Natural selection occurs where there is heritable variation in a population and where there are differences in survival and fecundity associated with this variation. Thus, in order for natural selection to operate there must not only be phenotypic variation, there must also be an underlying genotypic variation. The relationship between phenotype and genotype can be very complex (see Schlichting and Pigliucci 1998) but we will confine ourselves to simple situations where there is a one-to-one mapping of genotype to phenotype, or to cases of complete dominance. Genotypes, and thereby alleles, leaving the most descendants will tend to increase in frequency in the population through the process of natural selection. You will note that this last statement is not absolute, because if the heterozygous genotype leaves the most descendants, the proportions of the various genotypes may remain constant from one generation to the next (see section 11.2.2).

The various equations that quantify natural selection are largely developed intuitively by the use of empirical examples. For those who are interested, the mathematical details of the derivations are confined to a few text boxes and an appendix, but it is not necessary to be able to derive the equations yourself in order to understand them. Simulations are used to analyse and show the predictions of the equations, but their application to the natural world is left until Chapter 11.

The theory of natural selection is deceptively simple. We have seen in Chapter 2 that Darwin formulated the theory as a sequence of facts and logical deductions or inferences arising from these facts:

1. Individuals in a population vary in their characteristics, and these variations are heritable (i.e. genetically based) at least in part.

2. New variation is created generation after generation.

3. Parents produce on average more offspring than are needed to replace them, and so populations have the potential to increase exponentially. Resources are finite and so will be insufficient to sustain all offspring in the long term.

4. As a consequence, there will be a struggle for existence, and only a fraction (often a very small fraction) of the offspring will survive to reproduce.

5. Survival is not random with respect to variation, and some variations will be better able to survive and will produce more off spring than others. This results in the accumulation of favourable variations at the expense of variations that are less favoured, generation after generation. The characteristics of the population slowly change over time (i.e. evolve).

6. Given sufficient time, the accumulated change will be large, and over vast geological time periods could account for the production of all species from a single ancestor.

We will be examining many of these statements in more detail throughout this book. In this chapter we will amplify these six simple statements in order to discuss some of the popular misconceptions about the process of natural selection.

This chapter considers how an individual's chance of dying is influenced by its age and sex. After a preliminary discussion about age-specific death rates, we will review the various ways of constructing life tables, which tabulate the information on age-specific death rates in an orderly way, and finally we will compare some of the life tables of different species of mammals and birds.

Age-specific death rates

We can develop our understanding of age-specific death rates by considering the work of Peter and Rosemary Grant on the large cactus (ground) finch (Geospiza conirostris), in the Galápagos archipelago. During the period 1978–83 they marked 1244 nestlings and followed their subsequent survival year by year. The nestlings could not be sexed, and they made the reasonable assumption that half were male and half were female. Only 27 of the 622 female nestlings survived for one year, 20 for two years, 13 for three years, and so on, until all the females were dead by seven years of age (see Table 14.1 for full data set). If we plot the number of survivors versus age (Fig. 14.1) we obtain the shape of the survivorship curve.

The heavy early mortality obscures the shape of the curve beyond the first year of age. We can deal with this problem by plotting the number of survivors on a logarithmic scale (Fig. 14.2).

There are two conditions that are necessary for evolution to occur. First, the characteristics of an organism must vary in the population, and that variation must be related to differences in survival or reproductive success. Second, the variation must also have a genetic basis, at least in part. As a consequence, evolution changes the gene frequencies of populations. In Part 1, we noted that Darwin made a strong argument that natural selection was the main force driving evolution. However, the gene frequencies in populations can also be changed by other forces, such as mutation, migration, and even chance, and so we need to assess the importance of these factors on the evolution of populations.

The main purpose of the following eight chapters is to make a quantitative assessment of the various factors that affect the gene frequencies of populations. How do we measure the allelic and genotypic frequencies in populations, and how are they affected by sexual reproduction (Chapter 6)? How does genetic variation arise in populations and how is it maintained (Chapter 7)? How are gene frequencies in populations affected by mutation (Chapter 7), chance (Chapter 8), migration (Chapter 9) and selection (Chapters 10 to 12)? What are the relative strengths of these factors and how do they interact with one another (Chapter 13)? Thus, we will try to make an objective assessment of Darwin's theory of evolution by natural selection to see if it is supported by the theory of population genetics.

