Nevada State Museum Anthropological Papers No

Home , Oxytenia, Tule Springs

NEVADA STATE MUSEUM

ANTHROPOLOGICAL PAPERS NO. 13

PART 3

POLLEN ANALYSIS OF THE YULE SPRINGS AREA, NEVADA

PETER J. MEHRINGER, JR.

GEOCHRONOLOGY LABORATORIES UNIVERSITY OF ARIZONA TUCSON, ARIZONA

Carson City, Nevada October, 1967 •

TABLE OF CONTENTS

Page

INTRODUCTION . . • • • • • • 132 POLLEN ANALYSIS AND SOUTHWESTERN ARCHAEOLOGY 132 COLLECTION OF POLLEN SAMPLES . . 135 EXTRACTION OF POLLEN . . • • 136 INTERPRETING THE FOSSIL POLLEN RECORD . 138 THE STUDY AREA . 146

VEGETATION AND MODERN POLLEN SAMPLES. 149

THE FOSSIL POLLEN RECORD • 169

DISCUSSION. 186

A PLEISTOCENE (WISCONSIN) CORRIDOR IN THE EASTERN MOHAVE DESERT . 189

SUMMARY AND CONCLUSIONS. 193

REFERENCES CITED . 194

ILLUSTRATIONS

Color Plate. Pleistocene pollen from the Las Vegas Valley, Nevada. (ca. 640x) . Facing 129 Figure 1. Photomicrographs of pollen from human coprolites . . 133 2. Panamint Mountains modern pollen surface samples . . . 140 3. Modern pollen surface samples, Tule Springs and Corn Creek Sites 142 4. Lehner Ranch Arroyo. Pollen Profiles II, VII . . . 145 5. Map of Study Area...... 147 6. View of the Tule Springs Area . . . . 148 7. Pollen diagram of major pollen types from Kyle Canyon, Charleston Mountains ...... 151 8. Soil surface sample, Locality 25 . . 153 9. Modern spring sample, Locality 28 153 10. View from the Tule Springs Site . . . . . 153 11. Modern surface sample, Locality 1, Kyle Canyon . 153 12. Modern sample, Locality 27, at Corn Creek Spring . 154 13. Modern surface sample, Locality 5, on Kyle Canyon Fan . . . . 156 14. Vegetation between modern surface sample, Localities 9 and 10 in Kyle Canyon ...... 156 15. Modern surface sample, Locality 12, in Kyle . . . . . 156 16. Modern surface sample, Locality 15, in Kyle Canyon . . . . 156 17. A blackbush-joshua tree community on Cold Creek Road . . . 157 18. Modern surface sample, Locality 11, in Kyle Canyon . . . . 157 19. Modern surface sample, Locality 16, in Kyle Canyon . . 159 20. Modern surface sample, Locality 17, in Kyle Canyon . . . . 159 21. Modern surface sample, Locality 18, Charleston Peak, Kyle Canyon . 159

130 ILLUSTRATIONS (CONT.)

Figure Page 22. Modern surface sample, Locality 19, in Lee Canyon . . . . • 159 23. Photomicrographs of fossil pollen from the Tule Springs Site . • 162 24. Photomicrograph of fossil Abies pollen from Unit B2, Tule Springs Site • 163 25. Photomicrographs of fossil pollen from the Tule Springs Site. . . • 164 26. Photomicrographs of fossil pollen from the Tule Springs Site. . . • 165 27. Photomicrographs of fossil pollen from the Tule Springs Site. . . 166 28. Photomicrographs of fossil pollen from the Tule Springs Site. . . 167 29. Photomicrographs of fossil pollen from the Tule Springs Site. . • 168 30. Area-elevation graph for the Spring Range...... • 173 31. Pollen Profile I: pollen diagram from spring laid clay and silt, Tule Springs ...... 175 32. Pollen Profile II: A pollen diagram from buried lake sediments, Tule Springs ...... 177 33. Pollen Profile III: Pollen diagram of spring mound sediments, Tule Springs ...... 179 34. Spring mounds 4, 4A, and the Charleston Mountains . . . . . 180 35. Pollen Profile IV: Pollen diagram from the alluvial sequence of the Fenley Hunter Locality...... 181 36. Pollen Profile V: Miscellaneous fossil pollen counts from the Tule Springs Site Area ...... 182 37. A pocket of ponderosa pine, pinyon pine, and juniper at the mouth of Pine Creek Canyon, Spring Range ...... 188 38. Map of possible Pleistocene woodland corridoL . . . . . 190

TABLES

Table 1. Extraction procedure for Southwestern alluvial pollen samples . . 137 2. Relative frequency of Pollen types from Mono Lake, California . 143 3. Modern pollen samples from the east slope of the Spring Range and the Las Vegas Valley, southern Nevada ...... 160 4. Approximate age of the stratigraphic units shown on the pollen diagrams . 170 5. Typha pollen counts in tetrads, dyads, and monads . 171

131 INTRODUCTION The weathered sediments of the arid tant forest clearing is associated with a Las Vegas Valley are a most unlikely source decrease in tree pollen and an increase in of a fossil pollen record; therefore, the herb and shrub pollen, in addition to the pollen investigations were approached with presence of cereal pollen (Iversen, 1949, some apprehension, which was not entirely 1960). unwarranted. Most of the pollen samples In Europe the success of palynology extracted did not contain pollen and, in associated with archaeology is partly due many of the samples that did yield pollen, to the presence of artifacts in and the prox- the pollen was too poorly preserved for ana- imity of sites to bogs or other ideal deposi- lysis. In spite of the many problems inher- tional environments (Gray and Smith, 1962). ent in pollen analysis in arid regions, some In the arid Southwest such associations are useful and exciting information was obtained virtually unknown. Instead of peat bogs or from the pollen content of alluvial and spring lakes, which are ideal for palynological mound deposits. Most of the pollen results investigations, the pollen content of trash cannot be directly related to archaeological mounds, room fills, cave earth and alluvium finds, but a sequence of vegetational change must be studied (Martin and Gray, 1962). during the time that man occupied the Las New ideas about pollen and archaeology are Vegas Valley can be suggested. being developed and tested in bogless regions The conclusions given here vary from the where there are theoretical reasons why such preliminary report (Mehringer, 1965) only in work might have little hope for success and small details. Some new data and a more where there are still many unsolved problems. complete discussion of the fossil pollen The vast majority of pollen studies asso- record and the modern pollen samples are ciated with archaeological sites have been given in this report. Sections on the prob- approached with hopes of obtaining environ- lems and procedures of pollen analysis in the mental reconstructions, especially where arid Southwest are included, mainly as a ecological and cultural changes are assumed guide for Southwestern archaeologists. 1 to be related. For several reasons many of these attempts have not been entirely suc- POLLEN ANALYSIS AND cessful. Among these reasons are the lack SOUTHWESTERN ARCHAEOLOGX of pollen, the presence of pollen in a low Although pollen analysis has been density or in a very poor state of preservation, used as an archaeological tool for many years and the factor of cultural disturbance of the in northern Europe (Gray and 8mith, 1962), natural vegetation (Yarnell, 1965) and of the with few exceptions (Deevey; 1944: Sears, pollen content of beds in and around the sites 1932; Ogden, 1965, p. 495) its use in the New investigated. A pollen sample which is too World is relatively recent, with the great poorly preserved for routine analysis may con- majority of published work postdating 1960. tain archaeologically important economic pollen In northern Europe Neolithic man's effect on types even though the sample is essentially his environment has been well-documented useless for environmental reconstruction. by the pollen record, which shows that the Although some attempts at environmental introduction of agriculture and the concomi- reconstruction for periods encompassing parts

'This work was supported by National the present distribution of vegetation in southern Science Foundation Grants G-21944 and Nevada, C. V. Haynes for furnishing detailed GB-1959 to P. S. Martin, University of stratigraphic control, and P. S. Martin, T. L. Arizona, and GS-23 to R. Shutler, Jr., Nev- Smiley, and D. P. Adam for comments on the ada State Museum. I thank R. H. Hevly and manuscript. This article is Contribution No. P. 5. Martin for aid in pollen identification 116 of the Program in Geochronology, Univer- and counting, W. Glen Bradley for time spent sity of Arizona, Tucson. in the field and for information concerning 132 Figure 1. Photomicrographs of pollen from human coprolites. Photomicrographs of pollen from human feces collected from archaeological sites in Glen Canyon, Utah (identifications and photographs by P. S. Martin). A, Cucurbita pepo (640X); B, Cactaceae, Platyopuntia-type (600X); C, Oxytenia (1000X); D, Populus (1000X); E, Sarcobatus (1000X); F, Shepherdia (1000X); G, Cleome (1400X); H, Cucurbita moschata- type (575X); I, Gramineae (1000X); J, Zea mays (620X).

133 of the last two thousand years appear to (Fig. 1). While the pollen content of pre- have been successful (Hevly, 1964; Schoen- historic feces is of little direct use in paleo- wetter and Eddy, 1964), these studies were cological interpretation, it gives useful full-scale regional pollen investigations. information on the diet of the people and may It may be some time before an archaeologist add further economic plant types to the list can hope to collect a few samples from a of the ethnobotanist (Martin and Sharrock, single site and thereby gain meaningful 1964). environmental or cultural data with a minimum Not only does the content of human feces of time, effort, and expense, but gradually offer valuable information on the diet of peoples we are learning more about the problems and but the pollen content of the feces of domes- procedures of archaeological palynology in ticated animals is also of value. A case in the Southwest. The work of Hevly (1964) point is the analysis of turkey feces from the and Schoenwetter (1966) suggests the possi- dwellings at Mesa Verde (Martin, pers. comm.). bility of establishing regional pollen chronol- They contained abundant pollen of Cleome, Zea, gies which can be used as an independent and Opuntia -- pollen types which are commonly method for dating or correlating within and associated with archaeological deposits. In between sites. northern Europe, the presence of ivy (Hedera) Environmental reconstruction for longer pollen in the dung of domesticated animals has periods of time, as from the late-glacial, is been used as collaborative evidence suggesting also progressing rapidly and often in a con- that the wild plants were gathered for fodder text which has direct archaeological implica- (Troels-Smith, 1960). When examining the tions (Hafsten, 1961; Hansen, 1951; Heusser, pollen content of feces there is the possibility 1963; Hevly, 1964; Martin, 1963a; Mehringer, of identifying insect-pollinated types which 1965; Mehringer and Haynes, 1965; Wendorf might be important ecological indicators not and Hester, 1962). As the pollen evidence ordinarily recovered. continues to be added to the other new and Another possible use of palynology is the existing evidence for the paleocology of location of cultivated fields and the determina- regions occupied by man since he first mi- tion of what was being grown there. Most grated to and spread throughout the New economic pollen types are extremely rare in World, the archaeologists should have a non-archaeological deposits. For example, continually refined environmental framework the modern soil surface of an Indian or hybrid with which to mesh the history of subsistence corn field contains only 1% or less corn pollen. patterns, of abandonment and migrations, Thus, the presence of corn pollen in a shallow and of cultural change or stability. soil profile from a likely area near a village One of the most intriguing uses of could be an indication that the area was used palynology has been the study of the fossil as a cultivated field (Martin and Schoenwetter, pollen content of coprolites. This method 1960, Table 1). This has been demonstrated was first used in the Southwest at Gypsum at Mesa Verde where it was thought that crops Cave, Nevada (Laudermilk and Munz, 1934) were being cultivated behind stone check dams and later at Rampart Cave in the Grand Can- (Woodbury, 1961, Fig. 12; 1963, p. 63); soil yon of Arizona (Martin, Sabels, and Shutler, beneath the post-occupation alluvium behind 1961) in the investigation of the contents of the check dams contained the pollen of corn the dung of the Shasta ground sloth (Nothro- and beeweed (Cleome) (Martin and Byers, 1965). therium). The results showed that plants not A good example of close cooperation present as macrofossils were represented between the pollen analyst and archaeologist by fossil pollen and that the now extinct is found in the work of Hevly (1964), who Nothrotherium was feeding on desert plants. investigated the pollen content of several Human coprolites from the Glen Canyon of archaeological sites excavated by P. S. southern Utah are rich in the pollen of both Martin (Chicago Natural History Museum) in cultivated and wild species (Martin and central Arizona. By the careful analysis of Sharrock, 1964). The pollen types include floor samples from different rooms of a pueblo, corn, squash, and beeweed (Cleome), the Hevly was able to characterize the use of the most abundant ppllen type, as well as cactus, rooms. The difference in relative abundance grass, and Chenopodiaceae or Amaranthus of economic pollen types placed pueblo rooms

