Paleodemographic Change on San Clemente Island During the Medieval Climatic Anomaly

HISTOItY, PROCESS, Alro TRADITION: A SVMPOSWM FOil MAroro (OW-TA

PALEODEMOGRAPHIC CHANGE ON SAN CLEMENTE ISLAND DURING THE MEDIEVAL CLIMATIC ANOMALY

ANDY¥ATSKO

Two reglonal sampling surveys on San Clemente Island show a significantly reduced frequency of dated archaeological sites for periods of drought llSSociated with the Medieval Climatic Anomaly (AD 650 to 1250). Temporal and spatial relationships among dated sites indicate changes in paleodemography in apparent response to deteriorating paleoenvironmental conditions, including decreased numbers of sites, depopulation of upland terrain, and occupation near reliable water sources. Parallels exist between San Clemente Island and other areas of western North America where evidence for abandonment and population movement correlate with the Medieval Climatic Anomaly.

Participation in the 2002 SCA symposium significant influence on prehistoric cultural honoring Mark Kowta was occasion for me to again evolution in maritime southern California. The reflect on the influence his mentoring had on my research reported here used the products of the approaches to archaeological research. As a two sequential sampling surveys to address the graduate student at Chico State in the mid~ 1970s, probable role adverse paleoclimatic conditions seminars and lab classes with Mark included a firm played in influencing the size and distribution of grounding in regional sampling and analysis. I late Holocene human populations on San made use of some of these lessons in my thesis Clemente Island, one of the southern California research there, and have found many other Channel Islands. opportunities during my now 23 years as a federal archaeologist to apply them to cultural resources management. This has been especially true since ISLAND-WIDE SAMPUNG SURVEY 1984, when I came to work for the Navy and assumed management responsibility for the rich The first of the two island-wide sampling archaeological resources on San Clemente Island. surveys reported here was developed in response An example of Mark's intellectual legacy is to observations made in the mid-1980s during my reflected in two regional sampling surveys I've resurvey of areas of San Clemente Island conducted on San Clemente Island in the last previously inventoried by UCLA (McKusick and decade where the frequency of dated Warren 1959), San Diego Mesa College (Axford archaeological sites on the island is significantly 1984), and others (Zahniser 1981). These selective reduced for periods in the late Holocene resurveys assessed the adequacy of the previous associated with deteriorating paleoenvironmental surveys' recovery of site locations and the quality conditions. of the record data. In all cases, between 40% and 90% of identifiable sites had not been found by Much recent research on late Holocene the earlier surveys, with most previously recorded archaeology in the California Channel Islands has sites inadequately documented or poorly addressed questions on the influences such provenienced. conditions had on regional prehistoric popUlations. Archaeologists working in the region postulate In response to these findings, I designed a that at least two paleoenvironmental factors stratified, random cluster, probabilistic survey to changes in sea-surface temperature (Arnold 1992; gain a more representative inventory of island Arnold and Tissot 1993; Colten 1992; Glassow sites (Yatsko 1991). The island's terrain provinces 1996; Glassow et al. 1988; Walker and Snethkamp provided a logical basis for stratifying th~ sample 1984) and persistent drought conditions (Larson (Yatsko 1996) (see Figure O. Differences observed and Michaelsen 1989; Raab and Larson 1997; in archaeological site density and distribution Yatsko 2000) - separately or together exerted a among these topographically defined strata during

A.dy YaI.ko, N.,uralltesource. Office. N••y Ite/llo. Sou,hwe., fo.lr••men'al Deparlmen" So DI.,., CA 91147·5199 Pr.,eedloS5 of lhe So,'elr for Cd/for.l. ArcheoloSY. folume Iii. 1001. pp 1/19· J/9 110 PROCEEDINGS OF THE SOCIETY fOil CALIfORNIA ARCHAEOLOGY, VOL. 16, 1001

II Coeltl:lT*"'K.

fIIII1 Upland Maran. T.IT&C.

II EutMn itcarprnent

_ ...Canyon 11_0

Figure 1. San Clemente Island terrains .

