A Comparison of Sampling Methods to Estimate Production Metrics of Littoral Zone Fish Based on Multi-Spatial Habitat Variables of a Manitoba Reservoir

In recent years, government and industry have sought clear definitions and effective means in which to estimate and predict metrics of fisheries productivity.  This study seeks to address this question in part by comparing gillnet, seine and electrofishing methods to determine which method (or combination thereof) produce models having the greatest explanatory capacity of fisheries productivity metrics based on local, lateral and contextual habitat variables.  During the summer of 2012 and 2013, fish communities in 43 littoral sites were sampled from Lac du Bonnet, a reservoir in southeastern Manitoba (Canada).  In addition to fish sampling, habitat assessments were made at the local (e.g. depth, substrate composition, macrophyte cover), lateral (utilization of shores) and contextual (position relative to landscape attributes such as main channel and tributaries) scale in order to characterize sampling locations, and furthermore provide a base for our models.  Using these data sets along with a series of redundancy analyses, we seek to discover which statistical models hold the greatest explanatory power as a function of local, lateral and contextual habitat variables.  In addition, we seek to determine at which scale (local, lateral, contextual) habitat greatly influences metrics of fisheries productivity.