UC Merced Journal of California and Great Basin Anthropology

Title Prehistoric Human Land-use Patterns in the Alvord Basin, Southeastern Oregon

Permalink https://escholarship.org/uc/item/83g062fb

Journal Journal of California and Great Basin Anthropology, 6(1)

ISSN 0191-3557

Author Pettigrew, Richard M

Publication Date 1984-07-01

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California Journal of California and Great Basin Anthropology Vol 6, No. 1, pp. 61-90 (1984).

Prehistoric Human Land-use Patterns in the Alvord Basin, Southeastern Oregon

RICHARD M. PETTIGREW

WO ideas have inspired more discussion, occur has been strongly suggested by Fagan Tcontroversy, and research regarding the (1974) in his study of spring sites, but prehistory of the Great Basin than any others. population decline still cannot be ruled out These are the Neothermal chmatic sequence (such population dechne is argued for the proposed by Ernst Antevs (1948, 1955) and western Great Basin by Elston [1982:193]). the Desert Culture, or Desert Archaic, con­ Currently, the most widely held view cept of Jesse Jennings (1957, 1964). These appears to follow the suggestion of Bryan and proposals, as time-worn as they may now Gruhn (1964) that detailed archaeological seem, focused on two essential problems of studies of particular areas are needed to assess Great Basin prehistory that are as relevant and the effects of climatic change on culture. compelling now as they have ever been: Furthermore, it is now understood that cli­ environmental fluctuation and the nature of matic change within the past 10,000 years human cultural adaptation. was much more complex than the simple The nature of the connection between sequence proposed by Antevs (though he may climatic change and cultural change in the have been correct at his level of analysis), and Great Basin has long been argued. The propo­ that its effects, from both hydrological and sition of Antevs (1948), that a period of anthropological perspectives, must have dif­ distinct desiccation (the Altithermal, from fered from basin to basin in the desert West 5000 to 2500 B.C.) occurred, led to the (Mehringer 1977; Weide and Weide 1977). suggestion that the Great Basin was aban­ This view has stimulated interest in the study doned, completely or at least in part, by of particular basins or other spatially defined human populations during that time (Cress- areas of the Great Basin, such as Surprise man et al. 1942; Cressman et al. 1960; Baum- Valley (O'Connell 1975), Warner Valley hoff and Heizer 1965; Layton 1972; BedweU (Weide 1968), Owens Valley (Bettinger 1975, 1973). Opposed to these suggestions were 1977), Coffeepot Flat (Aikens and Minor those who argued either that the Altithermal 1977), Stinkingwater Pass (Pettigrew 1979), as a dry interval never existed to any appre­ Steens Mountain (Aikens, Grayson and Meh­ ciable degree, or that its effects were not ringer 1982), Monitor Valley (Thomas 1983), significant to Great Basin cultures and thus and the High Rock Country (Layton 1970, there was no abandonment, or both (Martin 1979). 1963; Jennings 1964; Fry and Adovasio 1976; The recent emphasis on sampling spatially Weide 1976; Grayson 1976). That abandon­ defined areas is a clear advance over the ment of the northern Great Basin did not earlier approach of basing archaeological re­ constructions on data from individual exca­ Richard M. Pettigrew, Oregon State Museum of Anthro­ pology, Univ. of Oregon, Eugene, OR 97403. vated sites. The picture offered by Steward

[61] 62 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

(1938) of scattered people and scattered the entire reconstruction. resources in the Great Basin at the time of Second, the area should be one where contact had a profound effect on the interpre­ moderate changes in annual precipitation re­ tations of prehistorians during the first few sult in very noticeable changes in water decades of archaeological study of the region. availability. This criterion depends on both Until the 1970s, archaeological evidence, ecological and hydrological assumptions that gathered mostly from dry caves where the are well founded. In any arid temperate artifactual density was very high and preserva­ region, the availability of water is the most tion of perishables was excellent, was widely critical factor limiting biotic productivity, so interpreted to reflect a culture (migratory) that the more available water becomes on or and an environment (arid) that had remained near the surface the more abundant will stable and mostly unchanged for the past become the plants and animals that depend 10,000 years (Cressman et al. 1942; Jennings upon it. Water availability depends upon 1957; Aikens 1970). There was little reason precipitation rates, of course, but also upon to doubt that the environment during that factors of hydrology and geology that vary time had remained dry enough to make the from region to region, but that are neverthe­ nomadic societies of the ethnographic period less relatively constant within a given basin. the best adaptive model to account for the Basins vary in their potential for collecting archaeological record. water and making it available to biota. Those Now, however, it is widely recognized regions that are more efficient in this regard, that cultural deposits at individual sites are particularly closed basins, are better candi­ too greatly biased by their specific environ­ dates for detecting prehistoric chmatic change mental settings to be used for environmental as well as for finding cultural effects of that and cultural reconstructions of broad geo­ change. graphic areas. It is now agreed that the most Third, the area should have a diversity of reasonable and potentially most productive kinds of water resources: lakes and streams way to approach these problems is to engage that are available during moist periods, and in surveys of substantial geographic zones, springs which are the only source of water and to analyze the resulting data in such a during dry periods. This criterion is postu­ way as to determine when and how specific lated so that changes in water availability will areas within these zones were utilized. Pre­ be observable by distinct changes in land-use sumably, a recognized shift in the settlement patterns that can be correlated with the pattern at a specific period of time would different kinds of water sources. Presumably, signal a change in the adaptive strategy, or an lakes and streams wiU attract human settle­ alteration in the environment, or both. ment or exploitation during moist periods To test paleoenvironmental and paleo- when they contain water, and the relative cultural propositions in the Great Basin, one intensity of habitation at springs will be could use four major criteria to select the greatest during dry periods when the lakes ideal geographic area. First, the area should and streams tend to be dry. contain an unbroken geo-archaeological re­ Fourth, the ideal study area should in­ cord, in which occupational hiatus cannot be clude both lowland and upland zones with explained geomorphically, spanning at least their different biota. The purpose of this the past 10,000 years. Obviously, a shorter criterion is partly to make it possible to record, or one that is discontinuous, presents detect changes in the relative use of upland problems of interpolation that may weaken and lowland zones, which not only tend to PREHISTORIC LAND-USE IN THE ALVORD BASIN

