontan. The bedrock is intrusive volcanic obtained from this unit in Area B. Above unit material of Tertiary age. The age of the rock C, and intergrading with it, is unit B, a layer shelter itself is not certain, although it pre of stratified, whitish gray sand and silt. These sumably dates to a time when the lake level sediments were interpreted by Antevs to have was relatively stable. It must have been cut been blown into the shelter from the Hum prior to the last Lahontan high stand, as the boldt Sink. This interpretation has since been inner wall is encrusted with calcareous tufa, confirmed by Dr. R. L. Hay of the Geology clear evidence of submergence (Fig. 6). Department, University of California, According to recent work on the chronology Berkeley. The unit B sediments are mineral- of Lake Lahontan (Benson 1978), the last high ogically the same as sediments from Humboldt stand (4360 ft.) occurred ca. 12,000 B.P., and Lake. Furthermore, the unit B sediments con prior to this (22,000 B.P. to 15,000 B.P.), tain diatoms and ostracods which also indicate the lake level was at approximately the same a lacustrine origin (R.L. Hay, personal com elevation as Leonard (4000-4200 ft.). It there munication, 1977). Few dates were obtained fore seems reasonable to conclude that these from unit B but dates from above and below dates provide maximum and minimum age it suggest that it was deposited between 6500 estimates for the formation of the shelter. and 4500 B.P. In terms of the Antevs model, After the lake dropped below the level of unit B is Altithermal in age. Above unit B is the rockshelter for the last time, tufa began unit A, a mixture of windblown sand and silt, to spall off the overhanging cliff face and tufa rockfall, bat guano, and packrat nest accumulate on top of the beach gravel (Fig. material. It was encountered in all four areas excavated. In Area D, it produced numerous 5, unit E). Resting unconformably on the artifacts, mostly of basketry and wood. No beach gravels and rock fall is a layer of bat radiocarbon dates were determined from guano (Fig. 5, unit D). A sample from the base Leonard unit A, but dates on similar artifacts of this layer was dated at 11,199 ±570 B.P. from nearby Lovelock Cave indicate that this (9249 B.C.) (Heizer 1951). The guano shows stratigraphic unit covers the time period from no evidence of having been disturbed or sub ca. 4500 B.P. to the present (Heizer and merged so this date can be taken as marking Napton 1970). Chronologically, this corres the beginning of the period during which the ponds well with Antev's Medithermal. shelter was open for human occupation. It was apparently soon occupied because obsid Unfortunately, the sediment samples ian flakes were recovered from the base of the taken in Area C do not include the whole of the THE ALTITHERMAL REVISITED 285

rn^M^

1^0^'iij Excavation below cliff face

o o Artifacts of Humboldt or Leonard o culture 20 Feet B Profile sample location -r 6 5 Meters Fig. 4. Site map of Leonard Rockshelter indicating sample areas analyzed.

DATUM.^

^ " • / • ^ mi^y^f vCô^.^:^•: .o°c^ c:>.^:< ^° o O' .^". oCfoV o^S',J^'oOoiuOp. O .O„oo°rô:o .' .;" °' o o • 0 • ° °. »n" \^. o^JcôOo . Q-.^P. :0' ^^ î-V °.," Volcanic°'°'° ° •••^^ -^Breccia/ " 0 '' •

o;ci-o. • o >'^A.O° Angular Rock, Sand and Siltx o'c>c^o5^^?Ci\TufaN 0 .

VPacingf " ° 0 "'" ^ orv- o o '

Sand and Silt

Sand and Angular Rock

o c 0 Dark Guano Layer ^ - Cj ^ <^ ^ Burned Guano ^ Rock and Guanob. ^ 0

Rock

0.5 1.0 15 2.5 METERS Fig. 5. Stratigraphy of the east wall of the trench in Area B. 286 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

