Development of shale gas resources in north central Pennsylvania is driving landscape modification in a region characterized by complex geomorphology and a long history of intense human land use. Thus, land managers are tasked with anticipating ecological responses to fragmentation of young (< 100 years), even-aged forests in variable settings. To address these challenges, we are adapting tools originally developed for rangeland management in the western US: Ecological Sites (ESs) with associated State-&-Transition Models (STMs). In this study, we are developing ESs, across a seven-county region of north central PA, using a geographic information system to partition landscapes by multivariate analysis of four, ecologically relevant geomorphologic variables (elevation, slope, aspect, and curvature) derived from 10-m resolution digital elevation models. We are developing STMs through spatially explicit analysis of two vegetation datasets. Pre-European disturbance vegetation of each ES is being assessed via analysis of corner trees in digitized maps of original land surveys (1780s through 1850s) of 49 townships. Within these same townships, modern forest composition data from 549 sampling plots are being used to determine current species distributions with respect to ESs, identify alternative forest composition states, and to explore potential drivers of changes between states.
Results/Conclusions
A boot-strapped analysis of possible clustering solutions indicated that six clusters (i.e. ecological sites) represented the most appropriate solution. The influence of each geomorphic variable on each ES was explored via contingency-table analysis, which indicated that cluster analysis divided landforms along logical gradients with respect to potential moisture availability. Though preliminary, inferences drawn from analysis of land survey data suggest that historic species distributions were strongly related to geomorphology. In contrast, modern data suggest that the relationship between species distributions and geomorphology has weakened relative to historic conditions. Furthermore, the modern data suggest extremely weak correlations between species abundances, suggesting little coherence in community structure. Species with broad ranges of ecological tolerance, especially red maple (Acer rubrum) and red oak, (Quercus rubra) are by far the most dominant species across the study region. We suggest that historic deforestation and associated modification of dynamic soil properties as the most likely mechanism driving these changes. In many respects, the modern forests of north central Pennsylvania represent novel community assemblages, in the early stages of reassembly following a massive resetting event a mere century ago. ESs and STMs could prove to be valuable management tools in these systems.