Fig. 1. Location of Alvord Project Area. have different biota but also tend to have but is limited in its variety of water resources. different precipitation rates and different The Lake Abert area (Pettigrew 1981a, 1985) distributional patterns of water sources. Fur­ has a good variety of water resources, and thermore, the presence of such distinct zonal upland and lowland zones, but so far has environments should make it more likely that shown a good cultural record for only about settlement pattern changes will be noticeable the last 4,000 years; a similar assessment in the survey data, assuming that sites or could be made for Surprise Valley (O'Connell artifact locations classified by environmental 1975) and Warner Valley (Weide 1968). How­ zone will demonstrate differences between ever, a study area that appears to satisfy all zones as reflected in terms of frequencies of the criteria is found in the Alvord Basin. those sites or artifacts. There are actually few areas in the north- THE ALVORD BASIN SURVEY em Great Basin (Fig. 1) that satisfy this strict In 1975 the University of Oregon signed a set of criteria. Fort Rock Valley (Bedwell contract with the Bureau of Land Manage­ 1973) appears to be one of the best candi­ ment (BLM) to undertake a cultural resources dates, with an apparently long cultural record, survey of the Alvord Basin in the southeastern but it lacks a significant upland component. corner of Oregon (Fig. 1). The Alvord Basin The Glass Buttes area (Mack 1975) has upland contains a cultural record that spans more and lowland areas and a long cultural record. than 10,000 years, its water resources are 64 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY varied and responsive to changing rates of of Steens Mountain to the west, and Trout precipitation, and the lowland area is flanked Creek, an intermittent stream with sources in by Steens Mountain, one of the highest the mountains to the south and east. The only mountains (9,670 feet) in the northern Great permanent streams in the project area are Basin. The BLM requested the archaeological those that emanate from hot springs, of which resources survey to comply with federal regu­ there are several, all non-potable. The avail­ lations prior to geothermal leasing, and the ability of potable surface water in the area resulting project presented a unique oppor­ varies greatly from year to year. As an tunity to map site and artifact distributions in example, Alvord Lake was completely dry in a previously almost unexplored basin that 1975, but was full in 1979. Alvord Lake promised to provide comprehensive informa­ apparently fills up enough at times to over­ tion relating to some of the most hotly flow, judging by the well-defined outlet chan­ debated issues in Great Basin prehistory. nel connecting it to the Alvord Desert playa The Alvord Basin project area (Fig. 2) is to the northeast. situated in a graben with an interior drainage The project area includes a total of nearly system between Steens Mountain to the west 180 square miles of BLM land, of which 61 and a much lower chain of mountains and square miles, or 34% of the total area, were ridges to the east. Elevations in the project chosen to be surveyed in a nonstratified area range from just below 4,000 feet to random sample of sections. An additional approximately 5,000 feet above sea level. four square miles were surveyed in areas Most of the area is below 4,100 feet, an chosen because of their expected high site elevation marked by a strand line that prob­ densities, for a total of 65 square miles ably represents the last major stand of the surveyed, or 36% of the total project area. Pleistocene lake that filled the basin. For Data collection was Umited to information ob­ purposes of this study, the area above 4,100 tainable from surface inspection of the sites feet is defined as the upland zone (above the found, and survey crews were instructed to flat bottomland), and that below 4,100 feet collect only temporally diagnostic bifaces, as the lowland zone. Four major playas are particularly projectile points, both at defined located in the lowlands: Alvord Lake in the sites and as isolates. Other artifacts were south; the Alvord Desert, by far the largest recorded but not coUected, and a wide range play a, in the middle; a small, unnamed play a of information about each site was recorded northeast of the Alvord Desert; and Mickey on standard site forms. Lake in the far northeast. In general, and this A total of 224 sites was found during the is important to remember, the lowlands slope survey, as well as 217 classifiable projectile down from south to north; the highest bot­ points, recovered both as isolated artifacts tomland elevations in the project area are and from sites. For the purposes of this south of Alvord Lake, and the lowest, just report, the sample also includes 109 classifi­ below 4,000 feet, are at Mickey Lake. Pre­ able projectile points found during a less sumably, then, when the pluvial Lake Alvord intensive survey of the same project area the dried up, the area of Mickey Lake was the last previous winter by a crew from Portland State to be exposed, and the area south of Alvord University. Dr. Tom Newman of Portland Lake was the first. State generously loaned that collection for The principal sources of surface water in this analysis. The total sample size of pro­ the project area, outside of the springs, are jectile points used for this report, then, is the intermittent streams that drain the slopes 326. PREHISTORIC LAND-USE IN THE ALVORD BASIN

Fig. 2. Alvord Basin (elevations in feet). 66 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

ANALYTICAL ASSUMPTIONS debris on the landscape in different periods. Similar distributions, then, can be strong It should be possible to detect changes in evidence supporting contemporaneity of pro­ the human land-use pattern of the Alvord jectile point types, but cannot by themselves Basin by an analysis of the distribution of be conclusive. temporally sensitive projectile points. Corre­ A potential problem affecting the sample lating these changes with known environ­ is the degree to which the project area was mental features of the area should provide surface-collected by artifact hunters prior to clues as to why these changes took place. the formal surveys. That a considerable These are the basic assumptions of the analy­ amount of such activity had taken place is sis to follow, but it would be wise to make beyond question—it is common knowledge in exphcit the potential problems of the pro­ the local area. The intensity of such erasure of cedure, so that the results will be understood evidence cannot be accurately assessed. It is in the proper context. First of all, the analysis assumed for present purposes, however, that to follow deals only with the distribution of prior collection activity may have reduced the projectile points, not the distribution of overall sample size, the total count of points archaeological sites. Finding a point at a that were recovered, but did not significantly particular spot does not reveal what the affect the relative proportion of types in the original user of the point was doing there or collection. Study of private collections from why he was there, although it may be the area will be needed to measure this assumed that most points were dropped by variable. hunters. One can never be sure that the Because of the chronological sensitivity of person who left a point at a particular projectile points, plotting them by temporal location did not find it elsewhere, though types, as defined by Thomas (1981: 13-14) such cases are assumed to be relatively rare. and as used by Pettigrew (1981b) for the What remains, then, is the assumption that Lower Columbia Valley, changes in the dis­ the locations of projectile points usually tributional patterns of points that may have record the unspecified use of those locations resulted from changes in the aboriginal land- by the original depositors, who probably use patterns should be observable. However, included hunting among their activities. Plot­ the difficulties of using projectile points as ting the distribution of points shows areas time markers are also very real; not the least used, but does not necessarily show how they of these difficulties is the dependence upon were used. Other evidence must be brought to often very imperfect chronological data for bear on the problem of determining specific point types. Some point types are not very land uses. rehable temporal indicators (witness the It is tempting, when comparing the spatial "Pinto point problem" as discussed by Warren distributions of point types whose age assign­ [1980] and Thomas [1981]) and, even with ments are not secure, to assume that similar types for which temporal placement is reason­ distributions imply contemporaneity. Such ably well documented in a given locality, conclusions must be drawn with care, how­ confidence in cross-dating declines with the ever. Though contemporaneity is expected to distance between that locality and the place result in distributional similarity, the reverse where the types are used for chronological need not be true. Other factors, particularly assessment. This is a particular problem for similar environmental adaptation and land the Alvord Basin, where there is still no firm use, can cause similar spatial patterning of cultural chronology, although good progress PREHISTORIC LAND-USE IN THE ALVORD BASIN 67 has recently been made by the Steens Moun­ Table 1 tain Project (Wilde 1984). ALVORD BASIN PROJECTILE POINT COLLECTION BY TYPE TYPOLOGY Type Count Percent Rather than rely strictly on the traditional 1 101 31.0 2 53 16.3 binomial classification of Great Basin projec­ 3 21 6.4 tile points, I chose instead to lump known 4 33 10.1 point types together on the basis of known or 5 9 2.8 6 31 9.6 suspected contemporaneity in an effort to 7 4 1.2 approximate as closely as possible the desired 8 8 2.5 temporal types. The potential confusion 9 19 5.8 caused by creating numbered types is counter­ 10 15 4.6 11 23 7.1 balanced by the improved temporal utility of 12 9 2.8 the resultant groups. This procedure was Total 326 100.1 followed with the knowledge that chrono­ logical placement of point types is beset with uncertainties and is often speculative. The whether neck width could be a useful dimen­ relative degree of error, however, in temporal sion in standardizing this distinction, as it has assignments is reduced when very long occu­ been on the Lower Columbia River (Pettigrew pational periods are under scrutiny. This is 1981b) and elsewhere in the Northwest (e.g., the case in the Alvord Basin, where approxi­ Pettigrew 1982). Plotting the frequency of mately 12,000 years of occupation are repre­ points by neck width produced some useful sented. Another mitigating factor is that a results. The separation of modes between knowledge of only the relative ages of the narrow-necked and broad-necked points point types would be sufficient to produce showed up, as expected, but the minimum significant indications about changing pre­ between the two modes is at 9 mm., as shown historic land-use patterns. We can discuss the by Coriiss (1972) for Great Basin points, period when point type "X" was in use rather than the 8 mm. expected on the basis without knowing exactly its age in years of the Lower Columbia data. For purposes of before present. this paper, those points with necks 8 mm. Twelve types of points resulted from this wide or less are narrow-necked, and those process, and the collection inventory is listed 9 mm. and more are broad-necked. by these types in Table 1. One important The twelve types of projectile points used result of lumping the many Great Basin styles for the present analysis are listed in Table 2 into a typology with a small number of types along with an estimated chronology and other is the advantage of increasing the sample size pertinent information. Without doubt the per group, so that patterns distinguishing the estimated periods represented by the types types become more noticeable, and the nu­ are a matter of opinion and controversy. The merical differences between them more statis­ lively debate about Great Basin projectile tically significant. point chronologies will probably continue in Since the criteria for distinguishing smah the future as it has in the past (for a stemmed points from large stemmed points summary, see Thomas [ 1981 ]), and new data had not been formally defined in the Great may improve the accuracy of cross-dating in Basin until very recently (with the work of the northern Great Basin. Recent excavations Thomas [1981]), I decided to determine during the Steens Mountain Prehistory Project 68 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

