Intraspecific Comparison of Ecology and Mating System of the Great Reed Warbler Acrocephalus arundinaceus: Why Different Results from Different Populations?

In the last 20 years intensive studies on the ecology and mating system of the Great Reed Warbler Acrocephalus arundinaceus have been conducted at five different localities in Japan and Europe. Although all the populations showed similar territorial polygyny, the studies showed partly different results and conclusions concerning the evolution of polygyny in this species. Here, we compare ecology and behaviour of the warblers from different populations based on published information, focusing on aspects of territorial polygyny inherent in the polygyny-threshold-model. Breeding results in one of the European populations were unique in that secondary females were far less successful than monogamous females compared to almost equivalent success in other four populations. Although causes of offspring mortality were not the same between populations, the European populations commonly showed higher rates of chick starvation in secondary nests than in monogamous nests, but this was not the case for the Japanese populations. Benefits which compensate for these costs of polygyny were found in three of the five populations. Despite variations in breeding results, all studies showed the existence of a potential cost of polygyny: reduced paternal care in secondary nests leading to a realized cost depending on climatic and ecological conditions. Evidences of female choice based on behavioural observations were available at some populations. In one of the populations the benefit of polygyny was linked to variance in territory quality caused by habitat heterogeneity and habitat partitioning pattern between males. But information on these aspects of ecology has not been collected enough for the other populations. Features of territory differed greatly between populations and it suggests the possibility that different mechanisms work to cause polygyny in different populations. Despite these regional differences, however, the polygyny-threshold-model is generally applicable to this species.