• 1 , , 4 S .... ff! P""""I!

the resurveys suggested that they presented an 1988; Glassow 1999), Assuming a regional sampling unequal range of residential and subsistence of 14C dates reasonably reflects all of this variation, opportunities to the island's prehistoric residents it should be possible to perceive general (Yatsko 1989). To investigate variability across population trends in the form of shifting date these terrains, I selected a 15% sample size for this frequencies. Probabilistic Survey. Random 500-meter-square (25 ha) sample units were proportionally drawn To examine this issue, Raab and I plotted 57 from the approximately 79% of the island Probabilistic Survey dates against 93 dates from encompassed by its three predominate an "all-island" sample of dates drawn from a topographic provinces - the Coastal Terrace, variety of sources (Breschini et al. 1996; Erlandson Upland Marine Terraces, and Plateau. et al. 1998; Gallegos 1994; Raab et al. 1994, 1997) (Figure 2). Because many of the "all-island" dates In collaboration with Mark Raab and the CSU lacked sufficient documentation to be corrected Northridge Center for Public Archaeology, the and calibrated to current standards, only mean Probabilistic Survey was conducted during the fall uncalibrated 14C ages were used, plotted in one and winter of 1991-1992. The survey documented century increments. 1,143 sites within the selected 73 twenty-five hectare sample units. The project also tested a The configurations of the frequency typologically stratified, random sample of these distributions shown in Figure 2 for the sites, recovering 14C samples from the 27 selected Probabilistic Survey and island-wide samples are that yielded a total of 68 radiocarbon ('4C) dates strikingly similar for the time range from about (Yatsko and Raab 1997). 5,000 RYBP to the present. Most important to this discussion are where the consistently high California researchers have suggested that 14C frequencies of dates in the Probabilistic Survey -date frequencies, like those from the and all-island samples after about 2,000 RYBP are Probabilistic Survey data, can be a reasonable punctuated by a dramatic and synchronous, drop proxy measure of general paleodemographic to zero at 800 RYBP. This decline is significant patterns, or at least a basis for hypothesizing about because it occurs in the midst of an overall trend such patterns (Breschini et al. 1996; Glassow et al. that is otherwise characterized by increasing date HISTOJ/Y, PJ/OCESS, AND TRADITION: A SYJIPOSIU.I FOR MAKaTO KOWTA 111

All Island Sample 4

Probabilistic Sample

10 20 30 40 50 60 70 so Number of C14 Dates

Figure 2. Frequency distributions of San Clemente Island radiocarbon dates.

frequencies. severe drought probably played a larger role than sea temperature in affecting the size and It was noted that this hiatus corresponded to distribution of marine-adapted populations in the the Middle-to-Late Period Transition between southern Channel Islands. This is partly because about A.D. 1100 and 1300 (ca. 650 850 RYBP) the southern Channel Islands' comparatively that has been linked to periods of late Holocene small, isolated precipitation catchments would paleoenvironmental stress in coastal southern have been extremely sensitive to such adverse California associated with elevated marine paleoclimatic trends. paleotemperatures (Arnold 1992; Arnold and Tissot 1993; Colten 1992; Glassow 1996). The Fundamentally, the goal of both these lines of hiatus also correlates wi th the period of research has been to understand how the widespread decreased precipitation and frequent interaction of particular paleoenvironmental prolonged drought between A.D. 650 to 1250 stresses and archaeologically detectable cultural known to climatologists as the Medieval Climatic responses affected cultural change, including Anomaly (Larson and Michaelsen 1989; Stine subsistence intensification, adoption of new 1994). Larson and Michaelsen (1989) documented technologies and settlement patterns, periods of very low rainfall and drought for coastal development of regional exchange networks and southern California between approximately A.D. long-term alterations to human health. However, 650 and 1250 with sustained droughts between the precise nature and cultural impact of forces AD. 750 and 770, AD. 980 and 1030, and A.D. like drought or elevated sea surface temperatures, 1100 and 1250, punctuated by a period of high individually or in combination, remains a matter rainfall between AD. 800 and 980. of debate (see Arnold et al. 1997; Raab and Larson 1997). On the basis of a broad range of available paleoclimatic data (e.g., Davis 1992; Enzel et al. As postulated by Larson (1987) for the 1989; Graumlich 1993; Larson and Michaelsen Southern California Bight, San Clemente Island 1989; :Mehringer 1986; Sti ne 1994), Raab and represents an extreme case of a 'high-risk Larson (1997) have argued that regional cycles of prehistoric littoral environment. It presents a context where the "epic droughts" (Stine 1994) 112 PROCEEDINGS Of THB SOCleTr fOR CAlifORNIA ARCHAEOLOGY, VOL. 16,2002

