Iguanids are visually-orientated lizards; the males of many species use head bobs and colored patches to signal territory ownership to conspecifics. Whereas the males of most species display patches with single, species-specific colors, males in a population of side-blotched lizard (Uta stansburiana), in the coastal mountains of Merced County, California, display three different color morphs. Some males have orange throats, some have dark-blue throats, and others have yellow throats. Orange males, which are very aggressive, vigorously defend large territories. Blue males, which are less aggressive, maintain smaller territories. Yellow males, which are nonaggressive, do not defend territories but instead sneak matings from females on the territories of other males. In considering this three-way color polymorphism we must ask the following question: Why has not natural selection culled two of the three color morphs from the population?

 Many species show polymorphisms in which different subgroups of a population display markedly different morphologies associated with differing strategies of reproduction (e.g., Brantley, et al. 1993). The hypotheses of how multiple morphs can coexist without one going to fixation include the following: (1) all the morphs offer equal fitness (Ryan et al. 1992); and (2) different morphs offer fitness advantages in different environmental conditions. A hypothesis that forms a subset of hypothesis number two above is as follows: all of the morphs offer a fitness advantage when they are rare in the population (Maynard-Smith 1982; Charnov 1993).

 In seeking to discriminate among these hypotheses, Program in Animal Behavior faculty member Barry Sinervo, in collaboration with Program in Animal Behavior faculty member Curt Lively, has modeled the fitness dynamics of a three-morph population using an evolutionarily stable strategy model (Sinervo and Lively 1995). Their model predicted that the abundance of the three morphs would oscillate, the oscillation driven by the advantages each of the morphs enjoyed when a complementary morph was common (Fig. 1). In field studies in California over the last six years, Sinervo has observed the precise pattern predicted by the model. Not only are there pronounced color polymorphisms in the Merced County population, but the relative abundance of the three morphs appears to cycle through time (Sinervo and Lively 1995). The abundance of orange males increased from 1991 to 1992, but decreased from 1992 to 1995. The abundance of yellow males increased from 1993 to 1994, whereas the abundance of blue males decreased from 1991 to 1994, but increased in 1995. Definitive evidence for cyclicity will depend on data from future field seasons, but preliminary results from 1996 suggest that the population is beginning a second cycle in the predicted direction. Whereas a two-way polymorphism has been observed in cichlid fish, this is the only documented instance of a three-way cycle of morph abundance within a species.

Sinervo and Lively (1996) found that morphs that were low in abundance had a higher chance of producing successful offspring. Thus, the "more successful when rare" hypothesis was supported. But even more interestingly, they found that not only were rare morphs more successful, they were successful at the expense of specific competing morphs, just as the model predicted. Yellow sneaker males were successful at out-competing territorial orange males because the orange males are too busy patrolling and fighting to keep track of the unobtrusive yellows. Blue males, which display mate guarding behavior, were in turn successful at out- competing yellow males, which could not access females guarded by the blue males. And orange males were then able to out-compete blue males by being more aggressive (Fig. 1). This circle of complementary competitive abilities sets up a dynamic "rock-paper-scissors" process (Maynard- Smith 1982) in which the frequencies of the three morphs in the population cycles in a predictable way.

 This rock-paper-scissors game is an extraordinary biological story. Yet even more extraordinary is how this rock-paper-scissors game evolved and how information on hormonal mechanisms in the side-blotched lizards provides us with an exhilarating glimpse into the precise workings of the evolutionary process. Evolutionary biology focuses on genetic, and concomitant phenotypic, variability within populations. The amount of variability in a population can be increased by mutation and gene flow; it can be decreased by natural selection and genetic drift. Thus, when pondering the evolution of the rock-paper-scissors game in the side-blotched lizard, we are faced with the following question: How did the alleles for three color morphs arise in the population, and how are they preserved (i.e., why are some of the morphs not eliminated by natural selection or genetic drift)?

 Because many mutations influence adult phenotypes by producing slight changes in the trajectory of the developmental program (see Raff and Kaufman 1983), a detailed understanding of the evolution of the three-morph system in Uta stansburiana will depend on an understanding of the developmental processes that produced the three morphs. Sinervo's research is particularly dramatic because it has discovered specifics about how the developmental program is changed by mutations that result in adults with the different male strategies. The specifics of the evolutionary scenario for how the three-way cyclical polymorphism originated is as follows. Males of other populations of Uta stansburiana display blue throats, so the ancestral condition in the Merced County population was probably blue. Then, a mutation arose in a blue lizard creating an aggressive orange morph. Testosterone (Fig. 2), a steroid sex hormone produced in the gonads, produces orange throat coloration and high levels of aggression in side-blotched lizards. Also, orange morphs possess high testosterone levels. Sinervo therefore believes that a mutation triggering a high rate of testosterone production transformed the ancestral blue morph into the aggressive orange morph.

