Chapter 9 Summary

Biological economics combine the benefits and costs to an individual of expressing a given anatomical or behavioral trait into a net payoff. The expressed value of a trait is part of an individual’s phenotype and is a result of both heritable components—the individual’s genotype—and environmental influences, including learning. Evolutionary economics shifts the focus from the individual to the average payoff of a given strategy when alternative strategies are also present in the population. Selection occurs within a generation when individuals adopting one strategy gain a higher payoff than do those adopting alternative strategies: the relevant payoffs for natural selection concern fecundity and survival, whereas those for sexual selection involve successful competition for matings.
The proportion of a population adopting a given strategy changes over successive generations. Some of these changes are due to selection in prior generations. The degree to which selection in one generation contributes to changes in strategy frequencies in the next depends upon the heritability of the strategy: higher heritability results in greater responses to selection in successive generations. Other factors contributing to the changes in strategy frequencies over generations include drift (random variation in individual survival and reproduction), genetic and cultural mutation, and migration into and out of the population. Change in the proportions of alternative heritable strategies over multiple generations is called evolution.
It ought to be possible to predict the trajectory of subsequent evolution, taking into account the payoffs of alternative strategies. This effort can become complicated when, as with communication, the payoffs for one strategy depend on how often other strategies are currently being employed. This is called frequency dependence. One approach is to use economic models that simplify the evolutionary process to the minimum needed for useful prediction.
Evolutionary models based on the phenotypic gambit assume that the details of genetic heritability rarely affect the direction of evolutionary trajectories. Simple optimality models make this assumption and focus on evolutionary trajectories in which the payoffs of a given strategy do not depend upon how many other individuals are adopting that or other strategies. Because signaling payoffs do depend on the actions of both senders and receivers, simple optimality models are rarely of use in the study of communication. In contrast, evolutionary game theory explicitly incorporates frequency dependence in the payoffs of alternative strategies. This method looks for equilibria at which it does not pay for any individual to switch to another strategy. An equilibrium strategy that cannot be invaded by rare alternatives when it is itself common is called an evolutionarily stable strategy (ESS). A third approach, adaptive dynamics, begins with an equation that describes changes in strategy frequencies between generations, including any necessary frequency dependence, and applies this equation recurrently to track evolutionary change over successive generations. Equilibrium is only one possible outcome of an adaptive dynamic model: other possible outcomes include limit cycles (repeating patterns), bifurcations (sudden instabilities and transitions), and chaos (completely unpredictable trajectories).
There are a number of genetic complications that might undermine the phenotypic gambit. The nuclei of different species and different sexes of the same species may contain different numbers of copies of each chromosome. Depending upon how finely the nuclear genome of an animal is divided into separate chromosomes, recombination and random assortment during meiosis and sexual reproduction may not fully randomize which genes are found together in offspring, resulting in persistent linkage disequilibria. In many animals, sex is determined at the chromosomal level: in haplodiploid species such as many arthropods, females are diploid, whereas males are haploid. Among birds and mammals one sex (females in mammals and males in birds) has two homogametic (similar) sex chromosomes; the other sex has a pair of quite different (heterogametic) sex chromosomes.
More subtle complications involve interactions between genes. One of the two alleles at a chromosomal locus in a diploid animal may be dominant, or the two alleles may combine their effects; which happens depends on the dosage levels of the enzymes they produce. A given allele may affect more than one trait (pleiotropy), and the expression of any one allele may depend on what alleles at other loci are doing (epistasis). Where sex is determined by chromosomal differences, special adjustments must be made in the dosage levels of active alleles to maintain normal biochemistry. This in turn can lead to quite different evolutionary forces on the two kinds of sex chromosomes, differences in their evolutionary rates, conflicts between the sexes, and complicated patterns of sexual selection and signal evolution.
Several evolutionary models incorporate at least some of these genetic complications. Polygenic quantitative genetic models allow for correlations between multiple traits caused by pleiotropy or linkage diequilibria. While this is less constraining than the phenotypic gambit, it still ignores epistasis and assumes that the additive genetic determinants of heritability (summarized in a G-matrix) do not change significantly between generations. Evolutionary equilibria using this approach are conveniently visualized as peaks or ridges in adaptive landscape maps. Extended quantitative genetics models minimally allow for certain levels of epistasis that can cause the G-matrix to evolve between generations. More ambitious approaches relax nearly all constraints imposed by other models, but are computationally intensive and not easily interpreted. Like polygenic quantitative genetics, extended adaptive dynamics models allow for genetic correlations between multiple traits. Unlike quantitative genetics methods, these methods can track multiple kinds of trajectories besides equilibria, and seek indicators that will predict which trajectory is likely in a given case. Price equation modeling partitions the response to selection into additive components, in contrast to the multiplicative components used for quantitative genetics models. This allows it to handle many of the genetic complications avoided by other methods. The cost is that other methods must be invoked to provide the information needed to apply it repeatedly over many generations.
