Climatically forced vegetation dynamics in eastern North America during the late-Quaternary

T. Webb III, B. N. Shuman, and J. W. Williams

2003 The Quaternary Period in the United States Chapter 21

Introduction —  Vegetation dynamics span multiple spatial and temporal scales, and the changes involved manifest themselves in a variety of ways depending upon the ecological unit (from individuals to biomes) and/or taxonomic level (e.g., species , genera, families and orders) of description. Many biotic phenomena contribute to vegetation change including 1) the establishment, growth, and death of individual plants within stands, 2) changes in the frequency, size, and genetic make-up of populations within landscapes, 3) changes in the distribution of species, genera, and plant functional types across regions and continents, and 4) the evolution and extinction of species. These biotic phenomena cause the vegetation to change in structure, density, extent, and composition, and they lead to and result from a variety of biospheric dynamics (such as variations in net primary production and carbon sequestration). Depending on scale, vegetation changes are caused by some combination of external (i.e., environmental) forcing and the biotic phenomena themselves. The multiple competing forcings (at work at different scales) and many nonlinear linkages (including feedbacks) can make the cause-and-effect explanations difficult to sort out at certain temporal and spatial scales. Across long-time spans, however, such as the late Quaternary, environmental variations are large and well known and their effect on vegetation history is relatively easy to recognize.

            In this chapter, we consider vegetation dynamics at regional to continental scales and across millennia, scales at which vegetation change is primarily forced by centennial to orbital scales of climate change. The vegetation changes show up as the changing abundance, geographic extent, location, and association of plant taxon populations, which we record as changing pollen percentages. Only by linking the forces and induced responses can we convert the study of vegetation change and history into an analysis of vegetation dynamics, because to do so we must relate the apparent “motion’ in these taxon populations to underlying forces, which is the very definition of dynamics. Motion by definition is temporal change in location, which requires temporal sequences of maps, difference maps, and/or isochrone maps to illustrate. Mapping temporal change in the vegetation is therefore central to studies of climatically forced vegetation dynamics. Here, we map fossil pollen data, as a proxy for vegetation data, from eastern and northern North America and compare both continental-scale and local records of the pollen-recorded vegetation change to maps and time series of independently observed or estimated paleoclimate data. These comparisons are key to our empirical understanding of late-Quaternary vegetation dynamics. We admit that the “motion” of taxon populations shown on our maps is an epiphenomenon of the differential carbon sequestration in the different taxa in different locations, but we focus here on the motion apparent in the time series of pollen maps and use it and other pollen-recorded changes to represent how the vegetation changed. Many studies show how well pollen data from surficial sediments represent plant taxon abundances today and thus underpin our interpretative step here (Webb, 1974; Bradshaw and Webb, 1985; Jackson, 1994; Webb, 1995).

            Datasets of lake-level variations, chironomid-inferred temperatures, and stable isotope ratios, as well as climate model output, help us to show the “forces” behind vegetation changes and to identify dynamics. We therefore take advantage of advances in paleoclimate data, analysis, and modeling that are providing an increasingly detailed picture of late-Quaternary climate changes. Just as radiocarbon dating freed pollen data from a correlation-based time frame, newly developed paleoclimate datasets now allow pollen data to be interpreted within an independently derived climate framework. We can therefore describe how the vegetation responded to multivariate changes in climate involving temperature, moisture, and seasonality.

            We use both time series and maps of pollen data and climate estimates 1) to illustrate a strong connection between climate and vegetation change, 2) to document continental- and regional-scale vegetation dynamics that result from millennial- and orbital-scale climate forcing, and 3) to demonstrate that the conditions held for dynamic equilibrium between vegetation and climate at orbital time scales and possibly at millennial scales. By mapping both individual taxa and assemblages of taxa, we describe vegetation responses to independently documented climatic forcing at several levels of ecological organization from taxon movements to shifts in biome position, extent, and composition. Our chapter focuses on examples from North American vegetation history that illustrate key climatically forced vegetation dynamics. In doing so, we aim to complement the discussion of vegetation history by Grimm and Jacobson (this vol.), Thompson (this vol.), and Anderson (this vol.), and build on the critical reviews written by Cushing (1965), Davis (1965), and Whitehead (1965) that Grimm and Jacobson (this vol.) so ably review in their introduction. Too few pollen diagrams with radiocarbon dates existed for mapping the data on an independent time frame in 1965. Since then palynologists have published over 500 pollen diagrams with radiocarbon dates in eastern and northern North America. Other researchers have generated data independent of pollen data for estimating past changes in climate, and climate modeling has yielded valuable simulations of late Quaternary climates and climate change (Wright et al., 1993; Webb, 1998). These developments allow a fresh understanding of vegetation dynamics  and testing of many of the hypotheses posed by Cushing (1965), Davis (1965), and Whitehead (1965).