From Pleistocene Mariners to Complex Hunter-Gatherers: The Archaeology of the California Channel Islands

Abstract

California’s Channel Islands were home to some of the most distinctive Native American peoples along the Pacific Coast. Never connected to the mainland during the Quaternary, the Channel Islands have an impoverished terrestrial flora and fauna, but some of the richest and most productive marine environments in the Americas, including diverse kelp forest, intertidal, and offshore marine habitats. Native Americans occupied the Channel Islands for roughly 13,000 calendar years until the early nineteenth century, providing one of the longest and best preserved records of maritime hunter-gatherers in the Americas. We provide an overview and analysis of Channel Islands archaeology, from the relatively mobile peoples who colonized the islands during the Late Pleistocene to the complex hunter-gatherers documented by early Spanish explorers. Our analysis demonstrates the importance of Channel Islands archaeology for enhancing knowledge on a number of broad anthropological issues, including coastal and aquatic adaptations, seafaring, cultural complexity, trade and exchange, and ancient human impacts on the environment.

Introduction

Over the last several decades, the mild coastal climate and the lure of fame and fortune have attracted extraordinary population growth in southern California. This has resulted in a booming economy and one of the most densely populated coastal landscapes in the Americas. Most residents of California and beyond, however, are unaware of the fact that this recent wave of settlement is part of a much longer trajectory of human occupation spanning at least 13,000 years. The Native peoples who inhabited coastal California left behind a remarkable archaeological record, including some of the earliest coastal sites in the Americas and thousands of sites of shell middens and lithic scatters with rich faunal and artifact assemblages. Unfortunately, population growth and urban sprawl along the mainland coast have destroyed or disturbed much of this extraordinary archaeological record. In contrast to the mainland, the eight Channel Islands are largely undeveloped, relatively isolated, and remain a microcosm of California before widespread historical development (Fig. 1).

Fig. 1

The Channel Islands and the southern California Coast

Sometimes called a North American Galapagos, the Channel Islands contain distinct flora and fauna and one of the longest coastal archaeological sequences in the Americas. The combination of a long and continuous archaeological record and limited historical development has made the islands a focus of archaeological research for over a century. During the last 20 years, however, research on both the northern and southern island groups has greatly expanded, resulting in some of the most impressive coastal research in the Americas. Channel Islands archaeological data have been used to examine a number of broad anthropological topics, for instance, including the emergence of cultural complexity among hunter-gatherers (e.g., Arnold, 1992a,b, 1996, 2000, 2001a,b,c; Arnold, Colten, & Pletka, 1997a; Raab & Bradford, 1997; Raab & Larson, 1997), human impacts on the environment and historical ecology (Erlandson, Rick, & Vellanoweth, 2004a; Hildebrandt & Jones, 1992; Jones & Hildebrandt, 1995; Porcasi, Jones, & Raab, 2000; Rick et al., 2006), the antiquity of coastal adaptations and seafaring (Arnold, 1995; Arnold & Bernard, 2005; Cassidy, Raab, & Kononenko, 2004; Erlandson, 1994, 2001, 2002; Fagan, 2004; Rick, Erlandson, & Vellanoweth, 2001a; Salls, 1991; and others), long-term trajectories in human social organization and environmental relationships (Kennett, 2005), subsistence and environmental change (Glassow, 1993a,b, 2002, 2005a), regional exchange and interaction (Arnold, 1991; Raab & Howard, 2002; Vellanoweth, 2001), and a variety of other issues.

While the occupations of the islands share a strong maritime orientation and technology, cultural variability is also evident in archaeological and ethnographic records. Historically, the northern islands were occupied by the Island Chumash, who spoke a distinct island dialect that is part of the larger Chumashan language family (see Jones & Klar, 2005, p. 472). The more widely dispersed southern islands were inhabited by Uto-Aztecan speaking peoples (Gabrielino or Tongva). Of the two, considerably more is known about the Chumash from ethnohistoric sources. Although the two groups are linguistically and culturally distinct, research on the two island groups and adjacent mainland shows many parallels and interconnections, as well as significant differences (see McCawley, 2002).

The archaeology of the Channel Islands has been discussed in recent syntheses of the North American Pacific Coast (Lightfoot, 1993; Moss & Erlandson, 1995), and California (Arnold, Walsh, & Hollimon, 2004; Fagan, 2003) or the California Coast (Erlandson, 1994; Erlandson & Glassow, 1997; Erlandson & Jones, 2002). Despite a long history of archaeological research in the area, a broad synthesis for all eight of the Channel Islands has never been published. In this paper, we provide an overview and analysis of the trans-Holocene archaeological record of Native American occupations of the Channel Islands, beginning with the earliest evidence for human occupation and following this through the Historic period. Our emphasis is on long-term cultural trajectories on the Channel Islands, demonstrating the dynamics of more than 10,000 years of Native American coastal and maritime lifeways. Due to the volume of research in the area, our focus is on work conducted during the last 25 years and is largely intended to promote interest in the area by non-Channel Islands specialists.

Environmental background

The Channel Islands, located off the California Coast between Point Conception and San Diego, are divided into northern (Anacapa, Santa Cruz, Santa Rosa, and San Miguel) and southern (San Clemente, Santa Catalina, San Nicolas, and Santa Barbara) groups (Figs. 2 and 3). The islands were never connected to the mainland during the Quaternary, making them somewhat distinct from the adjacent mainland. They are owned and managed by a variety of federal agencies and private organizations. The Northern Channel Islands and Santa Barbara Island form Channel Islands National Park, with western Santa Cruz owned by the Nature Conservancy and San Miguel owned by the US Navy. San Nicolas and San Clemente are both administered by the US Navy and contain naval installations. Santa Catalina contains the only formal city on the islands (Avalon), but most of the island is managed by the nonprofit Catalina Island Conservancy. Other than relatively small developments, and a long history of ranching, the islands remain largely undeveloped.

Fig. 2

The Northern Channel Islands showing topographic relief and major place locations

Fig. 3

The Southern Channel Islands showing topographic relief and major place locations

Ranging in size from about 2.6 to 249 km², the Channel Islands are between about 20 and 98 km from the mainland coast (Table I). The northern islands are an extension of the Santa Monica Mountains on the mainland (Weaver, 1969, pp. 9), forming an east-west trending line along the Santa Barbara Channel. The southern islands are considerably more dispersed and isolated. All of the islands have a Mediterranean climate, with mild summers and cool, wet winters. The relatively arid climate and generally alkaline archaeological soils have promoted good preservation of most archaeological constituents, although erosion by waves, wind, and running water have impacted many sites.

Table 1 General attributes of the Channel Islands^a

The Channel Islands contain a limited terrestrial fauna and flora, lacking many animals and plants common on the mainland (Schoenherr, Feldmath, & Emerson, 1999, pp. 7–17). The largest endemic land mammals–the island fox (Urocyon littoralis) occurs as a discrete subspecies on all the islands except Anacapa and Santa Barbara, and the island spotted skunk (Spilogale gracilis) found on Santa Rosa and Santa Cruz–are each about the size of a house cat. During the Pleistocene, pygmy mammoths (Mammuthus exilis) lived on the Northern Channel Islands, but until the Historic period the islands were devoid of the herbivores, carnivores, and rodents that dominate the mainland coast. This includes a number of burrowing animals (gophers, badgers, ground squirrels, etc.) that have mixed mainland archaeological sites for millennia. Until historic times, the island deer mouse (Peromyscus maniculatus) was the only rodent to occur on all the islands, with the harvest mouse also found on Santa Cruz and Santa Catalina and occasional reports of shrews and woodrats on Santa Catalina (Schoenherr et al., 1999, pp. 189–191). The only other indigenous land mammal on the islands is the ground squirrel found only on Santa Catalina. The dearth of mammals, particularly burrowing rodents, has generally promoted high stratigraphic integrity of island archaeological sites. While there are few terrestrial mammals, the islands are home to thousands of land and sea birds, along with several pinniped species that haul out onshore.

Vegetation communities on the islands are also distinct, including a number of endemic and relict species. The islands are currently dominated by introduced grasses, with considerable effort being devoted to reestablishing native species. A number of vegetation communities are found on the islands, including coastal sage scrub, oak and pine woodland, and chaparral. The greatest ecological diversity is found on the larger and more topographically diverse islands (Santa Catalina, Santa Cruz, and Santa Rosa). Ravaged by more than a century of historical overgrazing, island soils have periodically been exposed causing widespread deflation, gullying, and scouring of the landscape (see D. Johnson, 1972, 1980).

Channel Island marine environments are exceptionally productive, with the upwelling of nutrient-rich waters supporting large populations of pinnipeds, cetaceans, seabirds, shellfish, and fishes. The Channel Islands are located on a boundary between colder currents to the north and warmer currents to the south, providing a mix of cold and warm water marine fauna. The diversity of marine fauna and limited terrestrial resources have made the island’s Native American peoples highly maritime throughout their occupation. Island ecosystems are susceptible to El Niño/La Niña cycles, periodic droughts, and other perturbations on a variety of different spatial and temporal scales. Droughts may be particularly problematic on the smaller (Anacapa, San Miguel, Santa Barbara) or more southern islands (San Nicolas, Santa Catalina, San Clemente), where sources of surface water tend to be limited, especially during the dry season between about June and September.

During the last 20,000 years, the geography of the Channel Islands has changed dramatically. Sea level was considerably lower during the Late Pleistocene and Early Holocene, making the islands larger in area and somewhat closer to the mainland (see D. Johnson, 1983; P. Porcasi, Porcasi, & ONeill, 1999). At the height of the last glacial, around 18,000 years ago, the northern islands formed one large island land mass (Santarosae), with the islands reaching their current configuration shortly after the onset of the Holocene. A number of additional islets were also located throughout the area, most of which are now submerged (P. Porcasi et al., 1999). Due to dramatic changes in sea level, much of the early archaeological record may be submerged, a topic we return to below. Kinlan, Graham, and Erlandson (2005) noted that the dramatic changes in shoreline length, island area, and reef area following the close of the Last Glacial promoted reorganization of marine habitats, particularly kelp forests that influenced Native American settlement and subsistence strategies.

History, method, and theory

The Channel Islands have a long history of archaeological research, spanning over a century. Like most areas of North America, archaeological research on the islands progressed through phases of antiquarianism and exploration, culture history, and more recent paradigms (see Baldwin, 1996 for a detailed history). Some of the first relatively well documented projects were performed by Schumacher, Eisen, de Cessac, and Bowers in the late nineteenth century (Baldwin, 1996; Benson, 1997; Glassow, 1977; Schwartz & Martz, 1992). These researchers were largely concerned with cemetery excavations and obtaining unique specimens for museums such as the Smithsonian and other facilities. Research on the islands continued through the early nineteenth century led by researchers like David and Malcolm Rogers, Glidden, Sanger, Woodward, Olson, and Van Valkenberg. While much of this research continued to focus on acquiring museum specimens, the work of these early researchers also embraced the culture historical paradigm (e.g., Rogers, 1929). Although great time depth in the record was speculated, early research on the islands was largely devoid of direct temporal chronologies and the islands were often seen as cultural oddities. An exception to this was D. B. Rogers’ (1929) cultural historic framework for the Santa Barbara Channel region, which included Oak Grove, Hunting, and Canaliño peoples, aspects of which remain in use today.

Beginning in the late 1940s, research on the Channel Islands surged. Phil Orr (1968), working primarily on Santa Rosa Island, began a long-term project aimed at determining the lifeways and antiquity of the island’s earliest inhabitants. He was among the pioneers of radiocarbon dating on the islands, and although his sequence has been significantly revised, he determined that people had been on the islands since at least 10,000 B.P. Charles Rozaire’s (1959, 1978, 1993) survey and excavation projects during the 1950s, 60s, and 70s on San Nicolas, San Miguel, Anacapa, and Santa Barbara islands were also among the pioneering research in the area. Meighan’s (1959, 2000) research on Santa Catalina, San Clemente, and other islands brought new interest in scientific archaeology, prompting researchers to explore ecological paradigms. This surge in scientific archaeology persisted through the 1960s and 70s, as a series of researchers associated with various universities and museums worked on the islands, refined cultural chronologies, and explored new theoretical paradigms (e.g., Glassow, 1977; Greenwood, 1978; Meighan, 1959; Reinman, 1962; Rozaire, 1978).

In the 1980s and 90s, archaeological research on the Channel Islands increased again, with many of these projects persisting into the present day. Glassow’s (1980, 1993a,b, 2002, 2005a) work on Santa Cruz Island began looking at long-term patterns in human subsistence and settlement. Arnold (1987, 1992a,b, 2001a) initiated surveys and excavations of Late Holocene chert quarries and villages that would form the foundation for her research on emergent cultural complexity. Meighan, Salls, Yatsko, Raab, and others began excavations at Eel Point and other San Clemente sites (see Raab & Yatsko, 1992; Salls, 1990a, 1991). Work by Schwartz, Rosenthal, Martz, and others (see Martz, 2005; Schwartz & Martz, 1992; Vellanoweth, Martz, & Schwartz, 2002a) at numerous sites on San Nicolas Island gave new perspective on the prehistory of this most distant island. Erlandson et al.’s (1996; Erlandson, Rick, Vellanoweth, & Kennett, 1999; Erlandson, Rick, Largaespada, & Vellanoweth, 2004b; Erlandson, Rick, & Batterson, 2004c; Erlandson, Braje, Rick, & Peterson, 2005a; Connolly, Erlandson, & Norris, 1995; Rick et al., 2001a) investigation of Daisy Cave and numerous other early shell middens on the northern islands set the stage for establishing the antiquity and context of early human settlement on the islands. Kennett’s (1998, 2005) research on Santa Rosa and other northern islands provided important theoretical and contextual frameworks for investigating patterns in trans-Holocene human occupations and adaptations. These and other projects established a foundation for detailed theoretical and methodological inquiries that have given the Channel Islands international anthropological notoriety.