 Once orange males evolved, an opportunity was available for sneaker males to benefit from being nonaggressive and non-territorial, and thus being able to sneak copulations from females in the territories of orange males. Then, when a mutation for a nonaggressive yellow male arose in a population with a lot of orange males, there was a selective advantage for being a sneaking male and the yellow morph spread. Some males in the population are plastic in that they can switch from being yellow to being blue. Sinervo has shown that when these plastic males are given high levels of corticosterone (Fig. 3) early in their lives, they tend not to turn into blues. Therefore, yellow males may have arisen from blue males as a result of a mutation causing a high rate of corticosterone production. Interestingly, yellows have low testosterone levels, and when their testosterone levels increase, they are transformed into blues; thus testosterone could also have been involved in the evolution of the yellow morph.

 Once yellow males appeared, they invaded the orange strategy, the blue males in turn invaded the yellow strategy, and the orange males, in turn, invaded the blue strategy. The three- way oscillation had begun. Just as an animal's developmental program provides us with glimpses of the developmental stages of its ancestors (Raff and Kaufman 1983; Wolpert 1991), the year- by-year oscillation among the three morphs of lizard recreates, in each cycle, the evolution of the three-way system.

 In considering the above hypothesis about the evolution of the three-way polymorhism, you may be struck by the following question: Once a mutation arises creating the aggressive orange morph, why do not the oranges out-compete the blues and become the only strategy before the yellow morph arose? The answer is that the orange morph is self-limiting. Because high aggression, and high testosterone levels,extract a heavy toll on males, many oranges die after one year (conversely, blues can live up to five years). When the proportion of oranges in the population increases, the costs of being an aggressive orange rise; beyond a certain threshold of abundance the fitness of oranges drops dramatically and blues gain an advantage.

We now have a phenomenal picture of the evolution of an extraordinary cyclical polymorphism. Equally remarkable is the fact that female side-blotched lizards in Sinervo's population adjust their energetic investment in offspring of different sexes according to the fitness advantages offered by producing males versus females. Sinervo established that when orange females assess that there are a large proportion of orange females in the environment, they focus more of their energy to their female offspring. This makes evolutionary sense: orange females trying to survive and reproduce in an environment with a lot of aggressive orange females will need extra resources to be able to compete with other orange females. Conversely, Sinervo discovered that when yellow females assess that there a lot of orange females in the environment, they focus more of their energy to male offspring than to female offspring. This also makes sense; yellow females will do poorly against orange females, whereas yellow males enjoy a fitness advantage in an environment with many orange males because yellows can easily sneak copulations from the females on the orange males' territories.

 Just as Sinervo has discovered the hormonal basis for the evolution of orange and yellow strategies, he has also discovered the possible hormonal underpinnings of the allocation of resources among sexes in females. When corticosterone is added to orange females, they reallocate their energy to their daughters, whereas when corticosterone is added to yellow females, they reallocate their energy to sons.

 Sinervo has made many discoveries about the evolution of frequency-dependant sexual selection in Uta stansburiana. But his research is ongoing, and many exciting new discoveries about this dynamic three-way system undoubtedly await us. One future direction of research, which is being pursued by Program in Animal Behavior graduate student Yoni Brandt, is an investigation of the mechanisms of how the lizards assess the presence and density of different color morphs in their environment. Brandt's research will be highlighted in the next issue of the Program in Animal Behavior Bulletin.

LITERATURE CITED

Charnov, E. L. 1993. Life History Invariants: Some Explorations of Symmetry in Evolutionary Ecology. Oxford University Press, Oxford.

Maynard-Smith, J. 1982. Evolution and the Theory of Games. Cambridge University Press, Cambridge.

Raff, R. A., and Kaufman, T. C. 1983. Embryos, Genes, and Evolution. Indiana University Press, Bloomington.

Sinervo, B., and Lively, C. M. 1996. The rock-paper-scissors game and the evolution of alternative male strategies. Nature 380: 240-243.

Wolpert, L. 1991. The Triumph of the Embryo. Oxford University Press, Oxford.