--_... QiIf-OlII_... ---...... -...... ".. -.---

- Figure 3. San Clemente Island geology -_..... and supplemental survey study areas.

punctuating the Medieval Climatic Anomaly research at UCLA (Yatsko 2000). First, that the would have had direct negative effects on San island's pattern of settlement during the Medieval Clemente Island's marginal terrestrial ecosystems Climatic Anomaly would have involved a reduced by reducing both available water sources and number of residential localities increasingly primary production of the island's already limited concentrated around a small number of water terrestrial biological resources. Given these sources. Alternatively, prehistoric populations on environmental realities, prolonged drought would San Clemente Island would have migrated off certainly have caused the island's human island during the Medieval Climatic Anomaly in population significant subsistence stress. During response to worsening drought conditions, leaving periods of stress, the island's prehistoric occupants no residential localities. would have responded within an available range of "risk minimization" subsistence strategies from retreat to more productive or less stressed areas of ASUPPLEMENTAL REGIONAL SURVEY the island's environment to abandoning the island. Because some of the best indicators of To address these hypotheses, I developed a culture change are found in archaeological Supplemental Survey built on the earlier settlement patterning, how prehistoric cultures Probabilistic Survey, with an island-wide sampling deployed themselves on the landscape in strategy designed to investigate spatial and functionally differentiated modes of settlement temporal relationships between dated island sites can be highly informative (Binford 1980; Kelly and sources of available surface water. Based on 1995). As hunter-gatherer cultures organi;1:e considerations of alternative water source areas themselves to exploit vital resources, patterns of (e.g., surface-held runoff vs. groundwater) and settlement are likely to reflect related patterns of their geological associations, I defined a set of social organization, economics, and technology. hydrologically- and topographically-differentiated sampling strata. The presence or absence of fossil In 1994, I proposed testing two alternative dune deposits on the Plateau were used to hypotheses for the prehistoric San Clemente differentiate upland source areas, with Island population's response to periods of severe geologically mapped areas of the Plateau overlain medieval drought as the subject of my dissertation by these deposits constituting one sampling HISrORY, PROCESS, AND TRAOlfION: A SYMPQSIIIM FO/l MAloro KOIfl'A 113

Table 1. Supplemental Survey Study Area radiocarbon dates.

