Quantitative genetics of pigmentation development in 2 populations of the common garter snake, Thamnophis sirtalis.

ABSTRACT

The evolutionary importance of ontogenetic change has been noted since Darwin. However, most analyses of phenotypic evolution focus on single landmark ages. Here, we present an inheritance study that quantifies genetic variation in pigmentation across early-age (i.e., birth to 180 days) development in 2 populations of the common garter snake, Thamnophis sirtalis. The populations are phenotypically distinct and geographically isolated (Manitoba, CA and Northern California, USA). There were highly significant differences between populations for the developmental trajectory of mean pigmentation, with the Manitoba population exhibiting a mean pigmentation level that increased across ontogeny, whereas the California population exhibited mean pigmentation that was invariant across ontogeny. Subsequent quantitative genetic analyses revealed heritable variation at all ages in Manitoba but low levels of phenotypic and genetic variation in California at all ages. A quantitative genetic decomposition of the longitudinal genetic variance-covariance matrix for the age-specific pigmentation phenotypes in the Manitoba population revealed 2 primary orthogonal axes that explained most ( approximately 100%) of the pigmentation variation across ontogeny. The primary axis, explaining 93% of the genetic variation, is an axis of genetic variation whose principal value loadings change from positive to negative across development, suggesting that the most rapid evolutionary response to selection on pigmentation variation will occur in the direction characterized by a tradeoff in early-age versus late-age pigmentation phenotypes. Pigmentation is known to be ecologically important and subject to rapid evolution under selection. Our study shows that significant differences exist between these 2 populations for their capacity to respond to selection on pigmentation which is not only influenced by the population of origin but also by the developmental process. We suggest that developmental timing may be a potential explanatory mechanism for the difference between the populations.