Alternative Adaptive Peaks
A set of two or more phenotypic states that result in higher fitness than other states. This is in contrast to situations in which there is a single optimal phenotype, with all variation surrounding that optimum being maladaptive.
The fitness landscape is likely to be dynamic, such that the height and position of peaks are shifting when any aspect of the environment (including the social environment) changes (Lewontin 1978). When populations find themselves in valleys, adaptive processes should drive their phenotypes to the nearest peak. In some cases, populations can climb toward local peaks (local optima) that are shorter than the highest peak (global optimum), thus expressing phenotypes that are less optimal than other possible phenotypes. Traditional gradualist models of evolution predict that it will be difficult for populations on a local optimum to shift to a higher optimum as the population must first pass through an adaptive valley, in which phenotypes are less fit than they were at the local optimum. Sewall Wright argued that some combination of large mutations and genetic drift would allow small subpopulations to form phenotypes distant enough from local maxima that they reach the valleys (or saddles) of landscapes where selection will allow them to climb higher alternative peaks.
Alternative Adaptive Peaks
The fitness landscape metaphor has been used to help explain the persistence of adaptive genetic and phenotypic variation in populations, for example, the existence of personalities in humans (Nettle 2011). Populations with two or more behavioral strategies with equal fitness payoffs can be visualized as a landscape with two or more “alternative adaptive peaks” of equal height. There are two general situations in which such alternative peaks are expected to arise; where there is negative frequency or density dependent selection and where individuals within populations live in environments that vary across space and time.
Negative frequency-dependent selection occurs when the fitness advantage of a behavioral or phenotypic strategy decreases as it is more commonly used. Biological sex ratios represent an example of this type of selection, as whichever sex is more common tends to have lower fitness, thus maintaining a 50:50 sex ratio in most animal species (Maynard Smith 1978). Frequency-dependent selection can be visualized on a dynamic adaptive landscape as the lowering of peaks as they become occupied, and the concurrent rising of less occupied peaks (Arnold et al. 2001). The frequency of each strategy is expected to reach a stable equilibrium when the fitness of individuals using each strategy is equal, when peaks are of equal height.
A classic model of behavioral strategies maintained by negative frequency selection is the producer-scrounger polymorphism of foraging animals (Barnard and Sibly 1981). Producers invest energy to find or prepare food before consuming it, while scroungers rely on food found and prepared by scroungers. The first individual to scrounge will typically achieve higher fitness than producers, because it has potential access to the food produced by the rest of the population. However, as the frequency of scroungers increases (and the frequency of producers decrease), the fitness payoff for scroungers gets progressively lower, until it matches that of the producers. At this point, the height of adaptive peaks is equal and the frequency of each strategy is at a stable equilibrium.
It has been hypothesized that psychopathy (a cluster of traits including egocentrism and a lack of empathy) is maintained in human populations via negative frequency-dependent selection (Mealey 1995). As with the scrounger strategy, this hypothesis argues that the selfish/cheating behavior only thrives when a larger portion of the population is cooperative and susceptible to exploitation by psychopaths.
A related concept, “competitive disruptive selection,” describes situations in which what would be a single adaptive peak under low competition conditions becomes a stable local minimum when that peak becomes occupied at higher population densities. This would favor the diversification of phenotypes around that minimum. For example, in high population density conditions, where competition for prey is intense, stickleback fish with morphology that differs from the mean are at an advantage in specialized foraging (Svanback and Bolnick 2007).
Alternative peaks may also arise and maintain consistent variation when individuals within populations live in different environments. In this case, one horizontal axis might represent the habitat in which an individual found, while the other represents a trait that its utility depends on habitat. For example, the fitness payoffs for life history strategies, such as early investment in reproduction, depend on features of the habitat in which organisms are living. In a now-classic example of life history variation, Austad (1993) demonstrated that opossums living on islands where there was little risk of predation invested less in early reproduction and more in later reproduction compared to mainland opossums, where predation greatly reduces the probability of surviving long enough to reproduce later in life.
Adaptive Peaks and Personality
Applying adaptive hypotheses to explain human behavioral variation (e.g., personalities) is a relatively new endeavor, and thus, few studies have been published to date. Gangstad (2011) points out that low amount interculture variation in personality, compared to the consistently large intrapopulation variation, indicates that heterogeneous environments probably do not account for persistence of personality. This pattern instead favors the hypothesis that within environment, processes, like negative frequency dependence, maintain this variation. For example, Sulloway (2011) suggests that birth order effects may be driven by processes akin to competitive disruptive selection, as competition over roles and/or resources within human families favors to development of differences among siblings.
Adaptive peaks and the fitness landscapes on which they reside are useful metaphors for understanding adaptive evolution and development. However, as with most metaphors, the fitness landscape is limited in several ways. In particular, the fitness of individuals surely relies on more than two interacting traits, but our ability to visualize multidimensional landscapes that result when three or more traits influence fitness is limited. Moreover, the instability of the landscape – the degree to which its shape may change from moment to moment makes it difficult to use the landscape to predict where population phenotypes will be moving at evolutionarily relevant timescales (Kaplan 2008).
- Gangstad, S. W. (2011). Evolutionary processes explaining the genetic variance in personality: An exploration of scenarios. In D. M. Buss & P. H. Hawley (Eds.), The evolution of personality and individual differences. New York: Oxford University Press.Google Scholar
- Maynard Smith, J. (1978). The evolution of sex. Cambridge: Cambridge University Press.Google Scholar
- Mealey, L. (1995). The sociobiology of sociopathy: An integrated evolutionary model. Behavioral and Brain Sciences, 18, 523–541.Google Scholar
- Nettle, D. (2011). Evolutionary perspectives on the five-factor model of personality. In D. M. Buss & P. H. Hawley (Eds.), The evolution of personality and individual differences. New York: Oxford University Press.Google Scholar
- Sulloway, F. (2011). Why siblings are like Darwin's finches: Birth order, sibling competition, and adaptive divergence within the family. In D. M. Buss & P. H. Hawley (Eds.), The evolution of personality and individual differences. New York: Oxford University Press.Google Scholar
- Svanback, R. and Bolnick, D. I. (2007). Intraspecific competition drives increased resource use diversity within a natural population. Proceedings of the Royal Society, 274, 839–844.Google Scholar
- Wright S. (1932). The roles of mutation, inbreeding, crossbreeding and selection in evolution. In: Proceedings of the sixth international congress of genetics, Ithaca, Vol. 1, pp. 356–366.Google Scholar