1. Evolution: Micro versus macro
2. “The ecological theater and the evolutionary play”
3. Microevolutionary mechanisms
4. Contemporary microevolution
5. The unintended consequences of human technology
8. Genomics and microevolution
Microevolution occurs within and among populations of a species and usually involves changes in the mean value or relative frequencies of alleles and phenotypes that are shared by most populations of the species. Divergence among populations of a species (i.e., conspecific populations) is often associated with habitat differences, and such divergence often has important ecological consequences. Population genetics deals with evolution in terms of allele and genotype frequencies within populations, so it provides the theoretical foundation to study microevolution. Widespread species typically exhibit geographic variation, which has generally been thought to take thousands of generations to evolve. However, recent research on contemporary evolution suggests that geographic variation can evolve within a few generations after species colonize new habitats or experience environmental change. The high rate at which microevolution can occur is important because it means that pathogens, pests, and harvested natural populations can rapidly evolve traits that adversely affect people. DNA variation within and among conspecific populations can be studied as a product of microevolution, and it also provides powerful tools to tease apart the contributions of common ancestry and local adaptation to the evolution of geographic variation. Thus, previously intractable problems in microevolution and its applications to natural resource management can now be studied using the emerging technologies of molecular biology and genomics.
character displacement. This is the evolution of enhanced differences between species where they occur together as a result of selection against members of one or both species that use the same resources as members of the other species (i.e., ecological character displacement) or against individuals that tend to hybridize with members of the other species (i.e., reproductive character displacement).
cline. A cline is a geographic gradient in the frequency or mean value of a phenotype or genotype.
monophyletic group. This is a group of species that are more closely related to each other than any is to species outside the group.
phenotypic plasticity. A change in an individual phenotype that does not alter its genetic constitution and is not inherited by its offspring.
random walk. In population genetics, this is a change in allele frequencies from their initial values as a result of repeated episodes of genetic dirft.
taxon. A taxon (including higher taxon) is any named group (e.g., Vertebrata, Mammalia, Homo sapiens) at any taxonomic rank (e.g., Kingdom, Class, Species); higher taxa are more inclusive.
Biological evolution is change through time in the heritable properties of a lineage or monophyletic group (clade). Microevolution is generally confined to evolution within and among conspecific populations, and it occurs within relatively short time spans. In contrast, macroevolution involves changes in the number or characteristic properties (e.g., average body size) of the species of a clade. It depends on the variation among species generated by microevolution and unfolds over longer periods. Nevertheless, the definitions of microevolution and macroevolution have been controversial, and there is disagreement about their mechanistic relationships and even the value of the terms.
The division between microevolution and macroevolution is usually placed at speciation because members of different species do not routinely interbreed, and the evolutionary fates of separate species are largely independent. Microevolution involves changes in the frequencies of alleles and genotypes and of interactions between different genes. These changes are manifested as recognizable changes in the mean values or frequencies of biochemical, physiological, behavioral, developmental, and morphological phenotypes. A separate set of macroevolutionary mechanisms influences the probability of speciation and extinction. Thus, properties of species that promote speciation or impede extinction will tend to increase in a monophyletic group over time. Both microevolution and macroevolution contribute to biodiversity, but microevolution affects individuals and changes the properties of populations, whereas macroevolution alters the relative frequencies of species with different properties.
There are also practical reasons to distinguish microevolution and macroevolution. Microevolution can be studied using comparative, observational, or experimental methods to study individuals and populations over a few generations in the laboratory and field. Existing genetic properties and ecological conditions can be used to interpret microevolution. In contrast, macroevolutionary studies focus on differences among species. Careful species description, characterization of clades, and investigation of phylogenetic relationships among taxa are paramount in macroevolutionary research. The environmental factors and genetic properties that influenced speciation and extinction have typically been lost in the dim past and are hard to infer. Consequently, microevolution and macroevolution are generally studied using different methods.
