Selection Pressures as a Function of Age
KeywordsNatural Selection Late Life Inclusive Fitness Pleiotropic Gene Strong Supporting Evidence
The strength of natural selection on mortality and fecundity varies across ages.
The insight that natural selection should be weaker with age was originally proposed by Ronald A. Fisher (1930) in A Genetical Theory of Natural Selection. In this groundbreaking book, Fisher fused Charles Darwin’s theory of natural selection with Gregor Mendel’s theory of particulate inheritance – a fundamental insight in the modern synthesis in biology. Fisher, however, did not fully explore how the force of selection differs across ages. That exploration would have to wait for decades, when biological theorists (Medawar 1952; Williams 1957) realized that weakening in the force of selection with age could explain how evolution not only could – but should – result in aging: the decrease of fecundity or increase in mortality with age. Soon thereafter, William D. Hamilton (1966) provided theoretical clarity on the issue by explaining mathematically how the strength of selection varies across age groups. Age-dependent selection continues to be a topic of considerable interest, largely in research on aging and life-history theory, which addresses how organisms adapt to unpredictable versus stable environments.
Ronald A. Fisher (1930) and John B. S. Haldane (1941) first proposed that the strength of selection should vary depending upon age, and more specifically that it should be weaker as adults get older. Fisher (1930) suggested that the reduction of strength of natural selection with increasing age results from the related reduction in reproductive value – the number of offspring an organism is likely to have summed across the current age and all later ages. Reproductive value is at its maximum value until reproduction begins, at which point it reduces throughout adulthood.
Age-Related Selection and Aging
Inspired by Fisher’s insight about change in reproductive value over the lifespan, the biologist Peter B. Medawar (1952) made an early attempt to explain the evolution of aging in which he argued that selection should act less at later ages because of the cumulative effect of mortality across ages. That is, by assuming that death rates and reproduction rates are constant across age, a population will have a demographic structure that is comprised of more younger than older adults, with smaller and smaller proportions for older and older age groups. In this demographic structure, younger adults will produce a larger proportion of offspring each year than older members of the population – and therefore younger individuals will contribute more genes to the next generation – resulting in stronger evolutionary selection upon younger than older individuals. Medawar therefore argued that genes whose effects were expressed earlier in life would experience stronger selection, whereas genes expressed very late in life would experience little to no selection. Medawar observed that wild animals infrequently live so long as to suffer from aging, so selection would be unable to eliminate harmful mutations that would be expressed at ages later than death usually occurs in the wild (see also Haldane 1941). The resulting buildup of deleterious mutations would cause aging when expressed in individuals living in safe environments, such as in a lab or a zoo.
George C. Williams (1957) also attempted to explain aging based on age-specific strength of selection. He argued, in particular, that strong selection early in life but weaker selection later in life would result in populations carrying pleiotropic genes – specifically genes with beneficial effects early in life but detrimental effects later in life. The harmful effects of pleiotropic genes in late life, Williams argued, would result in aging.
Building largely on the work of Fisher (1930) and Williams (1957), William D. Hamilton (1966) made explicit the mathematics underlying the intuitions of others about how the strength of selection varies across ages and therefore results in aging. In particular, Hamilton improved upon Fisher’s proposal that evolutionary forces work on reproductive value. For example, Hamilton showed that a population that is decreasing in size fast enough has stronger selection on older than younger individuals (Charlesworth 1980). Hamilton demonstrated mathematically how the strength of natural selection on survival and fecundity varied across ages during the life cycle.
Hamilton’s Forces of Natural Selection
Hamilton’s (1966) analysis identified two scaling factors for how the strength of selection changes across the lifespan: a survival factor and a fecundity factor. Rose and colleagues (Rose et al. 2007, p. 1265) referred to these factors as “Hamilton’s Forces of Natural Selection,” equating their importance for biology with Einstein’s contributions for physics. The survival factor looks at “the fitness impact of an individual’s future reproduction” across different ages (Rose et al. 2007, p. 1266). Its value is at its maximum of 1 before sexual maturity, 0 once reproduction ends, and decreases with age between those values. This factor looks at how much of an effect natural selection can have on the survival probability of an age group. For example, natural selection has a strong effect on the survival probability of prereproductive ages and no effect on the survival probability of ages after the cessation of reproduction (the 0 value once reproduction ceases), should such ages exist in a species. As per Rose and colleagues (2007), this force “has to decline throughout adult life until the last age of reproduction, after which the force of natural selection is zero for all remaining ages” (p. 1272). The second factor is age-specific fecundity. As per Rose and colleagues (2007), “the force of natural selection acting on fecundity … declines with age until the last age of survival in the population’s evolutionary history” (p. 1272). That is, the second factor evaluates how much natural selection can influence the fecundity of different age groups. A common result is that selection can strongly influence the fecundity of younger age groups but has less effect on the fecundity of older age groups.
