Fluctuating Asymmetry in Tamarin (Saguinus) Cranial Morphology: Intra- and Interspecific Comparisons Between Taxa With Varying Levels of Genetic Heterozygosity

Abstract

Fluctuating asymmetry has been proposed as a measure of developmental homeostasis and an indicator of populations under stress. However, controversy surrounds not only the validity of an association between fluctuating asymmetry and levels of protein heterozygosity, but also whether fluctuating asymmetry can be used to identify populations under genetic and environmental stress. The relationship between levels of heterozygosity and developmental homeostasis is considered by comparing levels of cranial fluctuating asymmetry in three tamarin samples with contrasting levels of heterozygosity: (1) low heterozygosity cotton-top tamarins (N = 324), (2) presumably normally heterozygous Illiger's saddle-back tamarins (N = 208), and (3) relatively highly heterozygous hybrids between saddle-back tamarin subspecies (N = 31). All specimens originated at the Oak Ridge Associated Universities' Marmoset Research Center. A nested ANOVA design was used to separate out variation due to individual differences, side-to-side differences (fluctuating asymmetry), and measurement error. We found statistically significant levels of fluctuating asymmetry in nearly all of the traits surveyed and a negative correlation between levels of fluctuating asymmetry and genetic heterozygosity. Efforts to use fluctuating asymmetry to identify populations endangered by reduced genetic variability and/or under stress may be inhibited by small sample sizes, neglect of repeated measures, and lack of appropriate reference populations.

Fluctuating asymmetry and genetic variability in the roe deer (Capreolus capreolus): a test of the developmental stability hypothesis in mammals using neutral molecular markers

“…Studies on mammals, according to our knowledge, only yielded the expected negative correlation twice (Hutchison andCheverud, 1995 on tamarin monkeys andHartl et al, 1995 on brown hares). The study by Hutchison and Cheverud, however, is not directly comparable to similar studies because (1) their analysis comprises two different species and (2) they do not directly measure heterozygosity but rather provide qualitative ranks of genetic variability.…”

The Fitness of Manipulating Phenotypes: Implications for Studies of Fluctuating Asymmetry and Multivariate Selection

“…FA was estimated using differences between bilateral measurements (differences between right and left side measurements; R -L) and compared among groups using ANOVA (Palmer 1994;Hutchison and Cheverud 1995). To distinguish other forms of asymmetry such as anti symmetry and directional asymmetry from FA, we examined the distributions of the R -L measurements.…”

“…The distribution of differences was not bimodal, nor was there significant skewness or kurtosis, within any diet. Although none of the mean differences (R -L) differed significantly from zero, we corrected for the statistical influence of any directional asymmetry by adding the mean difference (R -L) for each group to the left side measurement of each individual within that group (for a discussion of this correction, see Palmer and Strobeck 1992;Graham et al 1994;Hutchison and Cheverud 1995). The corrected data were analyzed using a nested ANOVA with feet of diet on the shape of the matrix.…”

“…Reflectance spectra were captured in 0.3 nm increments from 400 nm to 700 nm. 0.5 replicate measure nested within sides and sides nested within individuals (Hutchison and Cheverud 1995). The presence of significant FA that was not simply due to measurement error was tested as an F-ratio of variance due to sides within individuals (FA) over the variance due to measurement error.…”

An Epistatic Genetic Basis for Fluctuating Asymmetry of Mandible Size in Mice

“…This distribution was found to be normal, suggesting that there was no apparent antisymmetry, a kind of asymmetry detected by significant platykurtosis (Palmer and Strobeck 1992). The mean of the right minus left differences of centroid size was subtracted from the signed difference between the sides to statistically correct for directional asymmetry, and the absolute values of these differences were used to measure FA (Leamy 1984;Hutchison and Cheverud 1995;Leamy et al 1997). The distribution of unsigned differences between sides for centroid size, however, was half-normal as expected (Palmer 1994), and these values therefore were subjected to Box-Cox transformations (Swaddle et al 1994) of the form (FA ϩ 0.0005) 0.33 , which were successful in achieving normality.…”