Empirical relationships for sport fish yield based on manipulable variables

Abstract

Although most authors accept that trophic status and angler effort affect fish yield, few predictive relationships use both variables because both are rarely available in the same data set. To fill this lacuna, we have developed a data set representing lakes of various trophic levels lying on both igneous and sedimentary drainage basins in Ontario. These data include estimates of sport fish yield, lake morphometry, lake chemistry, and fishing effort. With these data, we developed five models. None of these models were significantly better than the rest but all predicted sport fish harvest effectively despite the presence of ratios in three of the models. The model that is most attractive because it contains no ratios, explains 86% of the variation in sport fish catch, and uses manipulable variables, is: Log10catch (kgseason−1) = 0.685 log10angler hours (hoursseason−1) + 0.366 log10TP (total phosphorus)-area (mg m−1) – 0.124 (log10mean depth)3(m) −1.028. © 1994 Taylor and Francis Group, LLC.