Darwin began his ‘Transmutation’ notebooks in the spring of 1837 primarily because of John Gould's taxonomic findings on the birds of the Galápagos Islands. The fact that there were different closely related species of mockingbirds on different islands seemed at odds with the explanation that all species had been created by God (see Chapter 1). Why would a deity create different species, living much the same sort of lifestyle, on islands that were within sight of one another? To Darwin it seemed much more logical that one or more ancestral species had migrated to the islands from South America (where related species were known to occur), and that subsequently they had diverged to form different species on different islands. If that is what had happened on relatively young volcanic islands, imagine how much divergence would be possible worldwide over a much longer geological time period. This transmutation of species would also explain some of the observations he had made in South America. For example, he had found fossils of the giant sloth and the giant armadillo in shallow deposits which indicated that they had become extinct relatively recently. They were also very similar in body form to the present-day species. Perhaps the giant forms had given rise to the smaller species before their demise, or the larger and smaller species had diverged from a common ancestor and the giant forms had lost in the competitive struggle for survival.

In the previous seven chapters, we have examined various factors that influence the allelic frequencies in the gene pools of populations. It is easy to become mired in the details and lose sight of the broad, overall picture. The purpose of this chapter is to summarize how genetic variation is developed, maintained and directed in populations, a process which we call microevolution, so that we can develop an overview and general understanding of the interrelationships of the various factors or processes. The scheme that we will be following is summarized in Fig. 13.1.

Evolution can be considered to be a two-step process: first, the production of genetic variation by mutations and genetic recombination and, second, an ordering of that variation by natural selection which may be influenced by processes such as genetic drift and migration.

Mutations

Genetic variation is originally created by mutations, which cause changes in the precise sequence of DNA in the chromosomes. Mutation by itself is not an important driving force of evolutionary change because mutations causing the same phenotypic change occur at very low frequencies, somewhere in the order of 1 × 10-5 to 10-8 per gamete. It would take many thousands of generations to effect a substantial change in allelic frequency as a result of mutation pressure alone (see Chapter 7). Mutation creates genetic variation in a non-directed fashion, i.e. mutations are not created in relation to their need.

Prior to the time of Charles Darwin, there were many fine natural history studies that shed some light on the areas of population ecology and animal behaviour. Studies on population genetics were largely related to the breeding of domesticated animals and plants. Although considerable success had been made in breeding new varieties of many species, how the characteristics of organisms were inherited remained a mystery. Carl Linnaeus had developed the binomial classification system during the previous century and collectors were roaming the globe finding ever more species and plotting the distributions of many species. The astonishing variety of organisms was becoming more and more apparent. There had also been speculation about the evolution of organisms, in fact Charles Darwin's paternal grandfather, Erasmus Darwin, had written on the subject in his book Zoonomia, but undoubtedly the most famous theory on this subject was that of Jean Baptiste de Lamarck in 1801. However, these evolutionary ideas had little scientific credence at the time when Charles Darwin was receiving his education. So we may ask: what led Charles Darwin to conclude that organisms had evolved from a common ancestor?

Charles Darwin: some important early influences (1809–31)

Charles, born in 1809, was the fifth of six children of the physician Robert Darwin and his wife Susannah. When Susannah died in 1817 the household was ruled by the triumvirate of Charles' older sisters, whilst his father was a domineering presence who had little sympathy with the antics of a small boy.

Conceptually, it is a simple matter to estimate age-specific birth rates. The total number of live offspring produced by samples of females in different age classes is documented, and then the age specific birth rates (Bx) are calculated by dividing the total number of live offspring produced by the sample of females in each age class by the sample size of females in that age class. Thus, if a sample of 30 two-year-old females gave birth to 120 live offspring during the course of the year, the age-specific birth rates of females in age class 2–3 would be 120∕30, or 4.0. In practice, however, it may be difficult to estimate the number of live births. For example, some species of fish lay more than 1,000,000 eggs per female, and it may be extremely difficult to assess what proportion of the eggs are fertilized, i.e. are viable, under natural conditions. Other animals are secretive, and so one cannot observe the number of live births directly. In these cases, a sample of the population may be collected to measure the proportion of the population that is reproductively active in each class and to count the number of eggs, or embryos, or placental scars per female so that age-specific birth rates may be calculated. However, the problem of determining the viability of the offspring at birth still remains.