G. Evelyn Hutchinson’s famous 1965 book, from which the title of this section was borrowed, emphasized that evolution occurs within an ecological context. Although existing genetic properties of a population (e.g., presence of an advantageous allele) influence its microevolutionary response to natural selection, ecology is a major factor in microevolution and a crucial source of information to interpret it. Furthermore, if environmental differences cause microevolutionary divergence among conspecific populations, they will exhibit differences that must be taken into account in ecological studies. Studies of microevolution and ecology are intimately associated and reciprocally illuminating.
Because microevolution involves changes in the relative frequencies of heritable traits within populations of a species, it can be analyzed in terms of the behavior of alleles and genotypes within populations. This is the subject of population genetics, and the Hardy-Weinberg equilibrium is the starting point to develop the genetic theory of microevolution. The Hardy-Weinberg equilibrium describes the distribution of alleles among diploid genotypes in a population in the total absence of evolution. It will be sketched here only briefly, but most textbooks on evolutionary biology develop it in detail (see chapter I.15).
If no evolutionary forces impinged on a population, the relative frequencies of alleles and genotypes in the population would reach equilibrium values that would never change after one generation of random mating. Genotype frequencies under these Hardy-Weinberg equilibrium conditions can be calculated using the simple equation, 1 = (p + q)2, where p and q are the relative frequencies of two alleles of a gene and must sum to 1. Of course, no real population ever conforms to Hardy-Weinberg equilibrium conditions, although deviations from equilibrium frequencies are often so small that they are undetectable. Detectable deviations from equilibrium frequencies, however, indicate that microevolution is taking place and may suggest its causes. Potential causes for deviations from equilibrium frequencies include mutation, meiotic drive, assortative mating, gene flow, genetic drift, and natural selection, the most important of which are mutation, gene flow, genetic drift, and natural selection.
Mutations are heritable changes in DNA and are the ultimate source of variation for microevolution. However, mutation rates are so low (10–4–10–9 mutations/generation/trait/individual) that they do not usually produce measurable departures from expected Hardy-Weinberg equilibrium frequencies. Phenotypic changes caused by mutation are not biased (i.e., random) to produce adaptation.
Gene flow occurs when an individual is born in a source population and reproduces after migrating to a recipient population. Its effects depend on the magnitude of genetic differences and rate of migration between the donor and recipient populations. Gene flow is frequently a more important source of genetic variation than mutation, but it also tends to homogenize populations that otherwise would evolve differences.
Genetic drift is the change of allelic frequencies between generations just by chance (i.e., sample error). The magnitude of genetic drift is inversely related to effective population size (Ne), which, in turn, increases with the number of breeding individuals and evenness of the sex ratio and decreases as variation in the number of offspring per pair increases. When Ne is large, genetic drift causes very small differences between successive generations, but if Ne is small for even one generation, it can cause major changes in the genetic composition of a population. Even if Ne is consistently large, there will always be some drift, and its effects will accumulate, causing a “random walk” of small deviations that can add up to major changes in allelic frequencies over many generations. In populations with small N e, rare alleles tend to be lost by drift, and even traits that are disfavored by natural selection can drift to high frequencies. Because there is initially only one copy of a new mutant allele, even advantageous mutations tend to be lost by drift.
Natural selection depends on three components: survival selection, fecundity selection (ability to produce offspring), and sexual selection (mating success of individuals compared to other members of the same sex). Each component results from differences in the relative rates of success of different phenotypic classes, and selection can be quantified as Darwinian fitness or a selection coefficient. Darwinian fitness depends strongly on the interaction of the phenotype with the environment, and environmental changes can cause changes in fitness associated with a phenotype. If a phenotype with high fitness is heritable, alleles that produce it will tend to increase through time (i.e., evolve). New phenotypes may appear in a population by means of gene flow, mutation, and sexual recombination of existing alleles, and natural selection can increase their frequencies and cause a population’s phenotype to evolve beyond its previous range of variation. Thus, natural selection is the major cause for evolutionary adaptation and phenotypic divergence among conspecific populations.
