Future climate scenarios are expected to elicit significant regional vegetation responses. Understanding past vegetation dynamics provides a context within which we can evaluate future changes. Specifically, it is important to understand how disturbance processes (e.g., fire) and ecological processes (e.g., succession and dispersal) have interacted to produce regional vegetation patterns in response to changing climate. I propose that spatially explicit, regional-scale vegetation models can complement traditional paleoecological methods by demonstrating mechanisms behind spatiotemporal patterns observed in empirical data. As a case study, I used the LANDIS-II model to simulate dense hardwood forest expansion into surrounding prairies and savannas over the past millennium in the Big Woods region of south-central Minnesota. Isolating specific factors in the model demonstrated their relative importance in explaining the observed pattern of vegetation change in the region.

Results/Conclusions

Simulations confirmed the importance of fire in the region, and suggested that water bodies and rough topography served as firebreaks behind which the fire-sensitive hardwoods could establish. In this case, fire likely impeded the spread of hardwood forests into the prairie in response to climate change, which suggests that some systems may experience slow responses to future climate change. In addition, relatively long seed dispersal distances may have been required to explain the rapid development of the Big Woods forest.