摘要

Species distribution models (SDMs) are currently being used to identify essential fish habitat and guide fisheries management worldwide. We present SDMs based on generalized linear mixed models (GLMM) of the fall distribution or occurrence of juvenile American shad (Alosa sapidissima) and juvenile striped bass (Morone saxatilis) in the Hudson River estuary (HRE) based on data from a fishery-independent survey. The distribution of both species were modeled over a 6-year period (2000-2005) as a function of dissolved oxygen, salinity, water temperature, distance along the HRE denoted as river mile, time or Julian day, distance from submerged aquatic vegetation (SAV), and sediment characteristics. Salinity, river mile, and Julian day were the most important environmental determinants of juvenile American shad presence, and sediment type, salinity, river mile, and Julian day were the most important environmental determinants of juvenile striped bass presence. Calibration plots showed a high level of agreement between predictions generated by each model and actual observations of each species%26apos; occurrence. Based on this result, we mapped the predicted distribution of each species. We found the highest predicted probabilities of juvenile American shad presence in the upper HRE, but the highest predicted probabilities of juvenile striped bass presence were in the lower HRE. Our results suggest that habitat partitioning between these two species is present during the fall in this system but the mechanism is unclear.

  • 出版日期2012-11