Session A1- An Optimization Planning Model for Maximizing habitat connectivity for stream resident fish
Location
UMass Amherst
Start Date
27-6-2011 10:20 AM
End Date
27-6-2011 10:40 AM
Description
Systemic methods for prioritizing the repair and removal of fish passage barriers, while growing of late, have hitherto focused almost exclusively on meeting the needs of migratory fish species (e.g., anadromous salmonids). An important but as of yet unaddressed issue is the development of new modeling approaches which are applicable to resident fish species habitat restoration programs. In this talk, we develop a budget constrained optimization model for deciding which barriers to repair or remove in order to maximize habitat availability for stream resident fish. Habitat availability at the local stream reach is determined based on the recently proposed Cind metric, which accounts for the quality, distance and level of connectivity to different stream habitat types. We investigate the utility of the Cind metric, as well as assess the computational performance of our model using geospatial barrier and stream data collected from the Pine-Popple Watershed, located in northeast Wisconsin, USA. The Cind metric is found to be a statistically significant predictor of common resident fish species richness, indicating its usefulness in assessing the ecological impacts of barrier repair and removal decisions. We also find the optimizing model to be an efficient and practical decision support tool. Optimal solutions, which are useful in informing basin-wide restoration planning efforts, can be generated on average in only a few minutes.
Session A1- An Optimization Planning Model for Maximizing habitat connectivity for stream resident fish
UMass Amherst
Systemic methods for prioritizing the repair and removal of fish passage barriers, while growing of late, have hitherto focused almost exclusively on meeting the needs of migratory fish species (e.g., anadromous salmonids). An important but as of yet unaddressed issue is the development of new modeling approaches which are applicable to resident fish species habitat restoration programs. In this talk, we develop a budget constrained optimization model for deciding which barriers to repair or remove in order to maximize habitat availability for stream resident fish. Habitat availability at the local stream reach is determined based on the recently proposed Cind metric, which accounts for the quality, distance and level of connectivity to different stream habitat types. We investigate the utility of the Cind metric, as well as assess the computational performance of our model using geospatial barrier and stream data collected from the Pine-Popple Watershed, located in northeast Wisconsin, USA. The Cind metric is found to be a statistically significant predictor of common resident fish species richness, indicating its usefulness in assessing the ecological impacts of barrier repair and removal decisions. We also find the optimizing model to be an efficient and practical decision support tool. Optimal solutions, which are useful in informing basin-wide restoration planning efforts, can be generated on average in only a few minutes.
Comments
Jesse O'Hanley is a lecturer (assistant professor)in the Kent business School, University of Kent, UK. He obtained his Ph.D. in Environmental Science, Policy and Management from the University of California, Berkeley in 2005, and subsequently worked as a postdoctoral fellow in the Environmental Change Institute at the University of Oxford. He joined the Kent business School in 2006. His research focuses on the application of statistics, optimization and other operations research techniques to environmental planning and management. Recent and current lines of inquiry include river infrastructure mitigation and placement, nature reserve network design, and species distribution modeling.