Remote Areas as Potential Restoration Sites in Conservation Landscapes

摘要

The effects of roads and trails on species composition and diversity at landscape scales are uncertain. This leads to uncertainty about the best way to select sites for restoration. Three datasets were used in this analysis: a model that estimates theremoteness of a landscape by calculating the energetic cost required to walk through it, a set of 246 0.1 ha plant species abundance samples, and a remotely sensed plant community map. These data were used to ask three questions about the selection of restoration sites in Great Smoky Mountains National Park (Tennessee and North Carolina, US). First, does compositional similarity within a plant community change as a function of remoteness? Second, does plant species diversity, particularly of rare species, change with remoteness? Finally, how does the patch area of plant communities change as a function of remoteness? Compositional distance among pairs of sites increased as difference in energetic cost to reach the sites increased. Regression of Shannon-Wiener diversity against the multiplicative effects of energetic cost and community type revealed no significant effect of accessibility on diversity. Energetic cost also showed no significant effect on the proportion of species which occur only once in the dataset. Analysis of the landscape distribution of vegetation communities revealed that the average area of community patch does not vary by remoteness. These results suggest that accessible restoration sites are equivalent to remote sites for conserving diversity and landscape structure. Species composition of remote sites does help inform the acceptable variance in composition of restored communities.