Natural disturbance regimes as templates for the response of bird species assemblages to contemporary forest management.

Aim In managed forest landscapes, the tolerance of species to contemporary alteration of forest cover is often assumed to reflect their resilience to natural disturbances. We tested this central tenet of ecosystem-based management by comparing the structure of forest bird assemblages among four regions with contrasting historical natural disturbance regimes. Location Canada's boreal and northern hardwood forests. Methods Using point count data from four study regions across Canada, we first determined the relative sensitivity of individual bird species to the contemporary reduction of old forest cover at stand and 'landscape-context' (1-km radius) scales with log-linear models. The richness of species most sensitive to loss of old forest (hereafter 'sensitive species') was then modelled as a function of landscape-scale changes in old forest cover. Differences in the rate of decline in the richness of sensitive species with contemporary cover of old forest were compared among regions using ANCOVA. We then compared broken-stick regression models with linear models to detect thresholds, if present, in this relationship in each region. Results Bird assemblages from regions with relatively infrequent natural disturbances hosted more species sensitive to contemporary reduction in old forest cover. Those species were also more abundant than in regions with frequent natural disturbances, and the rate of decline in their richness with the loss of old forest was steeper in regions with infrequent natural disturbances than in those where they were frequent. However, we did not detect thresholds in this rate of decline in any study region. Main conclusions Our findings are consistent with the contention that historical natural disturbance regimes shape the response of biota to contemporary landscape alterations through evolutionary adaptation. We argue that forest management conducted within the natural range of variability in stand and landscape structure specific to a region is likely to be ecologically sustainable. [ABSTRACT FROM AUTHOR]