AUTHORS: Hossam E. M. Abdel Moniem*, Department of Biology, University of Toronto & Zoology Department, Suez Canal University; Ronan Marrec, Université de Picardie Jules Verne; Majid Iravani; Jahan Kariyeva; Branko Hricko; Christine Gray; Péter Sólymos; Ermias Azeria, Alberta Biodiversity Monitoring Institute, University of Alberta; Helene Wagner, Department of Biology, University of Toronto
ABSTRACT: Species-agonistic connectivity models based on how anthropogenic activities affect landscape features can add to our understanding of species assembly and community composition. These coarse-filter approaches rely on expert opinion to quantify the degree to which human activities affect habitat connectivity. Yet, such models have not been validated with biodiversity data at large spatial scales. We used the Alberta Biodiversity Monitoring Institute’s (ABMI) data on multi-taxa biodiversity and human footprint (information on anthropogenic land use patterns) to assess the power of species-agonistic connectivity models to explain community structure of vascular plants and birds. We aimed at: (1) quantifying landscape connectivity across Alberta and assessing local landscape context of key habitats; and (2) assessing the benefit of species-agonistic models for explaining community structure. We modeled landscape resistance to movement by categorizing 101 landscape feature types (due to physical human footprint and/or intensity of use), topography, and natural water bodies. We tested uncertainty around our metrics and compared 15 concurrent wall-to-wall resistance, and calculated connectivity maps computed with varying combinations of intensity indices. In the second part, we assessed the relative importance of our connectivity metrics, patch-level metrics, and landscape-level metrics, for explaining variation in community composition at 1100 sampling sites regularly distributed across Alberta. Results showed surprisingly low correlations (R2= 0.49–0.94) between connectivity maps. Within each map, correlations between different spatial resolutions (10m, 30m, and 100m) differed remarkably. Variance partitioning illustrated the power of species-agnostic connectivity maps as predictors of community structure, individually and when shared with both patch and landscape metrics. We conclude that the way resistance is conceptualized as a combined effect of the degree of human footprint and intensity of use matters for the validity of species-agnostic connectivity models. Our study could provide insights to land use planers and decision makers towards promoting biodiversity and ecosystem services.