From pattern to process: estimating expansion rates of a forest tree species in a protected palm savanna

Abstract

We assessed the possible influences of dominant tree density (Butia yatay palm trees) and fire on the expansion of a riparian tree population (Myrcianthes cisplatensis) over El Palmar National Park, a protected savanna in Argentina. Our approach is based on Skellam´s model of population expansion, which predicts that populations with density-independent reproduction and random dispersal will exhibit Gaussian-shaped expansion fronts. Using Poisson regression, we fitted Gaussian curves to Myrcianthes density data collected at varying distances from a riparian forest, within four environmental conditions resulting from combinations of palm density (dense and sparse) and fire history (burned and unburned). Based on the estimated parameters, we derived statistics appropriate to compare attained expansion velocity, mean squared effective dispersal distance, and density-independent population growth among environmental conditions. We also analyzed the effects of palm density, fire history, and distance from the riparian forest on local maximum size of Myrcianthes individuals. Gaussian curves fitted the data reasonably well and slightly better than two alternative front models. Palm density and fire history interacted to control Myrcianthes spread, making unburned dense palm savannas the preferential avenue for Myrcianthes population expansion across the landscape. Limitation of Myrcianthes expansion by fire appeared to result from low survival of small individuals to fire, whereas facilitation of Myrcianthes expansion by palm trees may have resulted from increased population growth. Our results stress the interactive role of fire regime and local biotic influences in determining propagule pressure and tree establishment at the forefront, and the overall vulnerability of savannas to colonization by forest species.