134 into two categories -- storage and habita- The usual procedure is to collect a tion. Some interesting information was also series of samples, from an arroyo or trench gained from the pollen content of soil washed wall, that will represent the fossil pollen from the surfaces and pits of metates. The rain through the period of time represented difference in economic pollen types suggests by the sediment. The sampling interval that different textured metates had different should be such that an important change in uses. A coarse basalt metate contained cac- the pollen record is not missed in sampling, tus pollen and a fine textured sandstone and is therefore dependent upon the sedimen- metate contained corn pollen (Hevly, 1964, tation rate. For example, fine sediments pp. 90-91, Table 4). might be sampled more closely than coarse Man's effect on the natural vegetation, sediments (Dittert and Wendorf, 1963). The which is well illustrated in the pollen rec- arroyo walls of southeastern Arizona are ords of northern Europe (Gray and Smith, 1962), ordinarily sampled at 10-centimeter intervals. can as yet only be inferred from the evidence This sampling interval is usually adequate. of possible secondary forest succession fol- In selecting a sampling plan the stra- lowing abandonment of some areas which were tigraphy of the site should be the guide. If formerly under cultivation (Leopold, Leopold, one locality in an arroyo has a greater thick- and Wendorf, 1963; Martin,1962). However, ness of one unit and at another locality a few in this case, the choice between cultural and meters away there is a greater thickness of climatic change cannot easily be made. another unit, each should be sampled at the While pollen analysis is widely used location which will give the most complete in environmental reconstruction, and ultimate- record. When the upper part of a stratigraphic ly the tree-ring and pollen chronologies may unit is located in one area and the lower part be correlated (Hevly, 1964, Fig. 20), there in another, both should be sampled and the are other types of information, some of which pollen diagrams crossed-matched (Fig. 4). have been mentioned, which may prove use- Samples should be collected so as to cover ful to the archaeologist. Untried approaches, each stratigraphic unit as completely as limited only by lack of imagination, may yet possible, and each unit should be treated as be fruitful. It is important that the archae- a separate sampling problem depending on the ologist be aware of the possibilities and type of sediment, its relative importance to the limitations so that he may ask realistic ques- sequence, and the associations with artifacts, tions, framed in an archaeological context, charcoal for C14-dating, or fossil remains. that are amenable to study by pollen analysis. A sample should be collected from near the top While the use of pollen analysis in South- and bottom of each stratigraphic unit but never western archaeology is still in its infancy on a contact. A few closer interval samples and in some cases still based on a trial and might be collected when the association is error approach, the results to date give every especially important (Fig. 4, Profile VIII). reason to expect a high degree of success in Once the wall has been mapped and a future studies. sampling locality established, the wall is thoroughly cleaned. The amount of material COLLECTION OF POLLEN SAMPLES removed from the wall will vary, depending on the degree of weathering and surface Whenever a site is to be sampled for cracks; usually 30 to 60 centimeters is suffi- pollen, the collector must first have in mind cient. On a fresh bulldozer cut, only a few the sort of information desired from the re- centimeters of material need be scraped from sults of the pollen analysis. Therefore, it the wall. After the wall is cleaned, the is difficult to proceed under any set of rules; section is measured and a red flag of surveyor's however, common sense is always the best tape is placed in the wall with a 16-penny nail guide. The following discussion is con- to mark where each pollen sample will be taken. cerned with the collection of pollen samples The sample number and depth should be written from alluvial profiles in the Southwest with on about every fifth flag so that the exact the purpose of reconstructing vegetational locality of any sample can be verified if the change, and the procedures given are those need arises in later work. Before collection, presently used by the author. each sample number and depth should be placed 135 on a stratigraphic diagram and a label pre- Potter and Rowley, 1960). pared for each sample with the same information. Samples are collected from the bottom EXTRACTION OF POLLEN of the profile to the top, and the prepared wall is again cleaned at the locality of a The following pollen extraction method, sample collection just prior to taking each which is in general use for alluvial sediments individual sample. at the Geochronology Laboratories, University About 200-300 grams of material is of Arizona, is a modification of methods used collected with a trowel or other suitable tool by Bernard C. Arms and is partly the result of and placed in a clean plastic bag and labeled. discussions with Richard Bennett of the geo- Whenever possible the sediment is collected chemistry section of the Geochronology Labora- in a single large ped or several smaller ones tories. Pollen has been difficult to extract so that before extraction each ped may be from some alluvium because of the formation washed with water until surface material has of resistant colloids, the presence of obscur- been removed. This serves as a further guard ing inorganic debris, and low pollen concen- against contamination which may occur during trations. Much of the time use of the following collecting. extraction method overcomes these problems. Surface samples are collected in order (For a comprehensive review of pollen extraction to obtain an approximation of the pollen con- techniques see Gray, 1965). tent of the soil surface. From 10 to 40 sub- In the following procedure the reagents samples (ca. 4 to 8 grams) of surface dirt and distilled water are stirred well with the (upper 0.5 centimeters) are collected on the sediment when added. After each step, the tip of a trowel and placed in a single clean sample is centrifuged for 2-3 minutes and the plastic bag. Before the sample is extracted, liquid is decanted. A nearly full test tube of the contents of the bag are thoroughly mixed. reagent or water (approximately 40 ml.) is All pollen studies should include modern added unless otherwise stated. The extraction samples from the different plant communities technique given here and summarized in Table in the area for use in interpreting the fossil 1 is only a general outline; steps may be added pollen record. At arroyo sites, these should or deleted as one gains experience with the include a sample from the arroyo floor. sediment of a particular age or area. Although A single subsample of surface Approximately 50 grams of sample is dirt will contain enough pollen for a count, mixed with 250 ml. of water in a 500-ml. there is sample variation among the sub- plastic beaker. Concentrated hydrochloric samples (Mehringer, unpubl.), and therefore acid is added, 25 ml. at a time and stirred several subsamples are necessary to obtain slowly. This is continued until the material a better approximation of the modern soil can be stirred vigorously with no further pollen. While collecting the surface sample, reaction. This step removes much of the car- the plant species in the area should be listed, bonate and assures the release of pollen and this list should accompany the sample. cemented in calcareous matrix (Anderson, To alleviate the problem of seasonal variation, 1955). The mixture is then stirred with a cir- it is desirable to collect the samples during cular motion until a strong vortex is formed the winter months when the fewest number of and all of the material in the beaker is in species are flowering (Hevly, Mehringer, and motion. The beaker is allowed to stand about Yocum, 1965). In any case, the date of 45-60 seconds until the heavier fraction such collection should be given, as well as notes as sand and gravel starts to settle out. While made on any species which were flowering the material is still in motion it is poured at the time of sample collection. It is as- slowly through a 4000-mesh screen into a sumed that samples collected in this manner 250-ml. beaker. Any sediment remaining on represent several years of pollen accumula- the screen is washed well with a strong jet tion and that the data is useful in interpreting of water from a plastic squeeze bottle. The the pollen record (Bent and Wright, 1963; material is again stirred in a circular motion Dixon, 1962; Hevly et al., 1965; King, 1964; until a strong vortex is formed, and then it is Maher, 1963; Martin, 1963a, 1963b; Martin allowed to stand less than two minutes. The and Gray, 1962; Mehringer and Haynes, 1965; material is then poured from the beaker into 136 a 50-ml. nalgene tube, which is centrifuged is removed from the vial with a capillary and the liquid decanted. The last step is pipette. A drop of the pollen-containing repeated until there are 2 to 4 cc. of sediment residue is placed on a microscope slide (1 mm. in the bottom of the test tube. The preceding thickness) with a drop of glycerol and stained part of the extraction technique is called the with basic fuchsin. The preparation is covered "HC1 swirl." with a No. 0 cover glass. The remaining res- After the 2 to 4 cc. of sediment are idue is stored in the shell vials in a mixture of accumulated in the test tube, it is washed with 2 parts glacial acetic acid, 2 parts glycerol, 10 to 14 percent hydrochloric acid followed by and 3 parts water. 25 to 30 percent hydrochloric acid and a water wash. The last two hydrochloric acid steps TABLE I. Extraction procedure for Southwestern remove the remaining carbonates and generally alluvium pollen samples. eliminate any vigorous reaction when the hydrofluoric acid is added. Next, 50 percent Approx. 50 r. of sediment hydrofluoric acid is added, starting with 10 ml., which is stirred gently. If no reaction HC1 swirl, 750 ml. beaker results after 2 minutes the remainder of the le hydrofluoric acid is added and stirred well. #100 mesh brass screen If there is a reaction, the remainder of the I, hydrofluoric acid is added in 2 or 3 steps and HC1 swirl, 250 ml. beaker stirred gently after each addition. The sample Li in 50 percent hydrofluoric acid is left unstop- transfer to test tube pered but protected from contamination by paper Li toweling and allowed to stand for 6 to 24 hours 10-15% HC1 in a well ventilated fume hood. Thirty ml. of le 70 percent hydrofluoric acid is added, allowed 25-30% HC1 to stand unstoppered for 6-24 hours and then 1( stirred and placed in a gently boiling water H20 wash bath for 20 to 30 minutes. While the 70 per- -V- cent hydrofluoric acid is in the water bath, a 50% HF, 6-24 hour stand container of water is heated and used for the 1, two hot water washes which follow the 70 per- 70% HF, 6-24 hour stand cent hydrofluoric acid step. If Next, 25 ml. of 20 percent nitric acid 70% HF, 20-30 minutes in is added and allowed to stand for 10 minutes boiling H20 bath and followed by two more hot water washes. Thirty ml. of concentrated hydrochloric acid 2 hot H20 washes is added and the tubes are placed in a boiling le water bath for two minutes. Again two water 20% HNO3, 10 minutes stand washes follow. The treatment with nitric acid Li oxidizes some of the organic colloids which 2 hot H21,0 washes become water soluble after the hydrochloric acid treatment. If necessary, acetylation conc. HC1 (37%), 2 minutes (Gray, 1965, p. 554) follows the water washes in boilil H20 bath after the preceding HC1 treatment. A 5-7 percent solution of sodium hydro- 2 H20 washes xide or potassium hydroxide is added and the ---E--- tubes are placed in a boiling water bath for 5-7% KOH, 2-3 minutes in boiling H20 bath 2-3 minutes. The sample is then washed 1/ with water until the decanted solution is water washes until decant is clear clear. This last step may require from 2 to I, as many as 12 water washes depending on transfer to vial the humate content of the sample. The re- If maining sediment is transferred to a 5-ml. mount on slide shell vial and centrifuged. Most of the water 137 INTERPRETING THE FOSSIL POLLEN RECORD are not. Pollen may also be carried by water from one ecological zone to another by both The use of fossil pollen in paleoeco- small and very large drainage systems (Muller, logical interpretation has the advantage that 1959; Potter, 1964b). pollen may be present in many types of sedi- The fossil pollen record consists mainly ments and when other fossils are entirely of wind pollinated types and therefore presents lacking. Pollen is also numerous and can a limited estimate of the plants which conceiv- therefore be treated statistically. While there ably could have contributed their pollen to any are the advantages of ubiquity and quantity to particular sediment. Few of the pollen types the use of fossil pollen, there are limitations which are commonly recovered in the Southwest to the method just as there are to the use of can be identified to species, and many of those any fossils in paleoecological reconstruction. are identifications "by default", with the iden- Before the pollen record can be adequately tification resting on the fact that a genus is interpreted, it is important to understand the monotypic, as Zea mays, or is represented in general limitations of the method and the speci- the study region by only one species. An edu- fic limitations for a given region. Some of the cated guess as to species can also be made more important factors for consideration include where two species have similar pollen types pollen production, dispersal, identification, but occupy different ecological niches, such and preservation, as well as counting methods as Shepherdia argentea and S. rotundifolia. and statistical treatment. Excellent discussions Despite notable variations in morphology, of these factors are given by Potter and Rowley many pollen types have only been identified to (1960) and Faegri and Iversen (1964). The sta- genus in Southwestern pollen studies. Others tistical assumptions of pollen counting are such as Ephedra and Polygonum may be easily presented by Mosimann (1965). separated into groups of species. In the case A discussion of pollen production and of the Gramineae (with the exception of corn) its effects on the fossil record is given by only the family identification is made; in the Potter (1964a). Since different species produce Compositae and Cactaceae, families may be different amounts of pollen, it usually is not divided into groups (Heyly et al., 1965,p. 128). possible to convert the relative frequencies of In the cheno-am category, identification in- pollen types to phytosociological parameters, cludes genera of the Chenopodiaceae and the although some approximations are possible genus Amaranthus of the Amaranthaceae. The (Leopold, 1964). The total pollen production of number of recognizable pollen types will in- a given species will be determined by the fac- crease with more morphological studies. tors which affect the physiology of the plant, Detailed morphological studies of Southwestern and thus the pollen production will vary from pollen are a prerequisite to a more refined year to year. There is also a periodic, pos- pollen chronology and essential to a better sibly cyclic, variation of pollen production in understanding of past environments. some genera (Hyde, 1952). Preservation of the pollen grains is Superimposed on the problem of differ- important in their identification. Divisions ential pollen production is the factor of which can be made within a family or genus on differential distribution and representation. the basis of minute morphological details or The vast majority of pollen grains recovered slight size differences cannot be accomplished in the fossil record are wind pollinated (anemo- in poorly preserved material. Thus, in the philous) types. Self-fertilizing and animal, alluvial fossil pollen record, in which the mainly insect (entomophilous), pollinated pollen is not commonly well-preserved, it species produce pollen types which are not would be unrealistic to attempt more precise adapted to wind dispersal, are not produced identifications and it is more practical and in abundance, and are rare or completely less misleading to work only with the major lacking in the fossil pollen record. Even pollen groups. among the types well adapted for wind pol- Preservation of pollen is dependent on lination there are differences in dispersal such factors as the method of transport, en- distance. For example, in the Southwest, vironment of deposition, and the history of the pine pollen is common many miles from its sediments after deposition. If the pollen grains site of production while spruce and fir pollen are transported in sediments to the site of 138 deposition, either by wind or water, they easily identified even in a poor state of may be mechanically eroded in transport preservation, the work to date has shown and in a poorly preserved state even before that much valuable information can be ob- being deposited. Once deposited, there tained. It has also been shown that the are still other factors which affect the pre- results can be duplicated and that they servation. It is assumed that a reducing follow the patterns that appear in the modern environment is conducive to pollen preser- soil surface samples (Martin, 1963a; Meh- vation, while an oxidizing environment ringer, unpubl.; Schoenwetter and Eddy, 1964). contributes to the destruction of pollen. The The presence of well-preserved pollen problem of poor preservation is compounded in a sample that also contains poorly preserved since not all pollen types are affected to the pollen of the same type may indicate contami- same degree. The pollen of Populus is nation of a sample with younger material. At fragile and easily destroyed, and thus it may the Lehner site in southeastern Arizona (Haury, be very poorly preserved or entirely lacking Sayles, and Wesley, 1959; Mehringer and in a sample where most of the other grains Haynes, 1965) this situation occurs when there are well-preserved. Juniper pollen is prob- is the mixing of sediments of two different ages ably preserved best in aquatic environments near the contact separating them. When younger, (Dixon, 1962). Pollen is also differentially better preserved pollen, some of the more obvious destroyed by phycomycetes (Goldstein, 1961). contaminants can be eliminated during counting, The history of the sediments after the but not with enough assurance to make this prac- pollen has been deposited is also a factor in tice reliable. The problem of mixing at contacts preservation. Pollen may be lacking on or in alluvial sediments seems to be common and below weathered surfaces and abundant thus a full understanding of the alluvial stratig- elsewhere in a section. This occurs even raphy is absolutely essential to the interpretation when the original depositional environments, of the pollen record (Fig. 4). such as lake deposits and organic spring Inferences can be made from the presence mats, should have been excellent environ- of specific pollen types in a sample. If cattail ments for pollen preservation. In the Las (Typha) pollen is abundant in a fossil sample, Vegas Valley, Nevada, spring mats buried the sediments probably represent a shallow- by two meters (about 6 feet) or more of water or near-shore aquatic environment. The eolian deposits and dated at about 10,000 abundance of greasewood (Sarcobatus) pollen radiocarbon years ago contain well-preserved might indicate high soil salinity. The presence pollen. Spring mats of the same age which of pollen types without reference to their rela- are exposed, or near the surface and weath- tive frequency is also useful in the analysis of ered, do not contain enough well-preserved coprolite samples (Martin et al., 1961; Martin and pollen for routine analysis. Another, pos- Sharrock, 1964), where the relative frequencies sibly related, case is the lack of pollen from are a result of artificial selection dependent on the upper one to two meters (3 to 7 feet) of the availability of certain plants or a food pre- Willcox Playa, southeastern Arizona (Martin, ference. 1963b). The changes in the relative frequencies The ideal situation in which all the of pollen types, which are assumed to indicate pollen grains present in a sample are well- a change in vegetation through time, are typi- preserved and in which there is little or no cally represented in the form of a Pollen Diagram differential destruction may be realized under which is composed of Pollen Spectra, each repre- the best conditions, but with few exceptions senting a single sample and plotted from older to these conditions have not been realized at younger from the bottom to the top of the diagram. lower elevations in the Southwest. Alluvium The pollen diagram is a graphic means of illus- may be the only source of a pollen record for trating and interpreting pollen counts. All the a given site or time period, and, typically, factors that affect the production, dispersal, the pollen present is not especially well- deposition and post-depositional history of the preserved. While such pollen samples do pollen grains are incorporated in the pollen not lend themselves to the sort of detailed diagram. morphological study which can be carried out What does the pollen diagram mean in with bog or lake samples, and while the types terms of past environment? There are two com- counted are those which preserve well or are monly used approaches to answer this question. 139 ES ERT SCRUB I WOODLAND (Artemisig - LAO - loirmitie)

50—

25 —

•• •• •• •• •• •• •• •• • • •• •• •• •• •• •• •• •• • •• •• • • 111

0 _

25—

0 — PERCENT

25 — —85

- 50

Lu ce tu -25 a. POLLEN COUNT

RELATIVE 75 —

50 —

25 —

0 - Feet 0 0 3000 3500 4000 4500 5000 5500 6000 6000 6500 6500 7000 7500 8000

91,0 10 a 2130 2 2,80 ,60 1 2 ,20 1 3,70 1520 16,70 18,2 0 18,2 0 a 1918 0 19, 8 0 , 241 30meters

Searles Lake Pluvial. ELEVATION Zone of Overlap 10000- 24000 BA pluvial & modern PANAMINT MOUNTAINS MODERN POLLEN RAIN

Figure 2. Panamint Mountains modern surface samples. A surface sample transect from the Panamint Mountains, California, showing the variation in the surface pollen counts with changing elevation and vegetation and the zone of overlap between the modern soil surface counts and the pluvial-age fossil pollen counts from Searles Lake. The averaged fossil counts from Searles Lake sediments of 24,000-10,000 B.P. are similar to the modern pollen rain between 1,820 and 2,130 meters. Used as a direct comparison, these counts suggest a lowering of the present woodland vegetation by at least 1,200 meters (from Martin, 1964, Fig. 25, Table 3).

