The Princeton Guide to Evolution

VI.6

Gene Flow, Hybridization, and Speciation

C. Alex Buerkle

OUTLINE

1. Gene flow leads to species cohesion

2. Gene flow and the origin of species

3. Hybridization: A common phenomenon

4. Evolutionary outcomes of hybridization

5. How to think about species in the context of gene flow and hybridization

The extent to which genetic material moves between divergent populations and species is a critical determinant of their evolutionary independence. High gene flow causes homogenization of populations and leads to their evolutionary cohesion, whereas low gene flow is more permissive of evolutionary divergence and independence. When divergent lineages mate or hybridize, there is the potential for genetic material to move between them. Gene flow through hybrids can erode evolved differences and can lead to stable hybrid zones, and to evolutionary novelty, including new species. The genetic, ecological, and evolutionary processes that affect the success of gene flow and hybrids are those that determine the conditions for the maintenance of reproductive isolation between species and for the origin of novel species.

GLOSSARY

Allele. An alternative nucleotide at any site in the genome, whether the locus is within a gene or elsewhere in the genome.

Gene. A region of an organism’s genome that codes for the chemical precursor of a protein.

Genome. The entirety of the genetic or hereditary material that is passed between parents and offspring, including chromosomes (in organisms that have them) and all nucleic acids that are inherited.

Hybridization. The process by which progeny are produced from matings between genetically divergent parents, including individuals from different species.

Hybrid Speciation. The formation of an independent evolutionary lineage through hybridization, either through the union of some combination of unreduced gametes, leading to an increase in the number of copies of chromosomes (polyploidy) in the hybrids, or through standard gametes and homoploid progeny.

Hybrid Zone. A geographic region in which two species come into contact and hybridize.

Introgression. Gene flow between species or lineages that moves foreign alleles into the native genetic background.

Reproductive Isolation. The lowered probability that members of different populations will mate with one another when they co-occur, relative to a randomly mating group of all individuals in the populations. Likewise, the lower fitness of progeny from crosses between populations, relative to randomly mated individuals.

1. GENE FLOW LEADS TO SPECIES COHESION

The dispersal of juvenile individuals from their parents potentially leads to the exchange of genetic material among populations, whether this process involves a seed that is transported by an animal or wind, or a juvenile animal that opportunistically settles in a favorable site. The movement of genetic variants, or alleles, among geographic locations and populations is referred to as gene flow. (Somewhat confusingly, when biologists refer to the movement of genetic variants [gene flow] among populations, they are referring to the movement of any genomic material, not just protein-coding regions. Given that genes constitute a very small fraction [about 1 percent] of the genomes of many eukaryotes, this is an important point.) The net effect of gene flow is to make populations genetically more similar to one another than they would be in the absence of this exchange, because novel alleles that arise by mutation in an individual at one location are passed on to potentially dispersing progeny, rather than being retained only at that location. Gene flow homogenizes differences among populations that arise due to chance fluctuations in allele frequencies (genetic drift) or to the action of natural selection in different populations. In other words, gene flow opposes differences that arise due to any evolutionary processes, by homogenizing allele frequencies among populations. Consequently, populations that are connected by gene flow evolve collectively to some extent. Additionally, gene flow will export adaptive mutations that arise locally and disperse them more broadly across the geographic range of a species. In contrast, populations that do not exchange genetic material, or do so only rarely, have the capacity to evolve independently along different trajectories.

The capacity for dispersal and gene flow varies widely among taxa. For example, the offspring of some plants with heavy fruits only fall passively to the ground beneath their seed-producing parent; in contrast, the seeds of other plants are carried by wind for thousands of kilometers between islands in the Pacific Ocean. Whereas marine turtles and salmonid fishes can move great distances across the globe in their lifetime, in many cases individuals do not disperse very far from their place of birth to breed but instead return to the same beaches and rivers where their parents reproduced. Thus, the movement of individuals during seasonal migratory periods or other life stages is not the same as dispersal for breeding. Furthermore, the capacity for dispersal is not equal to the capacity for gene flow, because gene flow requires not only dispersal but also successful establishment and reproduction in the new location. That is, for gene flow to occur, not only would a seed need to be dispersed to a new location but, additionally, the seed would need to germinate, mature to reproduction, and successfully leave progeny of its own in the new location.