During the last 20 years, archaeological research on the Channel Islands has become increasingly sophisticated. Many excavations are now conducted in natural stratigraphic levels, large numbers of ¹⁴C dates are available, surveys often include GPS and GIS technology, isotope analyses are performed on bone and shells, and there is a renewed interest in geoarchaeology, taphonomy, and site formation. Kennett’s (1998, 2005; Kennett & Kennett, 2000) analysis of stable isotopes in archaeological sites on the Northern Channel Islands, for instance, has proven important for analyzing site seasonality and ancient changes in sea surface temperature that may have impacted human subsistence and other aspects of society. The analysis of human remains by Ezzo, Kerr, Walker, Hawley, & Yoshida (2002), Goldberg (1993), Hollimon (1990), Kerr (2003), Lambert (1994), Titus (1987), Walker and DeNiro (1986), Walker and Erlandson (1986), and others also demonstrates the significance of these techniques for analyzing changes in human diet, health, interpersonal violence, or cultural affiliation. Arnold, Ambos, and Larson’s (1997b) use of ground penetrating radar and magnetometer work at the Prisoner’s Harbor site on Santa Cruz Island shows the potential for this work in complex shell middens. Trace element analysis of obsidian from Channel Island sites has also proven to be an important tool in documenting long distance exchange and interaction (Rick, Skinner, Erlandson, & Vellanoweth, 2001b; Scalise, 1994). The ongoing re-analysis of the Arlington Woman skeletal remains and Arlington Springs site is a multidisciplinary endeavor drawing on high precision ¹⁴C dating, CAT scans, ground penetrating radar, and other computer techniques to help document the context, chronology, and interpretation of some of the oldest human skeletal remains in the Americas (see Johnson, Stafford, Ajie, & Morris, 2002). Research on historical archaeology (Berryman, 1995; Braje, Erlandson, & Rick, 2006a), ancient human DNA (Potter, 2004), paleoethnobotany (Martin & Popper, 2001; Timbrook, 1993), radiocarbon methodology (Kennett, Ingram, Erlandson, & Walker, 1997; Rick, Vellanoweth, & Erlandson, 2005a), and direct AMS dating of key artifact types (Rick, Vellanoweth, Erlandson, & Kennett, 2002; Rick et al., 2005a; Vellanoweth, 2001) have also enhanced Channel Islands archaeology. Although underwater archaeological research on historic shipwrecks has been conducted around the islands (Morris & Lima, 1996; Russell, 2004), no systematic underwater archaeology for prehistoric remains has been performed.

These methodological advances have enhanced archaeological interpretation of a series of broad theoretical issues: 1) the antiquity of island settlement, maritime adaptations, and seafaring; 2) Holocene settlement and subsistence strategies; 3) exchange systems and interaction spheres; 4) complex hunter-gatherers and emergent cultural complexity; 5) contact period culture; and 6) human impacts on island ecosystems. These and other research goals have driven recent work and made the Channel Island archaeological record central to a series of key theoretical debates.

13,000 years of island lifeways

At the time of European contact, Native peoples on the Channel Islands led lives focused around the ocean, with complex exchange networks between the mainland and islands, sophisticated boats, fishhooks, and harpoons, and a heavy reliance on marine hunting, fishing, and foraging. Population densities were generally high for hunter-gatherers and many islanders appear to have been relatively sedentary, living in circular, semi-subterranean houses aggregated in sizeable villages. Much recent archaeological research has sought to understand when and how these developments evolved. It has become increasingly clear that most (if not all) of the islands were occupied throughout the Holocene, suggesting a long-term trajectory of cultural developments in the area. Although many of these regional developments may have been punctuated and rapid, many cultural practices also have their roots in long-term evolutionary trajectories.

In the sections that follow, we focus on these major developments, including changes in technology, subsistence, demography, social and political organization, and impacts on the environment. We rely on data from numerous key archaeological sites, usually referring to the ages of these sites in calendar years before present (cal B.P.). The ages provided for each site are generally 1 sigma calibrated age ranges provided by the original researchers and vary slightly depending on what was reported. Most dates from the northern islands and Santa Barbara Island were calibrated using Calib 4.3 (Stuiver & Reimer, 1993, 2000), applying a ΔR of 225±35 years to compensate for local upwelling (Kennett et al., 1997).

Maritime foundations: The Terminal Pleistocene and Early Holocene (13,000–7000 years ago)

California has one of the most extensive records of early coastal adaptations in the New World and many of the earliest sites are located on the Channel Islands. Early syntheses of the archaeology of the Santa Barbara Channel area identified Milling Stone (a.k.a. Oak Grove) cultures as the earliest well-defined cultural horizon on the mainland coast (Olson, 1930; Rogers, 1929; Wallace, 1955), and the lack of such sites on the Channel Islands led many to conclude that the islands were settled significantly later in time. Recent research has shown, however, that the earliest island occupations were roughly contemporary with the Folsom and possibly Clovis traditions (Erlandson et al., 1996; Johnson et al., 2002). No Clovis or Folsom points have been found on the islands, but a few isolated Paleoindian fluted points from mainland coastal or peri-coastal areas may provide cultural precursors to the first islanders. Alternatively, the Paleoindian and Paleocoastal traditions may have separate origins (Jones et al., 2002), an issue still to be resolved.

Determining when humans first reached the islands and how the earliest peoples adapted to island environments is difficult. According to Inman (1983), sea levels were roughly 120 m below present about 19,000 years ago, 50 m below present approximately 13,000 years ago, and 15 m below present about 9000 years ago. The bathymetry surrounding the Channel Islands is steep compared to many continental shelves around the world, so lateral movement of shorelines associated with sea level changes was relatively limited. The Northern Channel Islands were a single landmass (Santarosae) at the height of the Last Glacial, however, and some island shorelines moved as much as 15–20 km during the last 19,000 years (see Orr, 1968; D. Johnson, 1972; P. Porcasi et al., 1999). The inundation and erosion of former shorelines and coastal lowlands has destroyed or obscured those landforms where we might expect to find most of the evidence of early coastal occupations.

To compensate for these problems, recent work has focused on areas with relatively steep bathymetry (and limited shoreline movement), caves, and freshwater springs that may have enticed early peoples to move away from the coast long enough to leave behind archaeological evidence. If the earliest use of the Channel Islands was temporary or focused primarily on the milder and less stormy months from late spring to early fall, people often may have camped on the beach where archaeological evidence is unlikely to survive. Today, numerous sandy coves provide landing spots, shelter from the wind, easy access to freshwater, marine resources, and driftwood for fires. In this sense, upland sites may represent just a fraction of the earliest human use of the islands.

Five of the eight Channel Islands have now produced evidence for human occupation at least 8000 years ago. Santa Rosa, San Miguel, and probably Santa Cruz islands were settled by maritime peoples during the terminal Pleistocene; San Clemente Island was colonized at least 8500 years ago, and San Nicolas has produced provisional evidence for human occupation dating to as early as 8500 years ago (see Schwartz & Martz, 1992). Given the proximity of Santa Catalina Island to the mainland and San Clemente Island, and the intermediate location of Anacapa and Santa Barbara islands, it seems likely that all the Channel Islands were at least visited during the Early Holocene.

At least 44 Channel Island sites appear to be securely dated between about 12,000 and 7000 calendar years (Table 2). Most of these are from the Northern Channel Islands, where more archaeological research has been accomplished and concerted efforts have been made to identify early sites (see Orr, 1968; Erlandson, 1994; Erlandson et al., 2005a). For the Southern Channel Islands, the only well-documented early site is at Eel Point (SCLI-43) on San Clemente Island. Given the maritime character of the earliest assemblages, however, it seems likely that humans visited or settled all the islands by the Early Holocene and further research should greatly expand our knowledge of the geographic range and adaptive variability of the earliest islanders.

Table 2 Terminal Pleistocene and Early Holocene sites from the Channel Islands

Terminal Pleistocene sites

The earliest reasonably secure evidence for human occupation of the Channel Islands comes from the Arlington Springs site (SRI-173), located near a freshwater spring on the northwest coast of Santa Rosa. In 1959, Phil Orr identified a human femur eroding from a thin paleosol exposed in the canyon wall about 11.4 m below the rim of Arlington Canyon. Orr (1962) established that the bone was in situ and ¹⁴C dated two associated charcoal samples to 10,400±200 B.P. (∼12,340 cal B.P.) and 10,000±200 B.P. (∼11,440 cal B.P.). With no dates for the bones themselves, doubts about the age of “Arlington Man” persisted until Berger and Protsch (1989, pp. 59) reported a date of 10,080±810 B.P. for human bone collagen, a date consistent with Orr’s chronology. A team led by John Johnson and Don Morris has conducted additional field and laboratory work to help reconstruct the stratigraphy, chronology, and environmental context of the Arlington Springs site (Johnson et al., 2002). No further archaeological remains have been found in the terminal Pleistocene strata, but closer analysis of the human bones recovered by Orr found that Arlington “Man” was actually a woman.

A suite of AMS ¹⁴C dates for charcoal, mouse bones, and human bone extracts supports a terminal Pleistocene age for Arlington Woman. Charcoal from the bone-bearing paleosol was dated to 10,090±70 B.P. (∼11,620 cal B.P.), but mouse bone from the same soil was dated to 11,490±70 B.P. (∼13,450 cal B.P.). New dates on various bone extracts range from 6610±60 B.P. for osteocalcin to 9180±70 B.P. and 10,960±80 B.P. for decalcified collagen (Johnson et al., 2002, pp. 543). The osteocalcin date seems clearly erroneous and Johnson et al. have suggested that the age of Arlington Woman may lie between the ca. 11,000 B.P. date for human bone collagen and the 11,500 B.P. date on mouse bone. A precise calendar age for the skeleton cannot be calculated until the percentage of marine foods in the diet of Arlington Woman and an appropriate marine reservoir effect for the skeleton is established, however, which could affect the age of the skeleton by several centuries. For now, it seems most prudent to conclude that Arlington Woman died somewhere between 11,000 and 10,000 B.P. (ca. 13,000 to 11,500 cal B.P.). This uncertainty does not detract from the significance of the site, which demonstrates that maritime Paleoindians used watercraft to settle the Northern Channel Islands by the terminal Pleistocene.

Further evidence for a terminal Pleistocene occupation of the islands comes from Daisy Cave (SMI-261) on the northeast coast of San Miguel Island. With relatively steep offshore bathymetry, Daisy Cave remained relatively close to terminal Pleistocene shorelines and provided excellent shelter from the elements. Inside Daisy Cave, a single bone bead and two small chipped stone artifacts were found in a deeply buried soil dated to approximately 18,000 years ago. The interior of Daisy Cave has been heavily disturbed by previous excavators, however, and the context and chronology of these intriguing finds is not fully understood. Along the dripline of a small rockshelter just outside the cave, in contrast, a sequence of finely stratified soils accumulated since the end of the last glacial. Here the earliest secure evidence for a human presence is found in Stratum G, a low density shell midden dated to approximately 11,500 cal B.P. (Erlandson et al., 1996; Rick et al., 2001a). Other than abalone (Haliotis spp.), mussel (Mytilus californianus), and other shellfish remains, the evidence for human occupation in Stratum G is limited to a few expedient chipped stone tools and small amounts of tool-making debris. Like Arlington Woman, however, this early midden demonstrates that maritime Paleoindians colonized the northern islands by the terminal Pleistocene and that they subsisted–at least in part–on marine shellfish from rocky intertidal zones.

Early Holocene settlement

The discovery of Terminal Pleistocene archaeological remains associated with caves and springs on the Northern Channel Islands has led to additional survey of such locations on San Miguel and Santa Rosa islands, resulting in the identification of numerous Early Holocene shell middens located in caves or on upland terraces near springs (see Braje, Erlandson, & Rick, 2005, pp. 17; Erlandson, 1994; Erlandson & Rick, 2002a, Erlandson et al., 1999, 2004b,c, 2005a). Most of these middens are relatively small and inconspicuous and were not recorded by early surveys focused on larger and denser island sites.

The most extensive evidence for early settlement patterns comes from San Miguel Island, where at least three caves have produced evidence for Early Holocene occupation (Rick, Erlandson, & Vellanoweth, 2003): Daisy Cave, Cave of the Chimneys (SMI-603), and Seal Cave (SMI-604). On the northwest coast of San Miguel Island, a survey and dating program led by Erlandson has identified 11 small shell middens dated between about 10,000 and 8000 years ago (Erlandson & Rick, 2002a; Erlandson et al., 2004b,c, 2005a; Erlandson, Rick, & Peterson, 2005b). All these sites appear to have been located near freshwater springs that may have attracted early people to camp away from the coast, possibly during the dry season in summer and early fall. They range from small shell middens that probably were occupied for no more than a few days, to more substantial sites where a variety of activities appear to be represented. One of the latter, SMI-522, is dated between about 10,200 and 9000 years ago and has produced a relatively diverse array of tools associated with shellfishing, fishing, and marine hunting (Fig. 4). Another more substantial shell midden, SMI-608, is located on the south coast of San Miguel and dates to ca. 9500 years ago (Erlandson et al., 2005a). This shell midden, ranging from 10 to 40 cm thick, formed in a well-developed paleosol buried 1.5–2.0 meters below the modern surface. Five ¹⁴C dates on well-preserved marine shells suggest that the main site area was occupied between about 9600 and 9400 cal B.P., while a low density midden locus about 50 m to the west has been tentatively dated to 9750 cal B.P. (Erlandson et al., 2005a, pp. 679). Local bathymetry suggests that SMI-608 was located about 500–750 m from the contemporary shoreline, near the base of a steep escarpment where freshwater and shelter from the wind may have been more readily available.

