Journal of California and Great Basin Anthropology | Vol. 29, No. 2 (2009) | pp. 163–181 The Mitochondrial DNA Affinities of the Prehistoric People of San Clemente Island: An Analysis of Ancient DNA AMIEE B. POTTER Anthropology Department, Virginia Commonwealth University, 312 N. Shafer St., Richmond, VA P. SCOTT WHITE Biosciences Division, Los Alamos National Laboratory, Los Alamos, NM Native American mitochondrial DNA belongs to one of five haplogroups defined by lineage-specific markers. Haplogroup frequency distribution is non-random among cultures. At historical contact, the Gabrielino occupied the southernmost California Channel Islands and the adjacent Los Angeles Basin. The Chumash thrived on the northern Channel Islands and Santa Barbara mainland. The Gabrielino were linguistically, culturally, and possibly genetically distinct from the Chumash. Haplogroup frequencies were determined for two prehistoric populations from San Clemente Island (Eel Point and the Nursery Site) to investigate Uto-Aztecan migration onto the southern Channel Islands. Analysis included three measures of genetic distance and phylogenetic analysis, as well as Fisher’s exact tests. The Eel Point and Nursery Site frequency distributions were compared to one another, and to extant Uto-Aztecan and Great Basin/California populations. Results suggest that the prehistoric occupants of Eel Point and the Nursery Site were not closely related to one another or to the Chumash. t the time of historic contact, Southern ecosystems (Arnold 1992; Erlandson 1997; Masters A California was occupied by numerous distinct and Gallegos 1997; Raab 1997; Walker 1986). Despite cultures that spoke various languages within the Hokan, geographic proximity and similar cultural ecologies Penutian, and Uto-Aztecan stocks. Two California coastal at the time of European contact, the Chumash and groups are the central focus of the following discussion Gabrielino spoke mutually unintelligible languages and research—the Chumash and the Gabrielino. The and had distinct and separate cultural features in their Chumash spoke languages once thought to belong to religious and sociopolitical organization. Additionally, the Hokan language stock; however, it is now accepted although ethnographic data clearly describe active trade that Chumash is an ancient linguistic isolate (Mithun and intermarriage between the Chumash and Gabrielino, 1999). The Gabrielino spoke a Takic language within the there also existed trade networks and interaction spheres Northern Uto-Aztecan stock. for both groups that recognized separate cultural The ethnographically-described Chumash and boundaries (Byrd and Raab 2007; Raab 1997; Raab et al. Gabrielino thrived in the Santa Barbara and Los Angeles 1994; Vellanoweth 1995). These features, in conjunction Basin regions, as well as on the northern and southern with the limited skeletal data, suggest cultural and genetic California Channel Islands, respectively. Both groups distinctions between the Chumash and Gabrielino, had well developed maritime economies with similar as well as population movements in conjunction with technologies and subsistence patterns. Their resource language movements in the Los Angeles Basin, on the utilization was necessarily similar, as the environmental southern Channel Islands, and in the Great Basin. contexts for both groups were comparable, although Kroeber (1925) proposed that linguistic and there were slight differences in aridity and prevailing archaeological evidence supported a model of early 163 164 Journal of California and Great Basin Anthropology | Vol. 29, No. 2 (2009) Holocene settlement and occupation of southern and and Tubatulabalic (Koerper 1979; Lamb 1958; Moratto central California by homogenous Hokan-speaking 1984; Sutton 1988). At the time of historic contact, Takic populations (Kroeber 1925; Moratto 1984). By A.D. peoples occupied the Los Angeles coastal plain, the 1750, the Chumash, once thought to speak a language southern Channel Islands, and Orange and San Diego within the Hokan stock, occupied the larger Santa counties, while Numic speakers inhabited the Great Barbara Channel area. In the San Diego area, from Basin (Koerper 1979; Kroeber 1925; Miller 1991; Moratto the coast to the Colorado River, there were Yuman 1984; Raab and Yatsko 1992; Titus 1987). Gabrielino was tribes speaking languages that fell within the Hokan the name given by the Spanish to the Takic-speaking superfamily. The area between these two regions was populations inhabiting the San Gabriel Mission area, the occupied by Northern Uto-Aztecan peoples (Fig. 1). This Los Angeles Basin, and the southern Channel Islands at area included the Los Angeles Basin, Orange County, the time of historic contact. the Mohave Desert, the Owens