One might think that gene flow would necessarily be closely tied to sexual reproduction, and in many cases it is the union of gametes in sexual reproduction that introduces immigrant alleles into a new population. However, gene flow need not involve sexual reproduction. For example, in the case of asexual organisms, a population often consists of diverse individuals that reproduce clonally. The immigration of clonal lineages could introduce novel alleles into a population and otherwise alter the frequency of alleles, just as with gene flow involving sexual reproduction. For example, despite their showy yellow flowers, most common dandelions (Taraxacum officinale) make seeds without the need for sexual reproduction. Dispersal of seeds leads to gene flow among populations and shapes the evolution of dandelion populations, even in the absence of sexual reproduction. Likewise, in the case of bacteria, entirely new genes or allelic variants of an existing gene may be transferred between distantly related lineages by horizontal gene transfer without reproduction. Recent studies have found that horizontal gene transfer is not restricted to bacteria but is evident among eukaryotes, including transfers of genes and divergent alleles between distantly related plants (e.g., unrelated species of grasses, or between a flowering plant and a fern).

For species with small geographic ranges relative to the scale of their dispersal, gene flow is expected to thoroughly homogenize allele frequencies among all populations. Other species have much larger geographic ranges that dwarf the scale of typical single-generation dispersal of individuals. Large geographic ranges are likely to span a broad set of biotic and abiotic conditions that affect the performance of organisms, to encompass potential physical barriers to dispersal, and to include both suitable and unsuitable habitats. For example, consider small invertebrates that are restricted to tidal pools along the Pacific coast of North America, or butterflies that breed in mountain meadows that contain particular host plants on the high peaks of the Rocky Mountains. Each of these organisms possesses particular requirements for survival and successful reproduction, and dispersal of individuals and gene flow among populations requires traversing inhospitable sites.

Given that geographic ranges can be large relative to the scale of dispersal, there is the potential for relatively isolated populations of a species to diverge evolutionarily from one another. In the absence of homogenizing gene flow, mutations that confer fitness advantages in a local habitat will increase in frequency, and local adaptation of populations may occur. For example, novel mutations may lead herbivorous insects to shift to eating a novel host plant in particular locations. The evolutionary fate of the novel mutations will depend on a myriad of processes that affect their frequency, including the potential for gene flow from immigrants to eliminate local, novel alleles or to spread novel alleles among populations and lead to greater evolutionary cohesion among populations.

2. GENE FLOW AND THE ORIGIN OF SPECIES

It follows that if the evolutionary cohesion of populations is enhanced by gene flow, its diminishment allows for diversification of populations and lineages, and ultimately speciation. The origin of species is tied very closely to patterns of gene flow and reproductive isolation between ancestral and derived lineages. Although there are many criteria by which to recognize species (often referred to as species “concepts”), all of them share the idea that new species are formed when lineages evolve traits that reduce gene flow with other populations or lineages and thus become reproductively isolated and evolutionarily independent (see chapter VI.1).

Speciation can usefully be thought of as the accumulation of traits in diverging lineages that contribute to the diminishment, or complete cessation, of gene flow between them. Many evolutionary biologists study these traits because they promote reproductive isolation and increase a lineage’s potential to evolve independently and to harbor novelty relative to other species and lineages. The large diversity of biological features and mechanisms that serve to isolate lineages can usefully be divided into those that function prior to the formation of zygotes (prezygotic isolating mechanisms; a zygote is a fertilized egg) or after zygotes are formed (postzygotic isolating mechanisms; see chapter VI.1). These can further be subdivided into (1) prezygotic mechanisms that arise as features of the ecology, behavior, and reproductive biology of the organisms and that reduce the frequency of matings or fertilizations and (2) postzygotic mechanisms associated with the viability or fertility of the hybrid progeny that result from crosses between the lineages. It is important to recognize that in any given pair of evolutionarily independent lineages, a diversity of organismal traits and features of their environment may each contribute to the reproductive isolation of the lineages. For example, two closely related plant species could have a low probability of gene flow because of differences in pollinators, flowering time, the habitats they occupy, greater fertilization success of conspecific pollen, and some inviability and infertility of hybrid progeny.

