Spatial patterns of local dialects of the declining Corn Bunting ( <i>Emberiza calandra</i> ) in three regions with different population densities and small‐scale landscape features

Rapid dialect shift or disappearance due to anthropogenic disturbances such as habitat degradation and fragmentation is a common feature in farmland bird species. Here we explored the formation, maintenance and spatial structure of local dialect patterns in the Corn Bunting Emberiza calandra , a declining European farmland bird, in arable landscapes of northwestern and central Europe. This species displays a microgeographical song variation pattern, so‐called local dialects. This consists of small groups of nearby males sharing the same song repertoire, which is usually composed of one dialect component and several specific song type components. The song units are discrete and the spatial boundaries between local dialects are clear. Territorial males were recorded in three local Corn Bunting populations in Poland, France and Belgium. Pairwise song dissimilarity was used to analyse quantitatively the spatial pattern of microgeographical song variation in the three regions in relation to farming intensity, landscape features and population trends in each region. We found different local dialect patterns in the three regions. As predicted, in Poland, where the Corn Bunting was abundant, we observed the classical pattern of a spatial mosaic of distinct local dialects occurring at a local scale. This classical pattern was less apparent in France and even less apparent in Belgium, where the Corn Bunting was in sharp decline. Our results also suggest that a higher local density of territorial males was correlated with a greater song sharing between neighbouring males, which supports the hypothesis that local dialects are a density‐dependent phenomenon. Other results suggest that fragmentation of farmland habitats by patches of habitat unsuitable for Corn Buntings, such as woodlands and urban areas, contributed to the difference observed in the dialect‐specific component of songs. Our findings point to anthropogenic causes, but long‐term studies in the same areas and extension of the study to new areas would strengthen these conclusions and rule out other environmental factors or idiosyncrasies of individual populations.