140 The first is the use of pollen types of ities. species or genera which have known ecolo- Modern pollen samples are obtained from gical tolerances today and is based on the a variety of sources including air samples, moss assumption that these tolerances were the polsters (Hansen, 1949; King, 1964), bottom same during the time represented by the samples from lakes or bogs, cattle tanks (Haf- pollen record. This method has been widely sten, 1961; Martin, 1963a), and soil surface used in northern Europe and the work has samples. The latter source is presently the been reviewed by Dahl (1964), but, with the most widely used in the Southwest. It is exception of Wendorf (1961, pp. 127-129), it desirable to attempt to duplicate the deposi- has had only indirect use in the interpretation tional environment of the fossil record with the of past Southwestern climates from the fossil modern pollen samples. Thus, the soil surface pollen record. samples are suitable for interpreting alluvial The second method, which is discussed fossil pollen records but they may not be adequate in detail by Leopold (1964), attempts to re- for interpreting the fossil pollen content of lake late the relative frequencies of pollen types sediments. An example of the variation of pollen to the modern plant communities. A thorough counts from two different depositional environ- understanding of the modern pollen rain in ments -- soil surface and springs -- in the Las relation to the existing vegetation is essen- Vegas Valley, Nevada, is shown in Figure 3. tial in the use of this method. One assumes The apparent over-representation of pine that modern plant communities have fossil pollen in the fossil record of the arid Southwest equivalents and that the relative frequencies is a problem that plagues the pollen analyst. In of the pollen types which are found today in areas of high relief where the forests are separ- given plant communities have equivalents in ated from the deserts by several hundred meters the fossil pollen record. An attempt by Mar- in altitude but by only a few kilometers in dis- tin (1964) to apply this method directly is tance, high percentages of wind-blown tree shown in Figure 2. The advantage of this pollen are encountered in the modern desert method is that the interpretation does not pollen rain, as at Tule Springs. The relative require complete knowledge of the factors frequency of blown-in coniferous pollen increases that affect the distribution and preservation with the reduction in plant cover as one enters of every pollen type in the fossil record--in the desert (Martin, 1964; Martin and Gray, 1962, most cases this knowledge is lacking--but Fig. 1). If local vegetation iŠ entirely lacking, depends rather on the quantitative relations the long distance transport must, of course, among the various pollen types present in account for all pollen present in modern samples. the modern samples from various plant com- This situation occurs in two cases in the South- munities. For example, we know that 10 per- west--the center of barren playas (Martin, 1963b) cent pine pollen is not unusual in modern and in large lakes. soil surface samples in the Sonoran Desert Pollen counts by P. S. Martin of bottom even though the nearest pine trees may be samples from various depths in Mono Lake near many miles away (Heyly et al., 1965), while the California-Nevada border east of Lee Vining 10 percent fir (Abies)pollen is a good indica- California, and on Paoha Island (Table 2, Fig. tion of the close proximity of fir trees. 5) illustrate the over-representation of pine pol- The method is especially useful where len in a large body of water surrounded by shrub the modern plant community is not charac- vegetation below the forest border but near the terized by the pollen of its dominant species. source of forest pollen. Mono Lake is surrounded In parts of the Mojave Desert, where joshua by sagebrush (Artemisia) and Paoha Island is trees (Yucca brevifolia) and blackbush (Coleo- dominated by greasewood (Sarcobatus). Trees (Tyne ramosissima) are the characteristic and Artemisia were not observed by Martin while species, the pollen of Liliaceae and Rosaceae collecting the samples on Paoha Island. The might be expected in the modern soil surface coniferous forests of the Sierra Nevada are less samples. But in this community in the than 6 kilometers (about 4 miles) from the wes- Charleston Mountains, Nevada, I have not tern shore of Mono Lake. Within a kilometer of recovered Liliaceae pollen and Rosa ceae the western shore there is a dense stand of Pinus pollen is no more abundant than in the modern monophylla. The fact that the tree pollen does surface samples in many other plant commun- not increase with distance from shore suggests 141 MODERN POLLEN RAIN Tule Springs Site Corn Creek Springs 0 40 20 40 46 0 20 49 58 % i i I Charleston Peak Pine ■ feet) 3631 meters,(11,9I2 Juniper 1 1 2,000 - '\ Alpine Low-Spine Composites - 3500 High-Spine Composites MI Cheno-Ams 1 0,000 - Bristlecone Pine - 3 000 Grass 1 Limber Pine Cyperaceae White Fir 8, 000 - 2 500 Ephedra • s Ponderosa Pine Other La Pinyon Pine - 2 000 6,000 - SPRING- ' Juniper x RANGE Sagebrush elackbush, Joshua Tree TULE SPRINGS SITE 1 500 703 Meters ,(2307 feet) 4,000 - ltbushs Bursage - 1 000 Creosote Bush Saltbush 2,000 25 610 MILES 5 1 0 30 2.0 30 KILOMETERS 20 ,40 48,

Figure 3. Modern pollen surface samples, Tule Springs and Corn Creek Sites. A diagram showing the averaged modern pollen counts in two different environments in the Las Vegas Valley and the distribution of the vegetation zones with increased elevation from the Tule Springs Site to Charleston Peak. The Tule Springs counts are from soil surface samples, and the Corn Creek Springs counts are from spring pool bottom samples.

142

TABLE 2. Relative frequency of pollen types from Mono Lake, California.

w 2 ZW W. M ,Z ..,, , .., I am ,m o m o o Depth • 1 a ,..1 a w 3 E al E 4 00 '''' 0 Cyperaceae Sarcobatus Gramineae Sample no. (meters) Artemisia o 14 0 0 0

1. 34.0 2.1 75.5 3.0 1.5 13.5 1.5 0.5 1.5 1.0 -- 2. 27.0 1.5 71.0 3.0 2.0 15.0 4.0 2.0 0.5 0.5 0.5 -- 3. 27.0 1.9 68.6 1.4 3.3 15.0 0.4 1.9 1.4 3.3 0.9 1.0 4. 8.0 0.6 84.5 1.0 3.0 6.5 1.5 0.5 -- 0.5 1.0 0.9 5. 1.0 2.1 70.3 1.5 1.0 16.3 1.5 0.5 1.0 1.5 1.5 2.8 6. 0.7 0.6 61.8 3.4 1.0 11.3 18.3 1.0 1.5 1.1 Paoha Island 7. Soilsurface 0.5 20.2 0.5 -- 2.9 58.1 17.7 8. Pond 16.5 0.5 -- 3.5 64.5 14.0 0.5 -- 0.5 that there is thorough mixing of the pollen pollen is very abundant where the water types, probably in the surface layer by table is high, and ragweed (Ambrosia waves, before the pollen sinks to the bot- psilostachya, A. trifida) is common, as in tom (Table 2). cienegas, but it is also the most common The pine pollen percentages from pollen type from bajada communities of the Mono Lake are similar to those from the Sonoran Desert (Hevly et al., 1965) which modern pollen rain within the coniferous are dominated by bursage (Franseria del- forests or parklands (Martin, 1963b). The toidea , F. dumosa). The pollen of cheno- samples from Paoha Island show the ex- ams is more common along actively cutting pected frequency of pine pollen due to long arroyos or disturbed flood plains (Martin, distance transport. If pollen counts simi- 1963a, pp. 23, 59). Within the desert lar to those from the Paoho Island surface grassland the relative frequencies of cheno- samples were recovered in the fossil re- am or composite pollen are apparently depen- cord, the vegetation type represented could dent on local edaphic conditions and are be accurately determined; but the same useful for correlation of alluvial sediments, cannot be said for the lake samples, where but of limited use as indicators of regional the pollen is representative of coniferous climate. It was found that if these pollen forest and not the vegetation immediately types were excluded from the southeastern surrounding the lake. Sample No. 6 (Table Arizona postglacial fossil pollen records, 2) is near the shore of Paoha Island and in which they are usually dominant, grass shows the expected higher values of Sarco- or grass and Ephedra become the most abun- batus, but it is still much higher in pine dant types. A similar result is seen in the pollen than the island samples. modern soil surface samples of the desert The modern pollen rain samples from grassland and desert scrub (Mehringer, Mono Lake are used here to illustrate the unpubl.). maximum effect of long distance transport A method of counting, in which selec- of pine pollen in a large closed basin lake ted pollen types suspected of representing located outside the forest border yet very local or long distance over-representation near a large mountain range with extensive are excluded in the second count, is called forests. double fixed sums (Mehringer and Haynes, The proper interpretation of fossil 1965). A pollen diagram from the Lehner pollen records from playa lake cores is Mammoth Site (Fig. 4) illustrates the use dependent on a better understanding of the of double fixed sum counts. There is a present deposition of pollen within modern major change in the relative frequencies of lakes, and the samples from Mono Lake can cheno-am and composite pollen in the first serve as a basis for what might be expected count, which probably reflects local edaphic in a fossil pollen record where a large lake and water table level changes. By compar- was surrounded by sagebrush, but near a ison, little change is evident in the second source of coniferous pollen. In a similar count, which indicates that there was prob- manner, the modern pollen content of bottom ably no major regional vegetation change samples from Utah Lake could serve as a (a change in the elevational position of the starting point for interpreting fossil pollen vegetation zones of more than 300 meters - spectra obtained from playa sediments of about 1,000 feet) during the time represented central Utah. by the pollen profile. A similar method While the over-representation of pine might be used in the counting of playa lake pollen is due mainly to blow-in to areas cores where the high frequencies of pine where pollen production is low, there is yet pollen obscures the behavior of other pollen another problem--that of local over-repre- types. Instead of eliminating cheno-am and sentation. Local over-representation is composite pollen from the second count, pine especially apparent in the desert and desert pollen would be excluded. Adam (1965) has grassland of Arizona where cheno-am and used the double fixed sum counting method Compositae pollen may make up over 80 to expose the non-pine pollen record in the percent of the fossil and modern soil sur- Sierra Nevada of California. face sample counts. Low-spine Compositae The point to be made here is that the 144 LEHNER RANCH ARROYO PROFILE II SCALE C./ . 2.0 . 40%. 8/ % CI ..(43 Co mpositae .4 tl ei 4 Q 0 ■k; \A p . o e \o/ i; C o, ( v I5 A) e p .■`./ 0/0 e cIt' t, .is 0 0. ,.0 l e, 0' e 0 \Q ( CI ( ( ( 4 () C Q/ Q/ Q/ Q/ A- O V ST• 2 90 0 8 6 0 20 40 06 3 010 08 4 4 34 0.01

33 0.10

32- 020

31- 030

0.40

29- 050

28 - G3 060 .-

27 0.70

26- 0.80

25- 0.90

1 .00

23- 1.10 —• NO POLLEN I 40 EROSION

19- 5 I

18- 1.6 _ 2 F3 0 , CC z

Al I 17- '.7' • % • 0 0 w 0

16------1.8 e ■L` 1 I , z /

15- F2 • 41 ., I J • A_ 1

, Fi 3UCUS :RUCIFERAE • A / • - 4-378 12 0 1100E110,940 P / 1 13- 2,I PROFILE I= 6 1.60 F3 I1 gin R 5 - - - - - 1.70 3- F2 I 2 1.80 . Fi - - / - MAMMOTH BONE

Figure 4. Lehner Ranch Arroyo. Pollen Profiles II, VII. A pollen diagram illustrating the use of double fixed sums. The solid portion of the diagram represents the first count, including all pollen types. The lined portion indicates the second pollen count, excluding chenoams and composites (see Mehringer and Haynes, 1965). Profile II shows the pollen spectra from an arroyo wall. The time missing because of the erosion interval in Profile II is at least 5,000 years. The shaded area at the top of the profile indicates the mixing of fossil and modern pollen.

145 modern pollen counts should be treated grains per year per unit area, requires rea- so that the different vegetation types of a sonably precise estimates of sedimentation region can be distinguished; since the rates. For this reason, the method is not relative frequencies of all of the pollen suited for alluvial studies, where sedimenta- types may not be the best way to do this, tion rates may vary sharply. The method some other methods such as the double would be more readily applied to Southwest- fixed sums can be devised, using the modern mountain bogs or lakes. ern pollen samples and then transferring the The relation of the pollen sampling method to fossil spectra. This has been site to the vegetation zone borders is another done by subtracting certain pollen types factor which must be considered when inter- from the counts which are considered to be preting the fossil pollen record. If the site indicative of only local conditions and us- is situated in an ecotone area, even slight ing the frequencies of the remaining types shifts in the vegetation should be apparent for interpretation as Schoenwetter has done in the pollen record, but if the site is lo- with his "adjusted sum" (Schoenwetter and cated in the middle of a relatively homogen- Eddy, 1964), or by the use of ratios or eous vegetation type such as ponderosa pine other statistical treatment of specific pol- parkland, which may range with little floristic len types which are most indicative of change from 2,200 to 2,700 meters (about particular vegetation types (Maher, 1963). 7,200 to 8,850 feet) on the Mogollon Rim, The relative frequencies of a pollen Arizona, a shift of 100 meters (300 feet)-- type is a function of the abundance of all maybe even up to 300 meters (1,000 feet)-- other types. For example, if the numbers in zonal position may not greatly alter the of pine pollen grains in a sample increase pollen rain at the site (Hevly, 1964). By and the number of all other types remains analogy, the tree-ring terminology of compla- the same, their relative frequencies will cent and sensitive sites can also be applied decrease relative to pine, due to constraint to the fossil pollen sites (Adam, 1965). Thus, imposed on the count by the increase in the when a pollen record shows no apparent change, one type. In Figure 4, the pine pollen fre- it becomes important to evaluate the location quencies are too low to suggest anything of the site in relation to the areal extent of the but long distance blow-in. Further, it is vegetation zone which it occupies. impossible to be sure if the increases in pine pollen frequencies are the result of THE STUDY AREA expansion of the pine forests in the mountains near the site and thus a greater amount The Tule Springs Site, at an elevation of of pine pollen blow-in, or a decrease in the 703 meters (2,307 feet), is located in the Las local vegetation and pollen production, re- Vegas Valley of the Mohave Desert (Shreve, sulting in a relative increase in pine pollen. 1942), approximately 18 kilometers (12 miles) It is thus possible that in the fossil pollen north of Las Vegas, Clark County, Nevada profiles of the desert grassland of south- (Figs. 5, 6, 10). The Las Vegas Valley trends eastern Arizona (Martin, 1963a) an increase northwestward for about 125 kilometers (about in the relative frequency of pine pollen is, 80 miles). It is bounded on the north and east in some cases, the result of a change to sides by the Las Vegas, Sheep, Desert, and more arid conditions. Pintwater Ranges. On the west side it is The proportions of the different pollen bordered by the Spring Range, the highest cen- types in a pollen diagram may vary only be- tral portion of which is referred to as the cause of constraint imposed upon the counts Charleston Mountains. Charleston Peak when they are considered in terms of their (3,631 meters, 11,912 feet) is the highest relative frequency. The statistical methods point in the Spring Range and rises above of Mosimann (1962, 1963) can be used to timber line. show non-constraint correlations among At Las Vegas, Nevada (660 meters, pollen types (Martin and Mosimann, 1965). 2,165 feet), where the mean annual rainfall The use of absolute pollen counts is 11 centimeters (4.35 inches), the climate (Davis and Deevey, 1964), which yield is characterized by low and infrequent rain- information on the actual number of pollen fall, low humidity, and dry winds. About 146 _ Frenchman •-• '0! ( Flat A'- I- 2 7 •-• \- • • •-% ,,-- *' :•;1, • — /,'`.. cy, i - C.,;., -,--,. ------_-,' R .-.. . _---,,- A\ /in ,„,■ - 1 ''' 4 , ./// \ '‘' 1. 0 ;'. Qf Lr''' '..4.2 I NEVADA j bi A* ' ... -, 0) \ ■....-:. .. - ; v t"' !, I , -.-,---„,-5- ___ -. - \ 4 '- - .. 1 . ... \ (::.::';'.:R C, ,FA)I .'‘ ./N 7".—' 7// 7 2 / • ... - '',/ .•., / Chi \

• • ,•• --•-•' / tt• ./ „ •: • - H ayford y/ • Peak effk- ' o‘ - 0 M ono Lake - - . 3023 m 1.\X\- . /A V — %.(1.■ I/V• /7/1\ v\\ , ' • c \: 44. 74- / • cT,Q — \ 1 7 - ' (101 /4 , - 1 7 L„.•' 7.C / • I \t% , /0 \\ . . %1 ' CO \ \, /,A1 : /•"' • \\ O .• < • /P/7• / 4 \ 77_77 7__ • • / •.• \ - - • A 4' •\ ,\\ Y 40' \\ e„ \\\,\ (-//\r- \\-TV - • 40\ 47 • Z •-'s • 0.• - ••'•• - • / - • • ' Springs...7 , F- ttt Peakm \\\\ P4 --="! /7R/ I/ ; 4 / • • ,//„. - •■- ;,,,:-N //11\111V- S-, -...... •• TULE SPRINGS p \ V4- a. P p k -. -... 703m . z...- -; ///-;__,C.,, c \ canyon.'. :-_-_ IIL. --•. \ \ aW//,4 le ...... 1 .. _---/- ---k25%, ,, 3631 m- -,, Y. ..,,.*•-=',------. ypsum Cave 7'.. _"- --- MI ES • .-_;-, 0 -_- -4' 610 m

0 Lai VegaS KILOMETERS

FIGURE 5. MAP OF STUDY AREA. Figure 6. View of the Tule Springs Area. View of the Tule Springs Site area showing Las Vegas Wash and the Las Vegas Range. The more common shrubs of the site area include Larrea divaricata, Franseria dumosa, Atriplex spp., Ephedra nevadensis, E. funera Lycium andersonii, Dalea fremontii, and Krameria parvifolia. Modern surface samples 21-23 were collected from the site area.

148 one-half of the precipitation (57 percent) mountain snow pack (Deacon, Bradley, and falls during the months of October through Larsen, 1964) and in the daily and seasonal March. The months with the least precipi- temperature ranges. The mean maximum 0 0 tation are May and June with a combined and minimum temperatures are 105 and 76 total of only 6.7 percent of the yearly aver- F. for July and 55° and 33° F. for January age. There is great yearly variation in the (Brown, 1960, p. 8).

