anthropogenic perturbations to Earth's atmosphere, or a combination of factors (Rind and Overpeck, 1993). Because instrumental records rarely exceed 100 yr, paleoclimate reconstructions can establish which factors are most important over multidecadal time scales, distinguish anthropogenic and natural causes, and test climate-model simulations of decadal-and centennial-scale variability (Latif, 1998). Paleoclimatic records from polar and tropical regions suggest that decadal-and centennial-scale climate changes are characteristic of the past millennium, a period that includes the Medieval Warm Period (ninth through fourteenth centuries) and the Little Ice Age (fifteenth through nineteenth centuries) (Hughes and Diaz, 1994; Overpeck et al., 1997). Yet, except for tree-ring records (i.e., Cook and Jacoby, 1983; Stahle et al., 1998), multicentury records are sparse from most of the United States, and the ecological impacts of Holocene climate variability remain poorly known. We investigated the past millennium of paleoclimatic history of the mid-Atlantic region preserved in the sedimentary record of Chesapeake Bay. Chesapeake Bay, the largest estuary in the United States, is a 320-km-long, 20-40-kmwide, 6500 km 2 drowned river valley (average and maximum depths of 8.5 m and 53 m, respectively) inundated by the late Quaternary sea-level rise ca. 6-8 ka (Colman and Mixon, 1988). The bay's thick Holocene section (Colman and Halka, 1989) contains micropaleontological and geochemical evidence for changes in freshwater and sediment inflow, salinity, dissolved oxygen, and temperature (Cronin et al., 1999a). Its record of decadal-and centennial-scale variability holds promise for understanding impacts of past and climate change.
