Reasons for a restricted range: the relative influence of abiotic and biotic predictors of
specialist bee distributions
Erin Francispillai, Jessica Forrest, Ilona Naujokaitis-Lewis
Department of Earth and Environmental Sciences, University of Ottawa
specialist bee distributions
Erin Francispillai, Jessica Forrest, Ilona Naujokaitis-Lewis
Department of Earth and Environmental Sciences, University of Ottawa
Climate is a crucial determinant of a species’ range, continually shaping population distributions based on the effects of temperature, precipitation, light, etc. In order to study these shifts, species distribution models (SDMs) help estimate a species’ range based on data collected from sampled records and environmental datasets. Traditionally, SDMs have only used climatic factors to represent the environment’s suitability for the species in question, however the inclusion of biotic interactions has recently become more prevalent and deemed a necessary factor to consider. For instance, it is reasonable to assume a species’ food source would be a large contributing factor to its range, especially for dietary specialists who depend on a limited variety of resources. Dietary specialization is common among pollinator species, especially solitary bees, however our understanding of the factors that influence their range-wide distributions remains poor, particularly in relation to their resource dependency. On that account, the impact of these biotic predictors of specialist pollinator distributions was studied using the solitary bee species, Andrena erigeniae and Andrena violae, both of which specialize to one plant genera, Claytonia and Viola (respectively); both bee species are found mainly in the eastern United States and south-eastern Canada, despite their food plants (or host plants) being present throughout both countries. Using presence records from 1981-2020 (taken from the Global Biodiversity Information Facility) and the maximum entropy modelling software, MaxEnt, the predicted habitat suitabilities of both Andrena spp. were examined throughout the United States and Canada, assessing the relative influence of abiotic factors (climate) and biotic factors (their host plant’s predicted habitat suitability). The host plants provided large contributions to the model’s prediction, however their inclusion did not increase the model’s predictive power relative to using the climate variables alone. Possible reasons for this discrepancy are suggested, such as the study’s spatial scale; climate has largely been argued to govern distributions at broader macroecological scales. Regional, smaller-scale research may be required to fully analyze the host plant’s contribution on the Andrena spp.’s range margins.