Global patterns of plumage color evolution in island-living passeriform birds

Abstract

Island environments have the potential to change evolutionary trajectories of morphological traits in species relative to their mainland counterparts due to habitat and resource differences, or by reductions in the intensity of social or sexual selection. Latitude, island size, and isolation may further influence trait evolution through biases in colonization rates. We used a global dataset of passerine plumage color as a model group to identify selective pressures driving morphological evolution of island animals using phylogenetically-controlled analyses. We calculated chromaticity values from red and blue scores extracted from images of the majority of Passeriformes and tested these against the factors hypothesized to influence color evolution. In contrast to predictions based on sexual and social selection theory, we found consistent changes in island female color (lower red and higher blue chromaticity), but no change in males. Instead, island size and distance from mainland and other islands influenced color in both sexes, reinforcing the importance of island physiognomy in shaping evolutionary processes. Interactions between ecological factors and latitude also consistently influenced color for both sexes, supporting a latitudinal gradient hypothesis. Finally, patterns of color evolution varied among families, indicating taxon-specific microevolutionary processes in driving color evolution. Our results show island residency influences color evolution differently between sexes, but the patterns in both sexes are tempered by ecological, island characteristics, and phylogenetic effects that further vary in their importance among families. The key role of environmental factors in shaping bird plumage on islands further suggests a reduced importance of sexual and social factors in driving color evolution.

Fig 1. Islands are useful systems to study plumage coloration and the color of island species differs from their mainland counterparts.

Top panel: Islands are isolated, relatively small, and are replicated across a broad geographic scale, making them ideal systems to study evolutionary processes of trait variation. Dots represent global distribution of Passeriformes island species (n = 1,183) in our study. Bottom panels: Female colour variation between mainland and island passerine birds (n = 5,693). After controlling for phylogeny and ecological covariates, red chromaticity (±SE) was significantly higher (F = 57.2, P<0.0001) and blue chromaticity (±SE) was significantly lower on mainlands (F = 23.7, P<0.0001) relative to islands. Means from raw data are shown. Map was generated using the sf package [48] in R.

Fig 2. Island plumage color evolution was primarily influenced by ecology and geography.

Final path analysis models illustrating the effect of (A) habitat and land classification (island/mainland) on female red chromaticity, (B) geographic region on blue chromaticity, (C) latitude and land classification on male red chromaticity, and (D) male blue chromaticity. Solid lines indicate positive, while dashed lines indicate negative effects.

Fig 3. Patterns of island plumage color evolution varied by taxonomic scale.

While chromaticity generally varied between mainland and island populations, the direction of this effect varied among passeriform families (dots). Bold dots and solid lines represent increases in red (top panels) and blue chromaticity (bottom panels) and transparent dots and dashed lines indicate decreases. Note axis scales vary to avoid crowding of the lines. Means from raw data are shown.