Composition and structure of modern forests in Eastern North America have been influenced by past anthropogenic disturbances such agriculture, logging, and mining. Invasion by non-native species is a product of land-use history, but one which can be expected to differ between land uses with contrasting impacts on the physical and biotic environment. The goal of this project was to compare levels of invasion between forests with histories of agriculture, logging, and mining to test for contrasting land-use legacies. We tested the following hypotheses: 1) Levels of invasion would differ between land-use histories, with the highest abundance of non-native species in sites with histories of soil disturbance that opens habitat for invasion; 2) Invasion patterns differ between land-use histories because environmental gradients reflect land-use history; 3) Distributions of non-native shrubs differ between land-use histories according to individual species’ habitat preferences and life histories.
Results/Conclusions
Research was conducted at Powdermill Nature Reserve in the Allegheny Mountains of Southwestern Pennsylvania. Powdermill has been divided into 120 x120 meter grid cells using GIS technology, and the land-use history of each cell has been determined from aerial photos. We selected four cells each that have histories of agriculture, mining, and logging, and 14 continuously forest cells. Vegetation was surveyed using a modified Whittaker plot. We collected and analyzed soils from a subset of plots per cell. Total non-native cover and cover values for the most common non-native species were analyzed with generalized linear mixed effects models that used land-use history and site characteristics as predictor variables.
Non-native cover was significantly higher in post-agricultural and mined forests than logged stands and older second-growth. Older forests were least-invaded across all measurements. Presence of non-native species is most strongly linked to land-use histories that involve soil disturbance and open new habitats. Common species showed differing land-use associations. Berberis thunbergii was most abundant in post-agricultural forests with nutrient-rich, less acidic soils. Rosa multiflora was most abundant in mined sites due to its tolerance for rocky, nutrient poor soils. Distributions were species-specific and represented interactions of land-use history, life history, and environmental gradients.