In 1975, the site was visited again by a University of California field party and the opportunity was taken to collect surface samples. Four samples of 10 cm.^ each were collected from within 50 m. of the site. These were then combined to provide a composite surface sample. As a first step prior to pollen analysis each sample was well-mixed and sieved through a screen mesh of 250 microns. Subsamples of 5.2 cm.3 were then taken for processing. Standard extraction procedures were followed including HCl (10%), KOH (10%), HF (cone), and acetolysis. The residue was stained with safranin and mounted in silicone oil (2000 cts.). In most of the samples, pollen was re covered in good condition and in countable quantities. This was not the case, however, with the guano samples. Here the high con centration of chitinous material diluted the pollen concentration to the point that a statis tically reliable count could not be obtained. We therefore limited our analysis to those samples from above the guano. Fig. 6, Tufa deposits above the rockshelter in Area D, Counts were made on a Lietz Dialux microscope with a 40X planacromat objec Stratigraphic sequence. This is apparently due tive. The pollen was identified with the aid of to the fact that the excavation in Area C the University of California Museum of Pale extended back under the overhang. The sedi ontology Reference Collection, published keys ments are uniformly fine sand and silt and (Kapp 1969), and general accounts of South correspond to unit C in Area B (Fig. 5). As we western pollen types (Martin 1963; Martin and shall indicate later, this difference in stratig Drew 1969; Mehringer 1967). On taxonomic raphy is reflected in the pollen diagrams. matters, we followed the precedents estab POLLEN ANALYSIS lished by Martin and Mehringer. No attempt was made to identify pine The samples collected during the 1950 pollen below the generic level, although sev excavation were taken from the east wall of eral size classes were evident in the samples. trench B, Area B, and the south wall of Area C Similarly, Cheno/ Am pollen was very variable (Fig. 4). The exact provenience of the samples in size and this type undoubtedly includes is not known, but is close to the area indicated. several species. Typha pollen encountered as Each series was taken in consecutive 6-inch tetrads were listed as Typha, monads were increments from the surface to the base of the listed as Typha/ Sparganium. profile, and the samples numbered according The results of our analyses are presented to depth (e.g., 0"-6", 6"-l2", etc.). in two diagrams (Figs. 7 and 8). The pollen THE ALTITHERMAL REVISITED 287 sum includes all of the pollen counted and at Cheno/Am increases to equal and in some each level the total was at least 200 grains. The cases exceeds pine in the five intermediate Area B diagram shows marked changes in levels, and pine returns to dominance in the pollen frequencies. Pine and Cheno/Am three upper levels. pollen dominate the record and together None of the other taxa appears to correlate account for more than 50% of the total count well with the two major types, and we are at all levels. These taxa are also negatively therefore reluctant to establish pollen zones. correlated. Pine dominates the lower levels, For example, the three aquatic pollen types

Analyst Coim Bustiy. 1976 Compositoe ,^o^ A<^ -<^" .N^" 1^' \o^- oV^V'^ ^.o- c,0^^ tt-in o —I ' 1 r

216 Hl50

10 20 30 40 50 O 10 O 10 O O O lO 20 50 O lO O ID O 10 20 O ID Percent o* totol pollen Minor pollen types omitted from diagram Fig, 7, Area B percentage pollen diagram.

Analyst Roger Byrne, 1976 Compositoe -^^ i^^ o,^ E ?s^" .<^^^ o ^ -f>' -,r\^< • I 507 S 0 1360

496

514 50 451

388 397

885 • 100

1374

8 28

MA 150 480

20 30 40 O 10 O O 10 O 10 Percent of tofal pollen Minor pollen types omitted from diogrom Fig. 8. Area C percentage pollen diagram. 288 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

