The Evolution of Life Histories
Stephen C. Stearns
Simple explanations are rare in life history evolution. If one asks ‘why do humans mature at 16-18 years?’ or ‘Why do albatrosses lay one egg?’, no single approach gives a satisfying answer. Each of the next four chapters examine one approach in some depth. By combining all four --- demography, quantitative genetics and reaction norms, trade-offs, and phylogenetic analysis --- we can place variation in life history traits into balanced perspective. However, because of the specialization of science and the amount of work involved, no study exists that could be used here to illustrate the power of combined approach. (p. 11)
Our theory may provide a synthesis.
No general definition of fitness has been found (
From our theory, no such definition exists.
We could follow the population for a long time to see which clone dominate numerically or endures the longest before vanishing. The same clone might achieve numerical dominance and endure the longest, but a clone that was never dominant might last longer than a number that had been dominant for shorter periods. We are faced with several interesting decisions. Is the numerical dominant clone fitter than the others, no matter how short lived its dominance, or is the clone endures the longest before vanishing fitter than the others, no matter how relatively rare it might have been? A good argument can be made for each view, but it is easier to calculate numerical dominance than time to extinction, and that is the approach most used. (p. 15)
Fitness is a composite, relative measure of birth and death rates. The empirical view is that the fitness of a phenotype depends both on the other phenotypes present in the population and on the environmental conditions. Empirical fitness is a property of both an individual and a context. The theoretical measure of fitness most widely used, r, ignores many aspects of empirical fitness. It is called the Malthusian parameter in genetics and the intrinsic rate of natural increase in demography. Another fitness measure sometimes used is R0, the per generation rate of multiplication. (p. 16)
For recent work on demography in random environments, see Orzack and Tuljapurkar (1989). (p. 20)
Periodic cicadas (generation time = 13 or 17years). It is often claimed that insects don’t know prime numbers. Most insects don’t know prime numbers because there is no need to know. However, Periodic cicadas certainly know prime numbers because by emerging at periods of years of prime numbers, they have the least amount of danger to encounter predators.
The fitness of a phenotype varies across an environmental gradient, and for many physical and chemical factors, physiological ecology has shown that fitness is usually highest at an intermediate value and decreases toward extremes. … therefore conceive of phenotypic plasticity as the breath of environmental tolerance. (P. 64)
Male frogs calling for mates in the neotropics also attract predatory bats. The rate at which the fringe-lipped bats capture frogs is higher when the frogs are calling. The bats can discriminate the calls of different species and different sizes of frogs, and they prefer palatable species small enough to handle. The males trade off the attractions of reproduction with the risk of death (Tuttle and Ryan 1981). (p. 74)
The efficiency with which a female can graze, and the quality of the forage she eats, both depend on the dominance of her mate. Dominant males protect their females and get them favoured positions in ungrazed areas. Females pair with subordinate males cannot spend as much time foraging because they are frequently interrupted by dominate males, and the quality of the plants on which they graze is lower. The females of dominate males are nearly 10 per cent heavier when they fly north to breed.
This difference in foraging efficiency and fat stores in the spring has two important consequences. Females of dominant males return in the fall with more offspring; the females that were lighter in the spring sometimes return with none. And females paired with subordinate males change mates more frequently from year to year than those paired with dominant males. (p. 78)
The nest is an environment of good survival and poor growth; the sea is an environment of poorer survival and better growth. (p. 80)
The mixed evidence for a phenotypic trade-off between reproduction and survival suggests the following.
The fruit grower … knows that after a particular good crop he may expect one or two years of poor crop and that by reducing the load of fruit in a bumper year he can increase the chance of a good crop in the following year. (p. 85)
Why crop output varies each year? I think it is because the variation will decrease of the opportunity of large scale parasites or something like that.
Sectioned individuals that varied in number of ovarioles and measured the percentage of area occupied by gonads and by fat: the more gonad, the less fat. … In bighorn sheep, ewes that raised sons, and lactating ewes, had a higher faecal output of lungworm than did ewes that raise daughters or that were not lactating. (p. 86)
Species that delay maturity tend to be large, long-lived, and if they are birds or mammals, to have a few, large offspring, or, if they are reptiles, to have many, small offspring. (p. 123)
Can we explain this? The age of maturity may be approximately as fixed cost.
In most analysis, competition is not modeled. In reality when ecological niches are fully occupied, r may not be very important by itself.
The breeding structure of a species influences the difference in age at maturity between males and females. In polygynous species where males compete directly with one another for females, males tend to delay maturity, grow larger, and gain more experience before attempting to reproduce. … On the other hand, in species with promiscuous mating, external fertilization and indeterminate growth --- where males do not control access to females --- one expect the opposite. Because females gain fecundity with size at a higher rate than males, one expects the males to be smaller and younger at maturity than the females. That is the case in many fish. … three categories of social structure:
Category I: Promiscuous species that form leks
Category II: Promiscuous species with dispersed males
Category III: Species that form pair-bonds
Categories I and II involve polygyny of various types. The essential point is that in leks (I) males compete directly with each other, by fighting, for the mating sites preferred by females. When the males are dispersed (II), the competition is less direct and less intense. In pair-bonding species (III), there is little or no competition among males for mates following pair-formation. In all three categories, females mature at one yer, but male maturation is delayed in social systems that promote competition among males for mates (Table 6.5). (p. 128)
The same pattern occurs in human
society. In social systems where divorce is prohibited or discouraged, marriage
is early. In social systems where marriage or partnership are
unstable, males and females have to compete for partners all their life,
marriages are late.