Beta NRONo. I"C Agel fllC I3C Adj.l Callb. Calendar Age" No. YBr OAF Study Area 76785 OAF·79 220±60 ·24.9 220±60 308 - 271 (290) cal AD 1642· 1679 (1660) 51146 P4-B 320±60 ·23.9 340:1::60 478·303(381) cal AD 1472 - 1647 (1569) 76777 OAF·23 350± 70 -26.6 320:1::70 466 - 294 (403) cal AD 1484 - 1656 (1547) 76780 OAF-22C 450:1::70 ·25.9 440:1::70 526 - 340 (507) cal AD 1424· 1610 (1443) 51141 P4-2 490±70 ·23.1 520±70 552 - 506 (529) cal AD 1398· 1444 (1421) 76782 OAF·32 610:1:: 70 -24.6 620±70 659·539 (587) cal AD 1291 - 1411 (1363) 76778 OAF-22A 600±80 ·24.5 610± 80 659 - 539(601) calAD 1291 ·1415 (1349 51137 P3-3/4-C 670±60 -24.8 670:1::60 664 - 557 (650) cal AD 1286 - 1393 (1300) 51139 P3-3/4·D 1080± 70 ·24.0 1100 ±70 1065 - 936 (977) calAD 885· 1014 (973) 51132 P3-3/4-C 1570±80 ·20.1 1650± 80 1683 - 1415 (1535) cal AD 267 - 535 (415) MRB Study Area 120897 MRB-8 90±50 ·26.1 70±50 130 30 (NIl)' calAD 1820 - 1920 (Nil)' 120896 MRB-61 70±60 ·24.9 70±60 135 30 (N/Ii cal AD 1815 1920 (Nil)' 120900 MRB-54 160±40 -23.6 180± 50 225 - 135 (180) cal AD 1725 1815 (1770) 120898 WMR·) 2JO:I:: 50 ·24.6 210±50 285 - 270 (280) cal AD 1655 1680 (1670) 120899 MRB-A 260:1::60 -25.9 240:1::60 3 JO 275 (290) cal AD 1640 - 1675 (1660) 120901 P17-3 270±40 ·25.0 270±40 310 - 285 (300) cal AD 1640 -1665 (1650) 51121 PI4·B 1500 ± 60 ·24.8 1500:1::60 1413 - 1315 (1354) cal AD 537 - 635 (569) 76776 P4-6 3020±70 -13.6 3200:1::80 3275 ·3472 (3398) cal BC 1522 - 1325 (1448) SA Study Area 120906 SAN-19 80:1::50 -25.3 80±50 135 - 30 (NlIi calAD 1815 • 1920 (Nil)' 120910 SAN-21 380±50 -26.4 340±50 475 - 305 (380) cal AD 1475 - 1645 (1570) 120911 SA-74 51O±60 -24.9 51O±60 545 505 (525) cal AD 1405 1445 (1425) 120912 SA-61 520±40 ·23.1 550±40 550 525 (540) cal AD 1400· 1425 (1410) 120902 SAN-7 620± 70 ·24.6 630±70 660 - 540 (595) cal AD 1290· 1410 (1355) 120909 SAN·29 880±70 ·23.3 91O±70 920 - 730 (785) cal AD 1030· 1220 (1165) 93264 SAN· I 1 990 ± 110 ·25.2 990± 110 980 -760 (923) cal AD 970 - 1190 (1027) 120914 SA-62 1560±50 ·24.8 1560± 50 1515 1375 (1415) cal AD 435 - 575 (535) 120915 SA·59 3460±50 -21.0 3520± 50 3850 - 3705 (3755) cal BC 1900 - 1755 (1805) MP Study Area 120907 C3·051 130± 50 -26.4 100±50 140-20 (60) calAD 1810 1930 (1890) 120908 MPS·28 230± 50 ·26.4 220±50 300 - 270 (285) cal AD 1650 - 1680 (1665) 120904 C349 310± 50 ·26.5 280±50 320 - 285 (300) cal AD 1630 - 1665 (1650) 120903 C3-4 31O±50 -24.3 320±50 455 - 300 (405) cal AD 1495 1650 (1545) 120905 MPS·20 540±50 -24.0 550± 50 555 525 (540) cal AD 1395 1425 (1410) 51115 C3-68 1390±70 -25.2 1390±70 1333 1269 (1293) cal AD 617 681 (657)

Notes:

I. All from charcoal. processed by Beta Analytic. Dates are in RCYBP (radiocarbon years before present; present E 1950 A.D.) Errors represent I standard deviation (68% probability). 2. C13/CI2 ratios were calibrated relative to the PDB-I international standard. and the RCYBP ages were normalized to -25 per mil 3. Dendrocalibrated age in YBP (years before present; present = A.D. 1950), with los range, and mean intercept in parenthesis; calibrated by Calib rev. 3.0.3 (Stuiver and Reimer 1993) 4. Dendrocalibrated calendar years, with I-s range, and mean intercept in parenthesis; calibrated by Calib rev. 3.0.3 (Stuiver and Reimer 1993)

5. Nil E no intercepts. Calib rev. 3.0.3 (Stuiver and Reimer 1993) plotted no intercepts for these recent conventional dates. However, Beta Analytic's calibration analysis provided the indicated YBP & cal AD/BC range based on -s range for proximity to the dendrocalibrated curve. 114 PROCEEDINGS OF TIlt· SOc/ETI' FOR CAliFO RNItl ARCII.4EO I.OGY, VOL. 16,2002

