Palaeoecological reconstructions of the American Midwest during the last deglaciation suggest the expansion of parkland biomes lacking modern analogues. Despite their spatial extent and persistence over several millennia, the landscape configuration and environmental drivers for the ‘no analogue’ biomes remain speculative. Here we use regression analysis linking settlement-era forest composition and wetland extent to identify specific trees and forest units strongly indicative of high wetland prevalence. We then recompile a regional pollen time series to show transient increases in these flood-tolerant trees, with prominent peaks during the Bølling–Allerød interstadial (~14.6–12.8 thousand years ago). Taken together, the pollen records, and analyses of settlement-era forest composition and wetland prevalence, suggest the establishment of extensive deglacial wetlands in the American Midwest (40–60% of land cover). This extensive, yet transient, wetland was possibly supported by southward discharge of Laurentide Ice Sheet meltwater during the Bølling–Allerød. The timing of this wetland expansion and its mid-latitude location have implications for deglacial methane source dynamics; our estimate of ~11 Tg yr−1 of methane is comparable to the northern source enhancement modelled from ice-core records. Regional decline of the Laurentide Ice Sheet meltwater discharge at the onset of the Younger Dryas (~12.8–11.7 thousand years ago) explains in part why these no-analogue wetland-rich parklands sharply declined, weakening this potential methane source.
Bibliographical noteFunding Information:
We acknowledge funding for this research from the University of Toronto Centre for Global Change Science and the Connaught International Scholarship for Doctoral Students to E.B. and from the Natural Sciences and Engineering Research Council (Canada) to S.A.F. and S.A.C. We thank the many contributors to the Neotoma database for making pollen datasets available for further analyses, and we thank Y. Fan for sharing the water-table depth simulation results.
© 2020, The Author(s), under exclusive licence to Springer Nature Limited.
All Science Journal Classification (ASJC) codes
- Earth and Planetary Sciences(all)