Table 2 PROJECTILE POINT TYPES AS USED IN THIS PAPER

Type Fig. Description/Named Types Period Represented References 1 3 Desert Side-notched A.D. 1 to historic Baumhoff and Byrne 1959 Cottonwood Triangular Lanning 1963 Rosegate Series Heizer and Hester 1978 Thomas 1981 2 4 Elko Corner-notched 1500 B.C.-A.D. 500 Heizer and Baumhoff 1961 Elko Eared Heizer and Hester 1978 Thomas 1981 3 4 Pinto 2000 B.C.-A.D. 1 Heizer and Hester 1978 Gatecliff Split-stem Thomas 1981 4 5 Humboldt Series 4000 to 1000 B.C. Heizer and Clewlow 1968 Heizer and Hester 1978 Rice 1972 Northern Side-notched 5000 to 2000 B.C. Gruhn 1961 Leonhardy and Rice 1970 Hester 1973 Heizer and Hester 1978 6 6 Crescent 9000 to 5000 B.C. Tadlock 1966 Great Basin Transverse Clewlow 1968 Heizer and Hester 1978 7 6 Willow-leaf 8000 to 4000 B.C. Weide 1968 "Cascade" Leonhardy and Rice 1970 8 7 square base 9000 to 5000 B.C. Wormington 1957 Eden Sellards 1955 Scottsbluff Milnesand 9 7 tapered toward base 9000 to 5000 B.C. Campbell et al. 1937 Lake Mohave Butler 1965 Silver Lake Heizer and Hester 1978 Haskett 10 7 Windust 9000 to 6000 B.D. Daugherty 1956 Lind Coulee Rice 1972 Clewlow 1968 11 8 Black Rock Concave Base 9000 to 7000 B.C. Clewlow 1968 12 8 Clovis 10,000 to 8000 B.C. Sellards 1952 Hester 1973: 61-62

have produced new data (James Wilde, per­ depend on the chronology used, although the sonal communication 1984) that may make chronology has an important influence on the necessary some shift in estimates toward the interpretation of the analytical results. There­ "long" chronology of the eastern Great Basin fore, if new evidence or changed opinions (Aikens 1970). Despite these considerations, cause changes in the temporal estimates, the space limitations preclude a lengthy justifica­ results of the analysis undertaken in this tion of the temporal estimates used here, paper will be as valuable then as they are now. which would be a fine subject for another Some of the temporal types used here (and quite different) paper. Furthermore, as deserve further mention. Type 1 includes all will be seen, the analysis to come does not of the narrow-necked points. Thomas' (1981) PREHISTORIC LAND-USE IN THE ALVORD BASIN 69

o

T3 T3

•a .

O, D. O >.

H H

c 0^ fS g a" o BQ .S o. o ,>. ft. H

O ft. 70 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY PREHISTORIC LAND-USE IN THE ALVORD BASIN 71

Fig. 7. Point Types 8, 9 and 10. Top Row: Type 8; Second Row: Type 9; Third and Bottom Rows: Type 10. 72 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

Fig. 8. Point Types 11 and 12. Top Row: Type 11; Bottom Row: Type 12.

