Who has not dreamed of living forever? We know we won’t. Someone born in Japan today has a life expectancy of about 83 years. A white American born in 2010 is projected to live about 79 years on average, a black American 75 years. A French woman, Jeanne Calment, is said to have been mentally fully competent when she died in 1997 at the age of 122 years. But some other species live much longer. Greenland sharks (Somniosus microcephalus) were recently found to live for at least 272 years, and the largest individuals are estimated to be almost 400 years old. This is impressive, but hardly compares with a 5065-year-old bristlecone pine in California that is the oldest known unitary (“individual”) organism. The quaking aspens that grow on a mountain slope in the western United States appear to be separate trees but are often a clone: a single genetic individual, originating from a single seed, that may be more than 80,000 years old. Some organisms might well be immortal: they show no signs of senescence, the intrinsic changes that lower survival and reproduction with age. Some biologists think that clonal plants and fungal mycelia do not senesce, and this may be true of some sponges (estimated at more than 11,000 years old) and corals (more than 4000 years old). In contrast, many plants, insects, and even some mammals live for only 1 year, many small insects for only a few months, and some rotifers for fewer than 20 days. There exists enormous variation among organisms not only in maximum life span, but also in the process of senescence that foreshadows ultimate demise.
FIGURE 1 Species vary greatly in life span. (A) Bristlecone pines (Pinus longaeva), surviving in the punishing environment of desert mountaintops in California, are among the oldest known individual organisms. (B) Draba verna, a member of the mustard family, is an annual plant that germinates in early spring, sets seed within a few months, and dies. (C) Asexually propagating corals may not age, and persist for thousands of years. (D) Some rotifers live for only a few weeks. Clearly, life spans can evolve. It is possible that senescence does not occur in some clonally propagated organisms that retain primordial stem cells, which can differentiate into the organism’s diverse cell types [45]. But in most multicellular species, potential life span is one of several components of an organism’s life history that are intimately related to fitness, and that have evolved by natural selection. Fecundity (number of offspring) likewise varies. Many bivalves and other marine invertebrates release thousands or millions of tiny eggs in each spawning, but a blue whale (Balaenoptera musculus) gives birth to a single offspring that weighs as much as an elephant, and a kiwi (Apteryx) lays a single egg that weighs 25 percent as much as its mother.
FIGURE 2 Variation in fecundity (number of offspring). (A) Spawning oysters release clouds of minuscule eggs and sperm. (B) A coconut is a single enormous seed, and the coconut palm (Cocos nucifera) can produce only a few at a time. (C) Poplars (Populus) produce millions of tiny seeds, with fluffy hairs that enable dispersal by wind. (D) This X-ray of a kiwi (Apteryx) shows the bird’s enormous egg. (D, photo courtesy of Otorohanga Zoological Society.)
how low fecundity or a short life span can evolve by natural selection. In thinking about this, we must be clear about the level at which selection acts on these traits (see Chapter 3). Some biologists used to think that species such as codfishes produce hundreds of thousands of eggs in order to compensate for high mortality and ensure the survival of the species. Likewise, it has sometimes been suggested that animals die of old age to make room for a vigorous new generation. But we have noted that the future persistence or extinction of a species cannot affect, and is irrelevant to, the course of natural selection among individuals. So how can individual selection result in low reproductive rates or short life spans? The life history traits with which we are concerned are the ages at which reproduction begins and ends, fecundity at each age, and the average survival to each possible age. These traits affect the growth rates of populations and are major components of a genotype’s fitness. The age to which individuals survive in nature is often shorter than their potential life span, which would be attained only if extrinsic mortality factors, such as predation, disease, and food shortage, were not operating. The maximum life spans cited at the start of this chapter are closer to potential life spans than average realized life spans. An organism acquires from its environment a certain amount of energy and nutrients, which are allocated among several functions, especially self-maintenance (hence, survival), growth, and reproduction. (The “growth” portion is ultimately allocated to the other two functions.) We expect that there will be tradeoffs among functions: a fitness benefit of one function that is correlated with a fitness cost of another function. (This is another way of saying “there is no such thing as a free lunch.”) The fraction of energy and nutrients allocated to reproduction is sometimes referred to as reproductive effort. The trade-off between reproduction and all other functions is often called the cost of reproduction.
Fitness in age-structured populations
we defined fitness for the simple case in which individuals reproduce once and then die (a semelparous life history). Fitness was defined as number of offspring produced by an individual (taking into account the probability of surviving to reproductive age). For iteroparous species—those in which individuals reproduce more than once—the lifetime reproductive success is found by adding up the reproduction over all the ages at which individuals reproduce. To make the ideas concrete, take the example of an asexual lizard that starts to reproduce at age 2 and never lives longer than 3 years. We form a life table that shows the probability that a newborn will live to age x, symbolized as l x, and the average fecundity for a female at that age, symbolized as mx;
The last column of the table gives the product of the survival probability and the fecundity for the given age, l x mx. By adding those values over all ages, we find the expected lifetime reproductive success, symbolized as R:
In the life table shown above, R = 1. That means each female leaves on average one descendant, and the population is exactly replacing itself. If R is greater than 1, each female leaves more than one offspring, and the population size increases. Lifetime reproductive success is related to the intrinsic rate of increase, symbolized as r, which is widely used in ecology. If R is near 1 and time is measured in generations, lifetime reproductive success can be translated into the intrinsic rate of increase by the formula r ≈ ln(R). In a population that is stable in size, R = 1 and r = 0.
