Date of Award

5-2012

Document Type

Thesis

Degree Name

Master of Science (MS)

Legacy Department

Wildlife and Fisheries Biology

Advisor

Baldwin, Robert F

Committee Member

Homyack , Jessica A

Committee Member

Post , Christopher J

Abstract

Ephemeral wetlands are ecologically important freshwater ecosystems that occur frequently throughout the Atlantic coastal plain ecoregions of North America. Despite the growing consensus of their importance and imperilment, these systems historically have not been a national conservation priority. They are often cryptic on the landscape and methods to detect ephemeral wetlands remotely have been ineffective at the landscape scales necessary for conservation planning and resource management. Therefore, this study fills information gaps by employing high-resolution light detection and ranging (LiDAR) data to create local relief models that elucidate small localized changes in concavity. Relief models were then processed with local indicators of spatial association (LISA) in order to automate their detection by measuring autocorrelation among model indices. Following model development and data processing, field validation of 114 predicted wetland locations was conducted using a random stratified design proportional to landcover, to measure model commission (α) and omission (β) error rates. Wetland locations were correctly predicted at 85% of visited sites with α error rate = 15% and β error rate = 5%. These results suggest that devised local relief models captured small geomorphologic changes that successfully predict ephemeral wetland boundaries in low-relief ecosystems. Small wetlands are often centers of biodiversity in forested landscapes and this analysis will facilitate their detection, the first step towards long-term management.

Share

COinS