NONLINEAR EFFECTS OF FEMALE MATE CHOICE IN WILD THREESPINE STICKLEBACKS
Article first published online: 9 MAY 2007
Volume 58, Issue 11, pages 2498–2510, November 2004
How to Cite
Blais, J., Bernatichez, L. and Rico, C. (2004), NONLINEAR EFFECTS OF FEMALE MATE CHOICE IN WILD THREESPINE STICKLEBACKS. Evolution, 58: 2498–2510. doi: 10.1111/j.0014-3820.2004.tb00879.x
- Issue published online: 9 MAY 2007
- Article first published online: 9 MAY 2007
- Received March 18, 2004 Accepted August 3, 204.
- Gasterosteus aculeatus;
- mate choice;
- mean d2;
- parasite load;
- reproductive success;
- sexual selection
Abstract Sexual selection by mate choice represents a very important selective pressure in many animal species and might have evolutionary impacts beyond exaggeration of secondary sexual traits. Describing the shape and strength of the relationships linking mating success and nonsexual traits in natural conditions represents a challenging step in our understanding of adaptive evolution. We studied the effect of behavioral (nest site choice), immunological (trematode level of infection), genetic diversity (measured by mean d2) and morphological (standard length and pectoral fin size) traits on male mating success in a natural population of threespine sticklebacks Gasterosteaus aculeatus. Male mating success was measured by microsatellite genotyping of embryos used to infer female genotypes. First, we analyzed all territorial males (full analysis) but also considered independently only males with a nonzero mating success (reduced analysis) because some of the males with no eggs could have been part of a later breeding cycle. Multiple linear regressions identified a significant negative effect of parasite load in the full analysis, whereas no linear effect was found in the reduced analysis. The quadratic analyses revealed that nest location and parasite load were significantly related to mating success by positive (concave selection) and negative (convex selection) quadratic coefficients respectively, resulting in a saddle-shaped fitness surface. Moreover, there were significant interactions between nest location, mean d2 and parasite load in the reduced analysis. The subsequent canonical rotation of the matrix of quadratic and cross-product terms identified two major axes of the response surface: a vector representing mostly nest site choice and a vector representing parasite load. These results imply that there exists more than one way for a male threespine stickleback to maximize its mating success and that such nonlinear relationships between male mating success induced by female mate choice and male characteristics might have been overlooked in many studies.