{Typha, Typha I Sparganium, and Cyper- pollen diagram, particularly in the pine and aceae) all reach high values at lower and Cheno/ Am curves. For example, level 60"-66" intermediate levels. Conversely, Artemisia in Area C probably corresponds with level reaches its highest percentages at intermediate 54"-60" in Area B. Likewise, level 0-6" in Area and upper levels. C probably corresponds with level 18" to 24" Unfortunately, the uncertain provenience in Area B. Other than the shift from high pine of the Area B samples prevents us from estab to high Cheno/Am between levels 60" to 66" lishing a firm connection between the pollen and 54"-60", there are no major changes in data and the stratigraphic units described pollen frequencies. earlier. It is clear, however, that the Pine- Stratigraphically, the Area C diagram is Cheno/Am-Pine oscillation does not corres roughly equivalent to the aeolian silt layer pond exactly with the stratigraphic units (unit B in Fig. 3). Carbonized basketry from assigned by Antevs and Heizer to the Ana- the base of this layer in Area C was radio thermal, Altithermal, and Medithermal. carbon dated at 5779 +400 B.P. (3829 B.C.) In the Area B diagram (Fig. 7), the three and 5694 ±325 B.P. (3786 B.C.) (Heizer 1951). lower levels (54"-60", 60"-66", 66"-72") all Although there are some minor discrep have high pine values, and yet according to the ancies between the two diagrams, when their Antevs-Heizer interpretation they are chron stratigraphic relationship is taken into ologically equivalent to the early Altithermal account, they can be seen to be basically sim (unit C in Fig. 5). The basal guano layer, which ilar. Furthermore, the regularity of the curves was assumed to be equivalent to the Ana- for most taxa suggest that the sediments had thermal, is not represented in the diagram. not been seriously disturbed by the shelter's Similarly, the five high Cheno/Am levels prehistoric occupants, human or otherwise. represent 30 in. of sediment (24" to 54"), and The question now arises as to what the changes must therefore include not only the aeolian in pollen frequencies represent. sand/silt layer (unit B) but also some of the DISCUSSION OF RESULTS overlying mixed silt, rockfall, guano layer (unit A). We interpret the Leonard pollen record as Several radiocarbon dates were obtained being primarily a reflection of climatic change. by Heizer (1951) from the guano layer in Area The justification for this conclusion lies not so B, but none from the aeolian silt or upper much in the Leonard diagrams alone but in rockfall units. The date of 5088 B.C. from the their correspondence with other pollen dia top of the guano layer indicates that the dia grams from the Great Basin. gram represents approximately the last 7000 In southern Oregon, Hansen (1947:116) years. Unfortunately, the dates of the Pine- has documented the same postglacial sequence Cheno/Am-Pine oscillation can only be esti that is apparent at Leonard. Pine dominates mated by extrapolation from dates in other the early Hoiocene and is replaced by Cheno/ areas of the site. As a rough approximation, Am pollen in the mid-Holocene, and returns we would suggest that the pine minimum to dominance in the late-Holocene. More lasted from ca. 6000 B.P. to ca. 4000 B.P. recently. Bright (1966) has produced a similar The Area C diagram (Fig. 8) is basically diagram from Swan Lake in southern Idaho. a truncated version of the Area B diagram. His diagram shows how pine again dominates These samples were taken from a more interior the early Hoiocene, declines during the mid- location within the rockshelter and therefore Holocene (8000 B.P. to 3000 B.P.), and in represent less "time." This is reflected in the creases again during the last three thousand THE ALTITHERMAL REVISITED 289 years. In this area, however, Artemisia and contains 