Modeling potential fish densities using existing fish observations, TNC/NHDplus habitat attributes and neural-network models

This project involves the development of landscape scale habitat models for riverine fish in the Northeast US.  It is funded through the Regional Conservation Needs program with NEAFWA and focuses on large-scale prediction of fish habitat availability.  The study area covers 13 states from Maine to Virginia but modeling will initially focus on a selected watershed to establish protocols and test the results.  The goals of the project include: the collection of statewide fish databases from 13 states and the District of Columbia, the establishment of a rigorous quality control and standardization of these databases, the linking of fish data to NHDplus stream segments, intersecting The Nature Conservancy (TNC) derived and NHDplus habitat attributes and the prediction of individual species fish densities using neural-network modeling.  The expected products of the project include GIS-based maps of predicted species densities at the river segment level.  Future work includes the possibility to estimate target fish communities at the watershed and sub-watershed level and the inclusion of tools to model for anthropogenic and climate-based alterations to riverine habitats.