Multivariate fluctuating asymmetry in Israeli adults.

Fluctuating asymmetry serves as an indicator of developmental homeostasis; any genetic or environmental factor that destroys homeostasis disturbs bilateral symmetry. Many studies have attempted to correlate increased fluctuating asymmetry with measures of developmental homeostasis or of adaptation, but collective results have been equivocal. We still have much to learn about fluctuating asymmetry itself. We develop a multivariate treatment of bilateral asymmetry that allows us to measure and adjust for the effects of both directional (right versus left) asymmetry and antisymmetry (handedness) and for the effects of size and sex. We examine 29 morphometric traits (13 size and shape, 8 pairs of bilateral traits) on 400 Israeli adults. We show that, even if metric traits are logarithmically transformed at the outset so that nonlinear allometric relations among them are properly accounted for and if the resulting measures are adjusted for sex by subtracting the means and dividing by the standard deviations, the correlation matrices among traits are strongly heterogeneous across sexes. Although the bilateral traits as a set are modestly correlated with the size and shape traits, neither size nor shape is correlated with asymmetry. Moreover, asymmetry traits are virtually uncorrelated inter se. We also examine the often touted connection between asymmetry and departures from morphometric modality and show that asymmetry is not correlated with deviations from either average size or average shape. If properly measured, fluctuating asymmetry becomes a measure of the noise in development; in our study it correlates with nothing. We discuss these results in light of empirical results and in terms of developmental theory suggesting that fluctuating asymmetry should be correlated with measures of maladaptation. Some of the effects reported in the literature need to be reexamined with more refined morphometric and statistical techniques.