VEGETATION AND MODERN POLLEN SAMPLES1

The description of the vegetation of a. Lower Mohave Desert (150-1220 the study area includes the Las Vegas Val- meters, about 500-4,000 feet). Creosote ley and the east slope of the adjoining bush (Larrea divaricata), bursage (Franseria Charleston Mountains. Since the soil sur- dumosa), hop sage (Gravia spinosa), and face samples were collected from the same saltbush (Atriplex spp.) are very common. area, it is convenient to discuss the soil At higher elevations (915-1,220 meters, about surface sample pollen content with each of 3,000-4,000 feet) shrubs such as winter fat the vegetation units. The vegetation zones (Eurotia lanata), Mormon tea (Ephedra neva- used here are arbitrary, as they must be in deRsis), and Lycium andersonii are common, any such limited description, but they are along with the creosote bush, bursage, hop nevertheless a useful generalization and sage, and saltbush. Cottonwood (Populus fulfill the needs of this treatment. The fremontii), willow (Salix spp.), mesquite vegetation zones have been divided into (Prosopis spp.), catclaw (Acacia qreqqii), units which can be distinguished by their arrow weed (Pluchea sericea), Arizona grape modern pollen rain and which are conven- (Vitis arizonica), and salt grass (Distichlis) ient and practical to discuss in terms of are locally abundant around springs or in the vegetational history of the Las Vegas washes where the water table is high(Figs. Valley. 6, 8-12). The relation of the soil surface Modern surface samples from the samples to the vegetation zones in the lower Mohave Desert are dominated by either Kyle Canyon soil surface pollen transect low-spine composites (Ambrosiinae) or by is shown in Figure 7. All surface pollen cheno-ams. The low-spine dominated sta- counts, sample station numbers, locations, tions are those where Franseria dumosa is and elevations are given in Table 3. abundant. These include the areas of thin The following vegetation sketch of rocky soils or desert pavement. The surface the study area is mainly from correspon- samples of the bajadas and erosion remnants dence, discussion, and field work with of high areas or ridges in the valley are W. Glen Bradley (Nevada Southern Univer- characterized by low-spine composite pollen. sity). More detailed treatments of the Atriplex, Gravia, or Kochia grow on sandy, vegetation of southern Nevada can be found fine, or saline soils; the modern pollen con- in a number of sources (Allred, Beck, and tent of these soils is characterized by cheno- Jorgensen, 1963; Beatley, 1965; Bradley, am pollen. Corn Creek Springs (sample 26) 1964; Bradley and Deacon, Part 4 this re- and the extensive area of shadscale vegeta- port; Clokey, 1951; Deacon et al., 1964). tion on the floor of the valley near Corn A brief, simplified scheme of the Creek Springs (sample 25, Fig. 8) are the vegetative zones and discussion of the soil only soil surface samples from the lower surface pollen content of the east slope of Mohave Desert which are dominated by the Charleston Mountains and the Las Vegas cheno-am pollen. Valley is as follows: The samples from the springs them-

' Authorities for the binomials of plants (1951) or Munz and Keck (1963). given in this paper may be found in Clokey 149 selves represent a special case. In pine and juniper values with the increase sample 27 (Fig. 12) Cyperaceae pollen in elevation and proximity to woodland. is the most abundant type. In sample 28 Artemisia values increase sharply at about the high Salix pollen value (59.5 percent) 1,600 meters (about 5,200 feet). There is is the result of a single willow tree (Salix relatively little (less than 5 percent) wind- gooddingii) that grows on the mound (Fig. blown Artemisia pollen in the modern pollen 9) and covers a small pool of water less samples not collected in the immediate than 10 centimeters deep and 0.5 meters vicinity of where the plants are growing. in diameter. This sample is a good illus- For this reason I believe that Artemisia pollen tration of local over-representation of a is one of the more important indicators of single pollen type. Sample 29 is from a vegetational change in the fossil pollen re- pond at the Desert Game Range Headquarters cord at Tule Springs. at Corn Creek Springs. Samples 26 and 27 c. Juniper-Pinyon Woodland (1, 800- show a comparison of a soil surface sample 2,200 meters, about 6,000-7,200 feet). with a spring pool sample (Fig. 12). Sample Stands of pinyon pine (Pinus monophvlla) and 26 was collected from around the base of Utah juniper (Tuniperus osteosperma) are the spring mound and sample 27 was collect- found in the Spring, Sheep, and other ranges ed from the pool of the same mound. of sufficient elevation. Along the washes, The pollen of Larrea divaricata, the canyon bottoms, and lower slopes, desert most widely distributed and characteristic almond (Prunus fasiculata), mountain mahog- plant of the Southwestern deserts, is pres- any (Cercocarpus ledifolius), cliff rose ent in low relative frequencies in soils where (Cowania stansburiana), scrub oak (Ouercus Larrea is abundant. Larrea pollen is often turbinella), manzanita (Arctostaphvlos pun- poorly preserved is soil surface samples gens), banana yucca (Yucca baccata), Fal- and its morphology and size are variable. luqia paradoxa, Ephedra viridis, Eriodictvon The identification of Larrea pollen depended anqustifolium, and Chrysothamnus spp. are partly on the fact that the plants were grow- common (Figs. 15, 16). Small scattered ing nearby and its pollen was anticipated in stands of sagebrush are found on flats and the count. lower slopes. Mountain mahogany, man- b. Higher Mohave Desert (1,220- zanita, and scrub oak form limited areas of 1,800 meters, about 4,000-6,000 feet). On chaparral on the lower slopes in the pinyon- the bajadas blackbush (Coleogvne ramos- juniper zone and on south-facing slopes in issime), Joshua tree (Yucca brevifolia), the ponderosa pine-white fir zone. Mohave yucca (Yucca schidiqera), and salt- With the exception of the Cold Creek bush are very common; creosote bush is locality (sample 20, Table 3), the pollen still present but scattered at lower eleva- counts for the juniper-pinyon zone are shown tions. Shrubs such as rabbit brush (Chrvso- in Figure 7. The pollen rain is characterized thamnus spp.) bladder sage (Salazaria by the combination of high juniper and sage- mexicana), Apache plume (Fallugia paradoxa), brush pollen values. The relative frequency and Tetradvmia axillaris are more common of pine pollen is usually greater than in the than at lower elevations. Ephedra viridis, communities below pinyon-juniper, but it sagebrush (Artemisia tridentate), and scat- may be surprisingly low when the low-spine tered junipers occur at the higher elevations, composite counts are high (see sample 13, associated with joshua trees near their Table 3, Fig. 7). upper limit (Figs. 13-15, 17, 18). There is a rise in the relative fre- All of the pollen counts for the higher quency of low-spine composite pollen which elevation Mohave Desert are plotted in occurs in both the Kyle Canyon samples and Figure 7. With increase in elevation, the the Cold Creek sample. The most conspicuous pollen surface samples are characterized by Compositae are Chrvsothamnus and Gutier- a decrease in the low-spine Compositae rezia, both high-spine composites. In the pollen, as Franseria becomes less abundant, field I encountered no species whose pollen plus a relative increase in high-spine might account for the increased low-spine Compositae and cheno-am pollen. There composite values. It is possible that an is also a relative increase in the wind-blown annual, not growing when the samples were 150 Ephedra COMPOSITAE CHENO -AMS GRAMINEAE Erlogonum LU -J 0 4 w Juniperus Meters Artemisia 1500 2000 Quercus

woo Kyle CanyonFantoCharleston Parkat2,410meters. nev.-type Mountains. Apollendiagramofthemajor typesrecoveredfromsoilsurface sample numbers,andvegetationtypes.Thetransect extendsfrom900metersonthe samples inatransectKyleCanyon,Charleston Mountains,withtheelevations, 610 Pin Abies

us Figure 7.Pollendiagramofmajorpollentypes from KyleCanyon,Charleston Feet 2000

0 0 0 0

60 24= 20 40 24: IA- 0 o

KYLE CANYONSURFACETRANSECT 7 e 5

-■ _111w.

- 1

8 yellow White

1 ' 7

Fir'.. pine 1 l 6 1 1514

. . Pinyon .

• MAJOR POLLENTYPES -

Juniper-Sagebrusti -... _

`-, 151 Higher MohaveDesert Plantago

• Euphorbio- Mohave Lower Desert type 0 20 40% —•-•

Figure 8. Soil surface sample Locality 25, at 885 meters. The shadscale vegetation in fine soil on the valley floor west of Corn Creek Springs is dominated by Atriplex canescens, A. polycarpa, A. confertifolia, and Grayia spinosa. Prosopis juliflora grows on the dunes in the midground. View looking northeast with the Sheep Range in the background.

Figure 9. Modern spring sample, Locality 28 at Corn Creek Springs. The single large tree on the mound is a willow (Salix gooddingii). Some of the common plants on the spring mounds are Haplopappus sp., Kochia americana, Atriplex canescens, salt grass (Distichlis sp.), thistle (Cirsium sp.), arrowweed (Pluchea sericea), and Phragmites communis. Around the base of the mounds Larrea divaricata (in foreground), Franseria dumosa, Grayia spinosa, Hymenoclea sp., Kochia americana, and Atriplex spp. are very common.

Figure 10. View from the Tule Springs Site. A view from the Tule Springs Site area looking toward Kyle Canyon (center background), Charleston Mountains. Creosote bush and bursage are shown in the foreground. Modern surface samples 21-23 were collected from the site area.

Figure 11. Modern surface sample, Locality 1, Kyle Canyon. Modern surface sample Locality 1, on the Kyle Canyon Fan at 900 meters. The plants shown in the photograph are creosote bush, bursage, and Mohave yucca (Yucca schidigera).

152 -

CO 6 U-

153 - *NOV e.te-

• • 1k44,

Figure 12. Modern spring sample, Locality 27 at Corn Creek Springs. The spring pool on the top of a mound is surrounded by a dense mat of salt grass (also see Haynes, this report, Pl. 12, figs. 3, 4).

154 Figure 13. Modern surface sample, Locality 5, on the Kyle Canyon Fan at 1,090 meters. The plants shown in the photograph are Mohave yucca, joshua tree (Yucca brevifolia), and creosote bush; Lycium andersonii is also very common at this locality.

Figure 14. The vegetation between modern surface sample, Localities 9 and 10, in Kyle Canyon at 1,500 meters (4,920 feet). Banana yucca (Yucca bacata), joshua tree, blackbush, and Apache plume (Fallugia paradoxa) are shown in the photograph; Atriplex cansecens, Lycium andersonii, winter fat (Eurotia lanata), Tetradymia axillaris, and Ephedra nevadensis are also common at this locality.

Figure 15. Modern surface sample, Locality 12, in Kyle Canyon at 1,710 meters. Sagebrush, joshua tree, and Utah juniper are shown in the photograph; Atriplex canescens, rabbit brush, banana yucca, and Ephedra viridis are also common at this locality. The juniper is the lowest juniper on the floor of Kyle Canyon.

Figure 16. Modern surface sample, Locality 15, in Kyle Canyon at 2,105 meters. Pinyon pine, sagebrush, cliff rose, Ephedra viridis, and Opuntia sp. are shown in the photograph.

155 FIG.13 FIG .14

FIG.15 FIG.16 Figure 17. A blackbush-joshua tree community at 1,700 meters on Cold Creek Road.

Figure 18. Modern surface sample, Locality 11, in Kyle Canyon at 1,615 meters. Sagebrush, blackbush, joshua tree, banana yucca, and Atriplex canescens are common at this locality.

157 Figure 19. Modern surface sample, Locality 16, in Kyle Canyon at 2,165 meters. Sagebrush, mountain mahogany (Cercocarpus ledifolius), and ponderosa pine are shown in the photograph.

Figure 20. Modern surface sample, Locality 17, in Kyle Canyon at 2,285 meters. Ponderosa pine, sagebrush, mountain mahogany, and Quercus qambelii are common at this locality.

Figure 21. Modern surface sample, Locality 18, Charleston Park, Kyle Canyon at 2,410 meters. Ponderosa pine, white fir, and aspen are shown in the photograph.

Figure 22. Modern surface sample, Locality 19, in Lee Canyon. A snow covered meadow in Lee Canyon at 2,560 meters (about 8,400 feet) surrounded by ponderosa pine and white fir.

158 oc., (.9 EL:

159

Table 3. Modern pollen samples from the east slope of the Spring Range and the Las Vegas Valley, southern Nevada (N=200).

Sample Locality Elevation ,11 Cyperaceae Liguliflorae Soil Surface Samples Charleston Mtns. 1 Kyle Canyon 900 22 7 106 15 5 17 12 2 " " 905 25 10 120 6 7 14 11 3 11 11 945 23 7 110 14 5 24 8 4 " " 1000 25 4 1 110 23 4 13 13 5 I I I I 1090 29 5 117 5 1 22 17 6 " " 1120 26 5 1 130 12 3 6 9 7 " " 1160 20 9 110 18 3 13 19 8 " " 1305 31 15 58 44 6 22 15 9 11 11 1415 41 23 56 29 5 24 12 10 " " 1585 42 16 1 39 31 9 38 18 11 " " 1615 43 17 2 22 26 24 47 9 12 " " 1710 46 33 25 9 1 48 23 9 13 .. 11 1965 21 23 73 9 1 40 10 14 14 " " 2075 1 46 11 3 63 21 38 9 4 15 " " 2105 68 9 7 50 8 26 14 16 " " 2165 1 107 5 14 19 12 23 9 17 " " 2285 2 90 8 5 27 15 13 16 1 18 " " 2410 9 119 7 4 1 26 3 4 10 1 19 Lee Canyon 2560 2 84 9 2 40 15 2 13 2 20 Cold Creek Road 1905 1 44 19 1 52 10 48 7 10 Las Vegas Valley 21 Tule Sprgs. Site Area 705 20 6 1 102 15 1 28 3 22 " " '' " 705 16 6 101 12 2 39 6 23 " " " " 705 16 2 121 6 32 2 24 Eglington Scarp 675 16 12 1 105 8 3 33 3 25 2 k west of Corn Creek Springs 885 23 5 1 54 2 2 105 26 Corn Creek Springs 900 9 8 1 10 36 23 1 95 2 Spring Pool Samples Las Vegas Valley 27 Corn Creek Springs 900 14 76 5 20 1 68 1 28 " " " 900 4 119 27 5 3 26 29 " " " 900 6 2 3 8 5 1 165 1 Table 3. Modern pollen samples from the east slope of the Spring Range and the Las Vegas Valley, southern Nevada (N=200). (Cont.)

-Q ra 0 a) Sample Locality Elevation Ca. Malvaceae Gramineae (meters) Eriogonum Onagraceae Rosaceae Cruciferae Soil Surface Samples

Charleston Mtns. 1 Kyle Canyon 900 1 2 4 1 2 1 1 1 1 2 2 ' " 905 1 1 1 1 2 1 3 ,. .. 945 2 3 4 4 " " 1000 1 2 1 2 1 5 H .. 1090 1 1 1 1 6 " " 1120 1 1 1 1 1 1 2 7 .. .. 1160 3 5 8 " " 1305 2 2 1 2 2 9 t. .. 1415 8 1 1 10 " " 1585 2 2 1 1 11 " " 1615 6 2 1 1 12 " " 1710 2 2 1 1 13 " " 1965 6 2 1 14 " " 2075 2 1 1 15 " " 2105 12 3 3 16 " " 2165 10 17 " " 2285 16 3 1 1 2 18 " " 2410 11 2 3 19 Lee Canyon 2560 30 1 20 Cold Creek Road 1905 7 1 Las Vegas Valley 21 Tule Sprgs. Site Area 705 4 1 8 4 1 6 22 " " " " 705 10 1 2 4 1 23 " '1 II 11 705 3 1 4 3 1 9 24 Eglington Scarp 675 6 3 2 1 1 1 2 1 2 25 2 k west of Corn Creek Springs 885 1 5 1 1 26 Corn Creek Springs 900 3 7 3 2 Spring Pool Samples Las Vegas Valley 27 Corn Creek Springs 900 1 10 1 1 1 1 28 " " " 900 12 1 3 29 " " " 900 8 1 Figure 23. Photomicrographs of fossil pollen from the Tule Springs Site. A Typha tetrad (Typha latifolia); B, Typha monad-Sparganium; C Fraxinus- D, Liguliflorae; E, Salix; F2_, Betula; G, Cyperaceae.

162 Figure 24. Photomicrographs of fossil Abies pollen from Unit B2 of the Tule Springs Site.

163 Figure 25. Photomicrographs of fossil pollen from the Tule Springs Site. A-C, Abies; D-F, Pinus- G, Juniperus; H, Onograceae, Gaura-type.

164 A

6 0)-1

Figure 26. Photomicrographs of fossil pollen from the Tule Springs Site. A, Sium suave; B, Betula; C, Salix. D, Pinus; E, Malvaceae, Sphaeralcea-type; F, high-spine Compositae; G, Cyperaceae.

165 Figure 27. Photomicrographs of fossil pollen from the Tule Springs Site. A-C, Artemisia; ]J-E, Shepherdia argentea; F Juniperus; G Gramineae.

166 Figure 28. Photomicrographs of fossil pollen from the Tule Springs Site. A-B, Quercus; C, Sarcobatus; D Eriogonum; E, high-spine Compositae; F, cheno-am; Betula; low-spine Compositae.

167 40)1

Figure 29. Photomicrographs of fossil pollen from the Tule Springs Site. A-B, Ephedra nevadensis-type; C, Ephedra torreyana-type; D, Pinus• Onagraceae, Gaura-type.

