Size Selectivity of AFS North American Standard Gillnets

Gillnets are used to monitor fish populations, yet raw gillnet catches provide biased estimates of population size structure. This bias results from the widely-recognized fact that fish of different sizes have different probabilities of being caught in any particular gillnet. Therefore, calibrations are required to use gillnet catches for making meaningful inferences about fish populations. I developed a software package (in the R language) that provides a new tool for making such calibrations. This tool corrects for the size selectivity of two important processes in gillnetting: wedging and tangling. A novel model averaging technique is used to combine calibrations from several models into a single averaged calibration, which is more reliable than any single model. I used this tool to calibrate North American standard gillnets for several species of management importance in Ontario, Canada: Lake Trout, Walleye, Lake Whitefish, Yellow Perch, and Northern Pike. The reliability of these calibrations depend on the number of fish used to make them, which is low (< 500) for some species (e.g. Lake Trout and Northern Pike). To address this issue, we demonstrate that North American Gillnets can be reliably calibrated for poorly sampled species, by using data from different types of gillnet gangs.