Hormones and life histories: an integrative approach.

-This article, which focuses on hormones and the diverse effects they have on behavior and physiology, raises evolutionary questions that hormonal studies appear especially well suited to address. These include the endocrine basis for life-history trade-offs, the role of hormones in adaptive alterations in social organization and mating systems, and whether natural selection acts on traits or organisms. The article also shows how phenotypic engineering by hormonal manipulations can reveal the evolutionary significance of phenotypic variation. By generating rare or novel phenotypes, we can attempt to determine the shape of fitness profiles in nature. To illustrate phenotypic engineering, we manipulated plasma testosterone in a freeliving bird, the dark-eyed junco (Junco hyemalis), and measured the effects of the treatment on behavior, including allocation of time to mate attraction and parental behavior as well as use of space, and on physiology, including regulation of body mass, corticosterone, and molt. We compared treated males to controls for various correlates of fitness, including territory acquisition, mate acquisition, mate retention, physical condition of the mate, apparent reproductive success, extrapair fertilizations, and survival. The results to date appear to indicate that selection is relatively indifferent to a broad range of phenotypes, while extreme deviations from the norm are selected against. The objective of this article is to describe the potential that studies of hormones have to increase understanding of the consequences of variation in behavior and physiology and thus understanding of the evolution of life histories, social organization, and mating systems. We begin by reviewing some basic facts about the action of one hormone, testosterone, in birds. Our aim is to illustrate a few principles, including (1) that hormones are secreted in response to stimuli that are generated both internally and externally, (2) that hormones have multiple targets and diverse effects, and (3) that the complexity of hormonal action leaves ample room for variation on which natural selection can act. We then pose four interconnected questions that relate hormones to the evolutionary process and that provide fruitful avenues for research: (1) How might hormones form the mechanistic bases for trade-offs in life histories? (2) What is the role of hormones in the evolution of interand intraspecific variation in social organization and mating systems? (3) Does natural selection act on organisms or on traits? (4) What is the evolutionary significance of phenotypic variation in hormonally mediated traits? From the recent literature we give examples of studies that address or could be used to address each question. Finally, we describe an experimental method that we call phenotypic engineering (Ketterson et al., in press) and show how we and others have used this method to address the fourth question, the evolutionary significance of phenotypic variation. Our model organism is a small passerine bird, the dark-eyed junco Am. Nat. 1992. Vol. 140, pp. S33-S62. ? 1992 by The University of Chicago. 0003-0147/92/4007-0003$02.00. All rights reserved. This content downloaded from 129.79.32.220 on Fri, 21 Feb 2014 13:56:05 PM All use subject to JSTOR Terms and Conditions S34 THE AMERICAN NATURALIST (Junco hyemalis). We have manipulated plasma concentrations of testosterone in free-living male juncos in order to alter their behavioral and physiological phenotypes. Some of the changes we have induced are potentially beneficial, while others are potentially detrimental. Our objective has been to determine the overall effect these phenotypic changes have on fitness.