definitions are used for Type 2 (Elko Corner- correspond in terms of their horizontal distri­ notched and Elko Eared) and for Type 3 bution in the project area, a statistical device (Gatecliff Split-stem). Type 5 (Northern Side- was employed to measure the relative degree notched) is the only type whose estimated of co-occurrence of each possible pair of time span completely encompasses the pro­ types in each square-mile section that was posed period of the much-discussed Altither­ surveyed. This kind of analysis, called co­ mal. Type 6 (crescents) is most often found occurrence analysis, has been described else­ in Great Basin lowland locations on the where in much greater detail (Pettigrew margins of former lakes and marshes. This 1983). In this instance, the method answers evidence, combined with the distinctive for­ the question, "Does point type A co-occur mal characteristics of the style, points to their with point type B more or less frequently use in the hunting of waterfowl. The temporal than would be predicted in a random distribu­ estimate for Type 7 (willow-leaf) is very tion?" The result, for each pair of types insecure, because of their relative rarity, compared, is expressed as a z-score that is a particularly in excavated contexts, in the measure of the difference between the actual Great Basin. Types 8, 9, and 10 would be and predicted co-occurrence frequencies, and included in Layton's (1979) Great Basin is also a statement of the probability that Stemmed point series. point types A and B are randomly associated as well as a measure of the similarity or HORIZONTAL PATTERNS difference in their distributions. The z-score To find out how closely the 12 types matrix that results from a comparison of all PREHISTORIC LAND-USE IN THE ALVORD BASIN 73 possible pairs of point types can be summar­ ized in a more readily interpretable way by a dendrogram produced by an iterative or re­ computed linkage method, most easily accom- pUshed with the aid of a computer. The dendrogram produced by this analysis (dendrogram A in Fig. 9) divides the point population into two major groups, one made up of types 1, 2, 5, and 7, and the other composed of the remaining types. In the first group, Type 1 (essentially the probable arrow points) is quite divergent from the rest, and, in the second group. Type 3 (Gatechff Split- stem) stands out clearly from the rest while the sub-group composed of types 6 and 8 through 12 is very tightly clustered, leaving Fig. 9. Co-occurrence Dendrogram, Analysis by Type 4 (Humboldt series) somewhat distinct Square Mile Section. within the group. Because the tightly clus­ tered group (types 6 and 8 through 12) is and 8 through 12) that may reflect a fairly almost certainly the most ancient set of types stable settlement and resource utilization pat­ in the population, and Type 4 may or may tern during the early Holocene (approxi­ not be early in the local chronology, it was mately 10,000 to 5000 B.C.), and much decided to try the co-occurrence analysis looser clustering of later types (if the Hum­ again, but this time with only types 1 through boldt series is later, a possibility supported by 5 and Type 7. With the early group removed this analysis) that suggests more frequent the dendrogram (dendrogram B in Fig. 9) shifting of settlements. divided the population into two major groups, The significance of these results becomes one composed of types 1, 2, and 7, and the clearer when we look at the actual distribu­ other made up of types 3, 4, and 5. With the tions of point types found in the project area exception of Type 7, which has a sample size (Figs. 10 through 17). Types 11 and 12, the of only 4 specimens and shows no positive earliest group, cluster in the southern portion values in the z-score matrix (not shown), the of the project area (Fig. 10), south and east of first group is certainly younger than the Alvord Lake, on the margins of what was second. In the second group, types 3 and 4 probably a large marsh or shallow lake in are closely associated and link with Type 5 early post-Pleistocene time, about 10,000 to (large side-notched) at a fairly high value, 7000 B.C. Only one point of the group was in supporting the view held by Hester (1973:45) an upland area, and one of the group was that the Humboldt series dates primarily from found in the peripheries of the Alvord Desert 4000 to 1000 B.C., generally intermediate in and Mickey Lake playas, where the distance age between large side-notched points and between the lake bottoms and the surround­ Pinto (here Gatechff Split-stem or Type 3) ing basalt rims is short and was likely covered points. by water during this period. The results of the co-occurrence analysis Types 8, 9, and 10 form a group that is of the horizontal distribution of point types, similar but much more widespread (Fig. 11). then, show a tight clustering of early types (6 The majority of points is still from the 74 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY PREHISTORIC LAND-USE IN THE ALVORD BASIN 75 76 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY PREHISTORIC LAND-USE IN THE ALVORD BASIN 77 78 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY southern area, but the distribution surrounds water may have become scarce enough in the the Alvord Desert and Mickey Lake playas, Alvord Basin to seriously inhibit human utili­ which probably became separate bodies of zation of the area, particularly for a people water during this time, estimated to be whose traditional lifeway concentrated on between 8000 and 5000 B.C. Only three marsh and lakeside food resources. points of this group are from upland areas; The distribution of Type 4, or Humboldt habitation still apparently concentrated series points, appears rather surprisingly simi­ around bodies of water. lar (Fig. 14) to that of Type 6 and types 8 Type 7, the wihow-leaf-shaped style, has a through 12, the older styles. Just as with the very small sample size, but shows what older types. Type 4 points are mostly from appears to be a very different pattern from the southern half of the project area, particu­ those just described (distribution not map­ larly south and east of Alvord Lake. Only ped). Three of the four points in this group three are from the Mickey Lake Basin on the were north of the Alvord Desert playa, and all far north, and only two are from the upland were near the western margin of the lowland area. This distribution may suggest that zone. None were south and east of Alvord marshes and lakes were the primary extractive Lake, the southernmost playa, where types 8 zone while Type 4 was in use. Perhaps the through 12 were mainly clustered. This pat­ assumed temporal span of this style, between tern suggests that Type 7 is not as ancient as 4000 and 1000 B.C., is incorrect and should originaUy assumed, but the small sample size be shifted back in time by at least two or makes interpretations difficult. three thousand years. However, if the original Type 6, the crescent, foUows a pattern temporal estimate is correct, an explanation (Fig. 12) very similar to that seen for types 11 for the distributional pattern may be that and 12, which had the vast majority concen­ Type 4 was used during a later stand of the trated south and east of Alvord Lake. Only lakes in the Alvord Basin. Whatever the cause, one example came from the Mickey Lake the lowland concentration is marked. Basin in the north, and only two were found Type 3 (Gatecliff Split-stem or Pinto) in upland areas. Their distribution fits quite displays a pattern generally similar (Fig. 15) well with the notion that they were used on to that of Type 4, but has an increased the margins of shahow lakes and marshes to proportion of points in the northern half of hunt waterfowl during the period 9000 to the project area, surrounding the Alvord 5000 B.C. Desert, and a lesser proportion in the south- Type 5, the large, side-notched point, has em lowlands. This trend indicates a tendency a most pecular distribution (Fig. 13) in which for more dispersed and widespread land-use. six of the nine specimens are from the flats Type 2 (Elko series) points appear to between Alvord Lake and the Alvord Desert continue the trend toward dispersed land-use playa, two are from the margins of Mickey (Fig. 16), with specimens found throughout Lake, and one was near a spring at the edge of the project area, including the largest concen­ the lowlands east of Alvord Lake. This distri­ trations yet around Mickey Lake in the north, bution, combined with the low frequency of and an apparent increase in upland areas. A the type, fits well with the hypothesis of large proportion of this type was found near extreme desiccation for much of the time springs, a development that may suggest a between 5000 and 2000 B.C. At times during continued decline in the availabihty or de­ this period, the Altithermal of Antevs (1948, pendability of bodies of standing fresh water. 1955), water or resources dependent upon This distributional pattern continues into PREHISTORIC LAND-USE IN THE ALVORD BASIN 79 the period (approximately the last 2,000 years) represented by Type 1 (Fig. 17). As with Type 2, these points, mostly the small, narrow-necked styles generally considered to be arrow points, have a wide distribution (even wider than Type 2). A large proportion was found near springs, a particularly dense concentration was found in the Mickey Lake area, and points of this type were common in the uplands. Newly exploited areas, however, include the alluvial fan at the north edge of the Alvord Desert and the lowland dunes along a hne due south of Alvord Lake, the Fig. 18. Co-occurrence Dendrogram, Analysis by Ele­ southernmost playa. This distribution appar­ vational Intervals. ently represents an adaptation of native popu­ lations to an environment similar to current group of Fig. 18, dendrogram A, the only two conditions with the aboriginal technology types joined on the positive side of the scale documented at the time of contact. are types 4 and 5. In the second group, however, types 6 and 8 through 12 (as in the VERTICAL PATTERNS horizontal analysis) cluster at a very highly If the horizontal distribution of the point positive level, while Type 3 (Gatecliff Split- types shows changing land-use patterns stem) joins at a slightly negative value. through time, their vertical distribution may It is already clear, of course, that the be expected to provide complementary infor­ group of types 6 and 8 through 12 is the most mation of the same sort. In order to deter­ ancient of the lot, so the co-occurrence mine how the types varied in their elevational analysis was run again excluding that early distributions, a co-occurrence analysis was group, resulting in Fig. 18, dendrogram B. performed using as the co-occurrence units This new dendrogram, as before, divided the 5-foot intervals between 4000 and 4100 feet, population into two major groups, one com­ 50-foot intervals from 4101 to 4250 feet, posed of types 1, 2, and 5, and the other 100-foot intervals from 4251 to 4750 feet, made up of types 3, 4, and 7. The placement and the open-ended intervals below 4000 feet of Type 7, with its smah sample size (4), and above 4750 feet. These expanding inter­ needn't be given much consideration, but the vals were used to maintain comparably large joining of Type 5 (large side-notched) with sample sizes and to minimize the number of types 1 and 2 at a rather high level is intervals for analytical ease. The dendrogram somewhat unexpected. Inspection of the orig­ (A in Fig. 18) produced by this analysis inal z-score matrix from which the dendro­ divides the point population into two major gram was constructed (not shown) reveals groups, one made up of types 1,2,7, 4, and that Type 5 has a high positive z-score with 5, and the other including the remainder. This Type 4 (1.8) as well as with types 1 (1.5) and grouping is very similar to that resulting from 2 (1.9), pointing out the major weakness of the horizontal co-occurrence analysis (dendro­ dendrogram clustering: Type 5 shows tenden­ gram A in Fig. 9), the only difference cies to hnk with Type 4 as well as with types being that Type 4 (Humboldt series) is 1 and 2, but the dendrogram forces it to link now included with the first group. In the first with one and not the other. 80 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

A better understanding of the co­ occurrence clustering based on elevation may types 11112 be obtained by inspecting the elevational distributions themselves as summarized in Fig. types 8,9 £10 19. The earhest group, types 11 and 12, clus­ -TTT-rT^^'VTT^ ters very strongly between 4,051 and 4,075 type 6 feet with only less than 20% in the next VZZZ^^Z^^^ between 4,026 and 4,050 feet. The group types N--9 including types 8, 9, and 10 also clusters most U77A t////\ strongly between 4,051 and 4,075 feet, but the group is much more widespread in eleva­ type 4 N=33 tion, and appears to represent a time, between group is much more widespread in elevation, and appears to represent a time, between •!!'!!i m^^pzTT?^:^ 8000 and 5000 B.C., of declining lake and type 2 N-53 marsh levels. Type 6, the crescent, is contem­ typel ,__, N=101 poraneous with both previous groups, and kcltl 4000- 4014- 40S1- 407«- 4IOI- 4131- clusters at a somewhat lower elevation than 4000 403S 4050 40rS 4IOO 41SO SOOO they do, possibly because of its specialized usage along the margins of lakes and marshes. Fig. 19. Frequencies of Point Types by Elevation. Type 5, the large side-notched point, has the strangest vertical distribution of all, just as its horizontal pattern was aberrant. This may indicate a very limited usage of the Alvord filling of the lakes that accompanied a Basin bottomland during the estimated period changed way of life and an increase in of the proposed Altithermal, between 5000 population density. and 2000 B.C. That Type 5 showed tenden­ The vertical distribution of Type 3 (Gate­ cies to cluster in the dendrogram with both chff Split-stem), representing the period 2000 Type 4 and types 1 and 2 results apparently B.C. to A.D. 1, differs somewhat, though not from its substantial proportion from below a great deal, from that of Type 4. The most 4,000 feet (a factor it has in common with telling difference may be in its increase over types 1 and 2) and its high frequency between Type 4 in proportions above 4,050 feet, 4,000 and 4,025 feet (many Type 4 points are beginning a trend, that continues in later in this interval). types, toward more widespread utilization of Type 4, the Humboldt series, assumed to the landscape. date from 4000 to 1000 B.C., has its mode in Types 1 and 2, representing the past the same elevation as Type 6, but is generally 2,000 years, are remarkably similar in their more widespread in elevation than any of the vertical distributions, differing from the earli­ earlier types. This distribution is congruent er Type 3 in their considerably greater propor­ with the idea that Type 4 was the last style of tions above 4,050 feet and below 4,026 feet. point used during the first half of the post- With these two types, elevational diversifica­ Pleistocene time, when the lakes and marshes tion reaches its prehistoric maximum, an ap­ had reached their lowest levels before drying parent indicator of maximized economic up entirely, but also supports the notion that diversification and the final stage in a long the type was used during a late Holocene process of adaptational evolution. PREHISTORIC LAND-USE IN THE ALVORD BASIN 81

