Journal
WETLANDS
Volume 32, Issue 5, Pages 841-850Publisher
SPRINGER
DOI: 10.1007/s13157-012-0315-7
Keywords
Community distribution modeling; MaxEnt; Maximum entropy modeling; Patch-based ranking; Wetland mitigation; Wetland restoration
Categories
Funding
- Upper Susquehanna Coalition [CD97225309]
- Department of Environmental and Forest Biology at the State University of New York College of Environmental Science and Forestry [CD97225309]
Ask authors/readers for more resources
Current wetland mitigation practices do not fully recover wetland function, often due to poor mitigation site selection. Improved mitigation site selection methods are needed to efficiently assess the suitability and quality of potential wetland mitigation sites at broad spatial scales. We present a novel application of maximum entropy-based predictive distribution models coupled with a patch-based ranking scheme to identify potential wetland mitigation sites and contrast their effectiveness relative to a conventional expert opinion model. We used hydrogeologic and landscape features and widely available wetland community distribution data to predict locations of wetlands that were previously unknown, destroyed, or biologically rare in the Upper Susquehanna River Basin in the northeastern United States. An expert opinion model predicted wetland occurrence based on topographic slope and soil type. Maximum entropy-based models predicted an independent sample of wetland locations well (Area Under the Curve = 0.86-0.98; 92 % correct classification rate) and dramatically outperformed the expert opinion model (62 % correct classification rate). A patch-based ranking scheme, which incorporated additional influences on wetland quality, ranked sites with biologically important wetland plant communities highly among model-identified wetlands. We conclude that integration of maximum entropy-based predictive modeling and patch-based ranking can effectively identify high quality wetland mitigation sites.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available