Hamilton’s forces have some notable limitations and implications. When describing the death rates of humans, Hamilton (1966) suggested that the death rates should asymptote as the curve approaches the end of reproductive years. However, in humans the curve goes too far to the right (people live too long), especially for women. Hamilton agreed with Williams (1957) that the most likely explanation is parental care provided by mothers and grandmothers. This explanation is now referred to as the “grandmother hypothesis” and is also used to explain human menopause (Lahdenperä et al. 2004). By helping the survival and reproduction of kin in late life, selection would continue to work on mortality late in life and thereby increase the lifespan. Thus, although he did not incorporate parental care (or grandparental care) into his formulas for the strength of selection across ages, Hamilton acknowledged such care as another potential way for selection to act on different age groups.
Hamilton also addressed why infant mortality is so high in humans. He suggested that in general selection should drive deleterious effects to be expressed later in life (when they are less detrimental to fitness). The exception in the human case arises because offspring with deleterious genetic effects prohibit parents’ fitness and the children’s own inclusive fitness (the genes it shares with relatives, another idea for which Hamilton provided the definitive mathematical statement; Hamilton 1964). For example, Hamilton suggests that a child who dies at 10 years of age wastes more parental effort than one that dies soon after birth or even soon after conception. For children with a gene that is likely to result in death, by dying earlier in life the child can be replaced by another sibling and thereby increase inclusive fitness.
Finally, note that both forces of natural selection can have declined to the point of being zero, but that survival and fecundity rates themselves need not also be zero. This means that at sufficiently old age, aging itself stops – that is, mortality rates no longer increase – and fecundity rates no longer decrease (Rose et al. 2007). Given that both forces of natural selection are strong only in early life and eventually reduce to zero, natural selection should only result in somatic adaptations that are beneficial either for young individuals or for individuals across all ages; that is, bodily adaptations for late life are not expected (Rose et al. 2007).
Charlesworth’s (1980) review found substantial evidence supporting age-specific effects of genes that are consistent with Hamilton’s forces, but some anomalous findings as well. He considered five aspects of cross-species comparisons. First, strong supporting evidence is found for the presence of aging in species which separate soma and germ cells. Second, correlational evidence supports Hamilton’s scaling factors, such as species with greater external mortality – death due to uncontrollable dangers such as predators and disease – also aging faster. Third, species such as large reptiles, amphibians, and fish in which fecundity increases with age can live very long lives, as Hamilton’s theory predicts. Fourth, aging should begin immediately after reproductive maturity because of one of Hamilton’s scaling factors is constant before that point then decline. This is consistent with the reduced survivorship of humans immediately after puberty, the point in the lifespan of humans with the lowest mortality. However, the high juvenile mortality rates in many species are not anticipated by this theory. Fifth, for species in which reproduction ends before death, selection should allow deleterious genes to build up, but human females live after menopause. As mentioned above, Hamilton suggested this might be because they gain inclusive fitness by caring for children and grandchildren.
Diverse consequences result from the unequal strengths of selection across the lifespan and its relation to change in population size. One of the most important is that natural selection results in aging (e.g., Williams 1957). Another set of consequences are embodied in life-history theory, the idea that species (and individuals) that face relatively high rates of external mortality emphasize earlier-life reproduction at the cost of investing in growth, bodily maintenance, and parental investment (Del Giudice et al. 2015). This shorter-term strategy makes sense in unstable environments where long-term investments are unlikely to pay off.
Although Fisher (1930) had the insight that selection weakens as adults age, a full understanding of how natural selection on mortality and fecundity varies across ages did not arise until theorists attempted to explain how evolution would result in aging (Medawar 1952; Williams 1957). In particular, Hamilton’s (1966) mathematical treatment of the strength of selection not only provided a firm ground for addressing the evolution of aging, but laid the foundation for the major evolutionary perspective provided by life-history theory.
- Charlesworth, D. (1980). Evolution in age structured populations. Cambridge, UK: Cambridge University Press.Google Scholar
- Del Giudice, M., Gangestad, S. W., & Kaplan, H. S. (2015). Life history theory and evolutionary psychology. In D. Buss (Ed.), The handbook of evolutionary psychology (2nd ed., pp. 88–114). New York: Wiley.Google Scholar
- Haldane, J. B. S. (1941). New paths in genetics. London: George Allen and Unwin.Google Scholar
- Medawar, P. B. (1952). An unsolved problem of biology. London: H. K. Lewis.Google Scholar