Angler numerical response across landscapes and the collapse of freshwater fisheries.

Recreational angling opportunities in lakes are distributed across landscapes and attract anglers based on the combination of angling quality, travel distance, and availability of facilities. The relationship between angler density and fishing quality, as measured by catch rate, represents a numerical response that is analogous to a predator numerical response to variability in prey abundance. We quantified this numerical response of anglers to rainbow trout, Oncorhynchus mykiss, populations distributed over a large lake district in south-central British Columbia, Canada. We developed a harvest dynamics model by linking this empirical description of the spatial numerical response of anglers to a logistic population growth rate model. The model was parameterized for rainbow trout and simulated spatial patterns of angler density and catch rates over a landscape. At locations distant from urban centers, angler density is low and catch rate high, suggesting near pristine conditions; at intermediate distances angler density is higher while catch rates are lower and approximate maximum sustainable levels; and at short distances angler density is sufficiently high to harvest to local extirpation. We extrapolated the model to other lake districts varying in human population size using an empirically derived angling participation rate relationship. Extrapolation to lake districts with one-tenth the human population maintained viable fisheries close to the urban area, and districts with 10 times the human populations could not maintain viable fisheries across much of their lake district. Landscape-scale spatial patterns differed quantitatively for species varying in rates of intrinsic population growth and carrying capacity, but the qualitative spatial patterns were consistent among species, demonstrating the pervasive impacts of the angler numerical response. To achieve a management goal of sustaining fisheries across landscapes, a change in management perspective is necessary, from that of individual lakes to one of dynamic harvest processes across landscapes. This new approach makes it clear that a one-size-fits-all management approach must be replaced with a mosaic of approaches cognizant of landscape-scale processes.