Relative Brain Size Is Not Correlated with Display Complexity in Manakins: A Reanalysis of Lindsay et al. (2015)

Abstract

some species [Prum, 1990, 1994]. Lindsay et al. [2015] hypothesized that a more complex acrobatic display will result in an increased demand of neural computation and, therefore, an increase in brain size. Their analyses yielded a positive correlation between relative brain size and display complexity (fig. 2 in Lindsay et al. [2015]). However, their analyses have several potential problems, and alternative methods that correct these problems lead to different conclusions. The most significant issue is how relative brain size was calculated. The brain, like most organs and other physiological traits (e.g. basal metabolic rate) scales positively with body size [Gould, 1966, 1974; Martin and Harvey, 1986], and it is traditionally understood that the relative size of the brain (or any other trait) is one where the allometric effect of body size has been removed. The issue regarding the best methods for removing allometric effect in comparative studies has been given careful consideration in the literature [McCoy et al., 2006; Berner, 2011], including papers that specifically discuss the brain or its parts [Deaner et al., 2000]. Whichever method is deemed most appropriate, it is clear that the relative size of the brain (or any other organ) should not be correlated with body size [Deaner et al., 2000; Sol et Over the past 20 years, there has been an explosion of comparative studies looking to understand the driving forces behind brain evolution [Striedter, 2005]. Generally, in these studies variation in the size or morphology of the brain, or specific brain regions, has been associated with a long list of different behavioral, ecological and/or morphological characters [for a review, see Healy and Rowe, 2007]. Along with the surge in such comparative studies of brain and brain region sizes, there has been a parallel development of phylogenetic comparative methods [Garamszegi, 2014a]. Since recognizing that species do not constitute independent data points, an assumption that is inherent to most traditional statistical tests [Felsenstein, 1985; Harvey and Pagel, 1991], statistics that include phylogenetic information have been developed at an increasingly fast pace. This has resulted in disagreements on what are the best statistical methods to apply in different comparative studies, along with some degree of confusion on how to proceed. In the March 2015 issue of Brain, Behavior and Evolution , Lindsay et al. [2015] published a study on brain size variation in manakins (Pipridae), a family of passerine birds that exhibit remarkable variation in courtship displays, including complex, dance-like, acrobatic courtship displays in Published online: June 4, 2016