DISCUSSION lifespan in the northern Great Basin. This is suggested by the presence of fluted points in We cannot, of course, expect these data the Alvord Basin and elsewhere (Aikens from the Alvord Basin to inform us on more 1978), and paleontological evidence that than some of the details of the prehistoric now-extinct herd animals once roamed the cultural and climatic sequences. After ah, region (Cressman et al. 1942; Heizer and chronological estimates based on cross dating Baumhoff 1970). Each piece of such evidence are imprecise, and only a small upland area, moves us one step closer to recognizing a all below 5,000 feet, was sampled. Excava­ continent-wide horizon of late-Pleistocene tions of large, stratified sites would improve big-game hunters. This is not to assert that our chronological precision, and more sam­ fluted points necessarily imply speciahzed pling of higher elevations, such as that done big-game hunters, but only to suggest that by Fagan (1974) at spring sites and by the such big-game hunters were the first posses­ Steens Mountain Project' (Aikens, Grayson, sors of such points. I strongly suspect that the and Mehringer 1982), will give us a more first makers of fluted points in the northem complete picture of the seasonal round. Nev­ Great Basin were skillful big-game hunters; ertheless, we already have enough informa­ even if so, however, it is likely that their way tion, from the Alvord Basin survey and other of hfe very quickly became adapted to the sources, to propose a basic model of changing special conditions of that environment, which climatic and subsistence patterns for the was cool and moist, with abundant lakes and prehistory of the Alvord Basin and the north- marshes that offered a variety of resources em Great Basin. such as waterfowl, fish, deer, and edible seeds A very important test of any such model and tubers. By this reasoning, much of the is its ability to explain and encompass the daily lives of the later fluted-point makers and distributional data on the Alvord projectile their descendants was spent on the margins of point types. This criterion will be stressed in lakes and marshes, even if elephants, bison, the model proposed here. In the following horses, and other large mammals were taken discussions, I shall first present a behavioral whenever possible. The presence of fluted model to account for the pattems observed in points in the Alvord Basin in the same areas as the survey data, and then propose a more later, non-fluted points is the result, then, of generalized adaptational model to describe rapid adaptation to the local environment. the cultural evolution that was fundamental The first documented cultural tradition in to changing pattems of land-use. Finally, I the northem Great Basin, however, is what is shall compare the adaptational model to that aptly caUed the Westem Pluvial Lakes Tradi­ proposed by Elston (1982) for the westem tion (Bedweh 1973: 170-171), established Great Basin, as a way of exploring the possibly as early as 9000 B.C., and marked in potential for a region-wide interpretation of the archaeological record by several types of prehistoric cultural change. large, lanceolate projectile points and cres­ cents. Birds (possibly hunted with projectiles Behavioral Model tipped with crescents), fish, deer, and herds of Though there is little direct evidence yet large mammals, as well as significant vegetal to support the idea, the first well-recognized resources (most notably seeds and fruits), cultural tradition in North America, that of were dependable foods that required little the fluted-point big-game hunters, about seasonal movement, allowing small band ter- 10,000 to 9000 B.C., may have had a brief rirories as well as fairly dense human popula- 82 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY tions where winter game was abundant. Biotic carrying capacity. It would not be unreason­ resources that are now found only at high able to look for evidence of out-migration to elevations were available at low elevations more well-watered locations to the north and then, rendering hunting trips into the upland west, as suggested many years ago by Cress­ zones less necessary and less frequent than in man et al. (1942). The restricted, distinctive, later times. The distribution of the projectile and exclusively bottomland distribution of points of this period (ca. 9000 to 5000 B.C.), Type 5 (large side-notched) points, the only particularly the crescents, around the margins style whose estimated time span completely of lakes and marshes and only rarely in encompasses this period, is consistent with upland areas, such as at Last Supper Cave and the reconstmction presented here, in that this Hanging Rock Shelter in northwestem Nevada distribution clusters at lower elevations than (Layton 1979), argues for this kind of low­ any of the previous styles and in an area of land adaptation. The success of this way of the basin (between the Alvord Desert playa life may have contributed to its termination, and Alvord Lake) where earlier points are however, by reducing many of the large herd rarely found. In fact, Type 5 points are found animals, notably the mammoth, horse, camel, at the elevations and locations that in earlier and large bison, to extinction. This process times were probably covered by standing may have been quickened by the distinct water. The reconstruction proposed here warming and drying of the climate that would predict that Type 5 points should also appears to have taken place, and that reduced tend to be found at elevations higher than the extent of the lakes and marshes so those sampled by the Alvord Survey, and important to that mode of human adaptation. likely at the sites of dependable springs. Although it would be too simplistic to Between about 3000 and 2000 B.C., a believe that mid-postglacial times (5000 to much moister period began. This period, 3000 B.C.) were characterized by a single and dubbed the "Neopluvial" by Alhson (1982) uninterrupted period of desiccation (the Alti­ and by Curry and James (1982), saw the thermal of Antevs [1955]), it does appear reappearance of some lakes and previously that this temporal interval witnessed at least dry or nonexistent lowland springs. By this one of the most intense, and perhaps the most time the large, side-notched point of the culturahy significant, dry speUs of post- preceding period had probably been replaced Pleistocene time. Even if this period experi­ by the Humboldt series and then by the enced some wet intervals, it appears to have barbed, corner-notched point, and during the posed a very challenging problem to human Neopluvial period the human population in­ societies that had been highly dependent on creased and adopted a settlement pattern in herd animals and wetland resources. By 5000 which all elevational zones, from the basin B.C., the Western Pluvial Lakes Tradition had floor to the highest uplands, were exploited become extinct, and human groups were extensively. The broadly based nature of this forced to expand their territories to include adaptation is indicated by the increasing greater use of the high uplands above 5,000 diversification of the exploited territory and feet as well as lowland zones, to travel more elevations shown by types 4, 3, and 2 frequently and for longer distances in their (roughly in chronological order) (Figs. 14, 15, annual subsistence schedule, and to hunt 16, and 19). solitary game animals such as deer. Human By the time the narrow-necked points (all population density may have dropped signifi­ of which fit within Type 1) and, by inference, cantly in response to a lower environmental the bow and arrow arrived on the scene in PREHISTORIC LAND-USE IN THE ALVORD BASIN 83 southeastem Oregon about 2,000 years ago or have been four adaptive modes in the Alvord somewhat less, an extractive system generally Basin and perhaps adjacent areas. Each mode similar to that of ethnographic times was was an extractive economic system differing firmly established in the Alvord Basin and from the other modes in the evolutionary possibly in other, similarly dry basins. A sequence in terms of settlement patterns, different way of life, dependent on the preferred foods, scheduling of economic activ­ lacustrine ecosystem to a greater extent, ities, and (to a varying extent) technology. operated at Lake Abert at this time (Pettigrew The first adaptive mode was the Paleo-Indian 1981a), 1985), however, and may have charac­ (10,000-9000 B.C.) which has long been terized other basins with substantial lacustrine recognized as an early cultural stage in North resources, such as Summer Lake, Warner America. Evidence for Paleo-Indian occupa­ Valley and the Harney Basin. In the Alvord tion of the Alvord Basin is far from over­ Basin, however, the distribution of Type 1 whelming, as noted above, consisting of fluted projectile points, both vertically (Fig. 19) and points found in the Alvord survey. Neverthe­ horizontally (Fig. 17), shows an undiminish­ less, it is reasonable to propose that the ed, and in some ways increased diversity of human groups for which we have any evi­ exploited zones compared to previous styles, dence were fully nomadic foragers (Binford a pattern that appears to reflect little substan­ 1980, Madsen 1982b) speciahzing in large tial change in the extractive system. By 2,000 herd animals. Their tool kit was related to years ago or shortly thereafter, the Neopluvial that of the Upper Paleolithic of the Old in its most dramatic manifestations may have World, including large pointed bifaces hafted terminated, as suggested by Davis (1982: 67) on spear shafts and used as thrusting weapons for the Lahontan Basin, with more frequent as well as cutting tools. Tools for plant periods of drought, but there is no clear processing were not included, though a famili­ indication in the Alvord Basin projectile point arity with plant foods was probably part of data of a cultural response to such environ­ their tradition. mental change. It may be that the cultural The second adaptive mode was the West­ system as it had evolved previous to 2000 B.P. ern Pluvial Lakes Tradition (9000-5000 B.C.), in the Alvord Basin was adaptable to climatic which marks the beginning of what is proper­ fluctuations of the past two millenia, or that ly called the Archaic stage by the original the archaeological record as indicated by the definition laid down by Willey and Phillips survey is not detailed enough to reflect (1958). This mode, as indicated above, was cultural changes that occurred. This latter the first in which the regular use of a wide possibility is the most likely explanation for variety of food resources was practiced. What the lack of an archaeological expression of the brought about the end of the Paleo-Indian ethnic replacement connected with the Numic stage is not clear, though an appeal to expansion that probably took place within evolutionary ecology (O'Connell, Jones, and the last 2,000 years (Bettinger and Baumhoff Simms 1982) suggests that the change was 1982; Aikens 1982). brought about by the decimation of the primary prey of the big-game hunters, who Adaptational Model then chose a more reliable and diverse set of The Alvord Basin data appear to support resources to exploit. The folk of the Western the idea that there has been a series of major Pluvial Lake Tradition were collectors (Bin- adaptational shifts over the past 12,000 years ford 1980) who concentrated their economic of human prehistory. I propose that there effort on the wet bottomlands, probably 84 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY maintaining base camps on the margins of ized the subsistence economy, increasing the lakes and marshes and traveling to seasonal carrying capacity of the landscape even in the hunting and gathering camps. Upland areas absence of environmental amelioration. The were only sparsely used, since eco-zones were human population rebounded somewhat, con­ still depressed toward the lower elevations. centrating again in the lowlands. If there were house structures, they were The fourth and final adaptive mode was probably nonpermanent and unsubstantial, the Full Archaic (3000 B.C.-contact). The perhaps consisting of brush- or hide-covered adoption of plant foods (seeds and roots) as shelters. Heavy reliance on waterfowl is sug­ the staple created a subsistence strategy that gested by the frequency of crescents, though was very adaptable to changing conditions, a considerable amount of big-game hunting and less dependent on the fluctuating avail­ was practiced with the aid of large projectile ability of game. Wintertime food storage was points that may stih have been hafted to spear now facilitated, making it advantageous to shafts. Vegetable foods were doubtless used, maintain fixed winter settlements, ideally in though to an unknown degree. Manos and the lowlands where milder weather and easier metates may have been introduced during this access to game were critical factors. An period, but their degree of use is uncertain. increase in effective moisture after 3000 B.C. The rather high frequency of occurrence of (Madsen 1982a) only furthered the success of points from this period in the Alvord Basin this way of life, and the human population suggests that the human population was sub­ increased substantially, reaching a peak be­ stantial. tween 1000 B.C. and A.D. 1. The economic The third adaptive mode was what I caU system was based on a centrally based collec­ the Transitional Archaic (5000-3000 B.C.), tion strategy with seasonal camps used by which is poorly represented in the Alvord task groups at many localities throughout Basin. This mode is hypothesized to have each group's territory. The degree of seden- been an attempt by the human population to tism in the Alvord Basin is unknown; no adapt to a changed environment (notably, village sites with housepits have been found, without herd animals and substantial low- though they existed elsewhere in the northem elevation wetland resources) with a strategy Great Basin, particularly at Lake Abert (Petti­ that initially involved a shift to higher eleva­ grew 1981a, 1985). Hunting in the Full tions to collect the same foods that were used Archaic continued to be a very important by the people of the Westem Pluvial Lakes activity, as indicated by the high frequency of Tradition. Population density dropped to very projectile points. The adoption of the bow low levels, bands occupied large territories just less than 2,000 years ago may have been and moved more frequently, and people an adaptive improvement, but its only effect devised innovative techniques to maximize on the archaeological record seems to be an their their subsistence efficiency, including increase in the frequency of points, probably the hunting of solitary game mammals with caused by more frequent re-arming of projec­ notched projectile points. This was a time of tile shafts. shifting strategies from an emphasis on animal The Numic expansion that probably took products to plant foods. By the end of the place within the last millenium (Bettinger and period, the full-time adoption of the mano Baumhoff 1982 with references), whether it and metate for seed processing and possibly was the cause or the result of adaptive change the mortar and pestle for root processing in the northern Great Basin, did not consti­ (Ames and Marshall 1980-81) had revolution­ tute a new adaptive mode. The Full Archaic PREHISTORIC LAND-USE IN THE ALVORD BASIN 85 includes a range of possible adaptive systems work better for Elston's model to have higher (Madsen 1983b: 210-212), from those requir­ population densities for WPLT cultures, be­ ing frequent travel during much of the year cause it would help account for the demise of (Steward's Shoshonean model) to those with that way of hfe, as Elston himself (1982: sedentary villages occupied year-round (pos­ 193) points out. sibly at Lake Abert). This does not mean that Though Elston does not name a specific the Numic expansion was not accompanied period or stage identical with my Transitional by significant economic or adaptive change, Archaic (7000-5000 B.P.), he does mention but simply that the change was probably not the period 8000 to 6000 B.P., that the WPLT as fundamental as earlier shifts in adaptive terminated by 7000 B.P., and that his "Early mode. Archaic" began after 7000 B.P. A transitional period is imphed but not discussed, and no COMPARISON AND CONCLUSIONS earliest firm date for the Early Archaic is The model for prehistoric cultural evolu­ given. Both our models postulate a low tion outhned above is based on data from the population density for this period. Alvord Basin, and is intended for application There is general agreement between the at that locality. Its validity for the rest of the two models on the timing and nature of what northern Great Basin is a matter for future I call the Full Archaic and he calls the Archaic investigation. It may be useful, even so, to (Early, Mid-, and Late). Both models propose compare the model with that proposed by fairly low populations for the early part of Elston (1982) for the western Great Basin, this period, and both attribute the cause of since in my judgment the cultural records for the onset of the adaptive shift more to the northern and western Great Basin sub- culturo-technological factors than to environ­ areas show a great deal of similarity. Elston mental improvement. Both models also see a would lump my Paleo-Indian and Western population maximum attained between 3000 Pluvial Lakes Tradifion (WPLT) into his "Pre- and 2000 B.P., but the changes in subsistence Archaic" no doubt because of the paucity of and settlement pattems proposed by Elston Paleo-Indian evidence in the western Great (1982: 197-199) for the period following Basin and his belief that the WPLT lifeway 2000 B.P. find no counterparts in the Alvord was pre-Archaic, i.e., dependent on a few Desert model or data. major food resources at the top of the food The comparison with the Elston model chain, and thus similar in basic respects to the illustrates a substantial degree of similarity Paleo-Indian form of adaptation. I, on the between the cultural records of the Alvord other hand, see the WPLT as the earhest Basin and the western Great Basin, and a Archaic adaptive mode, even in the absence of convergence of views on the explanatory food-grinding tools. We differ also in our mechanisms responsible for the cultural interpretations of population density; Elston change that occurred. Perhaps most import­ (1982: 192) sees a very low density for WPLT antly, there is agreement that the Full Archaic peoples, while I interpret the large numbers lifeway did not evolve simply as a response to and spatial concentrations of points from the climatic amelioration fohowing the Altither­ period in the Alvord Basin as evidence of a mal, but rather was initiated during the arid fairly large population. Large numbers of mid-postglacial interval when seed-grinding early points in the Black Rock Desert, Nevada tools became habitually used. (Clewlow 1968), south of the Alvord Basin In summary, the projectile point distribu­ support this idea as well. It would actually tions recorded in the Alvord Basin have led 86 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY this paper to discussion of patterns and mode was one of fully nomadic foragers proposals that can be divided into three major specializing in large herd animals. The WPLT categories, ranging from the data-oriented to (9000-5000 B.C.) mode was characterized by the theoretical: collectors concentrating on the wet bottom­ 1. The distributional analysis of horizon­ lands, depending to a great extent on water­ tal and vertical patterns, using both co­ fowl and the remaining species of large herd occurrence analysis and graphical display, animals, as well as an undetermined variety of divides the projectile point population into other wetland resources. The Transitional groups that appear to show that point distri­ Archaic (5000-3000 B.C.) mode was an initial butions, and by inference land-use patterns, attempt to adapt to a changed environment changed significantly in the Alvord Basin without large herd animals and substantial during the documented span of human occu­ low-elevation wetland resources by shifting pation. The principal groups are (a) the oldest the extractive focus to upland areas and plant group, including fluted points and those foods as well as to the increased hunting of associated with the WPLT, (b) a middle solitary game mammals and to the regular use group, including large side-notched, Hum­ of milling stones. The Full Archaic (3000 boldt series and Gatecliff Split-stem or Pinto B.C.-contact) mode was reached when plant points, and (c) the most recent group, includ­ foods had become the staple of the diet, ing the Elko series, Rosegate series, and small facilitating wintertime food storage as well as side-notched as well as other recent point fixed winter settlements in the lowlands that styles. were the hub of a centrally based collection 2. The distributional data combined with strategy with seasonal camps through the the assumed ages of the point styles show a extractive territory. distinct trend from concentrated lowland use, A very important part of this evolutionary especially around the southern margins of the reconstruction is the position that the Full early Holocene pluvial lake, to a much more Archaic lifeway did not evolve as a response widespread and dispersed pattern. This shift is to climatic amelioration following the Alti­ particularly marked in those styles repre­ thermal, but instead was brought about dur­ senting post-mid-Holocene time, and is sug­ ing the arid mid-postglacial interval when gested to reflect a growing eclecticism in seed-grinding tools became habitually used. If subsistence and horizontal and vertical expan­ this is true, then we should expect to find sion of extractive territories as a result of the evidence of FuU Archaic culture that existed disappearance of both large herd animals and before the onset of anything resembling Neo­ the formerly extensive lakes and marshes. The pluvial conditions. distribution of fluted points coincides with The distributional patterns of Alvord Ba­ that of WPLT tools, suggesting that the sin projectile points and the explanatory fluted-point hunters adopted a WPLT-like model they have helped to generate represent way of life very soon after their appearance in an important step, but only a step, along the the region. No clear evidence of the late way of explicating the course of cultural Numic expansion is seen at this level of change in the Alvord Basin and adjacent analysis. regions. Interpretations will no doubt be 3. Four adaptive modes are proposed to influenced and to some degree changed by encompass the approximately 12,000-year re­ data others wih collect and analyze in the cord of aboriginal cultures in the Alvord years to come. Basin. The Paleo-Indian (10,000-9000 B.C.) PREHISTORIC LAND-USE IN THE ALVORD BASIN 87

