Possible resource partitioning in pairs ofPhylloscopusandAcrocephaluswarblers during autumn migration through a south Wales reedswamp

Abstract

Though there is evidence for resource partitioning within Acrocephalus and Phylloscopus warblers on their breeding and wintering grounds, fewer data are available on partitioning during autumn migration. In several respects, this is a time when segregation may be important. Patterns of habitat distribution (between different vegetation types and different capture heights), times of passage, and daily pattern of occurrence were therefore compared between two species within each of these genera during their autumn migration through a reedswamp in South Wales in 1988 and 1989. In Reed and Sedge Warblers, passage peaked contemporaneously, diurnal patterns of occurrence were almost identical, and capture height did not differ in either reedswamp or sallow carr (though vegetation type did significantly affect capture height). However, in both years, Sedge Warblers were significantly more numerous in reedswamp than Reed Warblers. By contrast, Willow Warblers and Chiffchaffs occurred always in the same habitats, and at similar heights in the vegetation, though there was evidence of different diurnal patterns of occurrence, and strong evidence of different seasonal patterns of occurrence. One inference is that the two Phylloscopus species partitioned resources by occurring at the site at different times, while the two Acrocephalus species partitioned resources by the use of different habitats (and different foods). Other reasons for these temporal and habitat patterns (e.g. contrasting morphology in Reed and Sedge Warblers, different migratory journeys in Willow Warbler and Chiffchaff) have been suggested previously, but they are not necessarily evolutionary explanations. The segregation of resources between the species in each genus should not be overlooked as reasons for their incompletely understood migratory strategies. Though governed by some difficulties of interpretation, this area is worthy of further study using data of the type generated by ringers.