Genetic drift also causes divergence among conspecific populations, and it is necessary to distinguish the effects of drift and selection. Phenotype–environment correlations indicate selection but are not sufficient to identify the environmental variable that causes it. Further evidence based on differences in function or Darwinian fitness of phenotypes is necessary to confirm inferences based on phenotype–environment correlations. It is surprisingly difficult to establish that natural selection has caused microevolution of a specific trait.
Not all phenotypic variation among conspecific populations represents microevolution. Phenotypic plasticity may result from conditions experienced by the individual. Genetically identical individuals or the same individual at different times may differ because of phenotypic plasticity. For example, muscles may grow larger from exercise, skin may become darker from exposure to sunlight, and learning may alter behavior, but such changes do not affect the genetic constitution of the individual that experiences the phenotypic plasticity or that of its progeny. Similarly, maternal effects, phenotypic differences caused by a female’s nongenetic contributions (e.g., messenger RNA, yolk, parental care) to its progeny, may influence the offspring’s phenotype but not be inherited. However, the individual’s ability to exhibit phenotypic plasticity (i.e., show a phenotypic response to environmental variation) may be heritable, and thus plasticity may evolve. Phenotypic plasticity may cause nonheritable but ecologically important phenotypic variation among conspecific populations. Much of the phenotypic change caused by human environmental disturbance apparently results from phenotypic plasticity and not microevolution.
It is widely believed that microevolution is rarely rapid enough to be observed in progress. For many years, industrial melanism in the UK stood as the lone well-confirmed example of contemporary evolution. The peppered moth, Biston betularia, and other moths and beetles evolved dark pigmentation where soot from industrial pollution darkened tree bark and killed light-colored lichens on which the moths rest during the day. Although the speed with which industrial melanism evolved was never in doubt, questions arose about evidence that bird predation selects against moths that contrast with bark color. Recent results seem to confirm this effect, and many other cases of rapid evolution in response to human-induced environmental change have been reported in recent years.
Initially, most of these cases involved evolution of resistance to insecticides by insects and to antibiotics by bacteria. It seemed possible that these cases of contemporary evolution might differ from typical microevolution. Insects have large populations and short generation times, both of which favor rapid evolution, and bacteria have even larger populations and shorter generation times. Additionally, selection imposed by human technology might be more severe than selection under natural conditions. However, it is also possible that evolution of resistance in insects and bacteria may not be atypical; it may just be more conspicuous because it has serious consequences for people. Consequently it is quickly noticed and carefully studied.
Figure 1. Variation in armor, size, and shape of threespine stickleback. The specimen in the middle is a completely plated (high Eda allele), anadromous (sea-run) stickleback, and those around the periphery are low-plated (low Eda allele), but otherwise phenotypically diverse, freshwater stickleback from western North America. All specimens were drawn to be the same size, and the scale bars equal 1 cm. Variable armor traits include the length and number of dorsal spines, expression of the pelvis (including absence), and number (including zero) and distribution of lateral plates. (Reprinted with permission from Bell and Foster, 1994)
Darwin’s finches are the classic case of adaptive radiation, and they have been closely observed for decades (see chapter I.19). They occur on the Galápagos Islands of Ecuador, which are relatively undisturbed. Dramatic evolutionary changes in body and bill size and in bill shape evolved in the medium ground finch, Geospiza fortis, and the cactus finch, G. scandens, in response to climate-induced changes in food availability during a 30-year period. Similarly, field mustard, Brassica rapa, grown from seeds collected after a 5-year drought in California exhibited higher tolerance to drying and flowered earlier than seeds collected before the drought.
Sea-run (anadromous) threespine stickleback fish, Gasterosteus aculeatus, which colonized a lake in Alaska, also evolved rapidly under relatively natural conditions. Freshwater stickleback are phenotypically diverse and differ strikingly from sea-run populations (figure 1). Within a dozen generations after the lake was colonized, several armor, body shape, and feeding traits evolved, and this young population has become indistinguishable from adjacent lake populations.
