Abstract
AbstractThe increased integration of ecosystem service concepts into natural resource management places renewed emphasis on prediction and mapping of fish biomass as a major provisioning service of rivers. The goals of this study were to predict and map patterns of fish biomass as a proxy for the availability of catchable fish for anglers in rivers and to identify the strongest landscape constraints on fish productivity. We examined hypotheses about fish responses to total phosphorus (TP), as TP is a growth‐limiting nutrient known to cause increases (subsidy response) and/or decreases (stress response) in fish biomass depending on its concentration and the species being considered. Boosted regression trees were used to define nonlinear functions that predicted the standing crops of Brook Trout Salvelinus fontinalis, Brown Trout Salmo trutta, Smallmouth Bass Micropterus dolomieu, panfishes (seven centrarchid species), and Walleye Sander vitreus by using landscape and modeled local‐scale predictors. Fitted models were highly significant and explained 22–56% of the variation in validation data sets. Nonlinear and threshold responses were apparent for numerous predictors, including TP concentration, which had significant effects on all except the Walleye fishery. Brook Trout and Smallmouth Bass exhibited both subsidy and stress responses, panfish biomass exhibited a subsidy response only, and Brown Trout exhibited a stress response. Maps of reach‐specific standing crop predictions showed patterns of predicted fish biomass that corresponded to spatial patterns in catchment area, water temperature, land cover, and nutrient availability. Maps illustrated predictions of higher trout biomass in coldwater streams draining glacial till in northern Michigan, higher Smallmouth Bass and panfish biomasses in warmwater systems of southern Michigan, and high Walleye biomass in large main‐stem rivers throughout the state. Our results allow fisheries managers to examine the biomass potential of streams, describe geographic patterns of fisheries, explore possible nutrient management targets, and identify habitats that are candidates for species management.Received May 20, 2014; accepted November 6, 2014
Highlights
The increased integration of ecosystem service concepts into natural resource management places renewed emphasis on prediction and mapping of fish biomass as a major provisioning service of rivers
We focused on total phosphorus (TP) because (1) streams in Michigan tend to be phosphorus limited (Hart and Robinson 1990); (2) TP concentrations in water are significantly correlated with total fish standing crop (Hoyer and Canfield 1991; Randall et al 1995); and (3) TP has been shown to drive positive responses in trout and bass fisheries
We used statistical models to map the capacity of riverine habitats in Michigan to support fish biomass
Summary
The increased integration of ecosystem service concepts into natural resource management places renewed emphasis on prediction and mapping of fish biomass as a major provisioning service of rivers. Boosted regression trees were used to define nonlinear functions that predicted the standing crops of Brook Trout Salvelinus fontinalis, Brown Trout Salmo trutta, Smallmouth Bass Micropterus dolomieu, panfishes (seven centrarchid species), and Walleye Sander vitreus by using landscape and modeled local-scale predictors. An important research question underlies the ability to map spatial variability in game fish availability to anglers: what factors constrain fish productivity at the landscape scale? Species–habitat models using as many as 25 habitat variables explained between 35% and 91% of the variation in abundances of 11 fish species in the Genesee River basin, New York (McKenna et al 2006), and other workers have successfully modeled fish abundances by using landscape and local factors (e.g., Gido et al 2006; Stanfield et al 2006). Synthesis of prior work suggests that nonparametric machine learning modeling approaches perform favorably in comparison with linear models (McKenna et al 2006) and that the inclusion of modeled local conditions (e.g., hydrology, nutrients, and temperature) with landscape variables can lead to greater predictive power (Zorn et al 2004)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