Human males are much larger than females, indicating that humans are polygynous.
The assumption that fecundity increases with size has been abundantly documented … Note that here fecundity increase with length within species but not among species. (p. 130)
Why?
Earlier maturing forms have shorter generation times, higher r, and a decreased period of exposure to mortality before maturity; later maturing forms may have increased fecundity and/or lower instantaneous juvenile mortality rates. At an intermediate point, an optimum should exist. The challenge is to estimate the relevant costs and benefits, model their interactions, then predict the optimum and see whether it matches the data. (p. 130)
Our theory may provide part of the answer that early maturation, being low fixed cost, is consistent with high r.
Not many quantitative predictions of phenotypic properties have been made in evolutionary biology. Those described here are among the most successful. The models do not move from assumptions about environments through selection pressures to predictions about phenotypes. They are much more narrowly focused than that, based almost entirely on the properties of whole organisms that are relatively disengaged from the environment, which enters only indirectly through its influences on mortality and growth rates. The predictions concern the adjustment of life history traits to one another, not to environmental conditions. (p. 148)
Our theory, by including sigma, may provide a direct link to environment.
That clutch size increases with latitude has been recognized at least since 1830s, … Geographic variation in reproductive investment is so well established that there is little sense in repeating such studies if the aim is just to document the exisitence of differences. It has been hard to use descriptions of such patterns to discover what causes the variation. (p. 154)
From our theory, r species thrive in high uncertainty environment. High latitude or altitude environment has higher variation in temperature than low latitude areas. That is why animals in high latitude areas produce more offspring.
Gadgil and Bossert found that a uniform change in mortality rates in all age classes has no effect on the optimal age at maturity or on the optimal age distribution of reproductive effort. … In Kozlowski and Uchmanski’s (1987) models, in which resources are represented, a uniform increase in mortality rates resulted in earlier maturity and a greater proportion of the growing season devoted to reproduction. Resource-based models are more realistic and get different results. (p. 162)
One effect is antagononistic pleiotropy, where the same genes that have positive effects on fitness components early in life have negative effects on survival late in life. Another is called mutation accumulation, where detrimental mutations that only act late in klife will not be as efficiently eliminated by selection as those that act only early in life and can accumulate in the population. (p. 182)
Our theory of project duration can give direct answer.
The cost of specialization into germ line and soma is death. (p. 183)
Sexual outcrossing allows exogeneous repair by the elimination of deleterious mutations present in some offspring but not others. Thus the evolution of sex can be viewed as the evolution of the mechanisms preventing the aging of the germ line. (p. 183)
Some semelparous organisms have evolved a hormonally controlled trade-off between resource allocated to reproduction and to repair. Pacific salmon that have been castrated live as much as 18 years longer than those that have not. Soya beans from which all fruits and flowers have been removed live long after their reproducing sibs are dead. The remarkable reproductive effort of male marsupial mice is associated with a strong rise in plasma androgen concentration and a breakdown in disease resistance. (p. 186)
Increasing the extrinsic adult mortality rate or decreasing the extrinsic juvenile mortality rate favors a shift towards semelparity by increasing the value of juvenile relative to adults. … Delayed maturity and lower fecundity favor a shift in the direction of iteroparity, a lengthening of the reproductive lifespan. (p. 188)
The experience with models for the evolution of iteroparity parallels that gathered for models of age and size at maturity and reproductive effort. The initial simplicity of the results attracted many to the field and is all that is remembered by some of those working in neighboring areas. Later we learned that there are no general predictions, for when the early intuitive generalizations were made mathematically rigorous, everything depends on quantitative details. (p. 198)
Thus the evolutionary answer to the question ‘Why do we age?’ has three parts. First, we have a clear separation of germ line and soma. Second, the force of selection declines with age until past some point, determined by factors discussed under the evolution of reproductive lifespan, older organisms are irrelevant to evolution. Third, under these conditions two sorts of genetic effects become possible, (a) the accumulation of more mutations with effects on older age classes than on younger ones and (b0 the accumulation of antagonistically pleiotropic genes that benefit younger age classes at the expense of older ones. (p. 200)
Our theory doesn’t depend on these three parts. Work out the details.
Repari is known to be costly. More than two percent of the energy budget of cells is spent on DNA repair and proofreading, on processes that determine accuracy in protein synthesis, on protein turnover, and on the scavenging of oxygen radicals. (p. 201)
Homozygous daughters lack ovaries and live much longer than normal females. (p. 202)
The cost of exposure to males in reduced lifespan is well known in wild-type flies and has been shown to consist primarily of the cost of mating. (p. 203)
The best evidence for the mutation accumulation hypothesis comes from the response to relaxation of selection for increased longevity. (p. 204)
Human longevity in today’s world may also be a response to relaxation of selection.
Physiology is moving closer to the centre of the life history evolution. This return to an attitude widely held two decades ago and never totally abandoned will give greater insight into phenomena already understood at other levels. (p. 204)
I totally agree.
Increase in the mean and variance of adult mortality should decrease the reproductive lifespan, eventually producing semelparity. Increase in the mean and variance of juvenile mortality can lengthen the reproductive lifespan, eventually producing very long lived organisms with high investment in somatic structures and their repair. (p. 205)
There is in general a positive correlation between adult and juvenile mortality. What is the combined effect in general?