Contemporary evolution is not restricted to life history and anatomical traits. The well-known fruit fly, Drosophila melanogaster, originated in the Old World but has been transported to every continent except Antarctica by humans. It lays its eggs in rotting fruit, where ethanol may reach lethal concentrations. Alcohol dehydrogenase (ADH) is one of the enzymes that detoxifies ethanol. There are two common alleles for the alcohol dehydrogenase gene, Adh-F and Adh-S. Adh-F is less stable at high temperatures and has higher activities at lower temperatures. It gradually increases in frequency (i.e., clinally) going away from the equator in Australia, Asia, and North America. The functional attributes of enzymes encoded by these enzyme alleles (i.e., allozymes) and other evidence suggest that these clines result from natural selection, and they must have evolved since D. melanogaster was introduced to these continents. Moreover, within the last 20 years, the position of an Adh cline in Australia has shifted southward by 400 km so that high Adh-F allele frequencies now occur farther from the equator than before, as would be expected from global warming. Thus, this recent microevolutionary change can be used to monitor global climate change.
Numerous additional examples of contemporary evolution have been described in recent years (see Hendry and Kinnison, 2001), indicating that natural environmental change and colonization of new habitats frequently cause detectable microevolutionary change on a human time scale and that microevolution is often so fast that close monitoring is necessary to catch it in the act.
In 1962, Rachel Carson’s Silent Spring sounded the alarm that insecticides, which had been in use since the late 1940s to limit crop pests and disease vectors, were also eliminating many desirable, nontarget species. She did not know that insecticides also caused rapid microevolution of resistance in the insect populations that were their targets. Of course, large population size, short generation time, and intense selection by insecticides should cause rapid evolution of resistance. By 1997, more than 500 insect species had evolved resistance to insecticides. Weeds were no different, and herbicide resistance had evolved in more than 200 weed populations by 2001. It has also become painfully clear that pathogenic bacteria, fungi, protozoans, and metazoans routinely evolve resistance to drugs that had previously produced cures. These microevolutionary responses adversely affect human health and agriculture, but they can sometimes be mitigated by using management strategies based on microevolutionary theory.
For example, genes from the bacterium Bacillus thuringiensis (Bt) have been inserted into the genome of cotton plants, conferring on them the ability to express Bt toxin, which kills pink bollworm, Pectinophora gossypiella. After 8 years of planting Bt-transgenic cotton, bollworm populations in Arizona still had not evolved resistance to Bt toxin. Failure of P. gossypiella to evolve resistance was apparently achieved by growing small plots of nontransgenic cotton, where nonresistant bollworms thrived. Because the nonresistant bollworms far outnumbered rare, resistant bollworms that survived in the Bt-transgenic-cotton fields, and they tend to disperse into the Bt-transgenic fields, most bollworms the following season are either nonresistant or hybrids between resistant and nonresistant parents, both of which are killed by Bt toxin. By sacrificing small plots of nonresistant cotton to bollworm infestation, natural selection favoring resistance to Bt toxin has been retarded by gene flow from nonresistant boll-worm populations.
Experience with the evolution of antibiotic resistance has been far less encouraging. By 1943, penicillin production was under way, and other antibiotics would soon follow. Even as the first antibiotics became available for clinical use, penicillin-resistant bacteria had already appeared, and resistance soon occurred in one bacterial pathogen after another. For example, penicillin could easily cure Staphylococcus aureus (staph) infection in the early 1950s, but by the late 1960s it had become ineffective. Methicillin still worked, but it became ineffective by the 1990s. Most staph infections can still be cured by vancomycin, but resistance to this “antibiotic of last resort” is spreading. New drugs continue to be developed, but this is an expensive arms race with tragic consequences and no end in sight.