168 collected or during a later trip when the southern Nevada. low-spine contributor was searched for, With the exception of the Lee Canyon is responsible for the low-spine composite locality, all the modern surface pollen pollen. Much of the low-spine composite counts for the ponderosa pine-white fir pollen at higher elevations is probably the vegetation zone are shown in Figure 7. result of Franseria pollen blown in from Photographs of the sample localities are lower elevations, but the frequencies in shown in Figures 19-22. The ponderosa samples 13-15 and 19-20 are too high to pine-white fir community is characterized be accounted for by blow-in./ by pine pollen in excess of 30 percent. d. Ponderosa Pine-White Fir (2,200- None of the surface samples below station 2,750 meters, about 7,200-9,000 feet). 15 (2,105 meters, about 6,900 feet) contains Ponderosa pine (Pinus ponderosa) is more more than 23 percent pine pollen. Fir pol- abundant at the lower elevations and white len is present in low percentages (less than fir (Abies concolor) at the higher elevations. 5 percent) and oak pollen (up to 7 percent) The area in which ponderosa pine and white is present in all samples. No soil surface fir are co-dominants is more extensive than samples were collected above the ponderosa the pure stands of white fir. Other coni- pine-white fir zone. ferous trees occurring at this elevation are e. Bristlecone Pine (2,750-3,500 pinyon pine, limber pine (Pinus flexilis), meters, about 9,020-11,500 feet). In the and pygmy juniper (juniperus communis). Sheep Range and Charleston Mountains some Other trees or shrubs often associated with pure stands of bristlecone pines (Pinus pine and fir include Quercus qambelii, Acer aristata) are found, and they are also asso- qlabrum, Sambucus coerulea, Amelanchier ciated with limber pine, white fir, pygmy utahensis, Ribes cereum, Rosa woodsii, and juniper, or ponderosa pine. Holodiscus microphyllus. Mountain mahog- f. Alpine (above 3,500 meters, about any and aspen (Populus tremuloides) occur 11,500 feet). The herbaceous vegetation in large stands on favorable sites. Spruce above treeline is limited to talus and other (Picea) and Douglas fir (Pseudotsuqa men- rocky sites which are unsuitable for the ziesii) are not present in the mountains of growth of trees at this elevation.

THE FOSSIL POLLEN RECORD

Introduction The pollen record at Tule Springs comes 1952; Erdtman, Berglund, and Praglowski, from a variety of sedimentary environments 1961; Erdtman, Praglowski, and Nilsson, including alluvium, a playa lake, organic 1963; Faegri and Iversen, 1964; Wodehouse, spring mats, and eolian spring mound depo- 1959) and the pollen reference collection of sits. In many cases pollen was poorly pre- the Geochronology Laboratories, University served, scanty, or entirely lacking in sedi- of Arizona. Photomicrographs of some of the ments, but 73 out of 225 (32.4 percent) fossil pollen grains are shown in Figures fossil samples extracted yielded sufficient 23-29. pollen for a minimal 200-grain pollen count. The pollen diagrams are broken where There were a small number of pollen types there is an erosional contact indicated in the (16 types in Profile II to 28 types in Profile stratigraphy or where there are sediments III). The percent of unidentified pollen types that lack pollen. It is assumed that a part was low (2.3 percent maximum in a single of the pollen record is missing at such breaks. sample). Unidentified types are not included It is possible that the missing sediments may in the pollen sum plotted on the diagrams. represent no more time than that normally Unless mentioned specifically in the skipped in the sampling intervals, or that an text, the pollen types recovered are dis- important vegetational change may be entirely cussed by Hevly (1964), Hevly, Mehringer, missing. and Yocum (1965), Martin (1963a), or are The sampling sites were chosen to self-explanatory. Pollen was identified by cover all the stratigraphic units as thoroughly the use of standard texts (Erdtman, 1943, as possible and to assure the proper associa- 169 tion with carbon-14 dates, stratigraphy, percentage gives an estimate of variation in and other features. The pollen samples were pollen preservation. The stratigraphy of the collected by P. J. Mehringer, P. S. Martin, Tule Springs Site and the C14 ages of the strati- and C. V. Haynes. Each sample was plot- graphic units shown on the pollen diagrams are ted on detailed stratigraphic diagrams by discussed by Haynes (Part 1, Appendix 1, this C. V. Haynes at the time of collection. report). Since the pollen studies are dependent Percentages for each pollen type, based on on the stratigraphic framework, I have refer- counts of at least 200 grains, are plotted to enced sections of Haynes' report where they the same scale on the pollen diagrams, and apply to specific pollen samples or profiles. the dots ( • ) indicate less than one percent In order to understand the relation between of a discontinuous type. The stratigraphic the pollen profiles and the stratigraphic se- units in the spring mounds are indicated by quences, it is imperative that the reader com- adding an "s" to the letter designations on pare these results with the stratigraphic stu- the pollen diagrams. dies. Radiocarbon dates associated with the Pollen Profile V (Fig. 36) is an accumu- pollen profiles are shown on the pollen dia- lation of miscellaneous samples from various grams (Damon, Haynes, and Long, 1964; areas of the site. The miscellaneous sam- Fergusson and Libby, 1964). The pollen re- ples are divided into spring and alluvial cord from Tule Springs will be discussed by samples and placed on the diagram in their the stratigraphic units in which pollen is proper stratigraphic sequence. In Profiles present (Table 4), starting with the oldest I and II (Figs. 31, 32), where pine is abun- unit (B2) and proceeding to the youngest dant, the percent of pine breakage is indica- (G). ted by the dashed line. The pine breakage

Table 4. Approximate age of the stratigraphic units shown on the pollen diagrams (C. V. Haynes, pers. comm.).

14 Unit C age B.P.

G1 ...... 300-1,000

F2 ...... 1,500-4,000

F 1 4,000-5,000 E 6,000-11,000 2 E 11,500-14,000 1 16,000-30,000

B > 40,000 2 ......

Cattail Pollen pollen diagram because it was not possible to distinguish broken tetrads from dyads or Tvgha (cattail) pollen was counted as monads. tetrads, dyads and monads (Fig. 23); the The tetrads of Typha pollen are refer- counts are given in Table 5. The tetrads, able to Tvgha latffolia and the monads to dyads and monads are not separated on the either T. angustifolia or T. domingensis. 170 Table 5. TvPha pollen counts in tetrads, dyads, and monads. (The figure numbers refer to the pollen diagrams.)

Stratigraphic Sample Figure Tetrads Dyads Monads- Unit Number Number Sparganium

G 182 36 4

E2 234 36 1

E2 161 35 1

E2 68 33 1

E2 234 36 1

E 1 ? 67 33 4

Ei 149 35 1

Ei 66 33 1

D 137 33 10 2

JD 41 32 1

JD 44 32 90 3

D 45 32 37

D 46 32 7

D 47 32 3

D 48 32 5 21

D 50 32 19 43

D 51 32 2

B 30 31 6 42 2 B 123 36 24 8 77 2 B 120 36 2 2 B 63 36 3 2 Panamint Valley (UCLA 990, 10,435+100 B.P.) 15 11 8 171 It is possible that the monads identified as (McNaughton, 1966). Typha are referable to Sparqanium, but this It is possible that the Typha pollen seems less likely except possibly during the record from a trench near the shore of Lake cooler and more moist periods of Units B2 Hill Island (Davis, in Rogers, 1966, p. 132), and D. The hybrids of T. latifolia and T. PanamintValley, California (Table 5) repre- anqustifolia have been reported to produce sents a hybrid population or that the dyads pollen grains in monads (Hotchkiss and and monads are just broken tetrads of T. lati- Dozier, 1949, p. 254) or in monads and folia. I have not found any Typha localities tetrads (Fassett and Calhoun, 1952, p. 376) at springs near the present playa in Panamint or in monads, dyads and tetrads (Smith, Valley, but T. anqustifolia does occur in the 1961, pers. comm.). canyons of the Panamint Mountains (Coville, In Arizona Typha dominqensis occurs 1893, p. 209). The presence of T. latifolia below 1,525 meters (5,000 feet) and T. lati- or T. latifolia hybrids on the present playa folia from 1,065 to 2,285 meters (about 3,500- edge about 10,500 years ago is especially 7,500 feet) (Kearney and Peebles, 1960, p. indicative of more moist and/or cooler condi- 64). The only reported locality for T. angus- tions. I have also recovered abundant Typha tifolia in Arizona is from the Colorado River monads from an undated core from Warm Sul- near Yuma (Mason and Hevly, 1961; C. T. phur Springs, Panamint Valley. Although there Mason, Jr., pers. comm.). In California is apparently ample Typha habitat at Warm all three species may occur together below Sulphur Springs, Typha does not grow there 915 meters (3,000 feet) and T. latifolia occurs today and its presence there in the past might above 915 meters (Mason, 1957). On the reflect a change in soil or water chemistry Nevada Test Site, T. latifolia grows at related to increased spring discharge, de- Whiterock Spring (1,525 meters, about 5,000 creased evaporation or some other climatically feet) and T. domingensis at Cane Springs controlled phenomena. (1,220 meters, about 4,000 feet) (Beatley, 1965, Fig. 1). T. latifolia has been reported The Pine Pollen Problem from the Charleston Mountains, but Clokey (1951) suggests that it was introduced. A The over-representation of pine pollen check of the Typha collections in the herbaria in the present desert areas and in large lakes of the University of California at Los Angeles has already been mentioned, and therefore and at Berkeley and the Rancho Santa Ana there is good reason to question any interpre- Botanic Garden showed that Typha latifolia tations of pollen diagrams from Tule Springs is now extremely rare at lower elevations based solely on an increase in pine pollen. in the Mohave Desert. All of the North Amer- There are yet other theoretical probelms which ican cattails tolerate some salinity, but T. should be considered before attempting to latifolia is the least salt tolerant (Hotchkiss interpret the pine pollen record from Units B2 and Dozier, 1949; McMillan, 1959). and D. The presence of Typha latifolia in the Using the area-elevation graph for the late Quaternary deposits of the Mohave Spring Range (Fig. 30), it is apparent that even Desert may indicate cooler temperatures, less a slight drop in the lower elevational limit of saline and alkaline waters and more acid soil a montane plant community will greatly in- conditions. With increasing aridity the T. crease the areal extent of the community. For latifolia populations might be replaced rap- example, if it is assumed that the present idly by T. anqustifolia or T. domingensis or distribution of conifers extends from about more slowly by hybridizations with these two 1,830 meters to 3,050 meters (about 6,000- species and the gradual elimination of T. lati- 10,000 feet) and that this zone is lowered 610 folia genes. The latter case might be reflec- meters (2,000 feet) so that the lower and upper ted by a gradual relative decrease in Typha limits would be about 1,220 and 2,440 meters, tetrads with an increase in the dyads and (about 4,000 and 8,000 feet), then the area monads, and finally by the complete elimina- occupied by coniferous forest would be about tion of tetrads and dyads. Any pattern of two and one-half times larger (Fig. 30). If it replacement might be further complicated by is further assumed that a given area of conifers ecotypic variation within a single population still produces the same amount of pollen, then 172 ELEVATION -9000 -12,000 -10,000 -11,000 5000 1500 7000 6000 8000 4000 1220 Feet meters). plot oftheplanareaaboveeach1,000-foot(305 meter)contourfrom4,000feet(1,220 Meters 2000 3500 3000 2500 Figure 30.Anarea-elevationgraphfortheSpring Range.Thecurveisdrawnfroma 0 Sq Kilometers Sq , Miles

500 250 AREA ABOVEINDICATEDELEVATION 1 000 173 500 1500 750 2000 2500 1 0 , 00 there would be a proportionate increase in rain of the northern Great Basin are not avail- the amount of coniferous pollen falling on able. Thus, it is not known if surface samples the valley floor at the Tule Springs Site as from the desert ranges are sufficient to inter- the area of conifers expands. Also, if the pret the pluvial Pleistocene pollen records actual number of wind-blown coniferous from the Mohave Desert. Since the changes pollen grains decreases away from the envisioned amount to a lowering of the present forest border (Tauber, 1965), there would vegetation zones by at least 1,000 meters be an even greater amount of pine pollen (3,300 feet) as well as a southward extension falling at the Tule Springs Site, since a of vegetation, the Pleistocene equivalents downward displacement of 610 meters (about may not be found in the modern surface sam- 2,000 feet) would bring the trees approxi- ples of the higher elevation communities of mately 12 kilometers (7.5 miles) closer to Southwestern mountains. In order to inter- the site (Fig. 3). Such theoretical consi- pret the pollen record of the Las Vegas Val- derations cannot be limited to the Spring ley during the deposition of Units B2, D, and Range. The increase in coniferous pollen E1, it might be preferable to have a transect production and the migration of conifers of surface samples from northern Idaho through closer to the site area would apply to all the Great Basin and Mohave Deserts to south- of the ranges surrounding the Las Vegas ern California to compare with the transect Valley, and, under this set of circumstances, through the various vegetation zones in both pinyon and ponderosa pines might ex- Kyle Canyon (Fig. 7). pand to smaller ranges that presently they do not occupy. A change to more effective Unit B2 moisture might also lead to a greater density of coniferous trees, thereby increasing The pollen samples from Unit B2 were the pollen production for a given area of obtained from spring-laid clay and silt and forest. from a spring deposit (Figs. 31, 36T. Although It is apparent that a slight reduction both deposits are related to springs, the in the lower limit of the present vegetation spring sediments are from the organic mat of zones in the mountains surrounding the a spring or were deposited within the spring Las Vegas Valley would greatly increase the pool (Haynes, Part 1,this report, Fig. 12). absolute numbers of pine pollen grains fall- The alluvial sediments of this unit were de- ing at the Tule Springs Site. However, the posited in spring-fed channels and are thus pollen counts are given in relative frequen- not directly associated with the springs. No cies. As previously stated, the main reason finite radiocarbon dates were obtained from for the over-representation of pine pollen in Unit B but it is thought to be of Early 2' the desert is that local plant cover is sparse, Wisconsin (Altonian) age (Haynes, Part 1, local pollen production is low, and therefore this report, Fig. 7). the relative frequence of wind-blown pine Sample 25 at the bottom of Profile I pollen is high. Under cooler and more moist (Fig. 31) has a sufficiently high value of conditions, which would allow a lowering of Artemisia to show that sagebrush occupied the elevation of the plant communities on the the valley floor, and indicates a Great Basin mountain slopes, I would also expect an Desert environment with slightly more mois- increase in vegetation density and increased ture and cooler temperatures than at present. pollen production on the valley floor thereby Through sample 28 there is an increase in reducing the over-representation of pine the pine and fir values, with little change in pollen, even though there is an increase in the relative frequency of Artemisia. The in- the absolute number of pine pollen grains crease in arboreal pollen is sufficient to reaching the valley floor. It is even possible indicate a major change in the vegetation. that a two-fold increase in pine pollen pro- The question is, how great a change? The duction, with the same increase in herb and pollen spectra of samples 27 and 28 are simi- shrub pollen production, in the valley would lar to the modern soil surface samples from not be detectable in the relative frequencies the ponderosa pine-white fir communities of of the pollen grains. the Charleston Mountains (Figs. 7, 21, 22; Adequate data on the modern pollen Table 3). The presence of Abies pollen 174 TULE SPRINGS. NEVADA: POLLEN PROFILE I

0 10 20 COMPOSITAE <," 4..%`• ■co \ 0 <0/2TC/ 16 LI194/ C2QN 8:1 ‘. ET§Y ET • 4C 1' 45 kN CO 41 .0 \O I`B TI/ .EFR \ 0 C, \ , 40 60 0 10 02 0 20 02 0 20 0 10 0 10 20 0 10 02 02 0 u 0 6 0 3 i t ? I t " 1 1 1 1 ■ 1 ■ 1 1 1 1 1 II i l i a I 1 II I 1 I I

