Developmental instability as phenodeviance in a secondary sexual trait increases sharply with thermal stress

Abstract We test for effects of thermal stress applied to pupal flies from Noumea (New Caledonia) and Taipei (Taiwan) on developmental instability (DI) in the male sex comb of Drosophila bipectinata, as well as on pre‐adult survivorship and adult body size. The temperature treatments were Low (25 °C), High (29 °C) and Variable (18 h at 29 °C, 6 h at 34 °C). Although the Variable treatment reduced survivorship and body size, absolute comb size and fluctuating asymmetry generally were invariant across treatments. In contrast, comb phenodeviance increased with stress in both populations. Phenodeviance in one comb segment (C2) increased sharply with stress, whereas phenodeviance in a second major segment (C1) also increased with stress but only in Noumea flies. A major conclusion is that phenodeviations induced in a secondary sexual trait reflect the developmental environment that also damages fitness components, a foundation stone of the hypothesis that expressions of DI reveal phenotypic quality in sexual selection.     

A primary role of developmental instability in sexual selection

In evolutionary biology, fluctuating asymmetry (FA) is thought to reveal developmental instability (DI, inability to buffer development against perturbations), but its adaptive and genetic bases are being debated. In other fields, such as human clinical genetics, DI is being assessed as incidence of minor morphological abnormalities (MMAs) and used to predict certain fitness outcomes. Here, for the first time, we combine these complementary measures of DI in sexual selection and quantitative genetic studies of a natural population. Comprehensive multivariate analyses demonstrate that FA and MMAs in a condition-dependent sexual ornament, the male Drosophila bipectinata sex comb used in courtship, are sole significant targets of selection favouring their reduced expression in New Caledonia. Comb FA and MMAs are positively correlated, confirming that each are linked to a common buffering system. Ornament size and DI (as FA and MMAs) are positively correlated, genetically and phenotypically, contrary to theoretical expectation of negative size-FA scaling under the assumption that FA reveals overall genetic quality. There exists significant additive genetic variance for MMAs, demonstrating their evolutionary potential. Ornament DI in New Caledonia is markedly elevated compared with populations where such selection was not detected, suggesting that the increased population-level DI is capacitating adaptive evolution.     

FLUCTUATING ASYMMETRY DOES NOT CONSISTENTLY REFLECT SEVERE DEVELOPMENTAL DISORDERS IN HUMAN FETUSES

Developmental instability (DI), as measured by fluctuating asymmetry (FA), may reflect fitness and facilitate the expression of morphological variation. Insights in the underlying mechanisms and magnitude of DI during early development would increase our understanding of its role in evolutionary biology. We studied associations between FA and congenital abnormalities of different origins and functional systems in deceased human fetuses. Major congenital abnormalities corresponded to severe, often-lethal developmental disorders disrupting normal development from early organogenesis onward, but only moderately increased FA. Lower FA with age also supported the hypothesis that more severe abnormalities, leading to an earlier death, increased DI. Although FA related significantly to measures of fitness or health, we anticipated stronger associations because fetal health problems were detrimental. Furthermore, elevated FA occurred in only 4 of 17 disorders (left–right patterning, limb defects, and problems of bronchopulmonary and urogenital system). Fetuses experiencing major abnormalities other than these four types did not show increased FA. This suggests that the functional importance of symmetry in limbs has resulted in strong selection for symmetry and reduced its sensitivity to stress. Finally, the observed patterns suggest that specific developmental pathways have a stronger effect on DI than others do.