Fig. 4

Early Holocene shell midden (SMI-522) on northwestern San Miguel Island, with deposits dated from ca. 10,000 to 9000 calendar years ago

On the northwest coast of Santa Rosa Island, at least six Early Holocene shell middens have been found (Erlandson, 1994). At SRI-6, situated on the bluffs at the mouth of Arlington Canyon, Orr (1968) described nearly 10 meters of stratified alluvium and windblown sands interdigitated with four buried middens at depths of roughly 2, 4, 5, and 7 meters. Orr reported uncorrected dates of 6820±160 and 7440±200 B.P. for the two middle middens, but never dated the others. In the 1990s, we found a shell midden exposed in the sheer sea cliff located in a steeply sloping, discontinuous deposit between 3.5 and 10 m below the surface. An abalone shell from this buried midden was dated to about 9300 cal B.P. (Erlandson et al., 1999). With sea levels roughly 20–25 m below present at the time, SRI-6 may have been located roughly 2.5 km from the coast. Another cluster of early sites is located on the southeast coast of Santa Rosa Island around the margins of the ancient Abalone Rocks Estuary (Rick, Kennett, & Erlandson, 2005b). One of the larger and denser of these estuarine sites, SMI-666 (a.k.a SRI-91-15; Erlandson, 1994, pp. 192), is dated to about 8000 years ago, covers an area about 140 m long by 50 m wide, and may have been an early base camp.

The best known Early Holocene site on Santa Cruz Island is located at Punta Arena on the southwest coast. Here, a large shell midden known as SCRI-109 caps a dune located near the base of a prominent rocky point. A perennial stream enters the sea about 200 m to the east and extensive rocky intertidal, kelp forest, and shallow reef habitats are located nearby. Most of these deposits date to the Middle Holocene, but an Early Holocene component was identified in the 1970s, when Glassow (1980, pp. 82) collected two column samples from eroding site profiles and obtained a basal date of about 7750 cal B.P.

The Southern Channel Islands have produced few Early Holocene sites so far. One of the most important island sites is located at Eel Point on San Clemente Island, situated 63 km off the San Diego Coast. Located on a point on the west coast of San Clemente, SCLI-43 contains a 3.5 m deep, stratified shell midden deposited during the Early, Middle, and Late Holocene. Dated between about 9000 and 8000 cal B.P., and encompassing an area about 30 m in diameter, the early component contains remnants of habitation structures, hearths, work areas, and extensive midden features (Cassidy et al., 2004, pp. 114).

The sites described above provide a general picture of the antiquity and nature of early human settlement on the Channel Islands. The known sites demonstrate that early maritime people used a variety of site types in both coastal and interior areas. Many of these early sites appear to have been short-term campsites or specialized shellfish processing sites, but several (Eel Point, Daisy Cave, SMI-522, SMI-608, SRI-666, etc.) appear to be the result of more substantial occupations. Although numerous early sites located along now submerged coastlines have probably been lost to sea level rise and coastal erosion, the number of known sites suggests that the Northern Channel Islands (and possibly the southern islands) were used by maritime peoples very early and relatively intensively.

Technology

At many early Channel Island sites, technologies were relatively expedient. In part, this may be due to small sample sizes or the seasonal or specialized nature of most site occupations, but the simple nature of early technologies at many sites may be due to the heavy economic emphasis on shellfish harvesting, which requires little sophisticated technology. Stone tools at most sites consist mostly of cores, flaked and battered core tools, and retouched and utilized flake tools. These are made from rock types available in the Santa Barbara Channel, with only limited evidence for long distance trade. Bifaces are relatively rare at most early island sites, most consisting of leaf-shaped or lanceolate forms similar to those from early mainland sites (Fig. 5). Several small and distinctive contracting stemmed points (“Arena” points) have been found in Northern Channel Island sites, however, and several eccentric crescents have also been discovered in recent years. Arena points probably were used as dart points to dispatch sea mammals and crescents may be transverse projectile points used in hunting waterfowl or sea birds (Erlandson 1994, p. 264). Bone and shell tools or ornaments are also relatively rare, but have been found at several Northern Channel Islands sites. Many early sites contain shell beads made from the purple Olivella snail, a precursor to the sophisticated shell bead trading network that developed later. Bone gorges, associated with thousands of fish bones at Daisy Cave, have been discovered at several Early Holocene island sites, representing some of the earliest evidence for hook-and-line fishing in the Americas (see Erlandson, 1994; Erlandson et al., 2005a, pp. 679; see Fig. 5). We know little about the weaving practices of early islanders, but almost 2,000 sea grass artifacts were recovered from Early Holocene levels at Daisy Cave and Cave of the Chimneys.

A large assemblage of Early Holocene artifacts has been found at Daisy Cave where chipped stone artifacts are common, but are dominated by debitage and expedient tools made from local raw materials. Also found were numerous small bipoints (fish gorges) made from bird and mammal bone, along with pieces of bone gorge-making debris. The assemblage contains several bifaces including an eccentric crescent, spire-removed Olivella biplicata shell beads, and a small knife fragment made of ground mussel shell. Remarkably, the artifacts also include hundreds of pieces of cordage and a few fragments of twined basketry made from sea grass (Phyllospadix sp.) (Connolly et al., 1995; Norris, 1997), probably preserved because they accumulated in seabird guano. In nearby Cave of the Chimneys, numerous well preserved sea grass knots, loops, and strands of cordage were recovered from strata dated between about 8400 and 7500 cal B.P., along with 27 spire-removed Olivella beads, a bone fish gorge, a few chipped stone artifacts (including a small obsidian flake), and chunks of asphaltum (Vellanoweth, Rick, & Erlandson, 2002b; Vellanoweth, Lambright, Erlandson, & Rick,

Fig. 5

Formal artifacts from the Channel Islands discussed in the text. Top row (left to right): chert microdrills (3), Olivella wall disk beads (2), Olivella callus beads (3), Olivella barrel beads (2), Olivella spire lopped bead (1). Second row (left to right): California mussel single-piece shell fishhooks (2), abalone single-piece shell fishhooks (3).Third row (left right): bone gorges (2), bone barbs (4). Fourth Row: chert projectile points

2003).

At SMI-608, a variety of chipped stone artifacts were recovered, including 11 bifaces, at least 35 flake tools, 7 cores or core tools, a hammer stone, and abundant tool-making debris (Erlandson et al., 2005a, pp. 679). A small, stemmed dart point with a triangular blade and barbed corners is similar to the “Arena” points found at Daisy Cave and Punta Arena. The chipped stone artifacts represent a variety of activities, from raw material procurement, to core reduction, and the manufacture, use, and maintenance of formal and expedient tools. Most of the chipped stone artifacts appear to have been made from rock types available on San Miguel Island, but several tools made from a chalcedonic chert likely came from the east end of Santa Cruz Island about 75 km away. Five worked bone artifacts (a bone gorge fragment, a small bone tool fragment, and three pieces of sawn or abraded bone) were also identified along with eight spire-removed Olivella beads, representing some of the oldest shell beads known from the Pacific Coast of North America (see Erlandson et al., 2005a, pp. 679–680; Fitzgerald, Jones, & Schroth, 2005). SMI-522, another relatively substantial midden dated between about 10,200 and 9000 cal B.P., has produced expedient core and flake technology, a serrated biface fragment, three bone gorges, and Olivella spire-removed beads that are generally consistent with tool assemblages identified at other early sites (Erlandson & Rick, 2002a).

On San Clemente Island, Early Holocene cultural deposits at Eel Point have produced evidence of two discrete occupational loci. At Locus 1, Cassidy et al. (2004, pp. 116) described “a low-lying arc of stones” that may have served as a wall or windbreak foundation, two overlapping hearths, ground-stone tools, lithic debris, and midden features. At Locus 2, a hard packed surface was identified along with associated domestic features including at least one posthole, ground stone and chipped stone artifacts, and midden debris. The early lithic assemblage includes cores, a burin, a chopper, a hammer stone, a stone anvil, a mano and a metate fragment, abraders, scrapers, flake tools, a wedge, drills, reamers, and tar applicators. Cassidy et al. (2004) suggested that many of these tools were part of a diverse toolkit used to build and maintain the relatively sophisticated watercraft needed to reach San Clemente Island.

Recent research has provided new insights into the technologies of early coastal peoples. Although early Channel Island assemblages are dominated by expedient flake or core tools, some sites provide evidence of more diverse technologies, including some of the oldest fishhooks, woven artifacts, and shell beads from the Pacific Coast of North America. The technologies represented at some key sites also hint at the existence of more diverse early subsistence strategies than previously recognized for the California Coast. The abundance of bone gorges and fish remains at Daisy Cave suggests that Paleocoastal peoples had a relatively sophisticated fishing technology, for example, and the abundance of bifaces at SMI-608 hints at a more sophisticated hunting technology and a greater reliance on sea mammal hunting than currently represented in the archaeological record.

Subsistence strategies

Subsistence data from several early Channel Island sites are consistent with general economic patterns identified for the southern California Coast, where shellfish and marine resources appear to have provided the bulk of the meat consumed by early coastal peoples (see Erlandson, 1994, pp. 260; Erlandson, Rick, Jones, & Porcasi, 2006). Shellfish are the most abundant constituent at nearly every island site for which quantitative data are available, but the dominant taxa vary from site to site. Mussels were a major resource at most sites, but black abalones dominated at SRI-6 and black turbans provided a major food source at Daisy Cave. Fish were secondary resources at most sites other than Daisy Cave, the only site where large numbers of fishing-related artifacts have also been found. Although many early studies of the Northern Channel Islands emphasized the significance of human hunting of pygmy mammoths (see Orr, 1968), no clear evidence of mammoth hunting currently exists (but see Agenbroad, 1998; Agenbroad, Johnson, Morris, & Stafford, 2005). Sea mammals were clearly taken by early islanders but there is little evidence, either faunal or technological, that they were major nutritional sources. Birds, too, appear to have been supplemental resources.

The dietary contribution of different animals varies from site to site, but marine resources clearly dominate the meat and protein intake of early islanders. For many early coastal groups, protein-rich marine resources probably were part of a dual economy in which terrestrial plant foods provided most of the calories consumed (Erlandson, 1991a). Since island plant foods were considerably less diverse and productive than along the mainland coast, marine resources must have been especially vital for early islanders. Nonetheless, plant foods may have been an important source of calories for early islanders. For example, among the charcoal recovered from the excavated sample at SRI-6 were two charred fragments of relatively large seeds. These are probably from a type of fruit seed, but were too fragmented and weathered for specific identification (Erlandson et al., 1999). During the terminal Pleistocene and Early Holocene, pine and oak forests were probably more extensive than the relict stands that persist on Santa Cruz and Santa Rosa islands today.

One of the largest Early Holocene faunal assemblages from the islands is from Daisy Cave. At least 19 shellfish taxa were identified, including California mussels, Tegula spp., black abalone, and sea urchins. The Early Holocene levels also produced roughly 27,000 fish bones from at least 19 different taxa (Rick et al., 2001a). The fish remains suggest that the Paleocoastal occupants of Daisy Cave fished in a variety of habitats, but the dominance of sheephead, sculpins, rockfish, and surfperch suggest that fishing took place primarily in kelp forest and other rocky nearshore habitats. Much smaller numbers of pinniped, sea otter, and bird bones were recovered from the Early Holocene levels and dietary reconstructions suggest that fish and shellfish dominated the meat diet of the cave occupants (Rick et al., 2001a).

The early deposits at Cave of the Chimneys also produced a rich assemblage of marine faunal remains. Rocky shore shellfish, especially mussels, black abalone, red abalone, and turban snails were the most common resource, but rockfish, surfperches, and California sheephead were also identified (Rick et al., 2001a). Excluding large numbers of tiny rodent and reptile bones, which were probably of natural origin, much smaller quantities of mammal and bird bone were recovered, including two fragments of probable deer bone that may have been transported from the mainland for bone tool-making purposes. SMI-608 produced faunal samples that are similar to Cave of the Chimneys and other early island sites, including marine shell (e.g., mussel, black abalone, owl limpet, and turban), land snail shell, and small amounts of fish (California sheephead, surfperch, and sculpin), mammal, and other bone (see Erlandson et al., 2005a, pp. 680–682). The data from Cave of the Chimneys and SMI-608 are also comparable to early assemblages at SMI-522, SMI-548, and SMI-606 (Erlandson & Rick, 2002a; Erlandson et al., 2004b,c).

