Models of the Abundance of Large Brown Trout in New Zealand Rivers

Abstract

Multiple-regression models of the abundance of brown trout Salmo trutta larger than 200 mm total length were developed from combinations of hydrological, catchment (89 sites), physical (59 sites), water quality, and benthic invertebrate biomass (43 sites) variables in New Zealand rivers. Three hydrological and catchment variables explained 44.4% of the variation in trout abundance, and benthic invertebrate biomass alone explained almost 45%. Together, invertebrate biomass and weighted usable area (WUA) for adult brown trout drift-feeding habitat explained 64.4% of the variation in trout abundance. These two variables, representing the food and space available to trout, appeared to be the major determinants of trout abundance, with winter water temperature acting as an overriding factor. Instream cover was also an important factor, whereas lakes, land development, site elevation, and river gradient were of lesser importance. Water quality (within the ranges measured) and presence of rainbow trout Oncorhynchus mykiss were not significant. The model best suited to the application of the instream flow incremental methodology (IFIM) explained 87.7% of the variation in brown trout abundance at 59 sites. Two variables were calculated with flow data: WUA for food production at median flow and WUA for adult brown trout drift-feeding habitat at mean annual low flow. Other variables within the model either did not vary with flow or varied little. This study demonstrates that WUA is an important determinant of adult brown trout abundance, refuting one of the major criticisms of IFIM.