In this last section of the book, we consider two different aspects of population biology. First, we examine some aspects of the interactions between different species. There are many ways in which species interact – symbiosis, commensalism, competition, predation, etc. – but we will only consider competition (Chapter 17) and predation (Chapter 18) because of space limitations. These two types of interactions have a very powerful effect on what Darwin termed ‘the struggle for existence’. Thus, it is likely that these two processes apply powerful selective forces on the characteristics of organisms. It will also be observed that in many cases, the behaviour of individuals plays an important role in these interactions.

Behaviour is considered in Chapters 19 and 20, and we return to some of the issues that Darwin raised in the fourth and seventh chapters of his book, The Origin of Species. After discussing the genetic basis of behaviour at the start of Chapter 19, the problem of altruistic behaviour is considered. In this type of behaviour, some individuals appear to reduce their fitness to help other individuals, and the most extreme example of this is the existence of sterile castes in insects. This type of behaviour appears contrary to the theory of natural selection, which states that only those traits that improve the fitness of an individual can evolve in populations.

This introduction to population biology is based on a 13-week course I have taught at the University of Saskatchewan since 1979. When I developed the course I was inspired by Wilson and Bossert's 1971 book, A Primer of Population Biology, by Emlen's 1973 book, Ecology: An Evolutionary Approach and by Wilson's 1975 book, Sociobiology. It was a revelation to me how these three books used an evolutionary perspective to synthesize such areas as population ecology, population genetics and behavioural ecology, because I had been educated in a tradition where such subjects were taught separately.

Over the past decade I became increasingly frustrated in my attempts to find an appropriate text for my course. There are many superb texts available: encyclopedic texts on either ecology or evolution; more specific texts dealing with population ecology or population genetics or behaviour; and a few texts that cover two of these more specific areas, but to cover the breadth of material I teach would require using parts of two or three of these books. What is disappointing, however, is the lack of any evolutionary perspective in most of the ecology books. This is surprising given that Darwin used various principles of population biology to develop his theory of natural selection: the potential for geometric growth of population numbers, and the limitation of resources that leads to a struggle for existence through the effects of competition, disease, and predation.

There must be genetic variation for evolution to occur. Mutation is the ultimate source of genetic variation, which is amplified by recombination during sexual reproduction. Mutations will only play a role in evolution if they are heritable. In most organisms this means that only the mutations occurring in the germ line leading to the production of gametes may have evolutionary consequences.

Gene mutations

The word mutation may refer to any change in the genetic material, ranging from a change to a single base pair in DNA, to changes in the structure and number of chromosomes. The discussion of mutation and genetic variation in this book will only consider mutations within a gene, and this gene mutation can be simply thought of as a change in the sequence of DNA. In principle the DNA must be sequenced to detect a mutation, but in practice most mutations are identified and named by their phenotypic effects.

The simplest kind of gene mutation is the substitution of one base pair by another. These point mutations, as they are called, may result in the replacement of one amino acid by another, but in many cases there is no change in the amino acid because of the redundancy of the genetic code (Fig. 7.1). In the example of isoleucine, two of the three substitutions in the third position do not result in a change of amino acid.

CHAPTER 4

1. The λ per year = 6000∕5000 = 1.2, and so the rm per year = ln (1.2) = 0.18232 (using Eqn 4.6). The population size after three years can be estimated using either Eqn 4.2 (Nt = 5000 × 1.23 = 8640) or Eqn 4.2 (Nt = 5000 × e0.18232 × 3 = 8639.96, or 8640).

2. The λ per century = 900 million∕600 million = 1.5, and so the rm per century = ln(1.5) = 0.4055 (using Eqn 4.6). The rm per year = 0.4055∕100 = 0.00405. The λ per year can be calculated, using Eqn 4.5, as e0.00405 = 1.004, or approximately 0.4% per year.

