Correcting for allometry in studies of fluctuating asymmetry and quality within samples

Abstract

Fluctuating asymmetry (FA) refers to deviations from perfect symmetry in a bilateral character and is believed to reflect an organism's quality. However, allometric relations between asymmetry and trait size may confound FA–quality relations. Larger traits may have more ‘opportunity’ to become asymmetric. Thus, researchers suggest that if allometric relations between FA and size are found, some correction for size is needed (typically FA scaled by size). I used a simulation model to examine potential consequences of allometry for detection of FA–quality relations within samples and examined efficacy of rules currently used to correct for size. Consequences of allometry can be severe, causing up to 77% type I errors or 98% power loss when true FA–quality relations exist (when FA–size relations are isometric), depending upon the strength and direction of size–quality relations. I illustrate why current rules to correct for size are inefficient and suggest alternative rules. Problems of allometry that require further attention are discussed.