On Santa Rosa Island our understanding of human subsistence is limited to data from just a few early island sites. At SRI-6 on the northwest coast, 17 shellfish taxa were identified, but were primarily from black abalone and California mussel. Vertebrate remains were much less common, but include sheephead, rockfish, and clupeids, a few undiagnostic bird and sea mammal bone fragments, and rodent and lizard bones that were probably of natural origin. Recent research has identified a unique cluster of early sites around the margins of an ancient estuary on the southeast coast of Santa Rosa. Two sites (SRI-84 and SRI-666) dated prior to 7500 years ago have been documented near this paleoestuary, both demonstrating a mixed reliance on rocky shore and estuarine shellfish, a pattern more similar to the adjacent mainland than other island assemblages (Rick et al., 2005b).

On Santa Cruz Island, the early levels at SCRI-109 produced a faunal assemblage rich in California mussels and other rocky coast shellfish remains, with smaller quantities of sea mammal, fish, and bird bones (Glassow, 1993b, 2002, 2005a). In his detailed analysis of shellfish remains, Sharp (2000, pp. 64) identified possible evidence for increasing predation pressure from the Early to Middle Holocene, including a decline in average mussel size from about 7.9 cm to 5.6 cm. Glassow recently completed a monograph on his work at SCRI-109, so further details on the faunal remains and subsistence technology should soon be available (Glassow, personal communication).

Analysis of Early Holocene faunal remains from the Eel Point site on San Clemente Island suggests a relatively heavy reliance on sea mammals such as seals, sea lions, and dolphins (Porcasi et al., 2000), but shellfish (California mussels, etc.) from rocky shorelines still appear to have dominated the diet of the earliest site occupants (Erlandson et al., 2006). Although fishing seems to have been of limited economic importance during the early occupation at Eel Point, zooarchaeological data from the site support the technological and geographic evidence for the existence of a relatively sophisticated maritime adaptation on San Clemente Island by at least 9000 years ago.

Summary

Once thought to have been occupied relatively recently (Rogers, 1929; Yesner, 1987), California’s Channel Islands are now known to have been settled by maritime Paleocoastal peoples at least 13,000 to 12,000 calendar years ago. Early island sites have produced some of the earliest evidence for maritime voyaging, shell middens, and marine fishing in the Americas, as well as the oldest fishhooks, basketry, and shell beads from the Pacific Coast of North America. These discoveries have contributed to the emergence of the coastal migration theory as a viable option for explaining the initial entrance of humans into the Americas. The antiquity and relatively large number of early sites now known from the Channel Islands also suggest that early island use was more intensive than once thought. This, in turn, raises the possibility that human impacts on insular and ecologically fragile island ecosystems may have begun earlier than previously believed possible.

Substantial progress has been made in understanding the nature of early human use of the Channel Islands, but many questions remain. When did humans first reach the various Channel Islands and when did the first relatively permanent occupation take place? Did humans contribute to the extinction of the pygmy mammoths on the northern islands and what effects did early islanders have on pinnipeds, sea otters, shellfish, and other marine species or communities? How important were plant foods in early island economies, which plants were used, and where were they obtained? What was the relationship of these early peoples with contemporary populations on the mainland or with the later islanders who developed into the Island Chumash or Tongva of historic times? To all these questions, at least for now, we have no clear answers.

Transition and expansion: Middle Holocene (7000 to 3500 years ago)

Between about 7000 and 3500 years ago, dramatic cultural and environmental changes took place on the Channel Islands. Increased evidence for permanent settlement, technological continuity and innovation, artistic elaboration, the beginnings of subsistence intensification, and the development of long-distance exchange systems all occurred during the Middle Holocene. It may also be the time when the ethnic divisions on the northern and southern islands first appeared. Some of the antecedents of social and political complexity seen later were probably first established during this time as well. Fueling these cultural changes may have been the relative abundance of food resources on the Channel Islands, seen in the formation of numerous shell middens containing a variety of shellfish and fish species and a probable increase in the use of sea mammals.

Many of these cultural developments coincided with environmental changes. The Altithermal, for instance, a 3000-year span of higher than average temperatures in western North America occurred during portions of the Middle Holocene (Antevs, 1952; Jones et al., 1999). Postglacial sea level rise slowed dramatically during the Middle Holocene, approaching levels similar to modern times. These long-term environmental factors had profound effects on Native populations living in the Americas and particularly strong consequences for Channel Island peoples. Although many intriguing cultural, historical, and ecological questions can be addressed with Middle Holocene data, this time period remains the most poorly documented and least understood for the Channel Islands (see Glassow, 1997).

We explore some of the salient issues of the Middle Holocene archaeological record of the Channel Islands, relying on data from several key sites to illustrate temporal changes and spatial variability (Table 3). The cultural developments of this time period involved changes in land use practices by the Native Islanders over time. These included when and how people settled the landscape, obtained food for subsistence, and developed innovations in utilitarian technologies concomitant with changes in land use strategies. The development of long distance exchange networks provides some of the best and earliest evidence for trade and exchange involving island peoples. In many respects, the archaeological record of the Middle Holocene appears to fit nicely between cultural developments that happened before and after, seemingly a transition from earliest settlement to the relatively complex, politically stratified, and linguistically diverse ethnicities encountered by Europeans beginning in the sixteenth century.

Settlement

During the Middle Holocene, people began to use the islands more intensively than in earlier times (Kennett, 2005, pp. 128–153). The overall distribution of sites suggests settlement focused on the most productive island habitats, although Middle Holocene sites are found across the Channel Islands, including coastlines (SNI-161), marine terraces (SCLI-1215), caves (SMI-603) and rockshelters (SMI-261), and interior hilltops (SRI-50). People took advantage of marine ecosystems that flourished after sea level stabilization. Intertidal and subtidal shellfish communities appear to have been particularly productive during the Middle Holocene due partly to changes in Holocene environments. Stable littoral cells formed with headlands providing sediments for sandy beaches and creating ideal locations for human settlement. People used the juxtaposition of sandy and rocky coasts to their advantage. Evidence in the form of large, multi-component sites attests to the importance of these areas for human settlement. Smaller sites indicative of more ephemeral occupation, like those reported for Santa Cruz (Glassow, 2005a) and San Clemente islands (Raab, 1997), suggest the intensity of site use was dependent on the availability and distribution of resources.

Sites located along the shores of headlands were particularly favorable for fishing and gathering activities in nearshore and kelp bed habitats (Glassow, 2005a). Examples of headland sites include Otter Point (SMI-481), Punta Arena (SCRI-109), Thousand Springs (SNI-11), and Eel Point (SCLI-43), all sites with evidence for prolonged and intensive use. Because of their large size and well-stratified deposits, these and other sites have been the focus of archaeological inquiry (see Bleitz-Sanberg, 1987; Glassow, 2002; Sharp, 2000; Raab, 1997). How they fit into the overall Middle Holocene settlement pattern is difficult to discern, although each contains evidence that a variety of activities took place, especially processing food remains. On Santa Rosa, some interior sites have been found on hilltops adjacent to plant resources and located in exposed areas with excellent view, possibly for defensive or territorial purposes (Kennett, 2005; Kennett & Clifford, 2004). The Celery Creek site (SNI-351), situated on the central plateau and almost directly in the middle of San Nicolas Island, suggests that Middle Holocene populations utilized interior locations while conducting many of the same activities seen at coastal sites (Martz, 1994). This site’s location adjacent to a freshwater source and with clear views of much of the island’s surface and surrounding ocean waters was an important factor in its use. In fact, radiocarbon dates suggest it was occupied for much of the Holocene (see Table 3). In general, however, human settlement during the Middle Holocene was focused around productive marine habitats, especially rocky coasts with relatively easy access to the ocean. On San Miguel and San Nicolas islands, the more heavily watered northwest coasts attracted heavy human settlement. These areas were also some of the most productive stretches of coastline.

Table 3 Selected Middle Holocene sites from the Channel Islands

On the east coast of Santa Rosa Island, the Abalone Rocks Estuary appears to have been a focus of human settlement for much of the Early and Middle Holocene, where several sites (SRI-77, SRI-191, SRI-192, SRI-667) contain dense occupations with a mix of marine and estuarine shellfish species (Rick et al., 2005b). Along the northwest coast of Santa Rosa Island a series of sites (SRI-3, 4, 5, and 6) excavated by Orr (1968) appear to contain cemeteries and/or dense concentrations of human settlement, most of which are dated to the Middle Holocene.

As the islands became a focal point for humans, land use patterns began to shift to more permanent settlements. The earliest evidence for permanent houses on the Channel Islands dates to the Middle Holocene. At the Nursery site (SCLI-1215) on San Clemente Island, four Middle Holocene house pits were excavated among a possible total of eighteen, although some of these may have been constructed and occupied later in time (Salls, Raab, & Bradford, 1993; Raab, 1997; Rigby, 1985). The four houses contained whale bone scapulae and ribs, presumably as structural members, storage pits, hearths, post molds, other household features, and were located in the lee of a fossil dune adjacent to a large shell midden. These house pits clearly illustrate that by the Middle Holocene, people settled the islands year-round and stayed in small villages with multiple households, probably representing extended families.

In addition to pit houses, indirect evidence suggests that people began to live on the islands more permanently with year-round occupation and more intensive use. For example, it is likely that people first transported island foxes (Urocyon littoralis) from the mainland to some of the Channel Islands during the Middle Holocene. A few San Miguel fox specimens may date to the Pleistocene (Guthrie, 1993), but no securely dated fox bones are known from the islands prior to 6000 years ago, with San Nicolas Island (the most remote Channel Island) receiving foxes as early as 5200 cal B.P. (Vellanoweth, 1998). If the islands were not being intensively used and permanently settled, it is doubtful that people would have invested the time and energy to capture, contain, and transport these animals across many kilometers of ocean water. Foxes may have been useful as fur-bearers, as scavengers to keep processing areas clean, as hunters to control disease-carrying mice populations, and as sources of edible meat during lean seasons or years.

Technology

In general, Middle Holocene technologies on the Channels Islands remained similar to those of the Early Holocene. However, more archaeological evidence of these technologies exists, suggesting island populations were growing and land use was intensifying. Technological advances in fishing and hunting gear appear to have occurred at the end of the Middle Holocene (Glassow, 1997; Raab, 1997). It is likely that population pressure, human induced environmental impacts, environmental or climatic perturbations, and changes in regional sociopolitical structure led to technological and cultural elaboration seen at the end of the Middle Holocene. The limited size of the islands, the discrete distribution of available resources, and the growth of human populations probably led to environmental circumscription, which forced people to adopt new technologies, settlement patterns, and subsistence pursuits at the close of this time period. Economic intensification, especially of fishing, promoted additional population growth and necessitated increases in trade, yet utilitarian technologies did not change much during the Middle Holocene.

Some archaeologists have suggested that there was a Middle Holocene “maritime optimum” on the Channel Islands; when people used existing technologies to harvest a variety of marine resources in a time of exceptional littoral productivity that required little technological innovation (Raab, 1997, pp. 32–33). The Little Harbor site on Santa Catalina Island (SCAI-17; Raab, Bradford, Porcasi, & Yatsko, 1995a), Eel Point on San Clemente Island (SCLI-43), and the Bird Blind site on San Nicolas Island (SNI-161; Vellanoweth, 1996; Vellanoweth & Erlandson, 1999) are shell middens that contain abundant evidence of intensive littoral and nearshore use by humans with little technological changes from the Early Holocene. Local quarries provided raw materials for stone tool production including contracting-stemmed, square-stemmed, and side-notched projectile points, knife blades, macro-drills, and core/flake or expedient tools. Stone projectile points were used to hunt sea mammals, while knives and simple core and flake tools were used to butcher carcasses, process flesh (muscle, fat, ligaments, and skin), and detach bones for tool-making material. Utilized flakes and expedient stone tool making debris, however, make up the vast majority of chipped stone tools found at Middle Holocene sites. Indurated sandstone was used to make abraders, perforators, bowls, anvils, net weights and other ground stone tools, items that the islanders traded later in time.

Like maritime peoples worldwide, Channel Islanders used a variety of shell and bone tools for everyday purposes. The bone gorge, made from bird and sea mammal bone, and similar to ones found in the Early Holocene, was the primary fishhook used in nearshore, kelp bed, and offshore fisheries. Sea mammal bone was also used to manufacture harpoon shanks, abalone pries, flakers, and spears, while bird bone was used to make needles, awls, and other tools for fiber-based technologies. King (1990, pp. 80) argued that during the Middle Holocene people in southern California first began to use composite bone fishhooks, but relatively little is known about this early fishing technology largely because it is unclear if bone barbs identified in archaeological contexts are from composite fishhooks, harpoons, or other technologies. Abalone and clam shells were processed into adzes, simple containers, and spoons, and mussel shells were sharpened and used as knives. It is also possible that single piece shell fishhooks, an important fishing innovation that transformed Late Holocene economies, first appeared near the end of the Middle and beginning of the Late Holocene. Most shells, however, formed the basis for an increasingly rich ornament and decorative industry that began in the Middle Holocene.