Figure 4. Frequency curves for supplemental Survey study area and Probablaistic Survey charcoal dates.

seo 1000 1500 2000 Deudrocalibrnted CYBP to Nearest 100 years

Eolian Dr~in ..ges tH ~ t : '>0 ••• .. iH •tH I "'. . .. ttH. ··t ,: Il! ,.. : u ". Figure 5. Temporal distribution of ~. ~ . ilJ. . • Supplemental Survey and Probabalistic Survey i: : !!! '000 - - e charcoal dates by hydrologic province. e • 1 11S(t ------'" +-t--- ~ ~ ---- ~: -- - Lt p n d M'runJ Drou;ht Periou ' t I ''''' I (..n.r Lt.rron & ltI id~etse" 1989) I tr ~. ,C" . I~ "l'" •..... ·r· 1>',50 f - ~.I6>"t '" I . ,;'t.",<:." t ' 000

Up la Ild Marine Coastal Tenacs Terraces Plateau

250

500 " tt

_~ ___-J. __ ~ 750 U Figure 6. Temporal distribution of i TtlT! Supplemental Survey and Probabalistic Survey " 1000 charco al dates by topogrphic province. ~e 1250 ---- :! +t

150 0 .....nd

~ . ,f' •. h~.. I'OlM 1750 ... ·r. '-....c. ,.. ... t ... • t . n ct r t

2000 HISTORY, PROCESS, AND TRADITIO!i: A SYMPOSWM FOR MAIOrO KOIfTA 115

stratum and Plateau bedrock areas another (see et al. 1997; Yatsko and Raab 1997). All or parts of Figure 2). For the island's lower elevations (i.e., each study area had been subject to earlier site the Coastal Terrace), I defined the sampling strata surveys, including portions encompassing on the basis of their corresponding upland between one and five Probabilistic Survey sample geology. Here I assumed that the geological units. Data collection involved systematic survey character of different drainages' upland source or resurvey of the collective 9.85 kmz within all areas (Le., aeolian vs. bedrock) created different four study areas, which cumulatively contained potentials for water availability along their lower 559 cultural loci, documented as 497 courses. For example, fossil dune source areas are archaeological sites. A stratified, random sample expected to have buffered the release of of fifty-two Supplemental Survey sites was tested precipitation to the associated downstream areas for the presence of charcoal for 14C dating. Thirty of the Coastal Terrace. Alternatively, drainages three charcoal samples from 33 sites were 14C originating within bedrock uplands, where rapid dated, fourteen of which derived from the runoff releases water more quickly downstream, Probabilistic Survey research. Mean intercept ages would be more episodic and less persistent in for these dates range from cal BC 1805 to cal AD flow. 1890 (Table 1).

