Convergent Ecosystems from Divergent Environmental Drivers: Revisiting Drivers of the Past Prairie-Forest Ecotone Across the North American Prairie Peninsula

Abstract:

The Prairie Peninsula in North America refers to the eastward extension of prairie and savanna ecosystems from western tallgrass prairie into deciduous forests of Minnesota, Wisconsin, Illinois, Indiana, and Missouri. Historically, the Prairie Peninsula border defined one of the largest shifts in vegetation and biomass in North America. Today, the fertile soils of the Prairie Peninsula have been transformed into a large agricultural center of national economic importance. The existence and location of prairie and savanna on the Northwest border of the Prairie Peninsula has previously been attributed to strong, dry westerly winds, low precipitation, and high fire frequencies. However, these drivers remain untested with historical vegetation across the whole Prairie Peninsula. We use recently digitized historical survey data of vegetation at the time of European settlement (the Public Land Survey (PLS)) to test these hypotheses with past structure and composition data of the entire Prairie Peninsula region. We demonstrate that commonly cited hypotheses for the existence of these ecosystems (westerly winds, low precipitation, and fire) cannot predict the extent and composition of the pre-settlement landscape in the Prairie Peninsula. Using the PLS data in a statistical model of biomass run with environmental covariates (precipitation, soil hydrology parameters, and topography), we test which covariates best explain the distribution of biomass and species composition across the region. Explanatory power of each covariate is spatially variable, suggesting that a single environmental factor alone did not drive prairie and savanna vegetation in the Prairie Peninsula. Rather, this work suggests that variable interactions between climate, soil hydrology properties, and disturbances promote similar vegetation structure across space.