600 2.00 - 7 —

2.25 -

2.50 -

2. 75- 211

3.00 - 10 —

3.25 - 200

3.50 -

12—

14.75 - 200

13— 4.00 -

4.25 - 14—, 583 Broken Pinus 0 20 40 60 z Total Pious

FIGURE 31. POLLEN PROFILE I: A POLLEN DIAGRAM FROM SPRING-LAID CLAY AND SILT OF UNIT B2. (NOTE, THE UMBELLIFERAE IN THIS DIAGRAM WAS INCORRECTLY LABELED AS EUPHORBIA-TYPE IN THE PRELIMINARY REPORT). (Figs. 24, 25), up to 5.2 percent, and water transport if the streams from the high pine pollen values suggest that white Spring Range flowed at least seasonally to fir and pines were growing at much lower the Las Vegas Valley. Streams probably did elevations. flow from the Charleston Mountains during If samples 27 and 28 are interpreted Pleistocene pluvial episodes, carrying arbo- only on the basis of the modern pollen counts real pollen to the valley floor, but this cannot from the Spring Range, it is possible that explain the 2 percent fir and 64.5 percent pine pine parkland occupied the valley floor. pollen in sample 121 (Fig. 36). Sample 121 is Also, if the vegetation zones were lowered from a fossil spring, so the pollen record would so as to bring ponderosa pine to the level not be affected by stream import, and the local of the Las Vegas Valley, it is conceivable pollen production around the spring would de- that parkland or forest could spread across crease the relative effect of wind transport. the valley. However, at higher elevations For these reasons I do not believe that the within yellow pine parkland or forests today, arboreal pollen record of any of the Unit B2 areas with deep soils are dominated by her- samples is seriously affected by water trans- baceous vegetation, as in meadows (Fig. port of pollen from higher elevations. 22), or by sagebrush (Fig. 19). It is there- The fluctuations in Pollen Profile I fore not reasonable to assume that in the could be the result of a climatic and vegetation past pines would have invaded all soil shift associated with an Early Wisconsin (Alto- types. nian) pluvial event. The pollen record suggests A similar interpretation was suggested a change from Artemisia-dominated Great Basin for the Pleistocene vegetation changes in Desert to more moist and cooler conditions, New Mexico. The pollen content of soils with a greater downward displacement of the in the Sandia Mountains, New Mexico, vegetation zones in the surrounding mountains, indicates a downward displacement of the followed by a return to Great Basin Desert. vegetation zones by 1,220 meters (4,000 All of the samples from Unit B2 indicate a feet) King, 1964, p. 33). King believes vegetation which presently exists under cooler that this displacement was limited to the and more moist conditions. A reduction in the mountain slopes and that the plains below present vegetation zones by 1,000 to 1,200 remained treeless, with changes on the meters (3,300 to 4,000 feet) seems required plains being of a "herbaceous nature." to explain the pollen record from Unit B2.. Considering the expansion of forest Harrington and Simpson (1961) reported area and increased coniferous pollen produc- charcoal identified as Quercus and either tion with a slight lowering of vegetation CuPressus or Tuniperus. It is now known zones (Fig. 30) and the fact that I would that this wood was a mixture from Units B2 not expect ponderosa pine or fir to invade and E1, so there is no certain identification areas of the Las Vegas Valley with deep of Cupressaceae wood from Unit B2 Fossil soils, even if these trees descended to Frayinus and Vitis wood was also recovered 700 meters (about 2,300 feet), I think that from Unit B2 (Mehringer, 1965, p. 183). the high pine and fir record from Unit B2 Fraxinus velutina grows around the desert can best be explained by an expansion of springs at Ash Meadows in Nye County, Nev- the forest, woodland, or parkland on the ada, 100 kilometers (62 miles) northwest of bajadas, with Artemisia, grass, and pos- Tule Springs at an elevation of 730 meters sibly scattered junipers on the valley floor. (about 2,400 feet), and in Clark County above Whatever interpretation is offered, a con- 1,130 meters (3,670 feet). Fraxinus anomala siderable expansion of the present distribu- occurs above 1,120 meters in Clark County. tion of white fir in the Spring Range seems The presence of ash trees at Tule Springs necessary to account for the frequencies of would indicate locally more moist conditions, Abies pollen in the fossil record. but not necessarily a major change in the pre- The modern pollen rain studies from sent vegetation of the area. Vitis arizonica the Southwest indicate that fir pollen is not occurs in the Las Vegas Valley today. greatly over-represented due to long distance The pollen of Sarcobatus (Fig. 28), a wind-borne transport. The presence of Abies genus which does not grow in Clark County pollen in Unit B2 could be accounted for by today, occurs in Unit B2 and sporadically 176 TULE SPRINGS, NEVADA: POLLEN PROFILE LE

COMPOSITAE / C.) \00 ‹,e, • 0 ,P P %\C° (4\ ,c°4 C) , 04 • •CA 10 + Q eD \<4 1/4 C?) CX.P .4X• •O‘Y\C%s Y C) E C5 4/ 0,0 VD q/ 2 20 40 60 80 1 20 40 0 8 N 1111111 I 1111,11 7., Th7 , ? , 7,T 7„,1?F ' I l l I I I I i i NO POLLEN UCLA 536 200

2.0

200 NO POLLEN 200 2.75

200

200 NO POLLEN — 200 0 A 462 31,300 cot ±2,500 I 200 NO POLLEN 4 / VI -„„-„,„, I Broken Pinus 0 20 40 60 80 / Total Pinue

FIGURE 32. POLLEN PROFILE II: A POLLEN DIAGRAM FROM BURIED LAKE SEDIMENTS OF UNIT D (SEE MEHRINGER, 1965, FIG. 6). throughout the record. The relative fre- near the shore of the lake. quencies of Sarcobatus are within the limits The only modern pollen samples which of what would be expected as the result of are suitable for comparison with Unit D are long distance transport (Maher, 1964). those from Mono Lake (Fig. 5, Table 2). With the exception of more Artemisia pollen Unit D in the Mono Lake samples, and considering that the pine pollen curve from Unit D is The fossil pollen content of mudstones, depressed by the abundance of Typha pollen, which represent buried lake sediments (Hay- the counts from the mudstones of Unit D and nes, Part 1 this report, Fig./ 14; Mehringer, those from Mono Lake compare rather well. 1965, Fig. 6), are shown in Pollen Profile A considerable expansion of the pine forests II (Fig. 32). The radiocarbon dates at the in the mountains surrounding the Las Vegas top and bottom of the profile were obtained Valley would be required in order to obtain from mollusc shells. With the exception of a fossil pollen record from Tule Springs com- an Artemisia peak (sample 45), the only parable to the modern bottom samples from major change shown in the diagram is the Mono Lake. replacement of pine pollen by Typha (cattail) Sample 137 (Fig. 33; Haynes, Part 1, pollen. Typha pollen does not disperse far this report, Fig. 17) is from the fossil organic (Durham, 1951), so it was probably growing mat of Spring Mound 4A, located approximately at or near the site where the pollen profile 40 meters north of Spring Mound 4 (Fig. 34). was taken and the high frequencies indicate The presence of 59 percent pine pollen with shallow water near the shoreline of the lake. juniper, oak, and birch pollen in a spring The replacement of pine by Typha through sample where the local production of non- the profile is an indication of the fluctuating arboreal pollen minimizes the effect of long lake level and does not necessarily reflect distance transport and may even mask the a regional vegetational or climatic change. local arboreal pollen, is especially indicative The levels without pollen are possibly the of the vegetation changes which occurred at result of the shoreline shrinking beyond the Tule Springs. The sample is placed below site of the profile toward the center of the Unit E because the pine pollen values are valley, thus producing a non-aquatic environ- higher lthan any sample from that unit, and ment of deposition less conducive to the Artemisia, characterizing Units Ei and E2, is preservation of pollen. It is possible that not abundant. The sample appears to be close there were climatic shifts during the time to the low Artemisia spectra of Profile II, Unit represented in Profile II which are not apar- D (Fig. 32); however, Abies pollen is low but rent because of the missing record or the consistent in Unit D and lacking in sample wide sampling interval. 137. On the basis of the radiocarbon dates Since all the pollen of Unit D is that bracket the sample and the pollen content derived from lake sediments, where the in relation to other pollen spectra, I would problems of over-representation of coniferous place sample 137 in Unit D. It may fall any- pollen from wind and water transport are where between 22,000 and 16,000 B.P. likely to be greatest, interpretation is difficult. When all factors are considered, I Unit E1 believe that the pollen record from Unit D is indicative of a major change in vegeta- Pollen spectra from Unit El are shown tion. A lowering of the vegetation zones in Figures 33, 35 and 36 (Haynes, Part 1,this by at least 1,000 meters (3,300 feet) is report, Figs. 8, 10, 15, 16). There were two indicated by the relative frequency of pine pollen types recovered from Unit E1 which I pollen and the low but consistent presence believe can be identified to the species level. of Abies, even when locally produced Typha These are Sium suave (Fig. 26), labeled as pollen is abundant. For example, sample 44 Umbelliferae in Profile III (Fig. 33), and contains 46.5 percent Typha pollen, and Shepherdia argentea (Fig. 27), labeled as still the pine value is 35 percent. In samples Sheoherdia in Profile III and V (Figs. 33, 36). 41 through 45 there is sufficient Artemisia Both of these species grow in damp ground pollen to indicate that sagebrush was common and the pollen types were only recovered 178 Pleistocene pollen from the Las Vegas Valley, Nevada.

TULE SPRINGS, NEVADA: POLLEN PROFILE Spring Mound 4 96 0 10 20 COMPOSITAE unit LI_L_LJ Q2' / ,,, c

B.P V 2 „se 0 .

C \ 00)/CP,„AVY \O . i. ,CPN P \ 0

years ,0 \./ 4s n.ii, ,:,c. .,..-. C° 1; s s c - \ ,2, s- X e 0 dY Tik V' •c •/ G CD 10 20 30 40 50 60 0 10 20 0 4 0 0 6 12 0 10 0 10 0 20 0 0 20 40 0 20 40 0 0, 04 Meters

Sample Number 0 0 20 Stratigraphic C" 1 I I I I 2 I I 3 2 N 1 1 1 . 1 , 1 1 1 ■ 1 ■111111,11 11111 1 1 1 1 1 1 1 1 1 1 1 i 1 1 1 1 I 1 1 - 5 89 S ° • IMi I . • I I I • • • 200 NO POLLEN EROSION ...... 80 200

2.00 - 11 .'11 79 tii .- 200

- l l 78 2.25 i ?il cl 2 00 S' FS 'c5 1 1 I I 77- • 1 I 2.50 EROSION 200 NO POLLEN 74- r 200 3.50 1 73- K I EROSION a WEATHERING 200 _ E2S NO POLLEN • 200 70_ UCLA 537 w ' a.. < i 200 Cc 69: t 150 W ,,, .2 250 :, 7 s e - _, 4 Lu ..2., 4 25 1 -1 - i e>i I 67 " -- ..., • • 22" E s' 4.50 66 l l 607 Spring Mound 4A 1 25 - 137- 4 II. I I • 0 NI I 0. 0 • MI • 400 * Sample 137 is stratigraphically below UCLA 529 - 9,200±250 B.P, and above UCLA 539 - 25,300±2,500 8.8

Figure 33. Pollen Profile III: Pollen diagram of spring mound sediments, Tule Springs. Sample 66 is from clay beneath the organic spring-mat. Samples 67 through 71 are from the organic spring-mat of sand, silt, clay, and plant remains. The remaining samples are from eolian sand and silt that accumulate on damp ground around the springs and form the mound (see Fig. 34). • .

1 . • • / . - • t • '7; ?

. • 4 4 , • ••••••.t. ' 0.-• ./- c, • -

••• "am ••• • e.• . •

4,00. , • '4•14A .. Figure 34. Spring mounds 4, 4A, and the Charleston Mountains. Spring Mound 4 in foreground, 4A in the midground, and the Charleston Mountains in the background (view looking northwest). Pollen Profile III (Fig. 33) was obtained from the west walls of the bulldozer trenches (see Haynes, this report, Pl. 9, Fig. 4). TULE SPRINGS, NEVADA: POLLEN PROFILE BE

% 0 10 20 0 0 1-1-1—a—I 0 CL ..

unit COMPOSITAE ta ■ . 4w g E w .4 :51 .1.1 .P ,NIi =I (.3 ILI .— ._„,0 %, C 4 Q ,,.... 0 4 ....0 cr 1 ■ ,6) 544 '1 °C yeiirs B.P .0 ° * I < a .2 .c 0 ...Nei° ...,-- l 0 v. s.,, s.," 0 a) qv-- K• W 02 WW 20 40 0 7 40 600 10 0 20 40 60 0 5 10 6 Cie 0 10 20 0 20 Sample Number Stratigraphic Meters 1 1 1 1 11 1 a 1 1 1 1, 1 . 1 1 1 1 , 1 . i 1 . 1 i 1 1 1 1 1 1 1 . 1 1 1 1 1 1 If N 0 7 NO POLLEN EROSION ----- 171 1• • • 200

0 170 • —200 2

.75 169 —200 1 168 1.00 —200

1.25 167 200

I 166- 1.50 —200 I 165 —200 1.75

164 —250 UCLA 519 2.00 7,480 7 ±120 163 2.25 20C

162 2.50 200 "F)I in 2.7 9 161 200 2

160- 3.00 10 —200

159 —400 3.25 II 158- 200

3.50 157. I 223 UCLA 510 EROSION 156- 8,000 400 t 1,000 1.25 1

1 50 155 200

», ::7, 1,75 a 154 it 200 a__.1 e: 2.00 7) a 7 Lal 7 o t> 153 , a ct 2.25 tu a. >- 8 . 2.50 152 200 9 EROSION and WEATHERING 151 2.75 —Q00

150 3.00 10 200 I

3.25 — 149 227 II UCLA 512 –WEAK CONTACT...... b...... 148 1 2.,4305 0 1 NO POLLE N .200 13

Figure 35. Pollen Profile IV: A pollen diagram from the alluvial sequence at the Fenley Hunter locality (see Haynes, this report, P 1 . 3, Fig. 4; Fig. 10). 181 Units F and G The Unit F pollen record from Profile The fossil pollen spectra from Units III (Fig. 33) has a very high grass count and F and G (Figs. 33, 36) are all indicative the lowest combined values of arboreal and of the present lower Mohave Desert plant Artemisia pollen in the Tule Springs pollen communities. Since the Tule Springs site record. I believe that the dominance of cannot be considered a sensitive pollen grass and high-spine composite pollen was locality, changes amounting to a 300 meter probably the result of moist conditions in (1,000 feet) lowering of the present vege- the immediate vicinity of the spring, with tation zones would probably not be appar- grasses such as Distichlis and Phragmites ent in the fossil pollen record. This fact very abundant on the spring mound. The is obvious from the modern soil surface local over-representation of their pollen sample counts (Table 1, Fig. 7). The would not be indicative of the regional vege- shifts in the dominance of cheno-am or tation of the Las Vegas Valley during the composite pollen in Units F and G are all deposition of Unit F. It is possible that probably the result of local edaphic changes. during the time represented by Units F and G Fossil wood of a Leguminosae, prob- there was an increase in the rainfall and a ably Acacia, was recovered from Unit F. subsequent increase in grass on the lower Acacia greggii is the only member of the desert, but there is no indication of such genus occurring in Clark County today and an event in any of the alluvial samples from its presence in Unit F would be in accord Units F or G. with the pollen evidence for desert conditions. DISCUSSION