NOTE 1981 sification on the Southern Columbia Plateau. North American Archaeologist 2(1): 25-52. 1. Recently reported results of the Steens Mountain Prehistory Project by Charlotte Beck Antevs, Ernst (1984) appeared too late to be included in this paper. 1948 Climatic Changes and Pre-White Man. Uni­ Once digested, these new data will no doubt enrich versity of Utah Bulletin 38(20): 168-191. the context of the data reported in this paper and put 1955 Geologic-climatic Dating in the West. Ameri­ to the test the proposals made here. can Antiquity 20(4): 317-335. ACKNOWLEDGEMENTS Baumhoff, Martin A., and J. S. Byrne 1959 Desert Side-Notched Points as a Time Mark­ My thanks go first to C. Melvin Aikens, who as er in California. Berkeley: University of Principal Investigator for the Alvord survey drafted California Archaeological Survey Reports me to lead the field crew and write the project report. 48: 32-65. I am indebted also to Tom Newman, who lent his Alvord collection for this analysis, and to Tom Baumhoff, Martin A., and R. F. Heizer Layton, Mel Aikens, and an anonymous reviewer, for 1965 Postglacial Climate and Archaeology in the their useful comments and suggestions. Lee Spencer Desert West. In: The Quaternary of the drafted the projectile point illustrations, and Joann United States; A Review Volume of the VII Brady typed the final manuscript draft as well as Congress of the International Association of earlier versions. Quaternary Research, H. E. Wright, Jr., and D. G. Frey, eds., pp. 697-708. Princeton: REFERENCES Princeton University Press. Beck, Charlotte Aikens, C. Melvin 1984 Steens Mountain Surface Archaeology: The 1970 Hogup Cave. University of Utah Anthropo­ Sites. Ph.D. dissertation, Department of An­ logical Papers 93. thropology, University of Washington. 1978 The Far West. In: Ancient Native Americans, Bedwell, Stephen F. Jesse D. Jennings, ed., pp. 131-182. San 1973 Fort Rock Basin: Prehistory and Environ­ Francisco: Freeman. ment. Eugene: University of Oregon Books. 1982 Archaeology of the Northern Great Basin: Bettinger, Robert L. An Overview. In: Man and Environment in 1975 The Surface Archaeology of Owens Valley, the Great Basin, David B. Madsen and James Eastern California: Prehistoric Man-Land Re­ F. O'Connell, eds., pp. 139-155. Society for lationships in the Great Basin. Ph.D. disserta­ American Archaeology Papers No. 2. tion, Department of Anthropology, Univer­ Aikens, C. Melvin, Donald K. Grayson, and Peter J. sity of California, Riverside. Mehringer, Jr. 1977 Aboriginal Human Ecology in Owens Valley: 1982 Final Project Report to the National Science Prehistoric Change in the Great Basin. Amer­ Foundation on the Steens Mountain Pre­ ican Antiquity 42(1): 3-17. history Project. University of Oregon De­ partment of Anthropology. Bettinger, Robert L., and Martin A. Baumhoff 1982 The Numic Spread: Great Basin Cultures in Aikens, C. Melvin, and Rick Minor Competition. American Antiquity 47(3): 1977 The Archaeology of Coffeepot Flat, South- 485-503. central Oregon. University of Oregon An­ thropological Papers 11. Binford, Lewis Alhson, Ira S. 1980 Willow Smoke and Dogs' Tails: Hunter- 1982 Geology of Pluvial Lake Chewaucan, Lake Gatherer Settlement Systems and Archaeo­ County, Oregon. Studies in Geology Number logical Site Formation. American Antiquity 11. Corvallis: Oregon State University Press. 45(1): 4-20. Ames, Kenneth M., and Alan G. MarshaU Bryan, Alan Lyle, and Ruth Gruhn 1980- Villages, Demography and Subsistence Inten- 1964 Problems Relating to the Neothermal Cli- 88 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