The samples surveyed in each of these ANALYSES hydrological strata were sized to cover portions of the Plateau and the Coastal Terrace sufficiently large to investigate the correlation of Analysis of the temporal site data presented in archaeological sites to likely ancient water sources Table 1 proceeded as a largely graphic exercise. within any individual drainage. To appropriately First, I plotted radiocarbon date frequencies by contain the scale of the Supplemental Survey, the time interval as markers of popUlation during Upland Marine Terraces topographic stratum and different periods of time. Then, after grouping its hydrology were excluded from the sample. these dates with regard to the island's topographic The design of survey areas was determined with and hydrological provinces, I evaluated their regard to average coverage of Plateau terrain geographic distribution for variability in encompassed within individual dendritic drainage settlement patterns through time. catchments originating on both fossil dune and bedrock upland geologies. Mapped extents for Figure 4 compares frequency curves for two both upland geological contexts indicated that suites of charcoal dates from the Supplemental survey areas of approximately 4.0 kmz were and Probabilistic surveys. These curves display required to cover both bedrock and eolian synchronous frequency depressions clearly drainage catchments on the Plateau (see Figure coincide with Larson and Michaelsen's (1989) 3). Because the Plateau terrain is roughly four drought periods. These chronological data define times larger than the Coastal Terrace (Yatsko a significant decline in the frequency of 14C dated 1996), appropriate proportional samples of the San Clemente Island sites between 1300 and 600 Coastal Terrace were approximately 1.0 kmz• To B.P., with dates apparently absent for the periods maximize dispersal of the Supplemental Survey's between 900 and 700 B.P. and between 1200 and sample across the island, the survey areas selected 1000 B.P. (Yatsko and Raab 1997). Along with the were drawn from different drainages in each "all-island" plot, this concurrence supports the terrain province rather than both ends of a single inference that, rather than being the product of drainage system. These criteria were used to sampling error (an opinion tendered by some define four study areas, two on the Plateau (Old observers), these San Clemente Island date curves Airfield (or OAF) and Middle Ranch-Box (MRB» likely represent actual, island-wide site and two on the Coastal Terrace (Shell-Abalone frequencies across the last 1,500 years. (SA) and Mail Point (MP) (see Figure 3). Figure 5 presents Supplemental Survey and Primary Supplemental Survey data collection Probabilistic Survey dates grouped by hydrologic occurred between 1994 and 1998, with selected province. The hiatuses shown suggest that spatial data sets also coming from work done prior to the distributions of dated sites are influenced by dissertation research, including investigations in factors related to their hydrological potentials. the OAF Study Area during 1991 and 1993 (Raab Dated sites from bedrock terrains are largely I 116 PROCEEDINGS OF THB SOCIBTY FOR CALIfORNIA ARCHAEOLOGY, VOL. 16,2002 absent for periods during the Medieval Climatic dates are present, if in very low frequency, across Anomaly, and show higher site frequencies the period of the Medieval Climatic Anomaly. outside the periods of medieval droughts. Consequently, the alternative hypothesis for total island abandonment during the overall Medieval Finally, because the regional sampling Climatic Anomaly cannot be supported. approaches used in both surveys were similarly topographically stratified, Figure 6 shows the The San Clemente Island settlement combined Supplemental Survey and Probabilistic chronology also suggests parallels with Survey charcoal dates from the two sampled archaeological sequences from other areas of topographic provinces (the Coastal Terrace and western North America where regional Plateau). Also shown are the Probabilistic Survey abandonments and major population movements charcoal dates from the Upland Marine Terraces, correlate with the Medieval Climatic Anomaly. the topographic province not sampled during the Cultural responses vary among these regions, but Supplemental Survey. There are clear each shows diachronic changes that, like those on interruptions in the distribution of dates across San Clemente Island, are best explained as the Plateau and Upland Marine Terraces responses to environmental deterioration and provinces. By contrast, we see a relatively related demographic stress (Jones et al. 1999). continuous sequence for the Coastal Terrace. The effects of the Medieval Climatic Anomaly on These latter dates are uniformly associated with aboriginal popUlations in the Mojave Desert of areas of the Coastal Terrace crossed by drainage California probably best parallel the San off upland eolian terrain. This observation Clemente Island example. Frequencies of supports a model where eolian-source catchments occupation dates for the Mojave Desert during are more likely to have available water during three periods between ca. A.D. 300-1800 show periods of drought than bedrock-source drainages. settlement patterns indicating a significantly reduced use of the desert between A.D. 800 and 1300 (Jones et al. 1999). This period matches the CONCWSIONS Medieval Climatic Anomaly defined by Stine (1994), Leavitt (1994), and others for the Great The two regional surveys presented here were Basin and Sierra Nevada. Of 84 radiocarbon-dated designed to complement one another and to avoid archaeological components spanning these 1,500 or minimize concerns of bias in the selection and years, 25 date to A.D. 300-800, 12 date to the use of 14C dates as a measure of population change Mojave Medieval Climatic Anomaly itself, and 47 (Glassow 1999). They present alternative date to the 500-year period that followed (A.D. approaches to sampling among the topographic, 1300-1800) (Jones et al. 1999). Spatial distribution ecological, and hydrological variables found on of these dated site components also shows that San Clemente Island. Used together, they can occupations between A.D. 800 and 1300 are portray the prehistoric island population's clpsely associated with major springs and oases responses to climatic stresses experienced during along the Mojave River. the Medieval Climatic Anomaly. The two surveys' cumulative chronological record provides some Viewed from the margin of the Southern empirical evidence for a punctuated decline in California Bight, the late Holocene San Clemente Island's prehistoric population paleodemography of San Clemente Island has during the Medieval Climatic Anomaly. This is wider implications for the study of prehistoric archaeologically expressed as a dramatic decline in human population dynamics in other Channel the number of sites dated during the period. Islands settings and across the region as a whole. Concurrent with this decline is a significanJ San Clemente Island's well-preserved geographic shift in the pattern of residential archaeological record offers a potentially finer settlement indicating a depopulation of San resolution of the influence of paleoenvironmental Clemente Island's upland bedrock terrains. stress on prehistoric regional cultures than may be Furthermore, fewer sites were occupied on lower currently available from some other Channel elevation terrain, usually in association with Islands. This is certainly the case when compared eolian-source hydrologic provinces. This to the more disturbed coastal mainland context. population decline appears not to have reached The ongoing study of these issues on San the point of island abandonment. Radiocarbon Clemente Island will continue to improve our HISTORY, PROCESS, AND TRADITION: A SYM'OSWM 10R MAlOTO KOIITA 117 understanding of these processes on the other Davis, O. K. Channel Islands and the southern California 1992 Rapid climatic change in coastal mainland. Southern California inferred from pollen analysis of San Joaquin Marsh. Quarternary Research 37:89-100. REFERENCES CITED

