Phylogenetic comparison and artificial selection. Two approaches in evolutionary physiology.

Interspecific comparison has a long and productive history in physiology. Conceptual and statistical advances over the last 15 years have demonstrated several ways in which comparisons can be enhanced by consideration of phylogenetic information, i.e., empirical estimates of the ways in which organisms are related (evolutionary trees). Choice of species to be compared should be informed by phylogenetic information. For example, a comparison of three species that inhabit high altitude with three that live at low altitude would be suspect if each of the two groups were composed of closely-related species (e.g., within single genera). To avoid such "phylogenetic pseudoreplication," one might instead study species from three different genera, each containing one high-altitude and one low-altitude inhabitant. Unfortunately, many studies have not been so carefully designed, sometimes because organisms were not accessible or because the studies incorporated data from the literature. Fortunately, several new statistical methods correct for problems caused by phylogenetic relatedness and descent with modification, the most common being phylogenetically independent contrasts. Another tool that can be used in comparative physiology is selective breeding, which has been practiced for millennia and applied in scientific contexts for over a century. In the last 20 years, ecological and evolutionary physiologists have begun using selection experiments to study processes of genetic adaptation in physiological and behavioral traits. For example, house mice have been maintained in the cold for multiple generations to see what adaptations may occur naturally in response to reduced ambient temperature ("laboratory natural selection"). Our own laboratory has used selective breeding to create four replicate lines of mice that exhibit high levels of voluntary wheel-running behavior, as well as various morphological and physiological characteristics that cause or allow the elevated locomotor activity. Similar experiments could be used to study adaptation to hypoxia.