In retrospect, microevolution of antibiotic resistance in bacteria is not surprising. Genetic variation for antibiotic resistance is common in bacterial populations, and their large size (i.e., Ne) and short generation time both facilitate the appearance of new mutants. Genes for resistance may protect bacteria from multiple antibiotics, and they can be transferred in plasmids between bacterial species. Natural selection for antibiotic resistance has been hastened by the misuse of antibiotics for diseases against which they are ineffective, failure of patients to complete antibiotic treatment, and widespread, chronic, low-dose antibiotic treatment to increase livestock productivity. All of these practices selectively eliminate less resistant bacterial clones, leaving behind more resistant ones to found new bacterial populations.
Under favorable conditions, it is also possible to select for reduction in the severity of disease (i.e., virulence). Many pathogens rely on their host’s social interactions to spread to new hosts before it dies or mounts an immune response that eliminates the infection. Consequently, if hosts with the most severe infections are isolated from other susceptible individuals, the most virulent pathogen strains will fail to spread. For example, installation of window screens in the southeastern United States during the first half of the twentieth century prevented mosquitoes from biting people who stayed indoors with the most serious cases of malaria, contributing to evolution of lower virulence in Plasmodium, the malaria pathogen. By separating malaria victims with the most severe symptoms from mosquitoes that spread Plasmodium, the most virulent strains could not spread, and the disease became less serious.
Human-induced microevolution may also play a crucial role in the loss of valuable commercial fish populations. Commercial fishing gear selectively captures larger fish, but smaller individuals slip through the net. Fisheries policies are intended to allow young fish to escape and grow to a larger size, at which they both reproduce and become more valuable as food. However, selective fishing for larger fish also favors individuals that stop growing at a smaller size and reproduce earlier in life. Because body size and reproductive schedules are heritable, size-selective fishing should cause evolution of smaller adult body size and earlier reproduction. After fishing is halted, the survivors should have genotypes for smaller body size and early reproduction. Many commercially fished populations never recover numerically, and those that do are often descendants of small individuals from which they inherit small body size. The conclusion that size-selective fishing has caused evolution of smaller body size in commercial fishes has been controversial because the quality of the marine habitats in which declining fish populations live has also deteriorated. Nevertheless, a growing minority advocates the incorporation of microevolutionary principles into fisheries’ management policy.
Variation that has evolved among populations in response to local ecological differences is a common phenomenon and an important source of evolutionary insight. Comparison of mainland or large central populations to populations on islands or peripheral habitat patches (e.g., mountain tops, desert springs) often reveals variation that is associated with environmental differences. Island populations may be isolated from predators, competitors, parasites, and pathogens that occur on the mainland, or they may encounter resources that are unavailable on continents, leading to evolution of unusual traits. The divergent properties of insular populations must be interpreted with care because insular populations are often small (i.e., low Ne) or were bottlenecked in the past, allowing genetic drift to influence their evolution.
Ecological character displacement may occur when closely related species that are usually allopatric occur sympatrically. In sympatry, the members of each species that most closely resemble those of the other one may compete poorly with it, and natural selection will tend to eliminate intermediate individuals and favor evolution of enhanced differences between them. Inference of ecological character displacement has been controversial, but some cases are well supported.
Clines, which were mentioned in passing before, are geographic gradients in the frequencies of genotypes or phenotypes or in phenotypic means. They may evolve in an initially homogeneous population that experiences an environmental gradient or even sharp differences in natural selection in different parts of its range. Clines may also form where previously isolated contrasting populations come into secondary contact and hybridize. Populations separated by a cline may eventually merge, or selection against hybrids owing to ecological or genetic differences may cause the populations to retain their differences. Similar clines may occur in multiple species and are recognized as bio-geographic rules. For example, in endothermic vertebrates, body size tends to increase (Bergmann’s rule), extremities tend to be shorter (Allen’s rule), and coloration tends to be paler (Gloger’s rule) toward the poles. Similarly, the number of vertebrae increases toward the poles in fishes (Jordan’s rule). Clinal variation among conspecific populations is a conspicuous and ubiquitous manifestation of microevolution.
