Although the extensive decline of North American grassland birds has been investigated, many of the identified causes of decline cannot be generalized across species.  Mechanisms behind decline are likely multi-causal and multi-scalar, yet most research focuses on single-scales and singular causes.  We investigated the decline of grassland birds in the largely undisturbed dry mixedgrass subregion of Alberta using 16 landscape and habitat variables grounded in previous research.  These were distance to body-mass edge, habitat and dietary specialization, dietary and foraging strategy, body mass, migratory strategy, and proximity to their edge of geographic range.  A body-mass aggregation is a group of species with similar body masses that interact with the environment at the same scale, and extinction risk may be associated with a species’ distance to the edge.  Aggregation edges define the transition to a different scale of resource structure and are areas of both opportunity and risk.  Aggregations were identified using a Bayesian CART and null models combined with simulations. Birds were defined as declining with a combination of provincial and federal status classification and Breeding Bird Survey data.  We built 20 simple models via multi-model inference techniques, assessing models using AICc values and model-averaged parameter estimates.  

Results/Conclusions

The best model was a positive relationship with distance-to-edge; birds farther from edges were more likely to be declining.  Other key variables in the confidence group of models include temperate zone migrants, smaller-bodied birds, and food and habitat specialists.  That smaller-bodied birds with body masses that place them farther from aggregation edges are more likely to be declining contradicts conventional wisdom.  Although counterintuitive, it is possible that extinction risks in stable and relatively undisturbed systems differ from highly disturbed and fragmented systems.  The migration results confirm that extinction risk is increased for temperate-zone migrants and reduced for non-migrants, indicating that the strategy itself increases vulnerability and not just that decline is driven by factors originating within Canada and the US as suggested in other studies.  These results suggest that the mechanisms responsible for shaping avian body mass distributions can also partially explain extinction risk.  They also indicate that drivers of decline are multi-causal and may vary between undisturbed and transformed landscapes.  Additionally, the combination of government status classification and BBS data to derive decline status may be a more robust approach to investigating species declines, as there are shortcomings to both methodologies.