Using Habitat Preferences Models, Ocean Forecasts, and Cooperative Research to Avoid River Herring Incidental Catches

Recent attention on declines in river herring (Alewife, Alosa pseudoharengus, and Blueback Herring, A. aestivalis) populations has brought attention to their incidental harvest in the Atlantic Herring (Clupea harengus) fishery. While catch caps were recently added to the Atlantic Herring Fishery Management Plan, strategic methods to reduce incidental catches are also needed (e.g., http://www.umassd.edu/smast/bycatch/). We developed generalized additive models (GAMs) to describe observed habitat preferences for river herring and Atlantic Herring using data from the NOAA Northeast Fisheries Science Center (NEFSC) bottom trawl winter and spring surveys. Observations for all three species (modeled as presence or absence) were related to bottom temperature, bottom salinity, depth, solar azimuth and elevation (proxies for season and time of day), and region. Preliminary model tests were conducted by omitting a subset of the data and comparing model predictions with observations; more than 72% of observations were correctly predicted. We plan to test the accuracy of these models for fishery-dependent data by coupling the habitat preference models with an ocean forecast model (FV COM), and then evaluating the predicted distributions for data collected from 2011-2013 by commercial Atlantic Herring vessels that work with the NEFSC cooperative research program.