Enzel, Y., D. R. Cayan, R. Y. Anderson, and S. G. Arnold, J. E. Wells 1992 Complex hunter-gatherer-fishers of 1989 Atmospheric circulation during prehistoric California: Chiefs, Holocene lake stands in the Mojave specialists, and maritime adaptations Desert: Evidence of regional climate of the Channel Islands. American change. Nature 341:44-47. Antiquity 57(1 ):60-84. Erlandson, J. M, L. Mark Raab and A. Yatsko Arnold, J. E., R. H. Colten, and S. Pletka 1998 Maritime origins on California's Channel 1997 Contexts of cultural change in insular Islands: A comparative study of early California. American Antiquity 62:300 human occupations at Daisy Cave and 318. Eel Point. Proposal to the National Science Foundation, NSF Grant Number SBR-9730749. Grant proposal on file, Arnold, J. E., and B. N. Tissot Natural Resources Office, Navy Region 1993 Measurement of significant marine Southwest, San Diego. paleotemperature variation using bJack abalone shells from prehistoric middens. Quarternary Research 39:390 Gallegos, D. R. 394. 1994 The archaeological investigation ofCA SCLI-847, San Clemente Island, California. D. R. Gallegos, editor, Axford, M. L. 1984 Four years of archaeological Gallegos and Associates. Report prepared for Natural Resources Office, investigations on San Clemente Island, Navy Region Southwest, San Diego. California. Manuscript on file, Natural Resources Office, Navy Region Southwest, San Diego. Glassow, M. A. 1996 Purisimeno Chumash prehistory: Binford, L. R. Maritime adaptations along the 1980 Willow smoke and dogs' tails: hunter Southern California coast. Harcourt Brace, Fort Worth. gatherer settlement systems and archaeological site formation. American 1999 Measurement ofpopulation growth and Antiquity 45: 4-20. decline during California prehistory. Journal of California and Great Basin Anthropology 21(1). Breschini, G. S., T. Harversat, and J. Erlandson 1996 California radiocarbon dates. 7th Edition. Coyote Press, Salinas. Glassow, M. A., L. R. Wilcoxson, and J. Erlandson 1988 Cultural and environmental change Colten, R. H. during the early period ofSanta Barbara 1992 Preliminary analysis of faunal remains Channel prehistory. In The archaeology from four sites on Santa Cruz Island. In of prehistoric coastlines, G. N. Bailey Preceedings ofthe Society for California and J. Parkington, editors, pp. 64-77. Archaeology, Vol. 5, edited by M. D. Cambridge University Press .. Rosen, L. Christenson and D. Laylander, pp. 247-267. San Diego, California. 118 PIIOCBEDINGS OF THE SoCIETY FOil CALIfOrNIA ARCHAEOLOOY, VOL. 16, 2002

