Can the use of landmarks improve the suitability of fluctuating asymmetry in plant leaves as an indicator of stress?

Abstract

Fluctuating asymmetry (FA), defined as the random deviations from symmetry in morphological traits, has repeatedly been suggested as a handy indicator of genetic and environmental stress experienced by organisms during their development. However, the value of FA for applied ecological and environmental research is now questioned by a growing number of studies that have found no increases in FA under stress. We used geometric morphometrics to test the hypothesis that FA measures based on naturally defined landmarks are better suited for detection of deviations from leaf symmetry than are traditional measures of FA. We compared two sets of mountain birch (Betula pubescens ssp. czerepanovii) leaves, obtained from trees growing in industrially polluted natural populations and from experimentally stressed seedlings, by examining three measures of FA: 1) a traditional distance-based measure calculated from the width of the left and right halves of a leaf at the middle of the lamina (FAwidth), 2) a distance-based measure reflecting differences in vein length calculated from three landmarks (FAveins), and 3) a multivariate FA measure describing asymmetry in leaf shape based on five landmarks (FAshape). These three FA measures were poorly (albeit positively) correlated with each other. The two distance-based measures reflected different parts of the multivariate measure FAshape, which had the highest sensitivity for detection of deviations from leaf symmetry. FAshape was also the only FA measure that weakly confirmed the expected increase in FA under experimental stress, and particularly under stress caused by application of heavy metals. However, very high levels of within-individual variation in asymmetry remain to be explained, and even FAshape did not show the expected effects of industrial pollution on leaf symmetry. We conclude that FA in birch leaves cannot be regarded as a reliable indicator of environmental stress, even when quantified using modern methods.