Fluctuating asymmetry (FA), which has been examined in a variety of plants and animals, is widely promoted as a useful bioindicator of exogenous stressors in habitats, whether of natural or anthropogenic origin. Wildlife managers and researchers often use a specific group of organisms as an indicator of the health of a given habitat. The purpose of this review is to demonstrate that FA can be an effective fish biomonitoring tool by presenting a vote counting meta-analysis of 81 fish FA studies published between 1966 and the first half of 2009. The vote counts were analyzed with the G test for independence to determine whether the probability of observing significant morphological asymmetry is determined by character type, exogenous stressor type, or fish order. The information obtained from these papers and their analysis is then used to outline areas in which FA studies can be improved: (1) carefully considering character choice; (2) distinguishing between asymmetry types; (3) determining the level of measurement error in between-sides character variation; (4) determining baseline FA levels in populations; (5) increasing the number of laboratory studies which corroborate field observations of FA; (6) conducting true replications of studies to validate previous findings. Only with more critical experimental design and data analysis can FA be used as a powerful tool for assessing environmental degradation.

Two species of fishes common to the Great Plains of North America (the western mosquitofish [Gambusia affinis] and the sand shiner [Notropis ludibundus]) were used to examine the relationship between individual time-to-death (TTD) in 96-h toxicity tests and average fluctuating asymmetry (FA). Previous studies have indicated that FA, the random deviation from bilateral symmetry in metric traits, is increased in organisms that develop in stressful and/or marginal environments. This association has led to the prediction that FA may be useful as a bioindicator. For example, if an individual's FA is negatively related to stress resistance, a population that has comparatively high values of FA may be more susceptible to further stressors. Adult fish used in these experiments were captured from natural populations in Kansas and acclimated to laboratory conditions for 1 month before use in toxicity tests. Each species was separately exposed to two insecticides: parathion, an organophosphate, and lindane, an organochlorine. For each TTD test, 200 individuals were exposed over a 96-h period to an LC70 concentration. Dead individuals were removed every 3 h. Individuals that survived the test were sacrificed for comparison to those that died. Measurements for seven external morphologic characters were obtained for each individual and used to estimate average FA. Three of four statistical tests showed a significant negative regression between TTD and average FA, indicating that individuals with smaller FA survived longer in the test conditions. These results support the use of FA as an indicator of susceptibility to pesticides.

Abstract-Two species of fishes common to the Great Plains of North America (the western mosquitofish [Gambusia affinis] and the sand shiner [Notropis ludibundus]) were used to examine the relationship between individual time-to-death (TTD) in 96-h toxicity tests and average fluctuating asymmetry (FA). Previous studies have indicated that FA, the random deviation from bilateral symmetry in metric traits, is increased in organisms that develop in stressful and/or marginal environments. This association has led to the prediction that FA may be useful as a bioindicator. For example, if an individual's FA is negatively related to stress resistance, a population that has comparatively high values of FA may be more susceptible to further stressors. Adult fish used in these experiments were captured from natural populations in Kansas and acclimated to laboratory conditions for 1 month before use in toxicity tests. Each species was separately exposed to two insecticides: parathion, an organophosphate, and lindane, an organochlorine. For each TTD test, 200 individuals were exposed over a 96-h period to an LC70 concentration. Dead individuals were removed every 3 h. Individuals that survived the test were sacrificed for comparison to those that died. Measurements for seven external morphologic characters were obtained for each individual and used to estimate average FA. Three of four statistical tests showed a significant negative regression between TTD and average FA, indicating that individuals with smaller FA survived longer in the test conditions. These results support the use of FA as an indicator of susceptibility to pesticides.