Human population movements into and around the outer Shumagin Islands of southwestern Alaska during the last 5000 years, and temporal gaps in Shumagin habitation correlate inversely with geologically inferred prehistoric earthquakes. Clusters of inferred seismic activity correlate with temporal gaps or with small, sparsely distributed archaeological sites; periods of relative seismic quiescence coincide with settlement florescence. Variations in terrace and archaeological site heights across the study area indicate differential uplift of the outer Shumagin Islands when corrected for the Holocene sea level rise and post-glacial isostatic adjustments.
INTRODUCTIONCoastal people in tectonically active areas can be dramatically affected by co-seismic events that may include permanent uplift or subsidence, severe shaking, tilting, and inundation by large tsunamis. Major food resources such as shellfish beds can be destroyed by increased or decreased shelf depths, landslides can bury settlements, collapsing seacliffs can take settlements and boats with them, and changes in harbor geometries by creation or removal of natural barriers can make a locality more or less favorable for future habitation. Therefore, following a great earthquake, major disruption of the human record, either a hiatus or marked reduction in size of sites, might be expected (Black, 1981).The interplay between tectonic uplift or subsidence and the early Holocene sea level rise not only predicts the location of desirable coastal habitats and the maximum age of archaeological sites, but also affects preservation of former settlements (Black, 1977). Therefore, recognition of temporal and spatial patterns of crustal movements and the timing and approximate size of geologically inferred prehistoric earthquakes and tsunamis is of critical importance to the evaluation of the prehistoric archaeological record. Conversely, interpretations of long term crustal behavior must not violate human