Early retreat of the Alaska Peninsula Glacier Complex and the implications for coastal migrations of First Americans

Abstract

The debate over a coastal migration route for the First Americans revolves around two major points: seafaring technology, and a viable landscape and resource base. Three lake cores from Sanak Island in the western Gulf of Alaska yield the first radiocarbon ages from the continental shelf of the Northeast Pacific and record deglaciation nearly 17 ka BP (thousands of calendar years ago), much earlier than previous estimates based on extrapolated data from other sites outside the coastal corridor in the Gulf of Alaska. Pollen data suggest an arid, terrestrial ecosystem by 16.3 ka BP. Therefore glaciers would not have hindered the movement of humans along the southern edge of the Bering Land Bridge for two millennia before the first well-recognized “New World” archaeological sites were inhabited.

Introduction

The timing and entry route of the first people to inhabit the Americas has been a controversial topic for decades. The debate as it pertains to a coastal migration route revolves around two major points: maritime adaptations/lack of archaeological sites along the coasts, and a viable landscape and resource base. The discussion about a viable landscape for migration includes the timing of access via a possible coastal route along the Aleutian Archipelago and Alaska Peninsula (controlled by retreat of the Alaska Peninsula Glacier Complex (APGC) along the Northeast Pacific margin) versus timing via a more terrestrial interior route known as the “ice-free corridor” and the dates of the earliest accepted archaeological sites in the Americas. Age models based on 22 radiocarbon dates from three lake sediment cores on Sanak Island in the western Gulf of Alaska (Fig. 1), located directly within the proposed coastal migration corridor, yield lake inception dates that place deglaciation ∼17 ka BP. These new dates are one to two millennia earlier than previous deglaciation estimates for the region (Mann and Peteet, 1994, Dickinson, 2011). Therefore glaciers would not have hindered the movement of humans along the southern edge of the Bering land bridge for two millennia before the first well-recognized archaeological sites in the Americas were inhabited (Mandryk et al., 2001, Erlandson et al., 2008, Waters et al., 2011).

Scholars have had difficulty determining the timing and nature of early coastal adaptations in North America, due to both postglacial sea level rise and erosion of coastal margins (Gruhn, 1994, Dixon, 2001). Intensive archaeological survey on Sanak Island, for example, has documented more than 120 archaeological sites spanning the last 7000 years, but did not result in any finds of terminal Pleistocene human activity. However, evidence shows that early peoples living along the coasts of the Americas relied heavily upon marine resources. Seaweed recovered from the Monte Verde site, Chile dates to ∼14 ka BP (Dillehay et al., 2008). Evidence from Huaca Prieta, Peru shows a reliance on maritime resources starting at 14.2 ka BP (Dillehay et al., 2012) and at Quebrada Jaguay, Peru, there is a predominance of marine resources ∼13 ka BP (Keefer et al., 1998, Sandweiss et al., 1998). Sites in the California Channel Islands contain remains of marine fish and mammals dating to ∼12 ka BP (Erlandson et al., 2011). The Arlington Springs site, also located in the Channel Islands and which contained human remains is dated to ∼13 ka BP (Johnson et al., 2002). Monte Verde remains the oldest well-accepted site in the Americas and was inhabited by 14.5 ka BP (Dillehay, 1997), prior to the opening of the ice-free corridor that was created when the Cordilleran and Laurentide ice sheets retreated 13.7 ka BP (Catto, 1996, Dixon, 2001, Dickinson, 2011). However, more detailed evidence shows that the ice-free corridor was probably not habitable by humans until close to 13 ka BP (Arnold, 2002, Dickinson, 2011). Together these data lend credence to the hypothesis that the Americas could have been populated first via the Pacific coast by people with maritime adaptations rather than through an interior ice-free corridor, although it is likely that there were later movements of people through the interior.

This is not a new argument. More than 30 years ago Knut Fladmark (1979) proposed that a coastal migration route to the America’s was a likely scenario. But this model was rejected by most scholars because ice sheets were assumed to have covered the North Pacific continental shelf from Puget Sound to the Aleutian Islands (Mann and Peteet, 1994, Mann and Hamilton, 1995, Dickinson, 2011). From the 1990s onward, a number of studies demonstrated clearly that parts of the Northwest Coast were likely an ice-free refugium (Heaton et al., 1996, Hetherington et al., 2003, Lacourse et al., 2005). Carrara et al. (2007) identified 8 areas from just south of Yakutat, Alaska to the southern extent of Graham Island that were potentially ice free and viable landscapes for humans to utilize on a coastal migration route. Ice retreated from the general area prior to 15.4 ka (Ager et al., 2010). Portions of Kodiak Island were also believed to have deglaciated somewhat early with radiocarbon dates at base of peat cores of 13,420 ± 20 (Beta-26607; Peteet and Mann, 1994:Table II) which using Calib 6.0 have median dates of 16.5 ka BP. The most significant remaining barrier to human coastal migration around the Northeast Pacific was the APGC, which was thought not to have disintegrated until 15–16 ka (Mann and Peteet, 1994, Mandryk et al., 2001, Goebel et al., 2008), too late to account for the early sites such as Monte Verde.