TEMPORAL DYNAMICS OF FOREST PATCH SIZE DISTRIBUTION AND FRAGMENTATION OF HABITAT TYPES IN PENNSYLVANIA
Open Access
- Author:
- Bishop, Joseph A.
- Graduate Program:
- Ecology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- November 13, 2007
- Committee Members:
- Wayne Lawrence Myers, Committee Chair/Co-Chair
Robert P Brooks, Committee Chair/Co-Chair
Richard Tucker Yahner, Committee Member
Shelby Jay Fleischer, Committee Member
Douglas Alan Miller, Committee Member - Keywords:
- forest fragmentation
GIS
response guild
land cover change
ecological landtypes - Abstract:
- Pennsylvania forests have been reported to be changing inefficiently in terms of developed land change per unit population. Change was quantified through two eras of satellite-derived land cover data separated by nine years, 1992 and 2001. Validation of these data revealed accuracies between forest and non-forest classes greater than 80%, 83% and 88%, respectively. Forest change that occurred between the two eras was analyzed to determine our ability to detect forest change and its impacts on avian habitats. Mapping revealed increasing forest fragmentation patterns, in some regions, that are potentially detrimental to avian communities. Forest fragmentation change was tracked using a variety of geographic information system (GIS) techniques. Changes among patch size and between edge and core forest were calculated. Forest cover was classed by collecting contiguous forest patches into size classes ranging from 1 ha to 25,000 ha for each era. Areas that changed patch class were analyzed. More than 60% of small patches (< 10 ha) converted to non-forest while total number of patches in small classes, nevertheless, increased at the expense of patch classes between 25 ha and 100 ha. In contrast, larger patch classes, > 500 ha, tended to remain in their original class. Core versus edge conversion was similar with 34% of previously edge forest changing to non-forest. Ecological landtype association (LTAs) boundaries were selected to analyze forest change because their size (mean = 1078 ha) and ecological relevance. Fragmentation change metrics were calculated within LTAs. When mapped, most forest change metrics illustrated clear regional fragmentation trends. Frequently, LTAs clustered together showing regions undergoing similar fragmentation. Notable regions include north-central Pennsylvania, showing evidence of fragmentation decrease, and northeastern Pennsylvania showing patterns of increasing fragmentation. To understand the impacts of forest fragmentation on natural habitats, breeding bird survey (BBS) data were used to create forest and grassland guilds to test avian responses. Avian guild richness responded predictably to fragmentation change for all guilds and the grassland area sensitive guild had significant results. With the majority of larger, more stable, forest patches under public ownership, consequently, results help to emphasize the management challenges Pennsylvania faces when managing smaller, privately owned, woodlots.