The fossil pollen record from Tule ing an Early Wisconsin (Altonian) ice advance Springs covers Wisconsin and Recent time and retreat could very possibly be represented (Haynes, this report, Fig. 7), a period by the pollen fluctuations in Unit B2 (Fig. 31). dating from some 70,000-50,000 years ago Part of the Late Wisconsin pluvial could be to the present (Flint, 1963; Flint and Brandt- reflected in the pollen spectra of Unit D, in ner, 1961; Frye and Willman, 1960). There the upper portion of Profile II (Fig. 32). On are several major and many smaller breaks the basis of radiocarbon dates, part of pollen in the pollen record, so that much of this Profile II also falls within Altonian and Farm- time is not represented in the pollen pro- dalian time. files. Some tentative correlations with There is a major change in the pollen other Pleistocene and Recent events are record at Tule Springs about 12,000 B.P. possible and the task is made somewhat (Martin and Mehringer, 1965, p. 439). This easier by the great number of radiocarbon event corresponds in time to the Two Creeks dates directly associated with pollen pro- interstadial of continental glaciation (Broecker files or single pollen spectra, and with the and Farrand, 1963). There is evidence from stratigraphic features of the site. Haynes the fossil wood and pollen that junipers were (Part 1, this report, Fig. 7) has attempted growing on the floor of Las Vegas Valley correlations of the radiocarbon-dated strati- 13,000 years ago. In Profile III (Fig. 33) graphic units of Tule Springs with pluvial there is a sharp decline in the juniper pollen lake, glacial, and alluvial chronologies. and a change to cheno-am dominance in With the possible exception of sample Unit E2. In Pollen Profile IV (Fig. 35) there 137 (Fig. 33), there are no pollen samples is also a shift to cheno-am dominance which fall within the time of the maximum following an abrupt decrease in the pine pol- advance of Late Wisconsin (Woodfordian) len in the uppermost part of Unit E1, above ice, about 18,000 to 20,000 years ago. a date of 12,400 B.P. At the Lehner Mammoth Therefore, the maximum of pluvial vegetation site a rapid climatic change, marking the end changes may not be represented in any of the of pluvial conditions in southeastern Arizona, pollen samples from Tule Springs. Pluvial occurred before 11,200 B.P. (Mehringer and changes south of the ice border accompany- Haynes, 1965, pp. 21-22). 186 Between 12,000 and 7,000 to 7,500 represented by pollen-yielding sediments. B.P., shadscale-sagebrush was probably There is nothing to be gained by attempting the principal vegetation in the Las Vegas to make the biotic evidence fit preconceived Valley. Within this time period there was notions about the Altithermal climate (Asch- a fluctuation to cooler and more moist con- mann, 1958, p. 23), so further speculation ditions about 8,500 to 8,000 B.P., as must await the evidence. However, I think indicated by an increase in the relative it is unlikely that the pattern of postglacial frequency of pine pollen (Fig. 35). The climatic change of the northern Mohave pollen record from southeastern Arizona Desert in Nevada and California would be also shows increased arboreal, grass, and much different from further north in the Great aquatic pollen types about 8,000 B.P. Basin where there is some evidence for a (Martin, 1963a, Figs. 20, 37). This is warm-dry Altithermal (Martin and Mehringer, about the same time as the Cochrane re- 1965, p. 442). advance in eastern Canada (Falconer, Using the modern surface samples for Andrews, and Ives, 1965). comparison, there was a lowering of the There is an indication of another vegetation zones by at least 1,220 meters possible fluctuation to more mesic condi- (about 4,000 feet) during the maximum of tions at about 10,000 B.P. In Panamint Wisconsin pluvial conditions represented Valley, California, bulldozer excavations in the pollen record. This estimate for the (under the direction of D. L. True and E. Tule Springs Site might be somewhat exag- L. Davis) into the edge of the playa sedi- gerated as an average for the northern Mohave ments exposed a dark organic mat (Davis, Desert because the site is located so near to in Rogers, 1965, p. 132). This mat was a large mountain mass where the downward dated at 10,435 ± 100 B.P. (UCLA-990). displacement of vegetation zones might be Pollen analysis of this material (Mehringer, expected to be greater (Martin, 1963a, Figs. unpubl.) showed that sedges and cattails 3, 4). Also, with the data now available, (TvPha latifolia or hybrids of I. latifolia) it is not possible to tell exactly what effect were locally abundant (Table 5). It is the expansion of the forests in the nearby probable that the me sic conditions indicated mountains would have on the fossil pollen by the organic mat and its pollen contents rain in the Las Vegas Valley. Added to this may be correlated with evidence for more is the fact that the discussion of vegetation effective moisture in the Las Vegas Valley, change by vertical displacement is only a and that there was a regional moist and descriptive tool. There are many exceptions cool period about 10,500 to 10,000 B.P. to present vegetation zonation in the South- The evidence is not yet sufficient to war- western mountains (Lowe, 1964, p. 87; rant correlations with other events of this Mehringer, 1965, pp. 185-186). An example age. of such an exception from the Spring Range The change to lower elevation Mohave is shown in Figure 37. Desert vegetation occurs in the upper part It is possible that the fossil pollen of Unit E2, This would correspond to the record is deceiving and that there was less beginning of Antevs' (1955) Altithermal. The change in vegetation than I infer. But the present distribution of hybrid oaks (Cottam, high juniper pollen values for samples 66 Tucker, and Drobnick, 1959; Tucker, Cottam, (Fig. 33) and 233 (Fig. 36) indicate strongly and Drobnick, 1961) may be the best botan- that junipers grew at the Tule Springs Site, ical evidence for a post-pluvial period of and the fossil Cupressaceae wood from two hotter-than-present climate in the Great different spring deposits dated about 13,000 Basin. There is evidence that the Altither- B.P. is even more conclusive. Using the mal vegetation at Tule Springs was like the present average lower limit of Juniperus in present vegetation, but no evidence that the Spring Range (1,700 meters, about conditions were either hotter or dryer. 5,575 feet), the presence of junipers at the Admittedly, even if such an event did occur Tule Springs Site (703 meters, 2,307 feet) it might not be detectable in the alluvial implies a lowering of the present vegetation pollen record at Tule Springs and most of the zones by at least 1,000 meters, (3,300 feet). time of Antevs' Altithermal may not be I think that a 1,000-meter displacement is an 187 „r774:

Figure 37. A pocket of ponderosa pine, pinyon pine, and juniper at 1,220 meters (4,000 feet) at the mouth of Pine Creek Canyon, Spring Range. The large conspicuous shrubs on the floor of the canyon are scrub oak (Quercus turbinella). View looking east. absolute minimum, but it is better to use Combinations of factors that determine the the minimum based on firm evidence than association of species could have been the figure of 1,220 meters (about 4,000 different in the past. A common plant com- feet) or more which is admittedly more munity of the present Great Basin might have tenuous. been the exception during the Pleistocene, Knowledge of Pleistocene climatic while other associations of plant species, change in the Southwest and Great Basin that are presently atypical, might have been is based on many lines of evidence such the rule. Different rates of plant dispersal as glaciation, pluvial lakes, soils, fossil and community succession coupled with plants and animals, and biogeography. I rapid climatic and geologic changes might have previously reviewed the literature result in "mixtures" of species from pres- relevant to the interpretation of the fossil ently different plant communities. Clearly, plant and pollen record at Tule Springs we need more information on the association (Mehringer, 1965, pp. 172-174), and of fossil species. Malde (1964) and Martin and Mehringer Because of the nature of paleoeco- (1965) have also summarized the evidence logical evidence, which is rarely if ever for Pleistocene climatic change in the arid complete, the interpretations must be based and semi-arid western United States. There on threads of evidence which are comparable is no doubt that there were changes in the to a few pieces of a jigsaw puzzle. The distribution of plants and animals accom- interpretations are usually made to a greater panying Pleistocene climatic change, but or lesser extent on one or more of the follow- the magnitude of biotic change is still open ing factors: (1) the fossil record; (2) the to question and should remain so, for the present distributions and ecological tolerances evidence is still meager. It would be naive of the fossil species and assemblages or to assume that the few pollen profiles from similar forms and assemblages; (3) untestable playa lake cores (Martin and Mehringer, ecological intuition; (4) a vivid, uninhibited 1965, Fig. 2; Roosma, 1958) fossil plant imagination. In many cases the reconstruc- remains from woodrat middens (Wells, 1964; tion of past environments approaches or Wells and Jorgensen, 1964), fossil plant exceeds our knowledge of the present ecology. and pollen remains from ground sloth dung For example, I have attempted to interpret the (Loudermilk and Munz, 1934, 1938; Martin, fossil pollen record from spring deposits and et al., 1961), and the fossil pollen and yet there are no comprehensive ecological plants from the Tule Springs Site can be studies of the present springs and spring more than indicative of the possible magni- mounds of southern Nevada. Fortunately, the tude of Pleistocene vegetational change in ecology of desert springs is now being studied the Mohave Desert. by W. G. Bradley and J. E. Deacon (Nevada Past vegetation change has been Southern University). These studies should described here with a bias imposed by a give a firmer basis for future interpretation conception of the association of familiar of fossil pollen and plant records from spring species in present plant communities. mound deposits.

A PLEISTOCENE (WISCONSIN) WOODLAND CORRIDOR IN THE EASTERN MOHAVE DESERT

There are not yet enough known Pleisto- woodland and forest communities is the south- cene fossil plant and pollen localities in the ward extension of fairly continuous mountain- Mohave Desert to draw any sweeping conclu- ous terrain from Potosi Mountain at the south sion about the vegetational changes for the end of the Spring Range, through Clark Moun- entire region. However, I have attempted to tain and the McCullough, New York, Provi- map (Fig. 38) what I believe are probable dence, and Granite Mountains. With a slight vegetational changes for a portion of the area change in the lower elevation of the plant from southern Nevada to the San Bernardino communities, this chain would give an almost Mountains, southern California. The main continuous Pleistocene woodland corridor, factor leading to the attempt at mapping the continuing through the Bristol, Bullion, and 189 117° 11 6° 115° ° ° 36 36 0B. LDER ...... Z50 411 •otosi C I Y L. Mesquite Mtn

- 231: L:Ivaripatt • • CALIF. Unmapped gstan ....

Ponderosa Pine-White Fir ...... ''''' NaPolio gh'''' ...... or. • .... SEAR H LIGH T Sagebrush-Juniper -Pinyon

878: .

Teutai°1o.Ait 22 4 L. Vt Mohave 295 AKelso Pk L. Maniz t;i14‘ wove R 50 35°- Harper" L. Bristol ft BARSTOW Mts 17 7 NEEDLE,rS\

...... itS TD. Granite Mts Ord Mtn :131- '' '' 4 • . A1920 ......

itifotinta...... :VMT: VILLE ...... 8...... 1623 01-d: ...... AS' ...... °No Woman. 17 ...Lucerne ...... Ws • .. Valley .... Bristol L

Cadiz

o TWENTY-NINE PALMS 340 1 34° ° 11 7° 11 6° 115

Figure 38. Map of possible Pleistocene woodland corridor. A map showing a possible Pleistocene woodland corridor through the eastern Mohave Desert, between the Spring Range, southern Nevada and the San Bernardino Mountains, southern California, and the possible southward extension of ponderosa pine-white fir into the New York-Providence and McCullough Mountains. The vegetation below the proposed lower limit of sagebrush- juniper-pinyon is not mapped, and, with the exception of the sagebrush-juniper-pinyon of the desert slope, no attempt was made to map the vegetation of the San Bernardino Mountains. Excepting minor fluctuations, the map is for the period from approximately 22,000 to 12,500 B.P. Such an expansion of woodland and forest toward the south and to lower elevations could have occurred several times during the Pleistocene. Elevations are given in meters.

190 Ord Mountains to the San Bernardino Mountains and northward. Only the juniper- Mountains of southern California. As map- pinyon is mapped in the San Bernardino ped, there is only one gap of less than Mountains, and even though there are a few 15 kilometers (between the Bristol and areas such as in the Ord and Old Woman Bullion Mountains) separating the postu- Mountains, which rise above the 1,300 lated continuous distribution of juniper- meters (4,300 feet) used as the lower limit pinyon woodland. of ponderosa pine-white fir, the areas are The map shows juniper-pinyon wood- small and discontinuous and are not indica- land above 800 meters (2,600 feet), 1,000 ted as ponderosa pine-white fir on the map. meters (3,300 feet) below its average lower The evidence on which the map is limit of 1,800 meters (6,000 feet) in the based consists of an extrapolation south- Spring Range, Nevada. Ponderosa pine- ward of the changes indicated primarily white fir is shown above 1,300 meters by the fossil plant and pollen record from (4,300 feet), 900 meters (3,000 feet) below the Tule Springs Site and also the following: its average lower limit at 2,200 meters (1) the joshua tree and possibly juniper (7,200 feet) in the Spring Range. No allow- remains from Gypsum Cave (Loudermilk and ances are made for less vegetation change Munz, 1934); (2) juniper remains from toward the south because the vegetation map Rampart Cave (Martin et al., 1961); of Johnson, Bryant, and Miller (1948) for the (3) radiocarbon-dated juniper remains from New York-Providence Mountain area shows the Frenchman Flat area north of Tule Springs the lower limit of juniper-pinyon woodland (Wells and Jorgensen, 1964); and (4) plant at approximately 1,550 meters (5,000 feet), remains from two other dated woodrat mid- 250 meters lower than the 1,800 meters I dens, that were discovered by P.V. Wells-- have used for the Spring Range. Also, the one from Negro Butte (UCLA-757, 9,140+ figure given for the Spring Range is higher 140) in Lucerne Valley and the other from than the average lower limit of juniper- the Turtle Mountains (UCLA-756, 13,900+ pinyon on the desert slope of the San Bern- 200). The Negro Butte woodrat midden ardino Mountains. The decreasing elevatior contains the remains of Juniperus osteo- of vegetation zones with decreasing latitude sperma. Both Juniperus osteosperma and in the Mohave Desert region was recognized Pinus monophylla are present in a midden by Loew (1876, p. 442). The fact that I from the Turtle Mountains (Berger et have used the Spring Range for comparison p. 367). has made the estimates shown on the map After the vegetation map was com- conservative for all of the ranges except pleted I discovered two other fossil wood- those of low elevation and small mountain rat midden deposits. They have not been mass. radiocarbon-dated or examined in detail, I believe that the 1,000-meter change but they do contain obvious evidence for shown should be considered minimum under vegetational change in the Mohave Desert. conditions where soil, slope angle, and One is in southern Nevada, 11 kilometers exposure allow for the growth of woodland northwest of Davis Dam, at an elevation today. There would be extensions of con- of about 730 meters (2,400 feet). This tinuous stands of sagebrush in the deep- midden contains both Tuniperus osteosperma soiled valleys and a mixture of sagebrush- and Pinus monophylla. The other is on juniper-pinyon above the 800-meter (2,600 Clark Mountain, California (Fig. 38) at an feet) contour used to map the lower limit of elevation of about 1,940 meters (6,400 juniper-pinyon, as well as an extension of feet). It contains needles of white fir some sagebrush and juniper below this and at least two species of pine that do level. No attempt has been made to map not now grow on Clark Mountain, includ- the communities below juniper-pinyon or ing limber pine (Pinus flexilis). above ponderosa pine-white fir. It is pos- Although, to the best of my knowledge, sible that species, which normally occur there is no fossil evidence for either pon- above ponderosa pine-white fir, could have derosa pine or white fir in the New York, colonized the higher peaks in the New York McCullough, or Providence Mountains, the

191 present distribution in nearby mountains probability that ponderosa pine and white indicates that suitable habitats would be fir could have occupied the area as indi- available for colonization by either pon- cated on the map (Fig. 38); I think it is derosa pine or white fir if conditions were probable that they did expand southward only slightly cooler and more moist than into areas that they do not presently occupy. they are today. Ponderosa pine and white Ultimately the matter can be resolved only fir presently grow on Potosi Mountain by fossil or biogeographic evidence. (Miller, 1945), and white fir grows on Clark Four Pleistocene lakes are shown on Mountain (Miller, 1940) and Kingston Peak the map. The lake areas--for the highest (Munz and Keck, 1963, p. 50; P. A. Munz, Pleistocene stand of each of the lakes-- pers. comm.). A lowering of vegetation are taken from Snyder, Hardman, and zones would greatly expand the present Zdenek (1964). Two dates of 19,300 and distribution of ponderosa pine-white fir, 19,500 B.P. on tufa (Fergusson and Libby, 'decrease the long-range dispersal distance 1962; Hubbs, Bein, and Suess, 1962) and between presently isolated ranges, and a date of 13,800 B.P. on Anodonta shells allow for an increase in the area suitable (Hubbs et al., 1965) from shorelines of for colonization of ponderosa pine or white Lake Manix, and ten dates ranging from fir introduced by long-range dispersal. For 8,350 to 13,670 B.P. (Hubbs et al., 1965, example, using the area-elevation graph p. 99; Warren and DeCosta, 1964) on Ano- for the Spring Range (Fig. 30), if the pre- donta shells and tufa associated with the sent area above 3,050 meters (10,000 feet) shore features of Lake Mohave are sufficient is suitable for colonization by a boreal to indicate that the lake levels shown on species introduced by chance long-range the map are close to the levels during the dispersal, and if, because of climatic maximum of Late Wisconsin (Woodfordian) change, the lower limit was reduced uni- pluvial conditions. There are no dates from formly by 610 meters (2,000 feet), so that Lake Mesquite or Lake Ivanpah, but based the lower limit at which an introduced on the radiocarbon chronologies from other boreal species could survive was 2,440 Pleistocene lakes (Broecker and Kaufman, meters (8,000 feet), then the "target area" 1965; Stuiver, 1964) and the dates from for long-range dispersal and colonization Lake Manix and Lake Mohave, the high of .a boreal species would be approximately levels of these two lakes, as shown on the nine times larger. The increase in space map, are also probably close to the levels and available microhabitats should also during the maximum of the Late Wisconsin allow for an increase in species diversity. pluvial conditions. 1 This sort of reasoning only indicates the

1 A part of this section of the report was junct populations of pinyon and juniper with sent to P. V. Wells (University of Kansas) in the larger bodies of existing woodland." January, 1966 after it had been submitted for In 1962, at the Fort Burgwin Conference publication. Wells informed me that he had on Paleoecology, P. S. Martin suggested presented a map of existing woodland vege- that Pleistocene biotic zones were at least tation and pointed out the likely woodland 800 meters lower on isolated desert mountain connections along the Spring-Clark-New York- ranges (Martin, 1964, p. 72) and clearly Providence-Granite Mountain axis at the illustrated his ideas on former continuous Ecological Society of America meetings in woodlands in the Southwest (Martin, 1964, August, 1964. The abstract of this talk Fig. 27). There seems to be general agree- (Wells, 1964) has been cited. It does not ment that many presently disjunct woodland give specific localities, but suggests former communities were continuous during the woodland connections "of the outlying, dis- Pleistocene. 192 SUMMARY AND CONCLUSIONS

