Abrupt Climate Changes During the Holocene Across North America From Pollen and Paleolimnological Records Konrad Gajewski and Andre E. Viau Laboratory for Paleoclimatology and Climatology, Department of Geography, University of Ottawa, Ottawa, Ontario, Canada Databases of ecological and cultural records, especially of pollen diagrams, record climate variability of several time scales during the Holocene and late glacial. Results from lake and wetland ecosystems geographically extend the evidence of rapid climate change obtained from ice cores and ocean sediments. Continental and regional climate curves for North America, based on pollen diagrams from the North American Pollen Database, illustrate abrupt changes on the order of every ~1000 years during the past 12 kyr, and major times of change in North American pollen records are coherent with vegetation changes across Europe. Novel analyses of the database show that even taxa that are widespread and with presumably broad climate tolerances were affected by abrupt climate changes such as the Younger Dryas and illustrate the complexity of ecosystem response to these changes. Reconstructions of freshwater as well as terrestrial ecosystems across northern Canada also show how climate variability affects terrestrial and freshwater ecosystem-level properties such as nutrient cycling. These results can be used to reconstruct the spatial patterns of abrupt climate change, as well as the impacts of climate change on ecosystem and cultures. 1. INTRODUCTION Although most of the “iconic” paleoclimate time series are from ocean [e.g., Imbrie et al., 1989] or ice cores [e.g., Johnsen et al., 2001; EPICA Members, 2006], a considerable amount of information is known about the terrestrial paleo- climates of the past 21 kyr [e.g., Bryant and Holloway , 1985; Wright et al., 1993]. Evidence from marine and ice cores suggests abrupt climate changes superimposed on the “Milankovitch-scale” variability that occurred throughout the last million years [e.g., Bond et al., 1997, 2001; MacMa- nus et al., 1999; O’Brien et al., 1995; Oppo et al., 1998; Raymo et al., 1998; Bianchi and McCave, 1999; Heinrich, 1988; Dansgaard et al., 1993; Rahmstorf, 2003]. However, one major uncertainty that arises in paleoclimate research is the lack of spatial corroborating evidence. For example, the Greenland ice core records offer a temporal record of past climate variations on several time scales; however, they remain a single point in space until results can be compared to records found in other regions. This lack of the spatial dimension has plagued large-scale paleoclimate research for many years. The last few decades have seen assembled large databases of paleoclimate information. The availability of large databases of pollen, tree ring, and other data enables syntheses of the terrestrial paleoclimate at several time and space scales. Considerable progress has been achieved in the past 30 years in producing paleoclimate estimates for the postglacial using pollen as a climate proxy for all of North America and Europe. This has come about due to the availability of a critical mass of fossil and calibration data and methodolog- ical developments suitable for large-scale reconstructions, Abrupt Climate Change: Mechanisms, Patterns, and Impacts Geophysical Monograph Series 193 Copyright 2011 by the American Geophysical Union. 10.1029/2010GM001015 173