Cultural conformity and persistence in the context of differing site fidelity

Abstract

Animal culture often shows geographic structure, with individuals in close proximity sharing more cultural features than individuals further apart. However, spatial extent of cultural features, along with the degree of conformity to local cultures, vary within and among species. Further, rates of cultural change presumably also vary, though documentation of temporal variability lags behind documentation of spatial variability. Understanding both spatial and temporal variation is essential to understanding cultural evolution, but mechanisms likely to be driving this variation have not been sufficiently explored. We hypothesized that conformity to local culture in Oscine songbirds and the persistence of culture over time and space are promoted by habitats that facilitate stable populations in which individuals show relatively high site fidelity. In contrast, sites in which habitat features cause rapid population turnover provide more vacant territories and so more opportunities for colonization. Colonization should drive more rapid cultural change, either through adult colonists importing foreign cultural variants or young colonists making errors as they learn the local song. To test this set of hypotheses, we examined temporal and spatial variation in vocal culture in a songbird (dickcissel, Spiza americana) in two distinct habitat types. As predicted, we found high site fidelity in relatively stable native grasslands and much lower site fidelity in nearby cropland sites which were disturbed by farming practices during the breeding season. We also found evidence of higher levels of song sharing and slower changes in vocal culture in our grasslands relative to croplands, though we found no evidence of different spatial scales of song sharing between these habitats. This is the first study we know of correlating the temporal rate of cultural change to differences in a demographic factor between habitats. Although our conclusions are based on many recorded songs, automated assessments of song similarity were imprecise and so our results here underestimate the overall degree of song sharing and thus possibly the strength of the effects of time and distance on this sharing. Further, because we examined song sharing at only seven sites, firm conclusions about site fidelity and song sharing will require larger samples in the future.