42% pine and yet there are no pines not Cheno/Am is the dominant mid-post in the West Humboldt Range. The nearest glacial pollen type. trees are in the Stillwater Range twenty miles For Nevada there are no diagrams yet to the southeast. It follows therefore that available that cover the whole of the Hoio changing pine percentages in the Leonard dia cene. Mehringer's (1967) well-dated Tule grams cannot be interpreted in terms of local Springs record is unfortunately incomplete changes in the importance of pine. after 7000 B.P. In southeastern Nevada, One possible explanation might be that the Madsen's (1972) diagram from the O'Malley Leonard record reflects changes in the impor Rockshelter is chronologically equivalent to tance of pine in the mountains to the west. A the Leonard Area B diagram and likewise indi cooling of climate, for example, would bring cates marked changes in vegetation. Here the about a lowering of tree lines and an expansion lower levels are dominated by sagebrush and of pine in the Sierran foothills. Alternatively, juniper, but at ca. 5200 B.P. there is a shift to a change in climate could mean a change in the grass and sagebrush, followed by a reversal to strength and direction of prevailing winds and sagebrush and juniper at around 3900 B.P. A this in turn would mean new patterns of pine similar trend towards more mesic conditions pollen dispersal. in the late-Holocene is evident in a diagram Another, perhaps more plausible, explan from Toquima Cave in central Nevada (Kautz ation is that the variation in pine pollen per and Thomas 1972). This diagram has a basal centages at Leonard is a statistical artifact. date of 3420 B.P. and shows a gradual shift As Mehringer (1967) has emphasized, changes from high grass and Compositae percentages in pine in Great Basin pollen diagrams may in the lower levels to high pine and juniper simply reflect changes in the local production near the surface. of non-pine pollen. In other words, a local The Great Basin pollen record is still increase in the production of Cheno/Am fragmentary, but when viewed as a whole it pollen would cause a percentage decline in clearly supports the Antevs model of climatic pine, even though the absolute concentration change. In different areas different species of pine did not change very much. Unfor are involved, but in all the sequences that we tunately, we cannot be conclusive on this are aware of there is an increase in xerophytic point as the absolute concentration values are types during the mid-Holocene. In other in part a function of the sedimentation rate, words, the controversial Altithermal was a and without better chronological control these period of warmer and drier climate. cannot be accurately determined. Less certain, however, is the magnitude of In the same context, it is interesting to note climatic change and the degree to which local that in both the Area B and Area C diagrams, environments were affected. Unfortunately, the high Cheno/Am levels correspond reason the pollen record is of limited value in this ably well with the wind-blown silt layer (Figs. context. The interrelationships between 7 and 8). One possible explanation for this is climate, vegetation, and the pollen rain are not that during the mid-Holocene falling lake yet well understood in the Great Basin and levels in the Humboldt Sink not only exposed detailed interpretation of pollen diagrams is the former lake floor to wind deflation but also therefore difficult. provided an extensive new area for coloniza A major problem is long-distance dispersal tion by plants. Furthermore, several of the of pine pollen. The Area B diagram (Fig. 7) invading species would be members of the illustrates the point well. The surface sample Chenopodiaceae or Amaranthaceae, and 290 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

