GC11E-1065
Vegetation and Pollen signals of the Anthropocene: Changing signals, changing scales

Monday, 14 December 2015
Poster Hall (Moscone South)
Simon J Goring, University of Wisconsin Madison, Madison, WI, United States
Abstract:
The signals of the onset of Euro-American settlement in North America are clear and pervasive, but there is a disconnect between pollen and vegetation signals, both prior to and following EuroAmerican settlement. Vegetation shifts with respect to climate are induced by biased land use patterns in the upper Midwestern United States, but because pollen can be transported long distances, and because the granularity of the pollen signal, obtained from lake sediments and interpolated across the landscape, is often much lower than for forest and land cover data, the apparent influence of climate in structuring vegetation is much stronger for pollen than for vegetation.

We examine the impacts of widespread human land use on taxon boundaries in the northeastern United States, both prior to widespread EuroAmerican Settlement and under modern conditions, for both pollen and forest cover and compare these distributions and the forest community ecotones to climate and physiographic features. We show that following EuroAmerican settlement vegetation ecotone sensitivity to both climate and physiography is much weaker than prior to settlement, while for pollen climate remains the dominant driver, but the presence of cosmopolitan “weedy” taxa produces a landscape that is more homogeneous, and again shows weaker sensitivity to climate and physiographic thresholds.

The drivers for decreased sentitivity are largely the result of broad-scale homogenization, the overall loss of strong vegetation ecotones in the region. Pollen is less sensitive to this decline because the influence of heavy producers has already resulted in an apparent homogenization of the landscape. This has implications for past land-cover reconstruction and our understanding of the strengths of climate and physiography in structuring vegetation cover over long time scales.