DNA sequence variation is most strongly influenced by genetic drift, and its evolution should be largely independent of phenotypic microevolution. It should carry a strong signal of evolutionary relationships or phylogeny. Heritable phenotypic differences among populations, however, may reflect both phylogeny and local natural selection (adaptation). Although gene flow complicates the analysis, it is possible to reconstruct the phylogeny of conspecific populations, which is called phylogeography, using DNA sequence data to distinguish the effects of phylogeny and adaptation. It is possible that different genes will indicate different relationships among populations of a species because one gene may have been present when one population split into two, and another may have entered one population by gene flow long after the two populations split. Variation of allozymes and restriction fragment length polymorphisms (RFLP) in mitochondrial DNA (mtDNA) were used in phylogeography until the late 1980s, when development of the polymerase chain reaction (PCR) enabled widespread use of DNA sequences from the nuclear genome.
Phylogeographic analysis of sockeye salmon, On-corhynchus nerka, illustrates the value of this approach. Sea-run sockeye salmon are widespread throughout the north Pacific, but rare, isolated lake-resident populations of O. nerka, called kokanee, also exist. Anadromous sockeye are about twice the size of kokanee but spend only 1 to 3 years at sea before spawning in fresh water. The smaller kokanee remain in fresh water and spawn after 2 to 7 years. The phenotypic similarity of isolated kokanee populations throughout their range suggested that they evolved in one place and spread from there, but their wide distribution suggested that they evolved repeatedly from local anadromous sockeye. Analyses of DNA sequence variation showed decisively that kokanee populations are genetically similar to adjacent migratory sockeye populations. They must have evolved many separate times from sockeye, and the similarity of isolated kokanee populations is a result of repeated (convergent) microevolution of similar adaptations to similar habitats.
Development of large DNA-sequence databases, including whole, sequenced genomes, laboratory methods to inexpensively obtain DNA data, and statistical methods to analyze them have created exciting opportunities to study microevolutionary processes, phylogeography, and the genetics of microevolution. Molecular markers, including RFLP, single-nucleotide polymorphisms (SNPs), and short tandem repeats (microsatellites) can be used to study the number of genes, the relative importance of different genes, their location in the genome, and even which parts of genes underlie phenotypic evolution. In the threespine stickleback, for example, at least four genes on separate chromosomes control variation in the number of lateral armor plates (see figure 1), but the Ectodys-plasin (Eda) gene accounts for more than 75% of the variation in plate number. Eda affects development of other vertebrate traits (e.g., teeth, fish scales, mammal hair, and sweat glands), so a change in the structure of the EDA protein would probably have adverse effects on other stickleback traits. As expected, the protein-coding region of Eda does not differ consistently between alleles for high-and low-plate-number phenotypes, implicating altered expression of Eda in evolution of plate number. Insertion of an Eda gene from a mouse into the genome of a low-plated stickleback caused an increase in the number of plates, confirming Eda’s role in plate number evolution. Remarkably, an ancestral Eda allele for low-armored phenotypes originated by mutation more than 10 million years ago, and alleles that have evolved from it have spread across the Pacific, Arctic, and Atlantic oceans from a single point of origin, providing the genetic variation on which natural selection acted to cause evolution of low plate number throughout this huge range. However, different genes with smaller effects may cause plate number differences in neighboring populations.