Cheno/Am pollen would therefore be blown pollen record is not a particularly sensitive into the rockshelter in increasing quantities. measure of Hoiocene climatic change. The Today one of the dominant species on the temporal resolution, for example, is less than former lake bed is shadscale {Atriplex con ideal, all short term (<500 years) oscillations fertifolia), and there appears to be no reason having been averaged out by the 6-inch for assuming this was not also the case during sampling interval. Furthermore, the pollen the Altithermal. In brief, we interpret the rise record in itself does not permit an accurate in Cheno/Am pollen as an indication of reconstruction of the magnitude of climatic lowered lake level and therefore an indirect change. On the other hand, the Leonard pollen reflection of climatic change. record does confirm Antevs's reconstruction If the lake level thesis is correct, the high of the environmental history of the site, and, percentages of aquatic pollens in the lower in the broader sense, his three-part division and intermediate levels of both diagrams of Hoiocene climate. We would also suggest present something of a paradox. None of these that it provides a logical explanation for the pollen types is effectively wind dispersed persistent confusion as to the significance of (Durham 1951), and the question arises Hoiocene climatic change. The Pine/Cheno- as to how they were deposited in the rock Am oscillation clearly indicates that during shelter if the lakeshore was no longer close by. the last seven thousand years there have been One possibility is that aquatic pollen was major environmental changes in the Hum introduced into the site by man. Cattail boldt Sink. It does not follow, however, that {Typha) pollen is known to have been eaten the environmental changes in the Humboldt by the Indians of the area in both historic and Sink were the result of major changes in cli prehistoric time (Napton 1970), and theoreti mate. As we indicated earlier, the present cally, it could have been stored in the rock climatic conditions in the Humboldt water shelter and accidentally incorporated into the shed are marginal as far as the existence of a sediments. lake in the Humboldt Sink is concerned. The An alternative explanation, that we favor, important point here is that in areas such as is that aquatic pollen was first deposited in this even small changes in climate can have the shallow water around the edge of the lake, major environmental consequences. incorporated into the surface sediment, and This problem of scale underlies much of then later blown into the rockshelter as the the confusion in the debate as to the signifi lake bottom was exposed and underwent de cance of Hoiocene climatic change in the flation. Perhaps significantly, the highest Great Basin. Antevs himselfcontributed tothe percentage values of aquatic pollen are en confusion when he characterized the Alti countered in the lowest levels of the wind thermal as "the long drought" and suggested blown silt, suggesting that the maximum that extremely dry conditions persisted from influx occurred when the lake margin was still 7500 B.P. to 4000 B.P. (Antevs 1955). Various quite close to the rockshelter. The rather lines of evidence now indicate that the climatic regular nature of the aquatic curves argues conditions of the Altithermal were not very against the idea of cultural introduction and different from those of the present (Mehringer suggests that wind was the primary mechanism. 1977). Probably the best evidence on the mag nitude of Hoiocene climatic change in the THE NATURE AND SIGNIFICANCE Great Basin is the Bristlecone Pine tree ring OF THE ALTITHERMAL record (La Marche 1974), and the comple We should emphasize that the Leonard mentary evidence of changes in altitudinal THE ALTITHERMAL REVISITED 291 position of the Bristlecones (La Marche 1973). site's chronology. The Pine-Cheno/Am-Pine Both lines of evidence cover approximately the oscillation corresponds reasonably well with last 6000 years, and indicate that summer the stratigraphic units assigned to the Ana- temperatures were l°-2°F. above the long- thermal, Altithermal, and Medithermal, and is term mean during most of the period 6000 B.P. best interpreted as a reflection of changing to 4000 B.P. This change in temperature was lake levels in the Humboldt Sink. not very great, but it represents a change in In the broader context, Antevs' climatic climate that was significant enough to cause model is also endorsed. The Area B diagram, the desiccation of most of the postpluvial lakes like several other diagrams from the Great in the Great Basin, including the one that occu Basin, clearly indicates that the controversial pied the Humboldt Sink. According to Ben Altithermal was a period of warmer and drier son (1978), geochemical analysis of lake sedi climate. Unfortunately, the pollen record in ments indicates that Pyramid Lake was itself does not permit an accurate estimate of probably the only lake in the Lahontan Basin the magnitude of climatic change. We would that did not desiccate during the mid- emphasize, however, that in areas such as the Holocene. Great Basin even small changes in climate can In the Humboldt Sink region, prehistoric have far-reaching consequences. Climatic con populations were heavily dependent upon ditions in the Great Basin during the mid- lacustrine resources such as fish, waterfowl, Holocene may not have been very different and aquatic plants (Heizer and Napton 1969, from those of the present, but they were dif 1970; Napton 1969, 1970; Cowan 1967; Ambro ferent enough to cause the desiccation of 1967; Roust 1967). It follows therefore that nearly all the postpluvial lakes. It follows, changes in lake level must have had an impor therefore, that prehistoric populations heavily tant impact on human population densities. dependent upon lacustrine resources would The relationship between lake level and carry have been drastically affected. ing capacity is clearly not a simple one, but by The real significance of the Leonard pollen definition a subsistence economy based on record is that it lends support to the thesis that lacustrine resources cannot exist without a during the Hoiocene the climate has changed lake. in a regionally coherent and recognizable way. It also seems likely that relatively minor The Antevs' model is in many respects over changes in climate would have had significant simplified, but in essence it appears to be valid. effects on non-lacustrine food resources. For example, a shift towards more mesic condi tions would lead to an increased growth of ACKNOWLEDGEMENTS grasses and forbs in upland areas and this in We thank Steven Brandt for help in the turn would lead to denser game populations. earlier phases of the study, and Larry Benson, According to Harper and Alder (1970), such a Ted Oberlander, Phil Wilke, and an anony change occurred in northwestern Utah during mous reviewer for their useful comments on the late Hoiocene, and conceivably similar the manuscript. changes characterized the Humboldt Sink area. REFERENCES CONCLUSION Aikens, C. M. The Leonard Rockshelter pollen record 1970 HogupCave. University of Utah Anthro largely confirms Antevs's interpretation of the pological Papers No. 93. 292 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

