Estimating fishing effort across the landscape: A spatially extensive approach using models to integrate multiple data sources

Measuring fishing effort is one important element for effective management of recreational fisheries. Traditional intensive angler intercept survey methods collect many observations on a few water bodies per year to produce highly accurate estimates of fishing effort. However, scaling up this approach to understand landscapes with many systems, such as lake districts, is problematic. In these situations, spatially extensive sampling might be preferable to the traditional intensive sampling method. Here we validate a model-based approach that uses a smaller number of observations collected using multiple methods from many fishing sites to estimate total fishing effort across a fisheries landscape. We distributed on-site and aerial observations of fishing effort across 44 lakes in Vilas County, Wisconsin and then used generalized linear mixed models (GLMMs) to account for seasonal and daily trends as well as lake-specific differences in mean fishing effort. Estimates of total summer fishing effort predicted by GLMMs were on average within 11 % of those produced by traditional mean expansion. These estimates required less sampling effort per lake and can be produced for many more lakes per year. In spite of the higher uncertainty associated with model-based estimates from fewer observations, the improvements associated with the addition of only three aerial observations per lake highlighted the potential for improved precision with relatively few additional observations. Thus, the combination of GLMMs and extensive data collection from multiple sources could be used to estimate fishing effort in regions where intensive data collection for all fishing sites is infeasible, such as lake-rich landscapes. By using these methods of extensive data collection and model-based analysis, managers can produce frequently updated assessments of system states, which are important in developing proactive and dynamic management policies.