Fiber-based technologies included nets, bags, baskets, twine, and rope commonly made from local sea grass (Phyllospadix sp.), although these perishables rarely survive in the archaeological record. More perishable artifacts are known from Early Holocene than Middle Holocene sites. This somewhat puzzling pattern is probably due to the emphasis of archaeologists on early sites, especially in caves and rockshelters (see Connolly et al., 1995; Vellanoweth et al., 2003). Indirect evidence for perishable technologies, however, is relatively abundant. The earliest evidence for coiled and twined basketry is found in Middle Holocene sites on San Miguel (SMI-87, SMI-396) and San Nicolas (SNI-11, SNI-161) islands. As early as 5000 years ago, people began water-proofing baskets with asphaltum (probably water bottles), leaving behind evidence in the form of asphaltum basketry impressions and tarring pebbles (Bleitz 1991; Braje, Erlandson, & Timbrook, 2006b; Rick, 2004a; Vellanoweth, 1996). The development of asphaltum-sealed water bottles may have been a response to Middle Holocene aridity and the growing need to store and transport water from springs that were less reliable and more widely scattered (Braje et al., 2006b). A variety of beads and ornaments were also produced during the Middle Holocene, when we see the first evidence for widespread trade and exchange networks.

Trade and interaction

The development of trade networks based on the flow of basic and luxury goods in and out of the Channel Islands connected islanders to each other and mainland peoples. As permanent settlement of the islands continued during the Middle Holocene, people had to maintain social and economic relationships throughout the region. Trade networks strengthened cooperative relationships, group cohesion, marriage alliance, and economic stability. Island resources, especially marine shell, chert, soapstone, sandstone, ochre, and other minerals and pigments formed the nexus of materials used to produce utilitarian and luxury artifacts for trade. Evidence for inter-island and mainland exchange increases in Middle Holocene archaeological sites. It is during this time period that the diversity of artifact types and style increased. These networks were probably mediated by shared linguistic and cultural identity but evidence for Middle Holocene long-distance exchange seemed to, in most cases, cross ethnic boundaries, suggesting that a complex set of social, political, and economic factors were in place.

Regional exchange involved the movement of specific resources between the islands and beyond. Santa Catalina soapstone, highly valued for its workability and resistance to heat, was a valuable commodity traded widely in the region (Howard, 2002; Meighan & Rootenburg, 1957; Romani, 1982; Rosen, 1980; Wlodarski, 1979). Although soapstone is also found on the mainland, Santa Catalina soapstone is unique for the Channel Islands and was the primary source in the region. Soapstone was manufactured into beads, bowls, comales (cooking pans/griddles), effigies, and other utilitarian and ornamental items. Santa Catalina soapstone, in the form of raw material, artifacts, and manufacturing debris, has been found in Middle Holocene sites on the southern and northern islands. For instance, all three Middle Holocene components at the Bird Blind site (SNI-161) on San Nicolas Island contain soapstone artifacts, including unmodified chunks, beads, and an incised charmstone (Vellanoweth, 1996). On Santa Rosa Island, soapstone artifacts, especially beads, have been reported for Middle Holocene sites as well (Orr, 1968; Kennett, 1998). The occurrence of soapstone on Santa Catalina Island helped stimulate regional exchange networks. Northern islanders likely traded high quality cherts, absent from the southern islands, in exchange for soapstone and other goods.

Although stone was an important trade item, it was the availability and abundance of shellfish on the islands that was the backbone of exchange networks. The Channel Islands were a center for shell bead and tool production that maintained trade routes throughout the region. As a testament to the growing importance of shell artifacts, shell beads and other ornaments became more elaborate during the Middle Holocene. Shell artifacts were made from a variety of species including abalone, mussel, clam (Tivela spp., Saxidomus spp.), money tusk (Dentalium pretiosum), and particularly purple olive (Olivella biplicata). While simple spire-ground Olivella beads were typical during the Early Holocene, the Middle Holocene witnessed a florescence of new styles including rectangular, oval, and round disks made from the wall portion of this small gastropod. Rectangle beads were some of the earliest forms of wall beads produced on the Channel Islands and have been found in Middle Holocene components on all but Anacapa and Santa Barbara islands (Orr, 1968; Raab & Howard, 2002; Scalise, 1994; Vellanoweth, 2001). This stylistic explosion probably reflects an increased demand for island products from mainland groups, but it may also be an outcome of increased sedentism that resulted in island regionalization and the formation of separate cultural or ethnic identities. The archaeological record for this time period suggests that people on the mainland increasingly desired island goods, especially Olivella beads.

The Middle Holocene witnessed an expansion of long-distance exchange networks that connected people on the Channel Islands to the coastal mainland and beyond. Shell beads may have been one of the island goods traded to interior groups in exchange for obsidian (Fitzgerald et al., 2005; Rick et al., 2001b) and other commodities. The distribution of Olivella Grooved Rectangle (OGR) beads may provide some of the earliest evidence for the formation of distinct cultural interaction spheres on the Channel Islands and adjacent mainland. OGR beads, a relatively rare (<275) bead type manufactured from the wall portion of the shell, have been found in archaeological sites across the Southern Channel Islands, the adjacent mainland coast, western Nevada, and central Oregon (Howard & Raab, 1993; Jenkins & Erlandson, 1996; Vellanoweth, 1995, 2001). These beads had sawn (rather than drilled) perforations that left a distinctive groove down the middle. Direct AMS radiocarbon dating shows they were made primarily around 5000±500 cal B.P. OGR beads originated on the Southern Channel Islands and moved rapidly throughout the historically documented territory of Uto-Aztecan speaking peoples. To date, this bead type has not been found on the Northern Channel Islands and only four have been found in Santa Barbara County. That these Middle Holocene beads are found primarily in areas occupied historically by Uto-Aztecan speakers suggest that this trade was at least partially mediated by social and cultural factors–and that the ancestors of the Tongva may have reached the Southern Channel Islands by at least 5000 years ago.

Subsistence strategies

In many ways, Middle Holocene peoples on the Channel Islands had a cornucopia of subsistence resources from which to choose, and people appear to have had a relatively balanced marine diet. Littoral foraging, nearshore and kelp bed fishing, and sea mammal hunting provided a dietary nexus that was supplemented by birds, wild plant foods, and trade with mainland peoples (Glassow, 1993b; Raab, 1997; Vellanoweth & Erlandson, 1999). Although plants likely provided key nutrients to the overall diet, little archaeological evidence for this has survived. The majority of dietary reconstructions have focused on the relative contribution of marine resources (Bleitz, 1993; Erlandson, Vellanoweth, Rick, & Reid, 2005d; Glassow, 1993a; Kennett, 2005; Martz, 2005; Raab, 1992; Salls, 1988; Vellanoweth & Erlandson, 1999; Vellanoweth et al., 2002b). These reconstructions suggest island peoples adopted a relatively balanced diet, although shellfish made up the majority of animal flesh consumed. Fishing appears to have become more important for sustaining growing and more sedentary populations, a pattern that continued into the Late Holocene.

Considerable variation has been documented for Middle Holocene subsistence patterns across the Channel Islands. On the northern islands and San Nicolas, red abalone (Haliotis rufescens) middens accumulated between about 7500 and 3500 years ago (Glassow, 1993a, 2005a; Sharp, 2000; Vellanoweth, 1996; Vellanoweth & Erlandson, 1999). Red abalones appear to have favored the cooler waters off the Northern Channel Islands, but were also abundant on San Nicolas Island (Fig. 6). Similar sites accumulated on the southern islands, although warmer water abalone species were harvested (Raab, 1992; Raab, Porcasi, Bradford, & Yatsko, 1995b; Salls, 1988, 1991, 1992; Vellanoweth & Erlandson, 1999). As Sharp (2000) noted, some red abalone middens actually contain a mix of pink, red, and black abalones, and a wide variety of shellfish taxa. These data suggest that while some Middle Holocene sites may be dominated by red abalone shells, Middle Holocene shell middens on the whole are considerably more diverse.

Fig. 6

Middle Holocene red abalone midden (SNI-161) eroding out of sand dune on northwestern San Nicolas Island

Middle Holocene shell middens vary in size from large, dense sites (likely villages or residential camps) located within dunes, to small, relatively thin single-component sites that were probably temporarily or seasonally occupied. Although abalones (H. cracherodii, H. corrugata, H. fulgens, H. rufescens) are sometimes the most conspicuous constituent in these middens, other shellfish taxa typically include mussels (Mytilus californianus, Septifer bifurcatus), turban snails (Tegula spp.), owl limpets (Lottia gigantea), chitons (Cryptochiton stelleri), and other rocky shore species. In decreasing dietary importance fish, sea mammal, and bird round out the food refuse commonly found in Middle Holocene shell middens. On Santa Cruz Island, Glassow (2005a) estimated the dietary significance of excavated Middle Holocene shell middens using the weight method, finding that mussels accounted for at least 60% of all shellfish in what he called red abalone middens. Our own research at Otter Point (SMI-481) on San Miguel Island showed that in some cases red abalone made up over 70% of the faunal remains and over 90% of the edible flesh represented in the sample (Vellanoweth, Rick, Erlandson, & Reynolds, 2006). At the same time in a roughly contemporaneous component, mussel dominated the shellfish assemblage, accounting for over half of all fauna by weight. When relative dietary reconstructions were calculated, however, mussel supplied slightly under 12% of the total diet while sea mammal (∼60%) and fish (∼17%) contributed the vast majority of meat consumed. Red abalone was negligible, providing less than 1% of the meat diet in this component.

The Bird Blind site (SNI-161) provides another example of the diversity of subsistence strategies employed by islanders during the Middle Holocene. This site contains four discrete components separated by dune sand (Vellanoweth, 1996; Vellanoweth, 1998; Vellanoweth & Erlandson, 1999). The components range in age between about 5200 cal B.P. and 3000 cal B.P. In the earliest two components and the latest (Late Holocene), fish comprised roughly 50%, 70%, and 85% of the meat diet, respectively. In the third component (∼3800 cal B.P.) the dietary contribution of animals was relatively balanced between shellfish, fish, and sea mammal, with red (∼36%) and black (∼13%) abalones dominating the shellfish taxa. Because San Nicolas is the outermost and most isolated of all the Channel Islands, we interpret the heavy use of fish at this site as an indicator of earlier than usual economic intensification based on fishing, a pattern more commonly seen during the Late Holocene (Vellanoweth & Erlandson, 1999). Other Middle Holocene sites with evidence for substantial use of nearshore fisheries include Eel Point, Big Dog Cave, and the Nursery site on San Clemente, as well as Little Harbor on Santa Catalina (Salls, 1988). California sheephead were more intensively exploited by people on the southern islands, perhaps reflecting more favorable conditions in the south (Raab, 1997, pp. 27–29). These and other subsistence-related differences between the islands should be expected considering the amount of environmental variation that exists in the southern California Bight.

Other cross-channel differences in subsistence patterns include the intensity of sea mammal hunting and offshore fishing. Hunting of sea mammals, including dolphins, appears to have peaked during the Middle Holocene on some of the Channel Islands (Glassow, 2005b; Porcasi & Fujita, 2000; Porcasi et al., 2000). The best archaeological evidence for Middle Holocene sea mammal hunting, however, comes from the southern islands including sites on San Clemente (Eel Point), Santa Catalina (Little Harbor), and San Nicolas (Thousand Springs). These sites contain an abundance of sea mammal bone, including whale, dolphin, elephant seal, sea lion, and sea otter (Bleitz, 1993; Porcasi, 2002; Porcasi et al., 2000). Ocean sunfish (Mola mola), a large, generally pelagic fish, and albatross (Diomedeidae) have also been identified in Middle Holocene components at some of these same sites (Porcasi, 1999a; Porcasi & Andrews, 2001). The Abalone Rocks Estuary on Santa Rosa Island, was also used during the Middle Holocene and appears to have been a focus of shellfish collecting during the Early and Middle Holocene (Rick et al., 2005b). More work on Middle Holocene sites would likely add to our understanding of the diversity of subsistence strategies that took place during this time period, when subsistence economies seemed to have transitioned from Early Holocene patterns focused on shellfish gathering to Late Holocene emphases on fishing. Clearly, the Middle Holocene was a time of great change, as human economies across the islands were expanding and diversifying.

Summary

In many ways, the archaeological record of the Middle Holocene fits nicely between the relatively mobile peoples of the terminal Pleistocene and Early Holocene and the complex, sedentary peoples identified on most of the islands during the Late Holocene. Temporally, culturally, and environmentally it was a time of transition, as populations probably grew and people used portions of the islands more intensively. The stabilization of sea level and apparent climatic variability also characterize the Middle Holocene and probably had profound effects on cultural developments, especially new and more diversified subsistence strategies. The cultural and environmental developments of the Middle Holocene were fundamental for paving the way for the rapid and dramatic changes that followed in the Late Holocene. Although we have amassed a great deal of information on Middle Holocene cultures of the Channel Islands, numerous questions remain and in many ways this is the least studied time period on all of the islands.

Diversity and complexity: The Late Holocene (3500 years ago to AD 1820)

The last 3500 years or so witnessed some of the most rapid and pronounced cultural developments on the Channel Islands and southern California Coast (see Erlandson & Jones, 2002). During this period, many of the developments or transitions that started in the Middle Holocene reached their zenith, as people were living in larger and increasingly sedentary villages, exchange relationships greatly intensified, and social and political hierarchies also increased. On San Clemente Island, there is only limited evidence for cultural complexity and a diverse array of site types, suggesting variability in social and political organization on the Channel Islands and southern California Coast (Raab, Yatsko, Garlinghouse, Porcasi, & Bradford, 2002). The Late Holocene also culminated with the first contacts with Europeans in the sixteenth century, the removal of most islanders to mainland missions by the early nineteenth century, and the ultimate occupation of the islands by Euroamericans. This was a period of great change for Native American peoples and the natural and cultural history of the Channel Islands. As part of a larger work on Late Holocene archaeology in California, several syntheses of Late Holocene island adaptations were recently published (Kennett & Conlee, 2002; Munns & Arnold, 2002; Raab et al., 2002; Vellanoweth et al., 2002a), where interested readers may find more details on specific islands.

