How reproducible are the measurements of leaf fluctuating asymmetry?

Abstract

<p> Fluctuating asymmetry (FA) represents small, non-directional deviations from perfect symmetry in morphological characters. FA is generally assumed to increase in response to stress; therefore, FA is frequently used in ecological studies as an index of environmental or genetic stress experienced by an organism. The values of FA are usually small, and therefore the reliable detection of FA requires precise measurements. The reproducibility of fluctuating asymmetry (FA) was explored by comparing the results of measurements of scanned images of 100 leaves of downy birch (Betula pubescens) conducted by 31 volunteer scientists experienced in studying plant FA. The median values of FA varied significantly among the participants, from 0.000 to 0.074, and the coefficients of variation in FA for individual leaves ranged from 25% to 179%. The overall reproducibility of the results among the participants was rather low (0.074). Variation in instruments and methods used by the participants had little effect on the reported FA values, but the reproducibility of the measurements increased by 30% following exclusion of data provided by seven participants who had modified the suggested protocol for leaf measurements. The scientists working with plant FA are advised to pay utmost attention to adequate and detailed description of their data acquisition protocols in their forthcoming publications, because all characteristics of instruments and methods need to be controlled to increase the quality and reproducibility of the data. Whenever possible, the images of all measured objects and the results of primary measurements should be published as electronic appendices to scientific papers.</p>