Measuring ecological connectivity with ecological distance and dynamic resistant kernels

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Context: Species-agnostic connectivity models are often used to inform management over broad spatial scales. The four main approaches to species-agnostic models parameterize resistance to movement based on naturalness, structural features, climate, or geodiversity variables. Though all four of these factors simultaneously affect species movement and flow of ecological processes, they are rarely combined. Objectives: We built upon an approach that uses all four of these factors to model current and future ecological connectivity for the Crown of the Continent Ecoregion, in Canada and the USA. Methods: We estimated resistance for each pixel on the landscape based on multivariate ecological distances to surrounding pixels. We then modeled connectivity with resistant kernels at different scales, and dynamically in response to future climates from 2020 to 2080. Results: Across the study area, we found median connectivity values decreased by 17–50% from 2020 to 2080 depending on the scale, with broader scales experiencing greater losses in connectivity. Though often considered natural conduits for movement, stream and valley bottoms generally lost connectivity through time. Wilderness areas had significantly higher connectivity values than unprotected lands for all time steps and scales, indicating their importance for maintaining future connectivity of ecological processes. Conclusions: We offer an updated approach for species-agnostic connectivity modeling that combines naturalness, structural features, and topo-climatic layers while considering multiple scales of ecological processes over a large spatial extent and dynamism through time. This approach can be applied to other landscapes to produce products for short- and long-term management of connectivity and ecological resilience.

This publication has 42 references indexed in Scilit: