Spatial Relationships between Vegetation and Forest Bird Boundaries in a Mixed-wood Managed Forest

A major current threat to the persistence of forest birds in Canada is forest fragmentation caused by industrial timber harvesting. The traditional silvicultural method of clearcutting creates forest fragments delineated by sharp boundaries. Conversely, selective logging techniques strive to mimic natural disturbance patterns and to produce smoother boundaries. Forest boundaries influence the ecological behaviour of forest birds in a species-specific manner. To assess the effect of forest fragmentation on bird occurrence and to derive management guidelines, I investigate the spatial association between boundaries of six bird species and forest features in a moderately harvested landscape (MHL) and an intensively harvested landscape (IHL) in New Brunswick. The focal bird species were used as indicator species for the effects of forest fragmentation. These analyses were carried out at both stand and landscape scales. I hypothesized that 1) the association of birds with forest variables would be i) species-specific and ii) stronger in the IHL than in the MHL; 2) that birds would have more common boundaries in the in the ML than in the MHL; 3) that both scales would yield complementary results, whereby the stand scale would refine coarse-scale ecological relationships of birds. By means of boundary detection algorithms and overlap statistics, I found that boundary spatial associations between vegetation and birds were more clearly defined at the stand scale in the IHL. I suggest that the spatial confinement of forest birds in clearcut forests fragments accounts for this relationships. Overall, understory forest components primarily influenced bird occurrence, yet bird response to forest features was species-specific. My findings were consistent at both analysis scales. Management efforts to preserve forest birds should strive to recreate the complexity and variability of natural ecosystems. A multi-scale approach is required in ecological research and monitoring of bird species in fragmented landscapes. For further research, I recommend that studies should include uncut controls, a broader range of treatments than the two that I analyzed, and some replication at all spatial scales. New spatial analytical tools such as boundary detection algorithms and overlap statistics are an asset to ecological studies. The methodology I used offers an objective means for evaluating land fragmentation and the effects of edgeslecotones on populations.