Variation in sexual size dimorphism and fit to Rensch's rule in 45 species of Costa Rican hummingbirds

According to Rensch’s rule, in species where males are the larger sex, increasing body size is coupled with increased sexual size dimorphism (SSD), whereas in species where females are larger, SSD decreases with body size. We describe the variation in SSD for body mass and its fit to Rensch´s rule in 45 species of Costa Rican hummingbirds. We examine SSD in body mass, bill length, and wing size and their fit to Rensch´s rule in nine species of hummingbirds for which we have detailed data. We expected that physiological constraints will limit the variation in SSD since hummingbirds have high metabolic rates, high dependence on energy-dense food, and high costs imposed by small size and hovering flight; these factors scale with body size and elevation. Large species should be physiologically capable of withstanding greater variation in SSD compared to small species, which are more energetically limited. Hummingbirds showed mixed allometry and fit Rensch's rule (slope of the RMA regression male vs female body mass = 0.8630). Eighty percent of the species showed male-biased allometry for SSD in body mass. Average variation in SSD was 12% and varied regardless of size. Physiological limitations act on hummingbirds as a group affecting SSD, independently of body size. SSD differences in body size, bill length, and wing area could influence ecological performance, sexual displays, food resource access, and foraging behavior. The SSD mixed allometry, and the positive relationship between the standard deviation of body mass, wing chord, and wing area with hummingbird size found only in males (9 species), suggests that selective pressures act differentially on the sexes and influence sexual niche segregation. Future research should examine intersexual variation in morphology, and quantify intersexual habitat use, niche segregation, and interspecific and intraspecific competitive interactions, targeting species located at the extremes of the Rensch's rule scatterplot.