INFLUENCES OF POPULATION-SIZE AND WOODLAND AREA ON BIRD SPECIES DISTRIBUTIONS IN SMALL WOODS

Distributions of individual bird species in 151 small woods (size rang e 0.02-30 ha) were investigated in 3 consecutive years during which th e abundance of certain species varied markedly. Relationships between the probabilities of certain bird species breeding and woodland area w ere described using incidence functions derived from logistic regressi on analysis. In general, for species which were largely dependent on w oodland and seldom occurred in other habitats (such as hedgerows and g ardens), the probability of breeding approached 100% only for woods of 10 ha and more, whereas species with less stringent habitat requireme nts occurred in the majority of woods, including those of 1 ha and les s. The sensitivity of incidence functions to changes in regional abund ance and the size distribution of the study woods was examined. For so me species, distribution patterns could not be distinguished from thos e expected if pairs had been distributed in proportion to woodland are a (random placement), but the majority did not conform to random place ment in at least 1 of the 3 years. This nonconformity was consistent a cross all 3 years for some species, such as wren (Troglodytes troglody tes), despite substantial fluctuations in population sizes between yea rs, while for others, such as robin (Erithacus rubecula), distribution patterns changed with changes in regional abundance. The results sugg ested that some species, such as wren and blackbird (Turdus merula), p referred small woods, while other species, such as chiffchaff (Phyllos copus collybita), preferred large woods. For several other species, in cluding robin, great tit (Parus major), long-tailed tit (Aegithalos ca udatus) and marsh tit (P. palustris), small woods appeared to be sub-o ptimal under at least some conditions.