Diverse responses of vegetation and fire after pleistocene megaherbivore extinction across the eastern US

Abstract

Megaherbivores are keystone species whose removal from landscapes can cause cascading ecosystem changes, yet the consequences of Late Quaternary megaherbivore extinctions remain uncertain. This paper tests the Megaherbivory Release Hypothesis (MRH), which posits that the decline and extinction of megaherbivores (body size >1000 kg) during the last deglaciation in eastern North America contributed to the expansion of more palatable hardwood tree taxa, the development of vegetation assemblages with no modern analogue, and increased fuel load and fire activity. Coprophilous fungal spores in lake sediment records are used as proxies for megaherbivore abundance and are essential to testing the MRH through analyses of lead/lag relationships among vegetation composition, megaherbivore abundance, fire, and climate. Although some prior analyses of coprophilous fungal spores from individual sites have supported the MRH, these interpretations have been complicated by 1) discrepancies in the timing of coprophilous spore declines versus megaherbivore extinction timing based on dated vertebrate remains, 2) reliance on a single fungal taxon (Sporormiella) rather than a full suite of coprophilous fungi taxa, and 3) uncertainties in the taphonomic processes that influence fungal spore abundances. To examine the spatiotemporal relationships among megaherbivory, vegetation, and fire, we developed five new multi-taxon coprophilous fungal spore records for comparison with existing pollen, spore, and charcoal records from 14 sites across eastern North America. The MRH was well supported in the northeast and central US, with most sites showing a coprophilous spore decline by ∼14.6 ka followed by a rise of hardwood taxa (∼14.4 ka). However, changes in fire regime varied widely among northeast and central US sites and may have preceded spore declines. The MRH was not well supported in the southeastern US, where a smaller rise in hardwood taxa (∼16.1–13.1 ka) generally preceded the decline in coprophilous spores at individual sites (∼15.8–12.7 ka). These site-level and regional differences suggest spatial variations in the strength of couplings among late-Quaternary megaherbivore extinctions, vegetation composition and structure, and fire regime. Possible explanations for the differences between the northern and southeastern US include (1) differences in landscape heterogeneity of canopy openness and palatability, (2) net primary productivity and sensitivity to top-down trophic effects, (3) megaherbivore density, and (4) climate trends and seasonality at orbital to millennial timescales.