The evolutionary modeling methods differ markedly in their assumptions and methods, yet they often produce similar general predictions. One advantage of multiple approaches is the opportunity to refine a general result by invoking that model whose assumptions are most met in the specific case. The generality of the predictions made using one method can be tested by invoking another method with quite different assumptions.
The primary currency used to compare the payoffs of alternative strategies is relative fitness. For populations of stable size, fitness is measured as the average lifetime reproductive success of individuals adopting a given strategy. Where population size is changing, the finite rate of increase is usually a better predictor. However, competition, spatial patterns, and temporal heterogeneity can all shift propriety from one measure to the other. Invasion coefficients are analogues of finite rates of increase that can be extracted from adaptive dynamic models.
None of the usual fitness measures may suffice for social behaviors since strategy adoption by one individual might affect the fitness of another carrying the same alleles for that strategy. Relative inclusive fitness takes this possibility into account is by combining the effects of an animal’s actions on itself and others, each discounted by a coefficient of relatedness. This coefficient compares the probability that the animal and a recipient of its actions share alleles with that computed for two randomly chosen animals in the population. A donor animal may adopt costly actions to help (altruism) or harm (spite) another animal, depending on whether their relatedness is above or below the population average. Hamilton’s rule specifies the relative benefit, cost, and relatedness values that allow such behaviors to evolve. Suitable recipients might be recognized by donors because they exhibited markers associated with shared alleles (genetic kin recognition and greenbeards), or because both donor and recipients grew up in viscous populations. Spite becomes more likely in small or sub-divided populations and in dispersing offspring than in retained offspring. Group selection accounting partitions selective effects in a different way from inclusive fitness, and while equivalent, can provide additional insights into evolutionary processes.
Evolutionary economics are most often used to identify optimal strategies. Life history strategies depend on trade-offs between fecundity and survivorship that determine fitness. Reproductive value at any age is the sum of an individual’s current and future fecundities, each discounted by the probability that the individual will survive to the relevant age. There is usually a trade-off between current fecundity and the remaining contributions to reproductive value. Reproductive effort is the expenditure of resources, usually scaled to body size, during a given life stage. An optimal life history is one that adjusts reproductive effort so as to maximize reproductive value at each stage in life.
Signaling puts demands on commodities used for multiple functions and thus generates trade-offs between signal benefits and costs. The development and maintenance of signaling organs impose energetic costs on the organism. Per-signal energetic costs range from minimal expenditures for singing birds to near maximal aerobic scope values in calling frogs and insects. Even for low per-signal costs, significantly repetitive signal emission can constitute 20–50% of daily energy budgets. The impact of signaling energy costs depends on whether the participant is an income spender or a capital spender.
Risks to physical integrity, such as disease or damage, impose additional costs on communicating animals. Although the ability to regenerate lost body parts is widespread among animals, it is most often limited to immature life stages. Most species practice self- or allogrooming to maintain the condition of their external surfaces. Damage risks come from predators, parasites, parasitoids, slave-makers, and conflict adversaries. Some of these may gain proximity to communicating victims by mimicking or eavesdropping on their signals.
Brains are costly organs; enlargement of one region to augment a specific function imposes trade-off costs on other brain regions, organs, and functions. Variation in signal code acquisition strategies often reflects this trade-off. In many arthropods, fish, amphibians, and reptiles, mate-attraction signal codes are largely inherited; in contrast, oscine birds’ song codes require extensive social learning. Many individual identity signals have a mixed strategy; the sender requires little if any learning to produce the signal, but the receiver must learn to identify and interpret the signal. Brain trade-offs likely contribute to the reasons that most animals use imperfect signaling rules of thumb for decision making.
The value of information is the difference in expected fitness between two alternative signaling strategies. Because receivers that ignore signals have default strategies for decision making that are still far better than chance, a switch to signals must do at least as well as the default strategy and even better if it is to make up the costs of participating in communication. The threshold at which a receiver should pay attention to a signal rather than ignoring it is determined in large part by the signal set reliability. Senders and receivers may disagree over the optimal level of reliability.