Effect of chromosomal reorganizations on morphological covariation of the house mouse mandible: insights from a Robertsonian system of Mus musculus domesticus

Abstract

Morphological integration and modularity depend on genetic covariation between traits, which emerges from pleiotropic effects of single loci and genetic linkage between loci. Since chromosomal reorganizations alter meiotic recombination, they might modify groups of linked genes and entail the fixation of new alleles with new pleiotropic effects. As a result, they could contribute to the intraspecific variation of the covariance structure of morphological traits. Although the mouse mandible has long been studied in terms of development and evolution, little is known about how its covariance structure varies in natural populations with chromosomal reorganizations. Consequently, here we analyzed the magnitude and patterns of morphological covariation of mandible shape in groups of mice with different karyotypes from a Robertsonian system of Mus musculus domesticus. Results: The organization of the mouse mandible into two main modules was confirmed in all chromosomal groups, since RV coefficients for the corresponding subdivision of landmarks were always significant. However, substantial variation in the magnitude of integration was detected between groups, especially when the effect of allometry was not removed. A significant positive correlation between differences in magnitude of integration of the symmetric component of shape and karyotypic distances between groups was detected when not correcting for size. Moreover, the degree of dependence of symmetric shape variation on size showed a negative association with the chromosome number and a positive association with the magnitude of integration. All groups showed similar patterns of morphological integration of the mandible, especially regarding the symmetric component of shape. However, the display of landmark displacements and the computation of vector angles highlighted some differences. In addition, distances between groups in terms of covariation matrices of the symmetric component were positively correlated with geographic distance. Conclusions: Robertsonian translocations do not alter the organization of the mouse mandible into two main modules, but do affect the magnitude of integration between them. This effect is mainly due to changes in the allometric relationship. In the 'Barcelona' Robertsonian system, geographically structured sources of variation seem to affect the patterns of integration by producing parallel variation in separate developmental pathways. Overall, our results suggest that Robertsonian translocations could play a role in intraspecific differentiation processes by producing changes in the covariance structure of morphological traits.