Genetics and geometry of canalization and developmental stability in Drosophila subobscura

Abstract

Background: Many properties of organisms show great robustness against genetic and environmental perturbations. The terms canalization and developmental stability were originally proposed to describe the ability of an organism to resist perturbations and to produce a predictable target phenotype regardless of random developmental noise. However, the extent to which canalization and developmental stability are controlled by the same set of genes and share underlying regulatory mechanisms is largely unresolved.

Results: We have analyzed the effects of clinical genetic variation (inversion polymorphism) on wing asymmetry by applying the methods of geometric morphometrics in the context of quantitative genetics using isochromosomal lines of Drosophila subobscura. For the analysis of overall size, developmental stability was positively correlated with levels of heterozygosity and development at the optimal temperature. For analyses of shape, the overall comparisons by matrix correlations indicate that inter- and intraindividual variation levels were poorly correlated, a result also supported when comparing the vectors describing patterns of variation of landmark position. The lack of similarity was basically due to the discrepancy between the genetic and environmental components of the interindividual variation. Finally, the analyses have also underscored the presence of genetic variation for directional asymmetry.

Conclusions: The results strongly support the hypothesis that environmental canalization and developmental stability share underlying regulatory mechanisms, but environmental and genetic canalization are not functionally the same. A likely explanation for this lack of association is that natural wing shape variation in Drosophila populations is loosely related to individual fitness.

Figure 1

Inbreeding and temperature effects on size Homokaryotipic averages for centroid size and centroid size FA (index FA1 in [39]) in inbred (black symbols) and outbred (open symbols) crosses. Small symbols give the average values for each of the three different homokaryotypes to appreciate the dispersion from the corresponding grand average (large symbols connected by lines). Squares give the values at 23°C and circles at 18°C.

Figure 2

Eigenvalues of causal covariance matrices for wing shape First 15 eigenvalues of the phenotypic (black bars), karyotype (hatched) and crosses (open) covariance matrices from outbred crosses.

Figure 3

Vectors of the landmarks displacements First two axes of wing shape variation for each effect in the two-way mixed MANOVA (individuals, individuals × sides interaction, and measurement error) for females from outbred crosses reared at 18°C. Also plotted are the percentages of total wing shape variation explained by the principal components for the corresponding covariance matrices.

Figure 4

Vectors of the landmarks displacements Same as Fig. 3 for males from outbred crosses reared at 18°C.

Figure 5

Vectors of the landmarks displacements Same as Fig. 3 for females from outbred crosses reared at 23°C.

Figure 6

Vectors of the landmarks displacements Same as Fig. 3 for males from outbred crosses reared at 23°C.

Figure 7

Vectors of the landmarks displacements First two axes of wing shape variation in the two-level nested MANOVA (karyotypes, crosses nested in karyotypes, and within crosses) for each causal component effect pertaining to the inter-individual variation in females from outbred crosses reared at 18°C. Also plotted are the percentages of total wing shape variation explained by the principal components for the corresponding covariance matrices.

Figure 8

Vectors of the landmarks displacements Same as Fig. 7 for males from outbred crosses reared at 18°C.

Figure 9

Vectors of the landmarks displacements Same as Fig. 7 for females from outbred crosses reared at 23°C.

Figure 10

Vectors of the landmarks displacements Same as Fig. 7 for males from outbred crosses reared at 23°C.

Figure 12

Left wing of Drosophila subobscura The image shows the thirteen landmarks (1 – 13) used in this work. I – VI longitudinal veins; cv-a and cv-p anterior and posterior crossveins; Co costal or marginal veins; L1 and L2 lengths of the proximal (Euclidian distance between landmarks 9 and 13) and distal (Euclidian distance between landmarks 13 and 5) segments of longitudinal vein IV, respectively. Wing shape index has been previously used to study shape clines in this species [30].

Figure 11

Wing shape index Averages of the relative length (with 95% confidence intervals) of the basal portion of longitudinal vein IV (L1) to the total wing length (WL = L1 + L2) versus karyotype for outbred crosses at the two rearing temperatures. A two-way factorial ANOVA using the shape index as , with karyotype and temperature as fixed effects, and crosses nested within karyotypes, detected statistically significant differences for the main effects (karyotype: female F_5,30 = 12.625, P < 0.001; male F_5,30 = 9.785, P < 0.001. Temperature: female F_1,390 = 30.219, P < 0.001; male F_1,390 = 61.835, P < 0.001) but no karyotype × temperature interaction (females: F_5,390 = 1.570, P = 0.168; males: F _5,390 = 1.111, P = 0.354).