Valley, Kern County, the The California Channel Islands lie off the coast of southern San Joaquin Valley, and the Great Basin. the Santa Barbara mainland and the Los Angeles basin. Within the Northern Uto-Aztecan language family, Populations inhabiting these islands shared cultural four linguistic subdivisions existed: Takic, Hopic, Numic, features and genetic ties with the mainland populations. Mono T Shoshone ubatulabal Y Esselen Southern okuts Valley Salinan Yokuts Chemehuevi Kawaiisu Kitanemuk Chumash Vanyume Tatavium Mohave Chumash Serrano Northern Santa Cruz Gabrielino Halchidhoma San Miquel Anacapa Santa Rosa Cahuilla Santa Barbara Luseño Channel Islands Gabrielino Cupeño Santa Catalina Quechan San Nicolas Ipai Tipai San Clemente Southern N 0 Miles 100 0 100 Kilometers Figure 1. Geographic locations of southern California ethnolinguistic groups at time of historic contact. Map reproduced with permission of Pacific Coast Archaeological Society, adapted from Sutton (2009:32). Shaded area indicates Takic languages. ARTICLE | The Mitochondrial DNA Affinities of the Prehistoric People of San Clemente Island: An Analysis of Ancient DNA | Potter / White 165 At the time of European contact, the islands were already occupying the area. The Fremont Valley, north divided between two cultural spheres, with the Chumash of the Antelope Valley, was claimed ethnographically by occupying the northern islands (Anacapa, Santa Cruz, the Numic-speaking Kawaiisu (Fig. 1). Archaeological Santa Rosa, and San Miguel), and the Gabrielino data from the Fremont Valley are limited, but those the southern islands (Santa Catalina, San Clemente, available suggest an extended occupation with some San Nicolas, and Santa Barbara). King (1990) has antiquity (Sutton 1988). As the above review shows, the reconstructed the origins of the historic island Chumash precise patterning and time when Uto-Aztecan intrusions through cultural developments extending back into occurred remain matters of continuing debate. the early Holocene, while a more complex picture of Current models of population replacement in the population replacement is evident in the southern islands Los Angeles Basin and on the southern Channel Islands and adjacent mainland (Kroeber 1925; Miller 1991; are based on interpretations of human osteological data Moratto 1984; Raab and Yatsko 1992; Titus 1987). that demonstrate subtle shifts in skeletal morphometrics Kroeber (1925) argued that “Shoshonean” and paleopathologies (Kerr 2004; Titus 1987; Titus and (Uto-Aztecan) groups expanded from a homeland in Walker 2000). These data support a model of population the southern Great Basin and intruded into the Los replacement on San Clemente Island, California during Angeles Basin area, splitting apart and displacing the the Late Holocene. Hokan-speaking occupants by what has been called Two temporally-separated prehistoric cemetery the “Shoshonean Wedge” (Kroeber 1925; Miller 1991; populations excavated from San Clemente Island, Moratto 1984; Raab and Yatsko 1992; Titus 1987). Kroeber California, have played a prominent role in the limited envisioned this expansion as involving a series of migrating number of biological anthropology studies conducted on waves, with the initial and most significant movement the southern Channel Islands. Skeletal remains excavated occurring approximately 1,500 years ago (Koerper 1979; from these two cemeteries are considered to represent Kroeber 1925; Moratto 1984). Drover and Spain (1972), two separate populations, based on radiometric dating and however, have suggested an early date of 6,435 ± 130 geographic separation (Potter 1998; Titus 1987; Titus and years B.P. for a Uto-Aztecan occupation of the Los Walker 2000). The two sites, Eel Point and the Nursery Angeles Basin (Koerper 1979). Others have suggested Site, have been argued to both predate and postdate the Uto-Aztecan intrusions as late as A.D. 700, but Koerper Uto-Aztecan intrusion onto the island (Fig. 2). The Eel (1979) found no evidence to support a Uto-Aztecan Point site dates to the late Middle or early Late Holocene, intrusion later than 2,000 years ago at a site (ORA-119-A) while the Nursery Site dates to the Late Holocene. Burials in Orange County (Koerper 1979; Warren 1968). Moratto from the Eel Point site have been argued to have been (1984) dated the intrusion of Uto-Aztecan peoples to more closely related biologically to Hokan or Chumash between 3,000 and 2,000 years ago, and Lamb (1958) peoples, while the later individuals at the Nursery Site argued that the expansion of the Northern Uto-Aztecan were more closely related to the Uto-Aztecan Gabrielino family occurred about 3,000 years ago, when it broke (Titus 1987; Titus and Walker 2000). up into its four subdivisions. Swadesh