Integrating Phylogenetic and Experimental Analyses: The Evolution of Male and Female Nuptial Coloration in the Stickleback Fishes (Gasterosteidae)

Abstract

Both male and female gasterosteid fishes display nuptial coloration. Male nuptial col- oration originated in the ancestor of the Pungitius + Culaea + Gasterosteus clade. Patterns of origin and diversification of characters involved in male-male, male-female, and parental interactions indicate that the evolution of male color was influenced by intersexual selection, natural selection during parental care, and intrasexual selection, in decreasing order of importance. This macro- evolutionary hypothesis was corroborated by examining changes in male color across the breeding cycle for two stickleback species, Gasterosteus aculeatus and Culaea inconstans. Female nuptial col- oration may have originated before the male signal. The phylogenetic diversification of the male and female signals are decoupled, suggesting that they have been subject to different selection pressures throughout their evolutionary histories. Macroevolutionary patterns and experimental studies indicate that color signal evolution has been more complex in this group of fishes than was previously thought. (Comparative biology; phylogeny; sexual selection; nuptial coloration; stickleback fishes.) The power of macroevolutionary pat- terns lies in their ability to falsify hypoth- eses of character evolution. For example, a proposal that the evolution of character X was influenced by character Y is falsified if X evolved before Y. Failure to falsify a hypothesis of character evolution based upon macroevolutionary patterns pro- vides, however, only weak support for the hypothesis. This statement holds whether you are investigating unique events or sta- tistically analyzing correlated patterns of character origin in different lineages be- cause neither macroevolutionary patterns nor statistical analyses are tests of mech- anisms. The only way to unambiguously demonstrate that mechanism X has had an effect upon the evolution of character Y is to perform the appropriate experiment. The comparative phylogenetic approach is thus a powerful tool for experimental ethologists because it allows construction of a set of predictions based upon macro- evolutionary patterns that can serve as the focus for experimental investigation (Brooks and McLennan, 1991; McLennan, 1991). In this paper, I explore the feasibility of using the comparative approach to study the evolution of nuptial coloration in stick- leback fishes. A comparative study is more aptly described as a feedback loop than as a linear process. The loop begins with the reconstruction of the phylogenetic tree for the study group and proceeds with the op- timization of specific characters onto the tree, delineation of the relationship be- tween the origin and the diversification of the characters, construction of a hypothesis about the forces that have shaped the ob- served patterns, use of the hypothesis to make a series of predictions that can be tested experimentally, and experimental investigation of the predictions. During the course of the experimental analysis, it is not unusual to discover information that may either reinforce or change the original phylogenetic hypothesis. Such feedback highlights one of the certainties of biology: the situation is usually more complex than originally thought. A comparative study is thus an ongoing process whose successive sweeps we hope resemble the ever short- ening strokes of a pendulum moving clos- er and closer to the central attractor rep- resenting truth.