Author: Jennifer E. Bontje,   (University of Toronto)

Description:
Species distribution models (SDMs) are useful tools for analyzing complex species and community data, potentially leading to strong inferences about habitat requirements and community structure. Species at Risk present a unique challenge when modelling distributions, due to the small sample sizes available. Additionally, most SDMs omit biotic factors like incidence of prey, competitors or predators. Co-occurrence modelling can use both abiotic (e.g. water temperature, depth) and biotic factors to give a more complete picture of a species’ distribution. The addition of these biotic factors can lead to substantial insight about the prevalence of antagonistic or synergistic interactions and their effect on a species persistence, as well as serving as surrogates for other unmeasured environmental data. We use co-occurrence modelling to predict the distribution of an endangered warmwater fish that inhabits that coastal wetlands of the lower Great Lakes, Lake Chubsuckers (Erimyzon sucetta ). We also evaluate whether the inclusion of biotic factors can increase the predictive validity of SDMs, focussing on whether this methodology can lead to new hypotheses about threats and limiting factors of Species at Risk fishes on Ontario. Results indicate that the dominant aquatic vegetation, Grass Pickerel ( Esox americanus vermiculatus ), and water temperature are key variables in predicting the presence of Lake Chubsucker.

Category: Scientific Program Abstract > Special Session > CS02 Management and Conservation of Aquatic Systems