Predicting coaster brook trout distribution based on landscapes, habitats, fish communities, and life history traits

Coaster brook trout, Salvelinus fontinalis, were once found throughout the Great Lakes until habitat degradation and over-harvest decimated populations. Coaster brook trout are currently the target of a number of restoration efforts and have recently been petitioned for listing under the Endangered Species Act. Little is known about in-stream factors associated with coaster brook trout stream habitat or the watershed context in which those streams are embedded. A series of analyses were conducted to predict the stream habitat and watershed-scale factors associated with coaster brook trout presence and absence. For all parts of this work, habitat data were collected through field surveys. Watershed-scale data were tabulated in a Geographic Information System (GIS) using available spatial data. Habitat-scale predictors of coaster brook trout were shallow stream depths, high proportion of riffles, and low conductivity. Watershed-scale predictors of brook trout presence were more conifers and hardwoods, well-drained soils, and few road crossings. Significant explanatory watershed variables such as proportion of hardwoods and proportion standing water explained 30% of the variation in habitat-scale predictors of brook trout presence. For the second part of this study, we hypothesized that brook trout growth rates were related to in-stream habitat, large scale habitat, inter and intra-specific competitors, or a combination of these factors. Growth of wild coaster brook trout differed among 22 tributaries to Lake Superior. The best predictors of growth rates were pebble size, conductivity, stream width, and proportion of riffles. The habitat features that influenced growth rates were related to landscape features such as conifer, standing water, and road crossings. The third part of the study was to determine shared variation among life history attributes and habitats to illustrate a mechanism by which landscape change shapes fish communities. We found that landscapes explain more variability in fish assemblage life history traits (38.2%) than smaller-scale habitat (23.2%). Disparity in variances may be due to temporal fluctuation in habitats at different scales. Because landscapes are less temporally variable than smaller scale habitats, landscapes may better describe (and predict) fish assemblage life history features. Results from this work may be useful for development of watershed-scale landuse management policy to protect and restore coaster brook trout.