Bird species distributions across woodland canopy structure gradients

The tree canopy characteristics of two broadleaved woods in southern England were quantified in terms of two independent measures of structure, canopy height (calculated using heights >= 1 m) and percentage canopy cover (derived using heights < 1 m), using airborne Light Detection and Ranging. The woods differed strikingly in structure due to their management systems; one was predominantly mature oak and the other coppice, comprising a patchwork of growth stages. Fine-scale relationships between breeding bird species distributions, determined by mapping censuses, and canopy height and canopy cover were assessed. Despite the differences in structure, species showed great consistency between the woods in their rank positions across gradients of canopy height (rank correlation between woods, r = 0.77, p < 0.001) and canopy cover (r = 0.61, p = 0.003). In both woods, and especially the mature oak (R-2 > 0.90, p < 0.001), there was a positive correlation across bird species between the mean values of canopy height and canopy cover associated with the mapped locations of each species. We suggest that canopy height acts as an effective surrogate of woodland structure and can be applied as a predictor of woodland bird composition and distribution, at least in lowland British conditions. Species associated with young growth were more restricted by habitat structure, as measured by differences in canopy height and canopy cover between the two woods, than were species associated with taller canopies. Remote sensing of canopy height potentially offers a simple, effective way of assessing habitat availability for many species, at both woodland and landscape scales. This may be especially relevant for species dependent on highly transient vegetation structures associated with the early pre-canopy closure stages of forest growth.