A predictive model for locating early Holocene archaeological sites in southern Southeast Alaska was developed based on shell‐bearing raised marine deposits. Fieldwork included coring of select‐raised marine strata, measuring their elevations, and radiocarbon dating the associated shell samples within the cores. A subset of the data was used to produce a relative sea‐level curve spanning the Holocene. The relative sea‐level curve suggests that sites favorable for habitation between 9200 and 7000 14C yr B.P. should be found 16–22 ± 1 m above present zero tide. The sea‐level curve and new high‐resolution digital elevation models allowed reconstruction of past shorelines at various elevations. Surveys to test the model found and recorded over 70 archaeological sites from present sea level up to 32 m above present zero tide. Eleven new sites were within the targeted elevation range and radiocarbon dated to 9280–6890 14C yr B.P. Initial investigations indicate these older sites are rich in microblade and pebble tool technology. The new early Holocene sites indicate more extensive early maritime settlement of Alaska than implied by previous studies and contribute to our understanding of the early movement of people into North America.

Recognition of marine reservoir effect (MRE) spatial and temporal variability must be accounted for in any radiocarbon-based paleoclimate, geomorphological, or archaeological reconstruction in a coastal setting. ΔR values from 37 shell-wood pairs across southern Southeast Alaska provide a robust local evaluation of the MRE, reporting a local Early Holocene weighted ΔR average of 265 ± 205, with a significantly higher ΔR average of 410 ± 60 for samples near limestone karst. Integration with our synthesis of extant MRE calibrations for the Northwest Coast of North America suggests that despite local variability, regional ΔR averages echo proxies for coastal upwelling: regional weighted averages were at their highest in the Bølling-Allerød interstade (575 ± 165) and their lowest in the Younger Dryas stade (−55 ± 110). Weighted ΔR averages across the Northwest Coast rose to a Holocene high during the Early Holocene warm period (245 ± 200) before settling into a stable Holocene average ΔR of 145 ± 165, which persisted until the late Holocene. Our quantification of local and regional shifts in the MRE shines a light on present methodological issues involved in MRE corrections in mixed-feeder, diet-based calibrations of archaeological and paleontological specimens.

We leverage a data set of >720 shell-bearing marine deposits throughout southeastern Alaska (USA) to develop updated relative sea-level curves that span the past ∼14,000 yr. This data set includes site location, elevation, description when available, and 436 14C ages, 45 of which are published here for the first time. Our sea-level curves suggest a peripheral forebulge developed west of the retreating Cordilleran Ice Sheet (CIS) margin between ca. 17,000 and 10,800 calibrated yr B.P. By 14,870 ± 630 to 12,820 ± 340 cal. yr B.P., CIS margins had retreated from all of southeastern Alaska's fjords, channels, and passages. At this time, isolated or stranded ice caps existed on the islands, with alpine or tidewater glaciers in many valleys. Paleoshorelines up to 25 m above sea level mark the maximum elevation of transgression in the southern portion of the study region, which was achieved by 11,000 ± 390 to 10,500 ± 420 cal. yr B.P. The presence of Pacific sardine (Sardinops sagax) and the abundance of charcoal in sediments that date between 11,000 ± 390 and 7630 ± 90 cal. yr B.P. suggest that both ocean and air temperatures in southeastern Alaska were relatively warm in the early Holocene. The sea-level and paleoenvironmental reconstruction presented here can inform future investigations into the glacial, volcanic, and archaeological history of southeastern Alaska.

No abstract

Despite the ubiquity of obsidian in early Holocene archaeological assemblages across Southeast Alaska, artifact sourcing using bi-plots and Principal Component Analysis has been hampered by the highly correlated geochemistry of two major sources: local Aguada Cove on Suemez Island, and distant Mount Edziza, in the Coast Mountain range. Partial Least Squares Discriminant Analysis (PLS-DA) models constructed with device-specific portable X-ray fluorescence (pXRF) source catalogs benefit from the precision of individual pXRF machines and the ability of PLS-DA to handle highly correlated data sets to provide a source classification system that complements existing methods. Of the known obsidian sources in or near Southeast Alaska, four are identified in early Holocene archaeological assemblages: Obsidian Cove and Aguada Cove on Suemez Island, Mount Edziza in the Coast Mountain Range, and a newly identified source on Zim Creek, Kupreanof Island. A comprehensive reanalysis of early Holocene microblade cores (101 cores from 13 archaeological sites) across Southeast Alaska confirms the presence of exotic obsidian from Mount Edziza in one of the oldest sites, alongside possible evidence of landscape learning: testing of the obsidian source on Kupreanof Island. The dominance of the Obsidian Cove source for microblade core production supports hypotheses framing Northwest Coast microblade core morphology as an adaptation to the small nodules of raw material available at the source. A simple distance-decay model indicates no relationship between core reduction intensity and distance from the source, suggesting high forager mobility characterized by regular trips to Suemez Island (a voyage of up to 380 km) throughout the early Holocene.