Evaluating the efficacy of window treatments to reduce bird–window collisions

Collisions with glass are a leading anthropogenic driver of avian mortality. Window treatments can be highly effective at reducing collisions, however, treating windows remains uncommon and many treatments are applied that do not follow evidence-based standards or guidelines. To evaluate the efficacy of window treatments that meet versus do not meet guidelines, and the drivers influencing collision rates, we conducted standardized collision surveys during spring and fall migration in 2024 and 2025 at window façades in Ottawa, Ontario, Canada. We documented 233 collisions involving 51 species, with 204 collisions at untreated façades and 29 at treated façades. Using a generalized additive mixed modelling approach, we found that glass area and the proportion of vegetation and tree canopy around façades were the largest drivers of collisions. Window treatments that met guidelines had a significant negative effect on collision risk (βtreatment met guidelines = -2.24, SE = 0.74, p < 0.01). Critically, we found no statistically significant difference in collision risk at façades with window treatments that did not meet guidelines and collision risk at untreated façades (βtreatment did not meet guidelines = -0.33, SE = 0.31, p = 0.28). Further, using a random forest model approach that used untreated window data to predict a “no treatment” collision rate at currently treated windows, we found a significant difference in reduced collision rates between window treatment groups. Window treatments that did not meet guidelines reduced collision rates by 76.3% which was significantly less than the 98.4% reduction when window treatments met guidelines (t = -2.796; df = 169.96; p < 0.01, one-tailed t-test). Window treatments should be prioritized at building façades with large windows near vegetation and canopy and to be effective, treatments must meet guidelines.