Genomics has also contributed an entirely new method to study natural selection. If a novel allele enters a population by mutation or gene flow, it may initially be represented by one copy. If this allele experiences strong positive selection, it will quickly be fixed (i.e., reach 100%), and the DNA sequence surrounding it will also be fixed before recombination with DNA surrounding other alleles for the same gene can occur. Until this region accumulates mutations, heterozygosity will tend to be depressed around the positively selected allele. Such depressed heterozygosity is the signature of a recent rapid selective sweep. Studies of the human and other genomes have detected numerous selective sweeps, and extended DNA sequences surrounding genes already believed to have experienced selective sweeps tend to be surrounded by regions of depressed heterozygosity. For example, an allele for the Let gene that confers the ability in adult humans to digest milk sugar (lactose) is surrounded by a long stretch of DNA with reduced heterozygosity, suggesting a recent selective sweep when humans became consumers of raw milk.
A wide range of methods continues to be used to develop new insights into the mechanisms for evolution within species. Since the mid-1960s, application of molecular biology to microevolutionary problems has revolutionized the field. Molecular methods and genomic data will continue to be used in phylogeography, and they will increasingly be applied to research on the genetics and development of phenotypes that vary within and among conspecific populations. As the power of molecular methods and number of sequenced genomes increase, new opportunities to use geographic variation to investigate basic problems in genetics and development will appear and create additional tools for research in microevolution. Increasingly, microevolutionary theory and molecular biology will be combined to address problems related to human health, agriculture, and natural resource management. See also chapters I.9 and I.13–I.16.
Avise, John. 2000. Phylogeography: The History and Formation of Species. Cambridge, MA: Harvard University Press. A review of methods to use molecular traits to infer microevolutionary history of conspecific populations.
Bell, Michael A., and Susan A. Foster, eds. 1994. The Evolutionary Biology of the Threespine Stickleback. Oxford: Oxford University Press. A review of microevolutionary phenomena and mechanisms in this microevolutionary model species.
Colosimo, P. R., K. E. Housemann, S. Balabhadra, G. Villarreal, Jr., M. Dickson, J. Grimwood, J. Schmutz, R. Myers, D. Schlüter, and D. M. Kingsley. 2005. Widespread parallel evolution in sticklebacks by repeated fixation of Ectodysplasin alleles. Science 307: 1928–1933. An extraordinary molecular analysis demonstrating that a single mutation created the variation on which natural selection has acted to produce a global pattern of geographic variation in a conspicuous skeletal trait.
Endler, John A. 1986. Natural Selection in the Wild. Monographs in Population Biology 21. Princeton, NJ: Princeton University Press. An encyclopediac review of natural selection.
Endler, John A. 1995. Multiple-trait coevolution and environmental gradients in guppies. Trends in Ecology and Evolution 10: 22–29. An overview of the microevolution of multiple phenotypes in guppies.
Ewald, Paul W. 1994. Evolution of Infectious Disease. New York: Oxford University Press. A ground-breaking book on pathogen microevolution.
Grant, Peter R. 1986. Ecology and Evolution of Darwin’s Finches. Princeton, NJ: Princeton University Press. An account of microevolution and speciation placed within exceptionally thorough biological and environmental contexts.
Hendry, A. P., and M. T. Kinnison. 2001. Microevolution: Rate, Pattern, Process. Dordrecht: Kluwer Academic Publishers. (Also published as Genetica, vol. 112–113.) An excellent collection of review papers concerning rates, patterns, mechanisms, and ecological contexts for microevolution in contemporary populations.
Lederberg, Joshua, and Esther M. Lederberg. 1952. Replica plating and indirect selection of bacterial mutants. Journal of Bacteriology 63: 399–406. A classic study showing that mutations are random with respect to adaptation.
Majerus, Michael E. N. 1998. Melanism: Evolution in Action. Oxford: Oxford University Press. An excellent review of the classic case of contemporary microevolution, it exposed some methodological flaws, which, however, do not invalidate this case.
Palumbi, Stephen R. 2001. The Evolution Explosion. New York: W. W. Norton & Company. An excellent popular review of microevolution in response to human technology.
Wilson, Edward O., and William H. Bossert. 1971. A Primer of Population Biology. Sunderland, MA: Sinauer Associates. A concise quantitative treatment of basic population genetics.