Many of the cultural hallmarks of historic island peoples probably first appeared in the Late Holocene. Items such as the circular shell fishhook, the plank canoe, Olivella cup beads, the bow and arrow, and toggling harpoons accompany significant social developments (Arnold, 1995; Arnold & Bernard, 1995; Erlandson & Rick, 2002b; Gamble, 2002; Glassow, 1996; Kennett, 2005; King, 1990; Munns & Arnold, 2002; Raab et al., 2002; Rick et al., 2002; Vellanoweth et al., 2002a). There were also substantial changes in subsistence, including increasing use of offshore resources like swordfish and tunas (Bernard, 2004; Rick, 2004a). As island populations greatly expanded and technologies became more sophisticated, we see some of the greatest human impacts on the environment (Erlandson et al., 2004a, 2005c; Raab et al., 1995b, 2002; Rick et al., 2006). Our knowledge of the complex archaeological record of each of these distinct topics comes to us from a series of archaeological sites that have been excavated on the eight islands (Table 4). The record is far from complete, with some of the bigger gaps on Santa Catalina Island and the smaller islands of Santa Barbara and Anacapa.

Table 4 Selected Late Holocene archaeological sites from the Channel Islands

Settlement

The changes in settlement that occurred in a few contexts in the Middle Holocene (e.g., increased sedentism and reduced mobility) reached their pinnacle during the last 3500 years when several large villages were established on many of the islands. Associated with the formation of these large, multi-family villages–many of which persisted into historic times–is a diversity of site types. Small lithic scatters and camps, a wide variety of shell middens, possible defensive sites, and the use of rockshelters are all part of the Late Holocene system of settlement and land use (Fig. 7). Our understanding of these practices has recently expanded due to survey and excavation projects conducted on Santa Cruz, San Miguel, Santa Rosa, San Clemente, and San Nicolas islands. One major bias in Late Holocene research on the northern islands, however, is a focus on large villages and cemeteries that date to the last 1500 years, with comparatively little work done on the previous half of the Late Holocene, or site types other than villages. Our understanding of Late Holocene settlement, particularly on the northern islands, is also enhanced by ethnohistoric data, which provide names and locations of historic Chumash villages, many of which were also occupied prehistorically (see Arnold, 1990; Kennett, 2005; J. Johnson, 1999a).

Fig. 7

Erlandson at dense stratified Late Holocene shell midden (SMI-525) on northwestern San Miguel Island. These deposits span roughly 3500 to 500 cal B.P (photo by Madonna Moss).

Some of the most comprehensive research on Late Holocene Chumash culture and settlement on Santa Cruz Island has been reported by Arnold (1987, 1992a, 1994, 2001a, 2004) and her colleagues, who worked at a series of villages dated from about 1500 years ago through the historic era. This research suggests that most Island Chumash were living in large, multi-family villages scattered around some of the most defensible and productive areas on the islands. These communities participated in large exchange and interaction networks with other villages on the islands and mainland. Depending on their geographic location, many of these villages focused on the production of beads and microblades. On eastern Santa Cruz Island, several interior chert quarries, used largely for microblade production, demonstrate broader patterns of human procurement and land use.

Perry’s (2003, 2004, 2005) trans-Holocene analysis of Santa Cruz Island settlement demonstrates that maritime subsistence patterns and cultural complexity increased during the Late Holocene. Perry’s (2004) work which involved 66 sites and 90 distinct temporal components also documented the use of several new chert quarries, rockshelters, and continued use of villages around ideal locations. One of the more unique sites Perry (2004) identified is SCRI-647, an interior site situated at about 1200 feet in an oak and pine woodland. Interpreted as a seasonal residential base with evidence of microblade production and use of plant resources, this is one of the few interior Late Holocene sites that has been described.

Late Holocene occupation of Santa Rosa Island is similar to those of Santa Cruz. A series of large villages (SRI-2, SRI-60, SRI-97, etc.) that were occupied into the historic period contain rich faunal and artifact assemblages and surface house features (see Kennett, 2005; Rick, 2004a; Fig. 8). Kennett (2005, pp. 169–170) describes a reduction in the use of interior residences after 3000 years ago across the northern islands, arguing that between about 3000 and 1300 years ago settlement was similar to the Middle Holocene with limited evidence for large and diverse site types and sedentism. One such site with a cemetery that may hint at sedentism is SRI-41 at Canada Verde that also has extensive Middle Holocene deposits (see Kennett, 2005, pp. 137, 170; Orr, 1968), suggesting that further research is necessary to better define its chronology.

Fig. 8

Historic and late prehistoric era Chumash village (SRI-97) on the south side of Santa Rosa Island (undulating hummocks and depressions in center of photo are the remains of semi-subterranean house pits surrounded by shell midden)

On San Miguel Island, populations appear to increase and, although there is greater settlement at large villages concentrated on the coast, a variety of site types were used, including rockshelters and interior sites (Kennett, 2005; Rick, 2004a). The settlement trend towards large villages after about 1500 years ago is apparent at a number of San Miguel sites, including SMI-163, SMI-468, SMI-470, and SMI-602. Prior to this time, large villages were generally rare or absent, although Rick’s (2002, 2004a,b) work at SMI-87 provided a picture of a large site in a prominent location with diverse faunal remains and human burials. This site probably indicates increased sedentism, but still differs from the large sites dated to after 1500 years ago. Rockshelters and lithic scatters dated to the last 3000 years, suggest that logistical sites and camps were also an important component of Late Holocene settlement systems on San Miguel (see Rick, 2004a).

On the Southern Channel Islands, there is also evidence for increased population growth and a diverse range of site types and settlement strategies. On San Clemente Island, several Late Holocene sites have been tested, including the trans-Holocene deposits at Eel Point (SCLI-43), small Tegula middens (SCLI-1319 and SCLI-1318), rockshelters (SCLI-119), and village sites. Raab et al. (2002, pp. 26) suggested that the cultural complexity apparent on some of the other islands may not have been completely manifested on San Clemente. The available data, however, suggest a diverse settlement and occupation of the island during the Late Holocene. At the Lemon Tank site (SCLI-1524), for example, there is evidence of ritual activities, including human, dog, fox, and raptor burials (Raab, Bradford, & Yatsko, 1994). The Nursery site (SCLI-1215) also provided evidence of a house pit and burials. Yatsko’s (2000) recent settlement and demographic analysis also indicates site abandonment during the Medieval Climatic Anomaly, a feature he attributed to climatic changes (e.g., drought).

On San Nicolas Island, a number of Late Holocene archaeological sites have been excavated and documented, with ¹⁴C dates from 45 components at 36 different sites (Vellanoweth et al., 2002a, pp. 87). As Vellanoweth et al. (2002a) and Martz (2005) noted, Late Holocene peoples appear to have used all parts of the island, a pattern that differs from the focus on the northwest coast during the Middle Holocene. These include numerous sites on the island’s upper plateau, suggesting that Late Holocene settlement may have involved movement to secondary areas. Sites such as Thousand Springs (SNI-11) suggest occupations spanning much of the Late Holocene (and earlier), and many of the larger shell middens and residential sites appear to be located in dunes (e.g., SNI-168, SNI-60). Similar to the other islands, San Nicolas’ population may have reached its zenith around 1300 years ago. Recent work at SNI-25, a village occupied during late prehistoric and historic times, documented numerous domestic features, hearths, ritual dog and fox burials, and diverse faunal and artifact assemblages, illustrating the relatively sedentary and complex lifestyles of the peoples of San Nicolas.

Although Santa Catalina Island is the most prominent of the southern islands, it has seen relatively limited archaeological research, with most of this work concentrated on soapstone quarries and associated sites (e.g., Howard, 2002; Rosen, 1980; Williams & Rosenthal, 1993). Consequently, many of the sites that have seen research (SCAI-26, SCAI-77, and SCAI-137) appear to be Late period or Protohistoric sites associated with soapstone procurement (see Howard, 2002). A few shell middens have also been excavated, such as SCAI-137, but our knowledge of the lifeways and settlement on Catalina is limited.

The small islands of Santa Barbara and Anacapa have also seen limited work. On Santa Barbara Island at least five sites dated to the Late Holocene have been identified and most of these appear to be camps (see Erlandson, Glassow, Rozaire, & Morris, 1992; Rick, 2001; Rick & Erlandson, 2001). Anacapa also contains about six shell middens and a rockshelter (ANI-8) that appear to date to the Late Holocene (Rick, 2006; Rozaire, 1978). Rozaire’s excavation of ANI-6 and ANI-8 on West Anacapa produced microblades that suggest that Anacapa Islanders were involved in broader exchange systems or that this island was used sporadically by peoples based on adjacent Santa Cruz Island. The presence of chert deposits (see Schoenherr et al., 1999) on West Anacapa suggest that people may have used this source for stone tool production, although much of it appears to be of low quality.

Technology and exchange

Early and Middle Holocene archaeological data suggest a great deal of continuity in the types of technologies that people used. Continuity is also present in the Late Holocene archaeological record, but many new technologies that represent improved subsistence strategies, exchange, or other activities first appear to have been used during the last 3500 years. One of the most controversial and heavily debated technological innovations is the use of the tomol or tiat, a plank boat that is seaworthy and relatively efficient (see Arnold, 1995; Arnold & Bernard, 2005; Davenport, Johnson, & Timbrook, 1993; Fagan, 2004; Gamble, 2002). These boats were a hallmark of historically documented peoples in the region. Seaworthy craft were obviously available since the terminal Pleistocene (Erlandson, 2002), but plank canoes may have been necessary for regular journeys from the islands to the mainland, to carry more people or larger loads, and to take swordfish and other large pelagic marine species (see Arnold & Bernard, 2005; Gamble, 2002). They were also part of the Chumash system of wealth, power, and prestige. Recent research has focused on the origins and antiquity of plank boats, with some suggesting that plank boats or some other similar boat have been around since the Early Holocene (Cassidy et al., 2004; Fagan, 2004) and others arguing that they probably appeared sometime during the last 2000 to 1500 years (Arnold, 1995; Arnold & Bernard, 2005; Davenport et al., 1993; Gamble, 2002). Jones and Klar (2005) recently suggested that plank boats (and other artifacts) may have been introduced to the Chumash and Tongva by Polynesians around 1500 years ago. A central problem in this debate is that boats or their parts (e.g., planks, canoe drills) rarely preserve or are difficult to identify in the archaeological record. The oldest identified plank in southern California, however, comes from Daisy Cave on San Miguel Island and was dated to ca. 1300 cal B.P. (see Gamble, 2002, pp. 308). Several other boat making parts have also been identified in island settings. Current evidence suggests that large plank boats may have first appeared during the Late Holocene, but islanders had efficient boats since their colonization.

A number of other technologies also first appeared or became widespread in the Late Holocene. New or refined subsistence technologies include the single-piece shell fishhook, toggling harpoons, composite bone fishhooks, and a variety of arrow points (see Fig. 5). The single-piece fishhook significantly improved fishing capabilities in the area and some of the oldest specimens in the region come from the Channel Islands (see Rick et al., 2002; Strudwick, 1986). At the Eel Point site on San Clemente Island, several fishhooks and blanks may date to about 3200 cal B.P., while others may be roughly 4000 years old (see Raab et al., 1995b, pp. 14), but none of these have been directly AMS dated, leaving questions about their precise age. Vellanoweth and Erlandson (1999) also reported an associated date for single-piece fishhooks of roughly 3000 cal B.P from SNI-161 on San Nicolas Island. The oldest directly dated hooks currently come from San Miguel Island where a hook from SMI-87 was dated to 2500 cal B.P. (Rick et al., 2002) and another hook from SMI-152 recently dated to about the same age. These island dates for fishhooks are roughly comparable with early dates from the mainland (see Koerper, Prior, Taylor, & Gibson, 1995; Rick et al., 2002), suggesting their use around 3000 to 2500 cal B.P. or earlier. After about 2500 years ago, circular shell fishhooks become one of the most common subsistence artifacts in island sites.

Stone and bone tool production, except microblade technology, has seen relatively limited attention on the islands. This is partly because stone tools are generally found in fairly modest quantities and are often dominated by expedient tools. The bow and arrow probably first appeared on the Channel Islands around 1500 cal B.P., altering some hunting strategies and may have also marked a period of increased interpersonal violence (Kennett, 2005, pp. 187). Arrows and other projectile points and bifaces have been identified in a number of Late Holocene sites across the islands (Martz, 2005; Pletka, 2001a; Rick, 2004a; Vellanoweth et al., 2002a). Microblades have been identified on many of the islands with a huge production industry identified on the Northern Channel Islands and probably centered near Santa Cruz, where a large source of raw material exists (Arnold, 1987). Microblades were fashioned into drills and then used to perforate shell beads, making them integrally linked to island bead production and exchange systems (Arnold, 1987; Arnold, Preziosi, & Shattuck, 2001; Preziosi, 2001; see Fig. 5). Wake (2001) recently reported on a large assemblage of bone tools including barbs, gorges, awls, and knives from Late Holocene Santa Cruz Island and Rick (2004a) reported many similar artifacts from San Miguel and Santa Rosa islands. Finally, perishable artifacts or their impressions on tar are also occasionally identified in Late Holocene sites and include fibers from bags, water bottles, fishing line, and other artifacts (see Bleitz & Salls, 1993; Martin & Popper, 2001; Rick, 2004a).