Graumlich, L. J. Mehringer, P. J. 1993 A 1000-year record of temperature and 1986 Prehistoric environments. In Handbook precipitation in the Sierra Nevada. of North American Indians, Vol. 11, Quaternary Research 39:249-255. edited by W. L. D'Azevedo, pp. 45-50. Smithsonian Institution, Washington. Jones, T. L., G. M. Brown, L. M. Raab, J. L. McVickar, W. G. Spaulding, D. J. Kennett, A. York Raab, L. M., K. Bradford, and A. Yatsko and P. L. Walker 1994 Advances in southern Channel Islands 1999 Environmental imperatives recon archaeology: 1983-1993. Journal of sidered: Demographic crises in western California and Great Basin North America during the Medieval Anthropology 16(2):243-270. Climatic Anomaly. Current Anthro pology 40(2):137-170. Raab, L. M., W. J. Howard, and A. Yatsko 1997 Report of the Old Airfield (OAF) Kelly, R. L. archaeological site evaluation project, 1995 The foraging spectrum: Diversity in San Clemente Island. Report on file, hunter-gatherer lifeways. Smithsonian Natural Resources Office, Navy Region Institution Press, Washington, D.C. Southwest, San Diego.

Larson, D. O. Raab, L. M., and D. O. Larson 1987 An economic analysis of the differential 1997 Medieval Climatic Anomaly and effects of population growth and punctuated cultural evolution in coastal climatic variability among hunters and Southern California. American gatherers and food producers. Ph.D. Antiquity 62:319-336. dissertation, Department of Anthropology, University of California, Santa Barbara. Stine, S. 1994 Extreme and persistent drought in California and Patagonia during Larson, D.O., and J. Michaelsen Medieval time. Nature 369:546-549. 1989 Climatic variability: A compounding factor causing culture change among prehistoric coastal populations. Stuiver, M., and P. J. Reimer Unpublished manuscript on file. 1993 Extended 14C data base and revised Department of Anthropology, Calib 3.0 14C calibration program. California State University, Long Beach Radiocarbon 35:215-230.

Leavitt, S. W. Walker, P. L., and P. Snethcamp 1994 Major wet interval in White Mountains 1984 Archaeological investigations on San medieval warm period evidenced in 13C Miguel Island: Archaeological and bristlecone pine tree rings. Climate physical anthropological research, Vols. Change 26:299-308. I and II. National Park Service, San Francisco.

McKusick, M. B., and C. N. Warren Yatsko, A. 1959 Introduction to San Clemente Island archaeology. UCLA Archaeological 1991 Project design/scope of work for the Survey Annual Report 1: 106-183. Legacy Program demonstration project: A probabilistic archaeological site survey, NALF San Clemente Island, California. Manuscript of file, Natural Resources Office, Navy Region Southwest, San Diego. HISTORY, PROCESS, AND TilADlTION: ASYMPOSIUM fOR MAKOTO KOII'TA 119

1996 Terrain provinces at San Clemente Island for the analysis of archaeological site distributions. Manuscript on file, Natural Resources Office, Navy Region Southwest, San Diego. 2000 Late Holocene paleoclimatic stress and prehistoric human occupation on San Clemente Island. Unpublished Ph.D. dissertation, Department of Anthropology, University ofCalifornia, Los Angeles.

Yatsko, A., and L. M. Raab 1997 Draft final report - Legacy Resource Management Program demonstration project No. 83, A probabilistic archaeological site survey, NALF San Clemente Island, California. Manuscript on file, Natural Resources Office, Navy Region Southwest, San Diego.

Zahniser, J., editor 1981 The cultural resources ofSan Clemente Island. Report on file, Natural Resources Office, Navy Region Southwest, San Diego. Prepared by Chambers Consultants and Planners.