| Abstrakt: |
AbstractTrout density and biotic integrity scores are central metrics used to guide trout stream management actions. However, it is unclear whether reach-scale habitat characteristics affect trout density and biotic integrity in a similar fashion. To determine the relative strength of relationships between reach-scale habitat characteristics and important biological metrics, we used artificial neural network models to examine the relationships between 11 reach-scale habitat variables and (1) catch per effort (CPE) for brook trout Salvelinus fontinalis, (2) CPE for brown trout Salmo trutta, and (3) coldwater fish index of biotic integrity (IBI) scores. The trout CPE models generally included habitat features related to physical characteristics; the IBI model generally included physical characteristics as well as those related to water quality. The brook trout CPE model included sinuosity and percent pool area. The brown trout CPE model included gradient and cover. The coldwater IBI model included gradient, percent fine sediments, buffer width, and width : depth ratio. Habitat restoration efforts that seek to maximize brook trout CPE, brown trout CPE, or IBI may benefit from consideration of habitat variables associated with each metric. We also performed quantile regression to evaluate whether any one of the three response metrics of interest was limiting to any of the other variables. Conditions that control IBI scores may also limit brook trout and brown trout densities; 90th-percentile regressions between brook trout or brown trout CPE and IBI score were significant. Although our findings suggest that IBI scores and trout densities can often be increased simultaneously, the fact that different habitat variables were included in each model suggests that the most appropriate habitat restoration efforts may also differ depending on the management goal. |
Plný text