Artifacts associated with exchange, personal ornamentation, and wealth greatly increased during the Late Holocene, particularly during the last 1000 years (Arnold & Graesch, 2001; King, 1990). While several types of Olivella and other beads have been identified in Early and Middle Holocene sites, the number and types of beads and ornaments identified in Late Holocene deposits increases exponentially, including numerous types of Olivella, red and black abalone, clam, mussel, Trivia, and other shell beads and ornaments (Arnold & Graesch, 2001; King, 1990; see Fig. 5). These beads were traded from islanders to mainlanders in a complex trade network that sometimes carried marine shell artifacts as far as the American Southwest.

During the last 1500 years, the Island Chumash manufactured beads by the millions (see Arnold, 1992a; Arnold & Graesch, 2001; Arnold & Munns, 1994; Graesch, 2004; Kennett & Conlee, 2002). The vast majority of these beads were made on the Northern Channel Islands, although beads were also made on the southern islands. Numerous island sites have produced these beads and associated artifacts, with some of the densest concentrations found in Late period and Historic villages on Santa Cruz, Santa Rosa, and San Miguel islands (Arnold, 2001a; Kennett, 2005). Some excavated levels at SRI-85 and SRI-97, for instance, contained over 40 kg of Olivella shell detritus per m³, suggesting a major emphasis on bead making at some Late Holocene sites (Kennett & Conlee, 2002).

Less is known about bead production on the Southern Channel Islands, but Late Holocene people on San Nicolas were also producing Olivella and other beads for exchange (Vellanoweth et al., 2002a). On Santa Catalina Island, Ripper’s Cove (SCAI-26), Miners Camp (SCAI-118), SCAI-77, and other sites have produced evidence of protohistoric specialization in soapstone olla and comal production for exchange (see Howard, 2002). Ground stone bowls intended for exchange may have also been mass produced on San Miguel (Conlee, 2000) and Vellanoweth et al. (2002a) and Thomas-Barnett (2004) also described a ground stone industry on San Nicolas Island. These data suggest that a variety of artifacts were made by Channel Islanders for trade. Much of this production-for-exchange on the islands may have resulted from the dearth of terrestrial resources on the islands and the need to acquire supplementary subsistence goods (e.g., acorns) from the mainland. Analysis of obsidian artifacts from Late Holocene island sites also suggests that these trade relations reached far into interior portions of California (Rick et al., 2001b). Although some Late Holocene artifacts may have been traded far and wide, the majority of beads and other artifacts were exchanged within Chumash and Tongva territory in a well defined network of interaction and trade.

Subsistence strategies

Correlating with many technological advances are a number of changes in Late Holocene subsistence. One of the most pronounced is an increase in the economic importance of fish and a decrease in the relative importance of shellfish (Kennett, 2005; Noah, 2005; Raab et al., 1995b, 2002; Rick, 2004a,b; Vellanoweth et al., 2002a,b; Vellanoweth & Erlandson, 1999). This change is probably related to population growth and the need for increased food yields provided by fish that other resource classes (e.g., shellfish) could not sustain. People continued to focus largely on nearshore fisheries, but during the last 1500 years there is an increase in fishing for deepwater fishes like tunas, swordfish, and mako sharks (Bernard, 2004; Bowser, 1993; Colten, 2001; Paige, 2000; Pletka, 2001b; Rick, 2004a). The increase in pelagic fishing may be related to the development of the plank canoe (Arnold & Bernard, 2005; Bernard, 2004; Davenport et al., 1993). The general increase in fishing efficiency is also related to the appearance of the single-piece fishhook.

At several sites on San Miguel Island (SMI-232, SMI-481, SMI-528), the densities of pinniped bones increase greatly around 1500 cal B.P., suggesting that some islanders were also intensively hunting pinnipeds during this time (Kennett, 2005; Rick, 2004a; Walker, Kennett, Jones, & Delong, 2002). After about 1500 cal B.P., the densities of pinniped remains in San Miguel Island sites are generally much lower. This reduction appears to be related to human hunting and other activities (see Rick, 2004a; Walker et al., 2002). At the Eel Point site on San Clemente Island, a trans-Holocene decline in marine mammal hunting efficiency was recently presented by Porcasi et al. (2000) who argued that during the Late Holocene people focused primarily on sea otters, with limited reliance on sea lions, fur seals, and dolphins that were important earlier in time. At several Late Holocene sites on Santa Cruz Island, Colten and Arnold (1998) and Colten (2002) noted that sea mammals were of supplemental importance to most economies. Collectively, these data suggest that sea mammal hunting varied on each of the islands, with some of the greatest hunting occurring on more distant islands (e.g., San Miguel, San Clemente).

Although their significance in human diets decreases, the variety of shellfish being used appear to increase in many Late Holocene sites, suggesting that people were intensifying their subsistence strategies to deal with population growth and greater social circumscription. On San Clemente Island, black abalones appear to decline in sites through time, while small Tegula shells appear to increase during the Late Holocene, a pattern demonstrated by the occurrence of numerous small Tegula middens (SCLI-1318 and -1319) that may be processing or satellite camps (Raab, 1992; Raab et al., 2002). On the Northern Channel Islands, the diversity of shellfish taxa is generally high in Late Holocene sites (Kennett, 2005; Rick, 2004a). California mussel appears to be the dominant shellfish species at most northern island assemblages and the contribution of black and red abalones generally appears to decline. Recent analysis of a trans-Holocene sequence of California mussel and red and black abalone shell sizes on San Miguel Island, including over 10,000 shell measurements, documents some declines in shell size through time that suggest human predation pressure, but no catastrophic change throughout the sequence (Braje et al., 2006a; Erlandson et al., 2004a; Rick et al., 2006). At the Prisoner’s Harbor site, a Historic and prehistoric village on Santa Cruz Island, Noah (2005, pp. 279–283) described a dense concentration of black abalone shells, associated with swordfish, sea turtle, sea mammal, a variety of other fish and shellfish remains, and metal crucifixes, that she argues may be evidence for a Chumash feasting ceremony.

Compared to other faunal classes, Birds are generally limited in Channel Island sites throughout the Holocene, and consequently less is known about archaeological bird remains than other faunal classes. On the Southern Channel Islands, however, the analysis of bird remains from the Little Harbor site on Santa Catalina, Eel Point on San Clemente, and Thousand Springs on San Nicolas has provided important information on bird exploitation (see Bleitz, 1993; Porcasi, 1999b). Porcasi indicated that birds were more important in Early and Middle Holocene sites than Late Holocene sites, where they generally provided only about 7 to 10 percent of the total diet (Porcasi, 1999b, pp. 51–53). Not surprisingly most of these bird bones were from marine species. Colten (2001) and Noah (2005) also demonstrated that birds were of supplemental importance to Late Holocene Santa Cruz Islanders, but a variety of taxa (gulls, pelicans, auklets, cormorants, etc.) were identified in their assemblages. Similar taxa were reported for Late Holocene sites on San Miguel Island by Guthrie (1980) and Rick (2004a). Ritual raptor burials have also been identified on San Clemente Island, demonstrating the significance of birds in island social and symbolic life (Raab et al., 1994).

The increase in marine fishing during the Late Holocene was associated with less diverse over-all diets and appears to have contributed to declines in general human health (see below) on the Northern Channel Islands (Lambert, 1993). According to Lambert’s (1993) analysis of skeletal assemblages from the Channel Islands and mainland coast, there are significantly higher rates of enamel hypoplasia, Harris lines, and interpersonal violence during the Late Holocene. Polluted water sources and increased crowding in villages may have also led to declines in health as human populations coalesced around the island’s fairly limited potable water supplies. Interestingly, research on the Southern Channel Islands has found only limited evidence of interpersonal violence and comparatively healthy populations during the Late Holocene, emphasizing variability in the archaeological record or differences in sociopolitical organization (see Ezzo et al., 2002; Kerr, 2003; Raab et al., 2002, pp. 25; Titus, 1987).

Coupled with increased subsistence efforts, these changes in human health on the northern islands suggest that during the Late Holocene Native peoples in the region were having a particularly pronounced impact on the environment and quality of life on the Channel Islands. Erlandson et al. (2004a, 2005c) recently synthesized aspects of human impacts on the San Miguel Island environment, documenting apparent impacts, but indicating that these were considerably less than those of the historic and modern periods. Rick et al. (2006) recently expanded on this research indicating that through time the Channel Islands and Santa Barbara Channel region were increasingly becoming anthropogenic landscapes shaped by the activities of Channel Islanders who had both positive and negative impacts. Raab et al. (2002), Porcasi et al. (2000), and others have argued for significant declines in human foraging efficiency during the Late Holocene on San Clemente Island. Guided by optimal foraging theory, they have argued for significant declines in marine mammals and shellfish starting in the Early and Middle Holocene resulting in the Late Holocene increase in fishing, a pattern that occurs with some variability on the Northern Channel Islands as well (Kennett, 2005; Rick, 2004a). In the San Francisco Bay area of California, Broughton (1994, 1997, 1999) has identified evidence for declining human foraging efficiency, or human hunting impacts on local animal populations. While Porcasi et al. (2000) and others have identified similar processes on San Clemente Island, such declines on the islands appear to be complex and warrant further research (see Kennett, 2005, pp. 220–223). This pattern is comparable to the variability noted by Butler and Campbell (2004) on the Northwest Coast of the Americas.

Cultural complexity

Associated with the dramatic changes in Late Holocene human subsistence, technology, health, and environmental impacts are a number of social and political developments, including heightened cultural complexity. In recent years, researchers have sought to explain when, how, and why Channel Islanders and Native peoples on the southern California Coast became increasingly complex and hierarchical, developments thought to correlate with the appearance of a simple chiefdom among the Chumash (see Arnold, 1992a, 2001b,c; Gamble, 2005; Kennett, 2005; Munns & Arnold, 2002; Raab & Larson, 1997; Vellanoweth et al., 2002a).

Much of this research has centered on understanding the evolution of cultural complexity among islanders and, in particular, the Chumash. This work has been guided by a number of theoretical paradigms including, Marxism, cultural ecology, and behavioral/evolutionary ecology (see Arnold, 1991, 1992a, 1993, 1997, 2001b; Arnold et al., 1997a; Colten, 1993; Larson, Johnson, & Michaelson, 1994; Raab & Larson, 1997). The Middle to Late period Transition (Transitional period; Arnold, 1992a) or Medieval Climatic Anomaly (Jones et al., 1999; Raab & Larson, 1997), ranging from roughly AD 1150 to 1300 or AD 800 to 1350, respectively, has received the most attention, although sites from this time period are generally limited. Arnold (1992a, 1992b, 1997, 2001a; Arnold et al., 1997a), Colten (1993), and others suggest a major cultural reorganization of Chumash society during the Transitional period, sparked by a period of elevated sea surface temperatures, variations in marine productivity, and drought. Aspects of this model have been challenged by Raab et al. (1995a) and Raab and Larson (1997) who argue that drought rather than changes in marine productivity were a catalyst to changes in Late Holocene social organization.

Recent research by Kennett (2005; see also Kennett & Conlee, 2002; Kennett & Kennett, 2000) has examined Island Chumash society across the Holocene. Grounded in behavioral ecology, Kennett (2005) employs a long-term perspective in looking at changes in hunter-gatherer societies. He emphasizes variability in the regional archaeological record, and a host of local responses to a variety of environmental and cultural variables. In his model, Island Chumash cultural complexity increases between 1500 and 650 years ago within the context of long-term population growth, climatic instability, resource intensification, and increasing patterns of interpersonal violence (Kennett, 2005; Kennett & Conlee, 2002; Kennett & Kennett, 2000). Raab (1996), Raab and Larson (1997), Raab et al. (2002), and others have also explored changes in cultural complexity, emphasizing the importance of population growth, resource depression, and other factors in promoting increased social hierarchy. Arnold (1992a) and Johnson (2000) have indicated the importance of inter-village trade in developing political hierarchies. All of these models document a series of abrupt changes occurring largely during the Late Holocene and culminating in the Late period.

Data from individual burials and cemeteries have also been used to examine changes in sociopolitical complexity in the region, particularly evidence for hereditary (ascribed) leadership (Gamble, Walker, & Russell, 2001; King, 1990; Martz, 1992). Burial data are valuable for interpreting social organization because they can provide information on the status of individuals. King (1990), relying partly on burial data from SCRI-333 on Santa Cruz Island, suggested that Chumash cultural complexity and hereditary status differentiation may have emerged roughly 2500 years ago or earlier. Munns and Arnold (2002, pp. 144–145) challenged King’s assertions for social ascription around 2500 years ago, suggesting the data were too limited to propose such an early date (but see Gamble et al., 2001). All the burial data from the Channel Islands and much of the data from the adjacent mainland comes from reanalyses of materials excavated 50 years ago or more by Phil Orr, Ronald Olsen, David Banks Rogers, and other early scholars working in the region. These studies often did not have the tight chronological or stratigraphic control necessary for defining changes in time, making the interpretation of burial data in the region somewhat problematic.