matic Sequence. American Antiquity 29(3): Society for American Archaeology Papers 307-315. No. 2. Butler, B. Robert Elston, Robert G. 1965 A Report on Investigations of an Early Man 1982 Good Times, Hard Times: Prehistoric Cul­ Site Near Lake Channel, Southern Idaho. ture Change in the Western Great Basin. In: Tebiwa8: 1-21. Man and Environment in the Great Basin, Campbell, Elizabeth W. C, W. H. Campbell, E. David B. Madsen and James F. O'Connell, Antevs, C. E. Amsden, J. A. Barbieri, and F. D. Bode eds., pp. 186-206. Society for American 1937 The Archaeology of Pleistocene Lake Mo­ Archaeology Papers No. 2. have: A Symposium. Southwest Museum Fagan, John Lee Papers 11. 1974 Altithermal Occupation of Spring Sites in Clewlow, C. William, Jr. the Northern Great Basin. University of 1968 Surface Archaeology of the Black Rock Oregon Anthropological Papers 6. Desert, Nevada. Berkeley: University of Cali­ Fry, G. F., and J. M. Adovasio fornia Archaeological Survey Reports 73: 1976 Human Adaptation During the Altithermal 1-94. in the Eastern Great Basin. Reno: Nevada Corliss, David W. Archaeological Survey, Research Paper 6: 1972 Neck Width of Projectile Points: An Index 68-73 of Culture Continuity and Change. Poca- Grayson, Donald K. tello: Occasional Papers of the Idaho State 1976 The Nightfire Island Avifauna and the Alti­ University Museum 29. thermal. Reno: Nevada Archaeological Sur­ Cressman, Luther S., Frank C. Baker, Paul S. Conger, vey, Research Paper 6: 75-102. Henry P. Hanson, and Robert F. Heizer Gruhn, Ruth 1942 Archaeological Researches in the Northern 1961 The Archaeology of Wilson Butte Cave, Great Basin. Carnegie Institute of Washing­ South-central Idaho. Occasional Papers of ton Publication 538. the Idaho State College Museum 6. Cressman, Luther S., David L. Cole, Wilbur A. Davis, Heizer, Robert F., and Martin A. Baumhoff Thomas M. Newman, and Daniel J. Scheans 1961 The Archaeology of Two Sites at Eastgate, 1960 Cultural Sequences at The Dalles, Oregon. Churchill County, Nevada. I. Wagon Jack American Philosophical Society Transactions Shelter. University of California Archaeo­ 50(10). logical Reports 20(4). Currey, Donald R., and Steven R. James 1970 Big Game Hunters in the Great Basin: A 1982 Paleoenvironments of the Northeastern Critical Review of the Evidence. Berkeley: Great Basin and Northeastern Basin Rim Contributions of the University of California Region: A Review of Geological and Biologi­ Archaeological Research Facility 7: 1-12. cal Evidence. In: Man and Environment in Heizer, Robert F., and C. W. Clewlow, Jr. the Great Basin, David B. Madsen and James 1968 Projectile Points from Site NV-Ch-15, F. O'Connell, eds., pp. 27-52. Society for Churchill County, Nevada. University of American Archaeology Papers No. 2. California Archaeological Survey Reports Daugherty, Richard D. 71: 59-101. 1956 The Archaeology of the Lind Coulee Site, Heizer, Robert F., and Thomas R. Hester Washington. Proceedings of the American 1978 Great Basin Projectile Points: Forms and Philosophical Society 100(3): 223-278. Chronology. Socorro: Ballena Press Pubh- cations in Archaeology, Ethnology and His­ Davis, Jonathan 0. tory 10. 1982 Bits and Pieces: The Last 35,000 Years in the Lahontan Area. In: Man and Environ­ Hester, Thomas Roy ment in the Great Basin, David B. Madsen 1973 Chronological Ordering of Great Basin Pre­ and James F. O'Connell, eds., pp. 53-75. history. Berkeley: Contributions of the Uni- PREHISTORIC LAND-USE IN THE ALVORD BASIN 89