Senescence
Our life table illustrates a general point: natural selection does not act to prolong survival beyond the last age of reproduction. In the life table above, a mutation that increases the chance of survival to age 4 from 0 to 0.25 has no effect on R, because females do not reproduce at that age. (There are a small number of interesting exceptions. In humans and orca whales, postreproductive parents care for their offspring, so postreproductive survival may be advantageous [15, 19].) But why should reproduction cease? Why do women experience menopause, and older men have lowered sperm production and sex drive? The answer is that, all else being equal, the selective advantage of reproducing declines with age. Increasing survival and fecundity at earlier ages has a larger effect on fitness than at later ages, simply because predators, disease, and all sorts of accidents make individuals less likely to survive to the later ages. In the discussion of the life table above, we saw that a mutation that increases fecundity at age 2 from 1 to 1.2 increases fitness, R, from 1 to 1.1. Now consider a second mutation that increases fecundity by the same amount at age 3 instead of age 2, increasing l3m3 from 0.5 to 0.55. Comparing the life tables for the two mutations, we have: ↓
We see that the second mutation increases R from 1 to 1.05. This is a smaller increase in fitness than that for the first mutation, because a smaller number of individuals survive to age 3 than to age 2. The principle that increasing survival and fecundity at earlier ages has a larger effect on fitness than at later ages is the basis of the two major factors responsible for the evolution of senescence and life span. One, first identified by Peter Medawar, is mutation accumulation: mutations that compromise biological functions reduce fitness less, the later in life they exert these effects. That is, selection against these mutations is weaker, and so they persist at higher frequencies in the population than if they affected younger individuals. Mutation accumulation will cause the genetic variance for reproductive success to be greater for older than for younger age classes. Exactly that pattern was found in a study of a laboratory population of Drosophila melanogaster
FIGURE 5 Evidence for two hypotheses for the evolution of senescence and life span. (A) The genetic variance for reproductive success increases with age in Drosophila melanogaster, as expected under the mutation accumulation hypothesis. (B) Several laboratory populations of Drosophila were selected for reproduction at a young age (by propagating offspring only from young parents) or for reproduction at an older age (by growing flies only from eggs laid by old females). Flies from the old-selected populations laid fewer eggs when they were young (1 week after reaching adulthood), relative to the nonselected baseline populations. This effect is expected under the antagonistic pleiotropy hypothes. Studies in several species of birds and mammals have found that mating between relatives results in greater inbreeding depression expressed at later than at earlier ages. Because inbreeding depression is caused by homozygosity for partially recessive deleterious alleles, this pattern is consistent with mutation accumulation. Because of allocation trade-offs, alleles that increase reproduction early in life are likely to have a pleiotropic effect on reproduction or survival later in life. The greater fitness impact of early reproduction causes the advantage of reproducing when young to outweigh the pleiotropic disadvantages at greater ages. Therefore, reproduction will be expected to dwindle with age, and perhaps to cease altogether. The selective value of surviving to later ages likewise declines, finally to zero. Based on this principle, George Williams proposed a second factor that can explain senescence and limited life span: antagonistic pleiotropy. Williams suggested that a great many genes are likely to affect allocation to reproduction versus self-maintenance—that is, they incur a cost of reproduction. Alleles that increase allocation to reproduction (reproductive effort) early in life will thus reduce function later in life. Antagonistic pleiotropy can cause a negative relationship between early reproduction and both longevity and later reproduction. This has been found in many selection experiments with Drosophila. For example, several investigators have selected Drosophila populations for higher reproduction late in life (and therefore were also selecting for long life). The flies evolved higher latelife fecundity, but their egg production at younger ages was lowered—exactly as expected under the pleiotropy hypothesis. Although both antagonistic pleiotropy and mutation accumulation contribute to senescence, many biologists think antagonistic pleiotropy is often the more important factor. Both factors can affect many genes, so it is unlikely that a single cause of senescence can ever be found.
Evolution of the Population Growth Rate and Density
The values of survival and fecundity in a life table are affected by ecological conditions. We have seen that mutations that increase R will spread through a population, causing the population to evolve a higher rate of increase. But this potential for population growth is often limited, as resources become depleted or as predation or disease become more common. These factors cause population growth to be density-dependent, and constrain population size. In the simplest ecological models, the per capita growth rate of the population (r) declines in proportion to the population’s size (FIGURE 11.6). If we measure population growth per unit of time, r declines from its maximum possible value, written as rm, which in most species occurs when density is very low. The reduction of population growth causes population size to reach a stable equilibrium number that is called the carrying capacity, symbolized by K.
FIGURE 6 A model of density-dependent selection of rates of increase. Assume that a population contains two different genotypes, A and B. Their instantaneous per capita rates of increase are r A and rB, both of which decline as population density (N) increases. The maximum intrinsic rate of increase for genotype B (rm,B)—its growth rate at very low density—is lower than the maximum rate of increase for genotype A (rm,A). Genotype B has a selective advantage at high density, however, and it attains a higher equilibrium density (K) than genotype A: KB is greater than KA.
FIGURE 7 Some semelparous plants reproduce once, after many years, and then die. (A) The cabbage palm Corypha utan, of southeastern Asia, may produce up to a million flowers. (B) Bamboos engage in highly synchronous reproduction and then die, resulting in years of great food scarcity for animals that eat the shoots (such as specialized insects and giant pandas) and great food abundance for those that eat the seeds (such as finches). (Photos by D. J. Futuyma.)
