Bird Song Sharing Declines With Distance, Meta-Analysis Finds

A meta-analysis of 58 studies finds that birds generally share more songs or song elements with individuals recorded nearby than with those farther away. The analysis combined 97 effect sizes, including results from 58 studies, and found an overall positive association between geographic distance and declining song sharing under the researchers’ sign convention. The pattern was consistent with local song traditions, but its strength varied among species, habitats and study designs, indicating that distance alone does not explain how birds divide their vocal repertoires across space.

The authors treated the focal relationship as a correlation between geographic distance and song sharing. They converted results from methods such as Mantel tests, Pearson and Spearman correlations, Mann–Whitney tests and comparisons of nearer and more distant groups into common correlation effect sizes, which were then transformed with Fisher’s Z. Positive values represented less sharing as distance increased, while negative values would indicate greater divergence among nearby birds. The overall intercept was 0.260, with a 95% credible interval of 0.114 to 0.391, providing evidence for the broad geographic pattern.

Not every part of the dataset behaved alike. The phylogenetic model found low heterogeneity among species and studies but medium heterogeneity among independent sampling groups, with an I² of 32.458 and a 95% credible interval of 9.254 to 55.933. This suggests that population-specific influences can be substantial even when a common tendency is visible across species. The analysis also detected a small but nonzero phylogenetic signal, with λ estimated at 0.153 and a 95% credible interval of 0.003 to 0.509, meaning that related species were somewhat more likely to have similar estimates than unrelated species.

Tests of publication bias did not reveal strong evidence that the literature was distorted by preferential publication. An Egger’s regression found no relationship between effect size and precision, and publication year did not significantly influence the results. A funnel plot based on marginal residuals also did not indicate serious bias. The relationship between effect size and standard error was only marginally nonsignificant, while variance was related to effect size at a conventional significance threshold. These diagnostics support the overall pattern but do not eliminate the limitations imposed by the available studies.

One of the most notable results ran against a commonly proposed explanation. Sexual dichromatism was used as an indirect proxy for the intensity of sexual selection, with dichromatic species expected to show conformity to local song traditions if sexual selection strongly favored similar songs. Instead, the estimated relationship between distance and declining sharing was stronger in monochromatic species than in dichromatic ones; the posterior contrast was 0.188, with a 95% credible interval of 0.008 to 0.352. One possible explanation is that species under stronger sexual selection may benefit from individually distinctive signals during competition for mates or social status. The result remains indirect, however, because plumage differences do not directly measure sexual selection, and some species may invest more heavily in visual signals than in song.

Habitat also modified the pattern. Birds recorded in suburban environments showed a steeper decline in song sharing with distance than those in natural habitats, with a posterior contrast of 0.264 and a 95% credible interval of 0.032 to 0.489. This did not support the expectation that suburban noise and habitat isolation would uniformly weaken local sharing. The authors instead proposed that a mixture of built and vegetated areas might restrict interactions or concentrate birds in remaining suitable patches, potentially increasing competition and strengthening local song traditions. They could not test this explanation because the included studies did not consistently quantify fragmentation, population density or competition.

Migration produced a less decisive pattern. Partially migratory species appeared to have a stronger distance-sharing relationship than either fully migratory species or residents, with posterior contrasts of 0.463 and 0.368, respectively, but only six estimates involved partial migrants, and residents and migrants did not differ significantly. Migration can weaken local traditions through movement and turnover, yet birds that continue learning after reaching breeding grounds may acquire songs from local neighbors. Migratory status therefore appears too coarse a variable to capture differences in learning schedules, site fidelity, movement and social turnover that determine which songs individuals encounter and retain.

Repertoire size and the unit of comparison also shaped the estimates. The distance-sharing relationship was not significantly positive for species with high repertoires or for migrants in the moderator models. Larger repertoires may make it harder for learning strategies to balance broad song diversity with the acquisition of socially shared material, although prior comparative work has also found no general relationship between repertoire size and sharing. Studies based on whole-song sharing showed a stronger pattern than those based on sharing individual song elements. This distinction may reflect the greater organizational complexity of complete songs and their susceptibility to change, but it was not consistently observed in studies that measured both levels.

The synthesis is constrained by the geography, taxonomy and methods of the available literature. The researchers searched the Web of Science Core Collection in 2024 and screened references in relevant reviews and primary studies. They excluded captive or translocated birds, comparisons between species, female song, and studies that did not quantify similarity in relation to geographic distance. Only studies measuring song or song-element sharing were retained, and direct geographic distances or territory-based separations were eligible. No recording site met the study’s strict urban definition, while habitat classification relied on maps and original descriptions, sometimes falling back on a species’ typical habitat because coordinates were unavailable. The result is therefore a broad association across available within-species studies, not a measurement of every bird population or a direct demonstration of a single evolutionary cause. The authors conclude that geographic variation in song is likely shaped by several interacting influences, with social processes, habitat structure, learning behavior and population-specific conditions all potentially contributing to the uneven geography of bird song.

Subject of Research: Meta-analysis of geographic variation in within-species bird song sharing

Article Title: Geographic Variation of Song Sharing in Birds—A Meta‐Analysis

Article References: Jablonszky, M., & Garamszegi, L. Z. (2026). Geographic Variation of Song Sharing in Birds—A Meta‐Analysis. Ecology and Evolution, 16(10), Article e74566. https://doi.org/10.1002/ece3.74566

Image Credits: AI Generated

DOI: 10.1002/ece3.74566

Keywords: bird song, song sharing, cultural transmission, geographic variation, bird dialects, sexual selection, migration, habitat effects, meta-analysis, social learning, Geographic, Variation