versify of California Archaeological Research Oregon Basin. In: Man and Environment in Facility 17. the Great Basin, David B. Madsen and James Jennings, Jesse D. F. O'Connell, eds., pp. 207-226. Society for 1957 Danger Cave. University of Utah Anthropo­ American Archaeology Papers No. 2. logical Papers 27. Martin, Paul S. 1964 The Desert West. In: Prehistoric Man in the 1963 The Last 10,000 Years. Tucson: University New World, Jesse D. Jennings and Edward of Arizona. Norbeck, eds., pp. 149-174. Chicago: Uni­ Mehringer, Peter J., Jr. versity of Chicago Press. 1977 Great Basin Late Quaternary Environments Lanning, E. P. and Chronology. In: Models and Great Basin 1963 Archaeology of the Rose Spring Site, INY- Prehistory: A Symposium, Don D. Fowler, 372. University of California Publications in ed., pp. 113-167. Reno and Las Vegas: American Archaeology and Ethnology Desert Research Institute Publications in the 49(3): 237-336. Social Sciences 12. Layton, Thomas N. O'Connell, James F. 1970 High Rock Archaeology: An Interpretation 1975 The Prehistory of the Surprise Valley. Ra- of the Prehistory of the Northwestern Great mona: Ballena Press Anthropological Papers Basin. Ph.D. dissertation. Harvard Uni­ 4. versity. O'Connell, James F., Kevin T. Jones, and Steven R. 1972 A 12,000 Year Obsidian Hydration Record of Occupation, Abandonment, and Lithic Simms Change from the Northwestern Great Basin. 1982 Some Thoughts on Prehistoric Archaeology Tebiwa 15(2): 22-28. in the Great Basin. In: Man and Environ­ ment in the Great Basin, David B. Madsen 1979 Archaeology and Paleo-Ecology of Pluvial and James F. O'Connell, eds., pp. 227-240. Lake Parman, Northwestern Great Basin. Society for American Archaeology Papers Journal of New World Archaeology 3(3): No. 2. 41-56. Leonhardy, Frank C, and David G. Rice Pettigrew, Richard M. 1970 A Proposed Culture Typology for the Lower 1979 Archaeological Investigations at Stinking- Snake River Region, Southeastern Washing­ water Pass, Harney County, Oregon. Univer­ ton. Northwest Anthropological Research sity of Oregon Anthropological Papers 15. Notes 4(1): 1-29. Mack, Joanne M. 1981a The Ancient Chewaucanians: More on the 1975 Cultural Resources Inventory of the Poten­ Prehistoric Lake Dwellers of Lake Abert, tial Glass Buttes Geothermal Lease Area, Southeastern Oregon. Association of Oregon Lake, Harney and Deschutes Counties, Ore­ Archaeologists Occasional Papers 1: 49-67. gon. Report of the Department of Anthro­ Linn-Benton Community College, Albany, pology, University of Oregon, to the Bureau Oregon. of Land Management. 1981b A Prehistoric Culture Sequence in the Port­ land Basin of the Lower Columbia Valley. Madsen, David B. University of Oregon Anthropological Papers 1982a Great Basin Paleoenvironments: Summary 22. and Integration. In: Man and Environment in the Great Basin, David B. Madsen and 1982 Archaeological Investigations in Mill Creek James F. O'Connell, eds., pp. 102-104. Valley, Crook County, Oregon. University of Society for American Archaeology Papers Oregon Anthropological Papers 26. No. 2. 1983 Archaeological Investigations at the Wagon- 1982b Get It Where the Gettin's Good: A Variable tire Site (35HA328), Harney County, Ore­ Model of Great Basin Subsistence and Settle­ gon. Oregon State Museum of Anthropology ment Based on Data from the Eastern Survey Report 83-4. 90 JOURNAL OF CALIFORNIA AND GREAT BASIN ANTHROPOLOGY

1985 Archaeological Investigations on the East Tuohy, eds., pp. 67-76. Special Publication Shore of Lake Abert, Lake County, Oregon, of the Idaho Museum of Natural History. Vol. 1. University of Oregon Anthropo­ Pocatello. logical Papers, in preparation. Weide, David L. Rice, David G. 1976 The Altithermal As an Archaeological 1972 The Windust Phase in Lower Snake River "Non-problem" in the Great Basin. Reno: Region Prehistory. Washington State Univer­ Nevada Archaeological Survey Research sity, Laboratory of Anthropology, Report of Paper 6: 175-184. Investigations No. 50. Weide, David L., and Margaret L. Weide Sellards, EUas H. 1977 Time, Space, and Intensity in the Great 1952 Early Man in America. Austin: University of Basin Paleoecological Models. In: Models Texas Press. and Great Basin Prehistory: A Symposium, Don D. Fowler, ed., pp. 97-111. Reno and Steward, Julian H. Las Vegas: Desert Research Institute Publi­ 1938 Basin-Plateau Aboriginal Sociopolitical cations in the Social Sciences 12. Groups. Bureau of American Ethnology, Bulletin 120. Weide, Margaret Lyneis 1968 Cultural Ecology of Lakeside Adaptation in Tadlock, W. L. the Western Great Basin. Ph.D. dissertation. 1966 Certain Crescentic Stone Objects as a Time Department of Anthropology, University of Marker in the Western United States. Ameri­ California at Los Angeles.' can Antiquity 31(5): 662-675. Wilde, James Thomas, David Hurst 1984 Hunter-Gatherer Settlement and Subsistence 1981 How to Classify the Projectile Points from in the Northern Great Basin: A Reevaluation Monitor Valley, Nevada. Journal of Califor­ of Cave and Surface Collections from Uni­ nia and Great Basin Anthropology 3(1): versity of Oregon Expeditions in the 1930s. 7-43. Paper given at the annual meeting of the 1983 The Archaeology of Monitor Valley: 2. Society for American Archaeology, Port­ Gatecliff Shelter. Anthropological Papers of land, Oregon. the American Museum of Natural History Willey, Gordon R., and Philip Phillips 59(1): 1-552. 1958 Method and Theory in American Archae­ Warren, Claude N. ology. Chicogo: University of Chicago Press. 1980 Pinto Points and Problems in Mojave Desert Wormington, H. Marie Archaeology. In: Anthropological Papers in 1957 Ancient Man in North America, 4th rev. ed. Memory of Earl H. Swanson, Jr., Lucille B. Denver Museum of Natural History Popular Barten, Claude N. Warren, and Donald R. Series 7.