3. A 15% increase per year = λ of 1.15 per year. When the population doubles in size, Nt∕N0 = 2. If we rearrange Eqn 4.2, we can see that 2 = λt. Taking the logarithm of both sides (i.e. ln(2) = ln(λ)t) we find 0.6931 = 0.14t, and so t = 4.96, or approximately 5 years.

4. First convert the rm per week to rm per day so that rm and t are in the same time units. So, rm of 0.14 per week is equivalent to rm = 0.14∕7 = 0.02 per day. Then use Eqn 4.4 to estimate Nt, setting N0 to 24, rm to 0.02, and t to 65. The answer is approximately 88 rats.

5. The multiplication rate (λ) over a four-week period is 5. Using Eqn 4.5 we calculate the rm per four weeks as ln(5) = 1.6094. The rm per day is 1.6094∕28 = 0.0575, and we may use Eqn 4.5 to calculate the λ per day (λ = e0.0575 = 1.059).

6. […]

So far we have considered characters determined by a single gene with two alleles, occurring in sharply contrasting states, which can have a major affect on the fitness of the organism. In some cases we are justified in modelling selection in this manner, but in many cases, probably the majority, we are not. It is possible to expand the basic theory to consider characters determined by two gene loci, but this approach is no longer useful when we consider characters that are determined by many genes. In these cases we may observe a general relationship between parent and offspring, which suggests that there is an underlying genetic basis to the trait, but we usually do not know how many genes are involved or how they interact. In addition, we may also be aware that the environment influences the trait to some extent. Consequently, in order to study these traits we examine their variability, and attempt to dissect this variation into its genetic and environmental components. This type of analysis is called quantitative genetics.

We can consider three types of quantitative traits (Hartl and Clark 1989):

1. Meristic traits in which the phenotype is expressed in discrete, integral classes. Examples include litter size or number of seeds produced per individual, number of flower parts, and kernel colour in wheat.

2. Continuous traits in which there is a continuum of possible phenotypes. Examples include height, weight, oil content, milk yield, human skin colour, and growth rate. In practice, similar phenotypes are often grouped together into classes for the purposes of analysis.

3. […]

In chapter six of The Origin of Species, Darwin showed that behavioural traits are evolutionary adaptations that have evolved by means of natural selection in just the same way as morphological and physiological traits. For this to be true, two conditions are necessary. First, variation in behaviour must be related to differences in survival or reproductive success, and second there must be a genetic basis to this variation in behaviour, at least in part.

It is not difficult to see that variation in behaviour can influence survival and reproductive success. For example, the success of lions in catching and eating animals like wildebeest (Connochaetes spp.) or zebra (Equus spp.) depends partly on their ability to stalk and get sufficiently close to the herd so that they are able to catch and bring down an animal when they make their final attack. If their hunting technique is good, they may be successful, but if they have a poor hunting technique they will probably see their intended prey escape before they can reach them and will go hungry as a consequence. On the other hand, the flight response of wildebeest and zebra depends on their ability to detect the lions before they attack, and this requires constant vigilance as we saw in Chapter 18. The least vigilant individuals, and those that are slow to respond when predators are detected, are the ones that are most likely to be killed in an attack.

We have come to the end of a journey during which we have introduced the four main areas of population biology. By now you will be realizing that a full understanding of the subject requires a grasp of the basic principles of evolution, population genetics, population ecology and behavioural ecology. A true synthesis of these areas is demanding because there are so many connections to be made as we shift between genetics, ecology and behaviour, and then try to make sense of it all from a Darwinian perspective. Nevertheless, it is important to attempt some form of synthesis because we will obtain a much more complete understanding of whatever process or phenomenon we are studying. Consider the following two examples.

First, the interaction between predators and their prey. There are numerous models of the growth of predator and prey populations that try to explain how predators affect the numbers and growth of their prey populations, and vice versa. Although these models help us to understand something about this type of interaction, a lot of questions remain. Why do some predators switch from eating one type of prey to another, and why can one predator limit the numbers of its prey but another cannot? Answers to these types of questions requires knowledge of an array of different factors. For example, the behaviour of both predators and their prey affect who is eaten and at what rates; the rates of energy acquisition versus energy expenditure may vary for different prey items, and predators may vary their diet accordingly; and genetic variation amongst the prey may make some more susceptible to predation than others.