Ambro, R. D. Vegetation des Zones Arides. UISB, 1967 Dietary-Technological-Ecological As Serie B (colloques), 9:101-102. pects of Lovelock Cave Coprolites. Bright, R. C. Berkeley: University of California Archaeological Survey Reports No. 70: 1966 Pollen and Seed Stratigraphy of Swan 37-48. Lake, Southeastern Idaho: Its Relation to Regional Vegetational History and to Antevs, E. Lake Bonneville History. Tebiwa 9(2): 1938 Rainfall and Tree Growth in the Great 1-47. Basin. Carnegie Institution of Washing Brown, M. ton Publication No. 469. 1960 Climate of Nevada. In: Climatography 1948 Climatic Changes and Pre-White Man. of the United States, No. 60-26. Wash In: The Great Basin, with Emphasis on ington, D.C.: U.S. Department of Com Glacial and Postglacial Times. Bulletin merce, Weather Bureau. of the University of Utah 38(20), Biologi cal Series 10(7). Bryan, A. L., and R. Gruhn 1955 Geologic Climatic Dating in the West. 1964 Problems Relating to the Neothermal American Antiquity 20:317-335. Climatic Sequence. American Antiquity 29:307-315. Arnold, J. R., and W. F. Libby 1950 Radiocarbon Dates (September 1). Cowan, R. A. University of Chicago Institute for 1967 Lake-Margin Ecologic Exploitation in Nuclear Studies. the Great Basin as Demonstrated by an Analysis of Coprolites from Lovelock Aschmann, H. H. Cave, Nevada. University of California 1958 Great Basin Climates in Relation to Archaeological Survey Reports 70:21-36. Human Occupance. Berkeley; Univer sity of California Archaeological Survey Cronquist, A., A. H. Holmgren, N. H. Holmgren, Reports No. 42:23-40. and J. L. Reveal 1972 Intermountain Flora: Vascular Plants of Baumhoff, M. A. the Intermountain West, USA, Vol. 1. 1958 Excavation of a Cache Cave in Pershing New York: Hafner Publishing Company, County, Nevada. Berkeley: University Inc. of California Archaeological Survey Reports No. 44:14-25. Durham, O. C. 1951 The Pollen Harvest. Economic Botany Baumhoff, M. A., and R. F. Heizer 5:211-254. 1965 Postglacial Climate and Archaeology in the Desert West. In: The Quaternary of Elsasser, A. B. the United States, H, E, Wright and 1958 The Surface Archaeology of Site 26-Pe- D. G. Frey, eds., pp. 697-707. Princeton: 5, Pershing County, Nevada. Berkeley: Princeton University Press. University of California Archaeological Survey Reports No. 44:26-51. Benson, L. V. 1978 Fluctuation in the Level of Pluvial Lake Elston, R., ed. Lahontan during the last 40,000 years. 1976 Hoiocene Environmental Change in Quaternary Research 9:300-318. the Great Basin. Reno: Nevada Archaeo logical Survey Research Paper No. 6. Billings, W. D. Fowler, D. D., ed. 1951 Vegetational Zonation in the Great Basin 1972 Great Basin Cultural Ecology: A Sym of Western North America. In: Les Bases posium. Reno: University of Nevada Ecologiques de la Regeneration de la Desert Research Institute Publications THE ALTITHERMAL REVISITED 293

