MODELING SPECIES ABUNDANCE WITH IMPERFECT DETECTION USING ANGLING

Loading...
Thumbnail Image
Date
Authors
Gehrke, Ty
Keyword
Occupancy , Abundance , Fish , Angling , Loughborough
Abstract
Modeling species abundance and predicting species distribution are important tools for fisheries management. Advanced modeling techniques have allowed for predictive models with increased accuracy to be developed because of their ability to account for imperfect detection. Novel sampling methods have been created to sample aquatic environments such as large bodies of water that cannot be easily sampled by traditional fisheries methods. I surveyed Loughborough Lake near Kingston, Ontario to develop abundance models that would predict species distribution based on depth and fetch habitat variables using angling as the sampling method. Loughborough Lake is a 2000-hectare lake that has interesting bathymetric and limnological characteristics with two distinct basins. An angling protocol was developed to be able to sample all environments of the lake while being able to efficiently sample sites. Abundance modelling was completed for Largemouth Bass, Smallmouth Bass, Rock Bass, Pumpkinseed, Bluegill and Yellow Perch. Results showed that a broad distribution of centrarchid species, and other species, was not observed throughout Loughborough Lake. In contrast, centrarchid distribution was characterized as basin-specific in nature with different species occupying either the eastern or western basin. This pattern was unexpected, but it may be explained by variation in fetch, which functions as a surrogate for vegetation. Angling was found to be an effective method that allowed an individual sampler to efficiently obtain count data. The valuable information gained from the combination of the abundance models and a novel sampling method could improve sampling programs and aid in future research.
External DOI