As stated above, analyses of patterns of health, disease, and violence from human skeletal remains from the Northern Channel Islands have also been used to explore the evolution of Chumash society (Hollimon, 1990; Lambert, 1993, 1994; Lambert & Walker, 1991). To Lambert (1993, 1994) and Lambert and Walker (1991), growing population densities and greater circumscription and territoriality promoted declines in health and increased violence. When compounded with environmental perturbations these events were important causes of cultural change and complexity. Hollimon’s (1990) study of gender roles, health, and status of the Santa Barbara mainland and Channel Islands confirms many of Arnold’s hypotheses about cultural changes during the Transitional period. Hollimon (1990, pp. 213–214) indicated that people living during the Transitional period suffered poorer health than people living before or after this time, and that cultural changes were probably instituted to buffer subsistence stress. Kennett (2005) and Raab and Larson (1997) also used some of these data to suggest that increased violence and subsistence stress around 1500 years ago may have promoted instability in the area and led to more competition and greater cultural complexity.

Collectively, the archaeological data from the Channel Islands suggest a great deal of spatial and temporal variability in human social organization and complexity. It is also clear that many of these developments (sedentism, exchange and interaction, etc.) have their roots in the Middle Holocene or earlier. The Island Chumash on the northern islands clearly had elaborate forms of social organization and exchange evident in the Late Holocene archaeological record. On San Nicolas Island, Vellanoweth et al. (2002a) documented similar patterns. On San Clemente Island, Raab et al. (2002) noted that there is limited evidence for cultural complexity, illustrating that the people who inhabited the various islands followed multiple trajectories and pathways throughout the past.

Culture contact

The final chapter of the Native American occupation of the Channel Islands began in the mid sixteenth century with the first contact with Europeans, intensified during the late eighteenth century, and came to a close by AD 1820 to 1835 when the last islanders were removed to mainland missions. The first Europeans known to contact Channel Islanders were members of a Spanish maritime expedition led by Juan Rodriguez Cabrillo in AD 1542, who made landfall on Santa Catalina and San Miguel islands (Wagner, 1929). Several other early explorers made sporadic and often poorly documented visits to the islands (Erlandson and Bartoy, 1995). In recent years, a debate has arisen over the nature of these early contacts during the Protohistoric period (AD 1542–1769) and the possible impacts such contacts had on Native American populations and culture. In a preliminary study of paleodemography based on the frequency of ¹⁴C dated sites, Erlandson, Rick, Kennett, and Walker (2001) speculated that population declines from introduced diseases may have occurred during the century after Cabrillo’s visit, then populations may have rebounded before the onset of the Mission period. J. Johnson (1988, 1999b, pp. 96), Arnold et al. (2004, pp. 7–8), and others have suggested that these early contacts had little or no impact on populations and cultures.

Although the impacts of Protohistoric cultural contacts remain uncertain, during the Mission period (AD 1769–1834) introduced diseases affected Native populations on the mainland and islands. Moreover, islanders were brought to the mainland in a campaign to convert all of the neophytes at the missions. Using mission records, J. Johnson (1982, 1988, 2001) has traced baptisms for several Northern Channel Island villages to the missions. These records have also been used to examine patterns of inter-island and inter-village marriage on the northern islands (Johnson, 1993). Over the last several years, Arnold (1990), J. Johnson (1982, 1988, 1999a), Kennett (2005; Kennett, Johnson, Rick, Morris, & Christy, 2000), and others have made detailed efforts to locate the 22 Chumash communities described for the northern islands in ethnohistoric sources. Most of these village locations are now relatively well established and excavations have also shown some of the dynamics of the contact period on the islands. After the secularization of the missions, many Island Chumash formed communities along the southern California mainland, illustrating that the distinct nature of island social life persisted later in time (Johnson & McLendon, 1999).

Compared to the Chumash, early historic and ethnographic accounts of the Island Tongva are extremely limited, leaving archaeology as the primary means of reconstructing contact era cultural processes. On the southern Islands, evidence for Mission era ceremonialism and ritual, including dog and raptor burials were identified at the Lemon Tank, Big Dog Cave (SCLI-119), and Ledge (SCLI-126) sites (Raab et al., 1994; Rechtman, 1985). Salls (1988, 1990b) also reported on the fishery at Big Dog Cave, which in many ways is similar to other precontact fisheries. Preliminary evidence on San Nicolas Island also provides evidence for Protohistoric occupation, including needle-drilled Olivella beads, European glass, and glass trade beads.

During the Historic period, the use of imported technologies (glass beads, iron needles, etc.) also had profound impacts on traditional cultural practices. Graesch (2001) described a variety of Historic period artifacts from Santa Cruz Island, including bronze crucifixes, metal fishhooks, nails, blades, and other tools, glass beads, worked bottle glass, and shell artifacts that had been perforated with iron needles. These needle-drilled shell beads are a hallmark of many Chumash sites dated to the Historic period (see Graesch, 2001; Kennett, 2005; Rick, 2004a). Olivella rough chipped wall disk beads appear in abundance during the Historic period and red abalone disk beads also greatly increase during this time. The increase in these beads and other artifacts may be responses to the influx of glass beads that prompted islanders to expand bead making efforts.

Changes in world view, technology, and health during the contact era profoundly transformed the lives of the Island Chumash and Tongva. The final removal of the last Chumash to the Missions by about AD 1820 and Juana Maria from San Nicolas by AD 1853 ushered in a new era of Channel Islands history. The islands were occupied sporadically by Chinese and other abalone fishers during the mid nineteenth century. Native Alaskans were also brought to the Channel Islands to hunt sea otters in the early portions of the nineteenth century and often had violent interactions with the Chumash (Ogden, 1941). All the islands also had fairly long and sizable ranching histories during the nineteenth and twentieth centuries that in many cases had devastating effects on the local ecology (see D. Johnson, 1980; Schoenherr et al., 1999; Swanson, 1993). Naval installations on San Nicolas and San Clemente islands, park service facilities on other islands, and a small town and other facilities on Santa Catalina, however, constitute the only sizable settlements on the islands today. Under federal and private management the islands have remained free of the urban sprawl that plagues the mainland. Nonetheless, by the early to mid nineteenth century the Channel Islands were transformed into a new landscape, bringing to a close more than 10,000 years of Native American island and maritime cultures.

Summary

During the Late Holocene, the archaeological record contains the full manifestation of the Channel Island societies encountered by early European explorers. Technologies like the plank canoe, circular shell fishhook, stone microdrills, and Olivella cup beads transformed island economies and social networks. Subsistence practices became increasingly focused on fishing, resulting in some cases in declines in human health and impacts on the environment. Although a great deal of information has been amassed on the Late Holocene environments and cultures of the Channel Islands, many questions remain. In particular, there are a dearth of data from the early phases of the Late Holocene (ca. 3500 to 1500 cal B.P.), little is known about the people who inhabited Santa Catalina, Santa Barbara, and Anacapa islands, and greater comparisons of the Southern and Northern Channel Island records are needed.

Summary and conclusions

The Channel Islands contain one of the longest, continuous records of coastal hunter-gatherers in the Americas. The Chumash, Tongva, and their predecessors were among the most sophisticated and elaborate hunter-gatherers on earth, with large and dense populations, complex exchange networks, social hierarchy and inherited leadership, and complex maritime subsistence strategies. The archaeological and ethnohistoric records of the region also demonstrate considerable variability in these practices (e.g., Altschul & Grenda, 2002; Kennett, 2005; Raab et al., 2002). Review of this trans-Holocene archaeological sequence demonstrates that many of the cultural practices described in early historic accounts have their roots in deep history, especially aspects of marine subsistence. Our analysis of this sequence illustrates several long-term trends in Channel Island prehistory that are of significance to broad anthropological issues, such as the origins of maritime Paleocoastal peoples, the emergence of sedentary communities, the development of exchange systems, interaction spheres, and shell bead currencies, the appearance of hierarchy and social ranking, changes in marine subsistence strategies, human impacts on the environment, and responses to natural and culturally induced environmental change.

Only limited evidence of human occupation during the terminal Pleistocene exists on the Northern Channel Islands, but after about 10,000 years ago the volume and types of archaeological sites increased considerably. Four of the islands (San Clemente, Santa Cruz, Santa Rosa, and San Miguel) contain evidence of occupation prior to about 8500 years ago and San Nicolas at 98 km offshore has evidence of human occupation possibly as early as 8500 years ago. These earliest occupants lived on islands where marine and terrestrial ecosystems were in a dramatic phase of reorganization, as rising sea levels reduced island area, separating Pleistocene Santarosae into the four separate northern islands, where pygmy mammoths had recently gone extinct. Most research currently suggests that these early occupations were by people who were relatively mobile and focused largely on shellfish, supplemented by marine mammals, fishes, and birds. There is great variability in this pattern (Porcasi et al., 2000; Rick et al., 2001a), however, and due to the destructive effects of sea level rise and coastal erosion our understanding of early island lifeways is far from complete. The Eel Point site, one of the most intensively studied early island sites, contains evidence of structural remains and a diversified maritime economy about 8500 years ago, providing limited evidence for sedentism and the more sophisticated marine foraging that was to appear during the Middle and Late Holocene. Although Olivella shell beads provide evidence of bead production during the Early Holocene, and mainland or other sources of island stone suggest trade or long-distance procurement strategies during this time, exchange systems were relatively limited during these early times. Clear evidence of social ranking is also absent in the archaeological record of the early Channel Islands and mainland, but some of the foundations of later developments (shell beads, basketry, fishhooks, seaworthy boats, etc.) were founded very early in Channel Island prehistory.

The Middle Holocene archaeological record demonstrates considerable cultural overlap with the Early Holocene, but new and unique cultural patterns also appear between about 7000 and 3500 years ago, setting the stage for many of the rapid cultural changes of the last 3500 years. Middle Holocene peoples, for example, had complex and far reaching exchange networks. The Olivella Grooved Rectangle bead, for instance, was made on the Southern Channel Islands and traded into mainland California, Nevada, and Oregon (Howard & Raab, 1993; Vellanoweth, 2001). There is also evidence for increased sedentism in the Middle Holocene, but many sites suggest that people were fairly mobile during this time. Although subsistence pursuits were focused largely on nearshore marine resources, new and unique subsistence strategies included dolphin hunting (Glassow, 2005b; Porcasi & Fujita, 2000), fishing for Mola mola (Porcasi & Andrews, 2001), and the taking of albatross (Porcasi, 1999a). Archaeological data from the Channel Islands currently suggest that the Middle Holocene was a time of transition, with a great deal of continuity with the Early Holocene, but glimpses of the rapid and pronounced cultural changes of the Late Holocene.

After about 3500 years ago, the pace of cultural change appears to accelerate, with the introduction of the single-piece fishhook, plank canoes, bow and arrow, new shell bead types, and several types of specialized craft production. Increased cultural complexity is perhaps the most heavily studied aspect of the Late Holocene archaeological record, as craft production, village organization, burial data, and other variables have been used to argue for the development of a simple chiefdom among the Island Chumash (Arnold, 1996, 2001c). These artifact forms and changes in social organization appear to correlate with sustained periods of drought and one of the coldest and most variable periods of sea surface temperature recorded for the Holocene (Kennett & Kennett, 2000), suggesting that people were increasingly forced to adapt to environmental fluctuations. Compounded by increased population density and territoriality, the Late Holocene landscape was one of great environmental and social uncertainty. For the Island Chumash and Tongva, this period of instability concluded with the arrival of the representatives of European colonial powers, a demographic and cultural catastrophe for the traditional cultures that had lived on the Channel Islands for more than 10,000 years.

The long Channel Islands archaeological sequence provides considerable insight into the evolution of coastal hunter-gatherer societies, illustrating the importance of investigating long term cultural histories to inform issues of broad anthropological importance. These data demonstrate that people with fully maritime capabilities were present in the Americas as much as 13,000 years ago, the contemporaries of Folsom and possibly Clovis peoples. Viewed over the long-term, the Channel Islands record of cultural and environmental change helps inform the evolution of maritime foraging strategies, the development of regional and long-distance exchange networks, and the movement of goods, ideas, and people. The emergence of cultural complexity among the Chumash and Tongva supplies important details on social and political dynamics in coastal and island settings where agriculture was not practiced (Arnold, 1996, 2001c; Sassaman, 2004). While this complexity has often been viewed as a relatively recent phenomenon on the Channel Islands, we have argued that many of the rapid and profound changes of the past 1500 years have their roots in much earlier cultural developments. The subsistence data also provide perspectives on historical ecology and human impacts on the environment that can be used as baselines to help remediate modern ecological degradation in the area (see Erlandson et al., 2004a, 2005c; Rick et al., 2006).

By comparing the full duration of the Channel Islands archaeological record, we see the incremental patterns of cultural evolution that are the result of millennia of interaction, environmental change, and population growth and decline. These changes appear to have been gradual at times and punctuated and rapid at others, but it is only within the long-term context of the record that the full manifestation, origins, and implications of these developments comes to light (Ames, 1991; Lightfoot, 1993). Ultimately, the 13,000 year history of humans on the Channel Islands is a testament to the resilience and diversity of the Chumash, the Tongva, and their predecessors. The archaeological record provides valuable insights into the lives and collective history of ancient Channel Islanders, providing an important backdrop for understanding the nature and structure of California and its islands today. The challenge is to ensure that this incredible record is preserved for the future and that the interpretations of island archaeology are increasingly applied to issues of broad social, cultural, and environmental significance.