1. Pollen analysis of the Tule Springs Excepting two major fluctuations, there is a Site was hampered by the absence of pollen trend toward warmer and dryer conditions from from most samples and poor preservation in about 12,000 to 7,000 B.P. By 7,000 B.P many others. Of the 225 samples extracted, the vegetation of the Las Vegas Valley was 73 yielded enough suitable pollen for minimal probably much like the present lower eleva- 200-grain pollen counts. tion Mohave Desert. A reversal of the warm- 2. The pollen content of Las Vegas ing-drying trend occurs in Unit E2 about Valley sediments indicates that the vege- 8,500 to 8,000 B.P. and about 10,500-10,000 tation underwent several major changes B.P. during Wisconsin time, but the pollen re- 7. Between 7,000 B.P. and the present cord is not complete for this time span. there are no significant changes in the pollen There are probably many minor vegetational spectra, but the pollen record is incomplete fluctuations which are not reflected in the through this time interval and there could have pollen record, and some major ones which been some important but probably minor cli- are missing due to the complete destruction matic and vegetational shifts. It is doubtful of the pollen in some stratigraphic units, if minor changes in the vegetation of the Las as well as missing sediments. Vegas Valley, amounting to less than a 300- 3. A major fluctuation which occurs meter (1,000 feet) variation in the elevational in the pollen record of Unit B2 indicates a position of the present vegetation zones, change from sagebrush-dominated Great could be detected in the fossil pollen spectra. Basin Desert to more moist and possibly 8. During the maximum of pluvial condi- cooler conditions, with an expansion of tions recorded in the Tule Springs fossil pollen forest and woodland communities in the record there was a minimal lowering of the ranges surrounding the Las Vegas Valley, present vegetation zones by 1,000 meters followed by a return to sagebrush desert. (3,300 feet). 4. AU of the pollen spectra from 9. The results of the pollen studies at buried lake sediments of Unit D, dating Tule Springs can be considered only as a from about 31,300 to 22,000 B.P., are first attempt, and preliminary to further study. indicative of a vegetation growing under The most important outcomes of the Tule more moist and cooler conditions than the Springs pollen studies are the broad framework present, but the variations in the pollen of vegetation change, which can be refuted or counts within Unit D could be mainly the verified and refined and expanded, and an result of fluctuating lake level. The pollen indication of the potentialities and problems spectra from the lake sediments and from a of further pollen studies in the Mohave Des- single spring mound sample indicate that ert. Because the potentials of spring mound there was an expansion of the present wood- deposits were not appreciated until these land and forest communities in the ranges pollen studies were almost completed, the surrounding the Las Vegas Valley. most promising single source of a fossil pol- 5. The most reliable evidence for len record from southern Nevada has barely vegetation change comes from Unit E1, about been investigated. Studies of the fossil pol- 14,000 to 13,000 B.P. High values of juni- len and plant remains from the spring mound per pollen along with the fossil wood of deposits of southern Nevada are continuing Cupressaceae leave little doubt that junipers and should add much to our knowledge of the were growing on the floor of Las Vegas Val- Pleistocene vegetational history of the Mohave ley. A southward vegetational extension is Desert. indicated by the presence of Shepherdia and 10. The possibility of an almost contin- Slum pollen in Unit El. uous Wisconsin-age woodland corridor between 6. A major change in the pollen record the Spring Range, southern Nevada, and the occurs at about 12,000 B.P. with sagebrush- San Bernardino Mountains, southern Califor- shads cale replacing juniper-sagebrush. nia, is suggested.

193 REFERENCES CITED

Adam, D. P. Billings, W. D. 1965 Exploratory Palynology in the 1949 The Shadscale Vegetation Zone Sierra Nevada, California. of Nevada and Eastern California Unpublished Master's Thesis, in Relation to Climate and Soils. University of Arizona. American Midland Naturalist 42, pp. 87-109. Allred, D. M., D. E. Beck, and C. D. Jorgensen Bradley, W. G. 1963 Biotic Communities of the 1964 The Vegetation of the Desert Game Nevada Test Site. Brigham Range with Special Reference to Young University Science Bulle- the Desert Bighorn. 1964 Trans- tin 2 (2), pp. 1-52. actions Desert Bighorn Council, pp. 43-67. Anderson, R. Y. 1955 Pollen Analysis, a Research Broecker, W. S. and W. R. Farrand Tool for the Study of Cave 1963 Radiocarbon Age of the Two Creeks Deposits. American Antiquity Forest Bed, Wisconsin. Geolog- 21, pp. 84-85. ical Society of America Bulletin 74, pp. 795-802. Antevs, E. 1955 Geologic-climatic Dating in Broecker, W. S. and A. Kaufman the West. American Antiquity 1965 Radiocarbon Chronology of Lake 20, pp. 317-335. Lahontan and Lake Bonneville II, Great Basin. Geological Society Aschmann, H. of America Bulletin 76, pp. 537- 1958 Great Basin Climates in Relation 566. to Human Occupance. University of California, Berkeley, Archaeo- Brown, M. logical Survey Reports, No. 42, 1960 Climates of the States, No. 60- pp. 23-40. 26, Nevada. U. S. Detrt. of Commerce, Weather Bureau, U. S. Beatley, J. C. Government Printing Office, 1965 Ecology of the Nevada Test Site Washington. 16 P. I. Geographic and Ecologic Distributions of the Vascular Clokey, I. W. Flora (Annotated checklist). 1951 Flora of the Charleston Moun- University of California, Los tains, Clark County, Nevada. Angeles, Biological and Medical University of California Press, TID-4500-39th ed., 69 p. Berkeley and Los Angeles. 274 p.

Bent, A. M. and H. E. Wright, Jr. Cottam, W. P., J. M. Tucker, and R. Drobnick 1963 Pollen Analysis of Surface Mate- 1959 Some Clues to Great Basin Post- rials and Lake Sediments from pluvial Climates Provided by Oak the Chuska Mountains, New Distributions. Ecology 40, pp. Mexico. Geological Society of 361-377. America Bulletin 74, pp. 491- 500. Coville, F. V. 1893 Botany of the Death Valley Expe- Berger, R., G. J. Fergusson, and W. F. Libby dition. U. S. National Herbar- 1965 UCLA Radiocarbon Dates IV. ium Contributions, Vol. 4, 363 P. Radiocarbon 7, pp. 336-371.

194 Dahl, E. - 1952 Pollen Morphology and Plant 1964 'Present-day Distribution of Taxonomy. Angiosperms. (An Plants and Past Climate," pp. 52- Introduction to Palynology I.). 61. In The Reconstruction of Chronica Botanica, Waltham. Past Environments. Fort Burqwin 539 p. Research Center Publication No. 3. Erdtman, G., B. Berglund, and J. Praglowski 1961 An Introduction to a Scandinavian Damon, P. E., C. V. Haynes, and A. Long Pollen Flora. Almquist and Wik- 1964 Arizona Radiocarbon Dates V. sell, Stockholm. 92 p., 74 pls. Radiocarbon 6, pp. 91-107. 'Erdtman, G., J. Praglowski, and S. Nilsson Davis, M. B. and E. S. Deevey, Jr. 1963 An Introduction to a Scandinavian 1964 Pollen Accumulation Rates: Esti- Pollen Flora. II. Almquist and mates from Late-Glacial Sedi- Wiksell, Stockholm. 89 p., 58 pls. ments of Rogers Lake. Science 145, pp. 1293-1295. Faegri, K. and J. Iversen 1964 Textbook of Pollen Analysis. Deacon, J. E., W. G. Bradley, and K. M. Scandinavian University Books, Larsen Munksgaard. 237 p. 1964 Ecological Distribution of the Mammals of Clark Canyon, Falconer, G., J. T. Andrews, and J. D. Ives Charleston Mountains, Nevada. 1965 Late Wisconsin End Moraines in Journal of Mammalogy 45, Northern Canada. Science 147, pp. 397-409. pp. 608-609.

Deevey, E. S., Jr. Fassett, N. C. and B. Calhoun 1944 Pollen Analysis and Mexican 1952 Introgression Between TyPha lati- Archaeology: An Attempt to Apply folia and T. anqustifolia. Evolu- the Method. American Antiquity tion 6, pp. 367-379. 2, pp. 135-149. Fergusson, G. J. and W. F. Libby Dittert, A. E., Jr. and F. Wendorf 1962 UCLA Radiocarbon Dates I. Radio- 1963 Procedural Manual for Archaeo- carbon 4, pp. 109-114. logical Field Research Projects of the Museum of New Mexico. 1964 UCLA Radiocarbon Dates III. Papers in Anthropology, No. 12. Radiocarbon 6, pp. 318-339. Museum of New Mexico Press, Santa Fe. 70p. Flint, R. F. 1963 Status of the Pleistocene Wiscon- Dixon, H. N. sin Stage in Central North America. 1962 Vegetation, Pollen Rain, and Pol- Science 139, pp. 402-404. len Preservation, Sangre de Cristo Mountains, New Mexico. Unpub- Flint, R. F. and F. Brandtner lished Master's Thesis, Univer- 1961 Climatic Changes Since the Last sity of New Mexico. Interglacial. American Tournal of Science 259, pp. 321-328. Durham, 0. C. 1951 The Pollen Harvest. Economic Frye, J. C. and H. B. Willman Botany 5, pp. 211-254. 1960 Classification of the Wisconsin Stage in Lake Michigan Glacial Erdtman, G. Lobe. Illinois Geological Survey 1943 An Introduction to Pollen Analysis. Circular 285. 16 p. Ronald Press, New York. 239 p. 195 1964 "Pollen Analysis and the Full- Mason, H. L. glacial Landscape," pp. 66-75. 1957 A Flora of the Marshes of Calif- In The Reconstruction of Past ornia. University of California Environments. Fort Burgwin Press, Berkeley and Los Angeles. Research Center Publication No. 3. 878 p.

Martin, P. S. and W. Byers McMillan, C. 1965 Pollen and Archaeology at Weth- 1959 Salt Tolerance within a Typha rill Mesa, Colorado. American population. American Journal Antiquity 31(2), Part 2, pp. 122- of Botany 46, pp. 521-526. 135. McNaughton, S. J. Martin, P. S. and J. Gray 1966 Ecotype Function in the Typha 1962 Pollen Analysis and the Ceno- Community-type. Ecology zoic. Science 137, pp. 103-111. Monograph 36, pp. 297-325.

Martin, P. S. and P. J. Mehringer, Jr. Mehringer, P. J., Jr. 1965 "Pleistocene Pollen Analysis and 1965 Late Pleistocene Vegetation in Biogeography of the Southwest," the Mohave Desert of Southern pp. 433-451. In H. E. Wright Nevada. Journal Arizona Academy and D. G. Frye (eds.), The Qua- of Science 3, pp. 172-188. ternary of the United States. Princeton University Press, Mehringer, P. J., Jr. and C. V. Haynes, Jr. Princeton. 1965 The Pollen Evidence for the Environment of Early Man and Martin, P. S. and J. E. Mosimann Extinct Mammals at the Lehner 1965 Geochronology of Pluvial Lake Mammoth Site, southeastern Cochise, Southern Arizona, III. Arizona. American Antiquity 31, Pollen Statistics and Pleistocene pp. 17-23. Metastability. American Journal of Science 263, pp. 313-358. Miller, A. H. 1940 A Transition Island in the Mohave Martin, P. S. , B. E. Sabels, and D. Shutler, Desert. Condor 42, pp. 161-163. Jr. 1961 Rampart Cave Coprolite and 1945 Birds of the Yellow Pine Associa- Ecology of the Shasta Ground tion of Potosi Mountain, Southern Sloth. American Journal of Nevada. Condor 47, pp. 130-131. Science 259, pp. 102-127. Mosimann, J. E. Martin, P. S. and J. Schoenwetter 1962 On the Compound Multinomial 1960 Arizona's Oldest Cornfield. Distribution, the Multivariate Science 132, pp. 33-34. B-distribution, and Correlations Among Proportions. Biometrika Martin, P. S. and F. W. Sharrock 49, pp. 65-82. 1964 Pollen Analysis of Prehistoric Human Feces: a new approach 1963 On the Compound Negative Multi- to ethnobotany. American nomial Distribution and Correlations Antiquity 30, pp. 168-180. Among Inversely Sampled Pollen Counts. Biometrika 50, pp. 47-54. Mason, C. T., Jr. and R. H. Hevly 1961 Additions to the Aquatic Flora 1965 "Statistical Methods for the Pollen of Arizona. Madrono 16, p. 32. Analyst: Multinomial and Negative

198 Multinomial Techniques," pp. 636- Schoenwette r, J. and F. W. Eddy 673. In B. Kummel and D. Raup 1964 Alluvial and Palynological Recon- (eds.), Handbook of Paleontological struction of Environments, Nava- Techniques. W. H. Freeman and jo Reservoir District. Papers in Co., San Francisco and London. Anthropology, No. 13. Museum of New Mexico Press, Santa Fe. Muller, J. 155 p. 1959 Palynology of Recent Orinoco Delta and Shelf Sediments. Sears, P. B. Micropaleontoloqv 5, pp. 1-32. 1932 The Archaeology of Environment in Eastern North America. Amer- Munz, P. A. and D. D. Keck ican Anthropologist 34, pp. 610- 1963 A California Flora. University 622. of California Press, Berkeley and Los Angeles. 1681 p. Shreve, F. 1942 The Desert Vegetation of North Ogden, J. G., III America. Botanical Review 8, 1965 Pleistocene Pollen Records from pp. 195-246. Eastern North America. Botani- cal Review 31, pp. 481-504. Smith, S. G. 1961 Natural Hybridization and Tax- Potter, L. D. onomy in the Genus TyPha, With 1964a "Limitations of Palynology to Particular Reference to California Paleoecological Reconstruction," Populations. Unpubl. Ph.D. pp. 39-43. In The Reconstruc- Dissertation, University of tion of Past Environments. Fort California, Berkeley. Burqwin Research Center Publi- cation No. 3. Snyder, C. T., G. Hardman, and F. F. Zdenek 1964 Pleistocene Lakes in the Great 1964b Stream Transport of Pollen. Basin. Miscellaneous Geological Bulletin Ecological Society of Investigations, Map I-416. U. S. America 45, p. 76. Geological Survey, Washington, D. C. Potter, L. D. and J. Rowley 1960 Pollen Rain and Vegetation, San Stuiver, M. Augustin Plains, New Mexico. 1964 Carbon Isotopic Distribution and Botanical Gazette 122, pp. 1-25. Correlated Chronology of Searles Lake Sediments. American Tournal Rogers, M. J. of Science 262, pp. 377-392. 1966 Ancient Hunters of the Far West. Union-Tribune Publishing Co., Tauber, H. San Diego. 207 p. 1965 Differential Pollen Dispersion and the Interpretation of Pollen Dia- Roosma, A. grams. Danmarks Geoloqiske 1958 A Climatic Record from Searles Undersoqelse Series 2, No. 89, Lake, California. Science 128, 69 p., 2 pls. p. 716. Ting, W. S. Schoenwetter, J. 1961 On Some Pollen of Californian 1966 A Re-evaluation of the Navajo Umbelliferae. Pollen et Spores Reservoir Pollen Chronology. 3, pp. 189-199. El Palacio 73, pp. 19-26.

199 Troels-Smith, J. Llano Estacado," pp. 115-133. 1960 Ivy, Mistletoe and Elm Climatic In F. Wendorf (ed.), Paleoe- Indicators-Fodder Plants. Dan- cology of the Llano Estacado. marks Geologiske Undersogelse Museum New Mexico Press, 4(4), 32 p., 6 pls. Fort Burgwin Research Center Publication No. 1. Tucker, J. M., W. P. Cottam, and R. Drobnick 1961 Studies in the Quercus undulata Wendorf, F. and J. J. Hester Complex II. The Contribution of 1962 Early Man's Utilization of the Quercus turbinella. American Great Plains Environment. Amer-

Tournal of Botany 48, pp. 329-339 • ican Antiquity 28, pp. 159-17/. -Wodehouse, R. P. Warren, C. N. and J. DeCosta 1964 Dating Lake Mohave Artifacts and 1959 Pollen Grains. Hafner Publish- Beaches. American Antiquity 30, ing Co., New York. 574 p. pp. 206-209. Woodbury, R. B. Wells, P. V. 1961 Prehistoric Agriculture at Point 1964 Pleistocene Vegetation in of Pines, Arizona. Memoirs for Mohave Desert: Some Woodrat the Society of American Archae- Midden Evidence. Bulletin Eco- ology No. 17, 48 p. logical Society of America 45, p. 76. 1963 "Indian Adaptations to Arid Environments," pp. 55-85. In Wells, P. V. and C. D. Jorgensen C. Hodge (ed.), Aridity and Man, 1964 Pleistocene Woodrat Middens American Association for the and Climatic Change in the Advancement of Science Publica- Mohave Desert: a record of tion No. 74. Horn-Shafer Co., Juniper Woodlands. Science Baltimore. 143, pp. 1171-1174. Yarnell, R. A. Wendorf, F. 1965 Implications of Distinctive Flora 1961 "An Interpretation of Late Pleis- on Pueblo Ruins. American tocene Environments of the Anthropologist 67, pp. 662-674.

200