in the Social Sciences No. 8. Jennings, J. D. 1977 Models and Great Basin Prehistory: A 1957 Danger Cave. University of Utah Symposium. Reno: University of Nevada Anthropological Papers No. 27. Desert Research Institute Publications in Kapp, R. O. the Social Sciences No. 12. 1969 Pollen and Spores. Dubuque: Williams Hansen, H. P. C. Brown Company. 1947 Postglacial Forest Succession, Climate Kautz, R. R., and D. H. Thomas and Chronology in the Pacific North 1972 Palynological Investigations of Two west. Transactions of the American Prehistoric Cave Middens in Central Philosophical Society 37(n.s.):l-130. Nevada. Tebiwa 15(2):43-54. Harding, S. T. La Marche, V. C. 1965 • Recent Variations in the Water Supply 1973 Hoiocene Climatic Variations Inferred of the Western Great Basin. Berkeley: from Tree Line Fluctuations in the University of California Water Re White Mountains, California. Quater sources Center Archives, Archives Series nary Research 3:632-660. Report 16. 1974 Paleoclimatic Inferences from Long Tree Harper K. T., and G. M. Alder Ring Records. Science 183:1043-1048. 1970 The Macroscopic Plant Remains of the Loud, L. L., and M. R. Harrington Deposits of Hogup Cave, Utah, and 1929 Lovelock Cave. University of California Their Paleoclimatic Implications. In: Publications in American Archaeology Hogup Cave, by C. M. Aikens, Appendix and Ethnology 25(1). I, pp. 215-240. University of Utah Anthropological Papers No. 93. Madsen, D. B. 1972 Paleoecological Investigations in Heizer, R. F. Meadow Valley Wash, Nevada. Reno: 1938 A Complete Atlatl Dart from Pershing University of Nevada Desert Research County, Nevada. New Mexico Anthro Institute Publications in the Social pologist 2:68-71. Sciences No. 8:59-65. 1951 Preliminary Report on the Leonard Martin, P. S. Rockshelter Site, Pershing County, 1963 The Last 10,000 Years. Tucson: Univer Nevada. American Antiquity 17:89-98. sity of Arizona Press. Heizer, R. F., and A. D. Krieger Martin, P. S., and C. M. Drew 1956 The Archaeology of Humboldt Cave, 1969 Scanning Electron Photomicrographs of Churchill County, Nevada. University Southwestern Pollen Grains. Journal of of California Publications in American the Arizona Academy of Science 5:147- Archaeology and Ethnology 47(1). 176. Heizer, R. F., and L. K. Napton 1969 Biological and Cultural Evidence from Mehringer, P. J., Jr. Prehistoric Human Coprolites. Science 1967 Pollen Analysis of the Tule Springs Area, 165:563-568. Nevada. Nevada State Museum Anthro pological Paper 13:130-200. 1970 Archaeology and the Prehistoric Great Basin Lacustrine Subsistence Regime as 1977 Great Basin Late Quaternary Environ seen from Lovelock Cave, Nevada. ments and Chronology. Reno: Univer Berkeley: Contributions of the Univer sity of Nevada Desert Research Institute sity of California Archaeological Re Publications in the Social Sciences No. search Facility No. 10. 12:113-168. 294 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

Morrison, R. B. Roust, N. L. 1964 Lake Lahontan: Geology of Southern 1966 Archaeology of Granite Point, Pershing Carson Desert, Nevada. U.S. Geological County, Nevada. Berkeley: University Survey Professional Paper 401. of California Archaeological Survey Reports No. 66:37-72. Napton, L. K. 1967 Preliminary Examination of Prehistoric 1969 Archaeological and Paleobiological Human Coprolites from Four Western Investigations in Lovelock Cave, Nevada Caves. Berkeley: University of Nevada. Kroeber Anthropological California Archaeological Survey Society Special Publication No. 2. Reports No. 70:49-88. Stockmarr, J. 1970 Archaeological Investigations in Love 1971 Tablets with Spores Used in Absolute lock Cave, Nevada. Unpublished Ph.D. Pollen Analysis. Pollen et Spores 13:615- dissertation. University of California, 621. Berkeley. Swanson, E. H., Jr. 1966 The Geographic Foundations of the Ranere, A. J. Desert Culture. Reno: University of 1970 Prehistoric Environments and Cultural Nevada Desert Research Institute Pub Continuity in the Western Great Basin. lications in the Social Sciences No. 1: Tebiwa 13(2):52-73. 137-146.