Data from: No support for solar radiation as a major evolutionary driver of malar stripes in falcons

The malar stripes of falcons (Falco spp.) are often hypothesised to function by reducing the amount of solar glare reflected into the falcon's eyes while hunting, thereby aiding foraging efficiency in bright conditions. This "solar glare hypothesis" is supported by intraspecific trends in Peregrine Falcons (Falco peregrinus), in which populations inhabiting regions of higher average annual solar radiation exhibit larger and darker malar stripes on average. Here, we extend the methodological approach previously used in Peregrine Falcons to examine both intra- and interspecific relationships between solar radiation and malar stripe morphology across all extant falcon species, thereby providing a more robust test of the hypothesis that falcon malar stripes evolved as an adaptation against negative visual effects of solar glare. We obtained web-sourced photographs of all extant falcon species, taken across each species' geographic range, and related mean breeding season solar radiation at each photograph location to the size and darkness of the birds' malar stripes, simultaneously testing for intraspecific and interspecific relationships between malar stripe characteristics and solar radiation, and including phylogeny and relevant ecological traits as covariates. We found no consistent interspecific relationship between solar radiation and malar stripe characteristics Likewise, in 38 out of 39 species, malar stripe characteristics were not positively intraspecifically related to solar radiation, with only Peregrine Falcons showing trends towards larger and darker malar stripes in brighter regions. Falcon malar stripes are thus unlikely to represent an adaptation against visual effects of solar glare, and their adaptive significance is more likely to be explained by crypsis or social signalling, if indeed they do represent an adaptive trait. Malar stripes may have become co-opted for solar glare reduction in Peregrine Falcons due to the species' specialisation for high-speed aerial hunting, although the intraspecific patterns observed may alternately be explained by phylogeography.

Photograph metadata for user-submitted photographs of 39 falcon species (including the name of the photographer, date and location at which the photograph was taken, GPS coordinates, photographer comments, and species, sex, and subspecies identification) were downloaded from the Macaulay Library (https://www.macaulaylibrary.org/) and GBIF (https://www.gbif.org/) online databases, using the websites' export tools, with additional photographs sourced from Flickr (https://www.flickr.com), Oiseaux (https://www.oiseaux.net), BirdGuides (https://www.birdguides.com), the African and Oriental Bird Club Image Databases (currently hosted at https://africanbirdclub.org/afbid and https://www.macaulaylibrary.org/oriental-bird-images/, respectively), and private photographers via email correspondence. All photographs were then downloaded using the list of image URLs and filtered to include only photographs of adult birds and those taken within the breeding or resident range for the species. For species in which resident and wintering populations overlap in geographic range, photographs taken within the area of overlap were also filtered to only include those taken within the migratory population's breeding season, using species range maps provided by BirdLife International (https://datazone.birdlife.org/). For each species, a random selection of up to 1000 photographs was then selected for analysis, with this selection process repeated until the maximum number of 1000 usable photographs for each species was reached, or until there were no more photographs available. Mean monthly solar radiation (W/m²) measurements at each photograph's GPS location were then extracted from the TerraClimate dataset (http://www.climatologylab.org/terraclimate.html) using Google Earth Engine, and averaged over the six-month period representing the species' or population's breeding season (March-August for boreal summer breeders and September-February for austral summer breeders), representing the 30-year (1991-2020) average breeding season solar radiation for the species or population at the photograph GPS location. For photographs which did not have user-submitted GPS coordinates available, the GPS coordinates for the written photograph location were sourced from Google Maps. Photograph metadata and solar radiation data were collected between March 2020 and May 2021.

For each individual bird in each photograph, a single observer then scored the characteristics of the malar stripe according to an eleven-point visual scale constructed for this purpose. Three aspects of the malar stripe were quantified: length, width, and darkness. The same observer also scored the angle and position of the bird's head in each photograph according to a visual scale (see ReadMe for details). Photographs were cropped to include only the bird's head, pooled into a single folder, and shuffled into a random order prior to scoring, with the observer unaware of the photograph location or species during the scoring process. Birds were also assigned a binary score of 0 or 1 based on whether the species or morph in question had uniformly-coloured facial plumage without definable malar stripes, to enable the analysis to be rerun on a subset of the dataset with these birds excluded.

Ecological and biometric data for the 39 falcon species (body mass (g), hand-wing index), prey preference, and habitat openness) were then sourced from AVONET (https://opentraits.org/datasets/avonet.html) and supplementary sources in the literature. Species were assigned a "prey preference" score of 1-4 based on the proportion of bird prey in the diet, with a higher score representing a higher degree of specialisation on bird prey, and a "habitat openness" score of 1-5 based on the degree of dependence on forest habitats, with higher scores representing more open (less forested) habitats. Species trait data were collected between January and April 2022.