3. HYBRIDIZATION: A COMMON PHENOMENON

Many species lack complete reproductive isolation from other evolutionarily independent lineages and, instead, hybridize. Hybridization is a common phenomenon across the diversity of life, with hybrids between species known to occur in most familiar groups of organisms. The commonness of hybrids might seem to contradict the concept of species as evolutionary independent lineages. However, hybridization does not necessarily lead to a complete loss of independence; hybrids might be restricted geographically to a hybrid zone along an ecological gradient, or the hybrids might fail to contribute to gene flow between lineages because they are largely inviable or infertile. If one takes the view that complete reproductive isolation and evolutionary independence lie at one end of a gene flow continuum, then progeny from crosses between any two divergent lineages can be considered a type of hybrid (e.g., hybrid corn or maize varieties dominate North American agricultural production and are the result of crosses between divergent lineages of the same domesticated species). Crosses between divergent lineages are likely to occur, or are at least possible, in most organisms, which leads to the conclusion that hybridization is pervasive. Furthermore, isolation and evolutionary independence are not all-or-nothing characteristics but are quantitative attributes that vary by degrees between lineages. Isolation may also vary across the genomes of a pair of species, because exchange of alleles is more effectively counteracted in some regions of the genome than others (see chapters VI.1 and VI.9). Consequently, hybridizing lineages might possess recognizable trait differences that are maintained by divergent natural selection while experiencing substantial gene flow in portions of the genome that do not underlie important trait differences.

There are many examples of naturally occurring hybrids in various taxonomic groups. Some of these have been studied extensively because hybrids play a prominent or notable role in the evolution of these groups. For example, species of sunflowers (Helianthus) commonly hybridize, as do ragworts and groundsels (Senecio), and researchers have studied the genetics and ecology associated with their hybridization (see further reading). Likewise, tree species in several genera have a high propensity for hybridization, including oaks (Quercus), cottonwoods and aspens (Populus), and spruce (Picea), and have been studied extensively. The commonness of hybridization is by no means restricted to plants. For example, researchers have examined the role of hybridization in the invasion biology of fish (sculpins, suckers, sticklebacks) and salamanders (Ambystoma). And it has been estimated that approximately 10 percent of primate species hybridize in the wild. Hybrids between species belonging to the same genus, and even to different genera, are common in birds. The evolutionary and ecological significance of hybrids has been studied in butterflies, ants, corals, mussels, and many other animal groups. Laboratory techniques that allow molecular assays of genetic variation have been instrumental in detecting hybrids and confidently distinguishing them from variants within parental species. As biologists discover and study hybrids in more taxonomic groups, there is a growing appreciation that we can learn about the nature of species boundaries and reproductive isolation by studying instances in which isolation is incomplete and hybrids are formed.

4. EVOLUTIONARY OUTCOMES OF HYBRIDIZATION

The fitness of hybrid progeny can exceed (as in hybrid corn or maize), be equivalent to, or be lower than that of parental forms (e.g., sterile hybrid progeny). This variation can result from intrinsic features of the organism, including genetic and developmental determinants of viability and fertility. Likewise, variation in fitness of hybrids can be shaped by the ecological context in which hybridization occurs. Intrinsic and extrinsic factors interact to affect the outcomes and dynamics of hybridization. For example, the relative abundance of hybrids and parental species, spatial and temporal variation in abiotic and biotic determinants of fitness, and genetic variation among hybrid phenotypes all contribute to determining the fate of hybrids. Biologists can learn a great deal about the ecology and evolution of species by studying the role of different factors in determining the outcomes of hybridization. Three categories of outcomes may be recognized as heuristic points of reference among the complex and dynamic states that occur in nature.

First, it is likely that the most common outcome of hybridization between divergent lineages is their homogenization and the loss of any evolutionary genetic novelty (sets of mutations and trait differences) that had accumulated and been restricted previously to one of the diverging lineages. This must happen frequently when divergent lineages come into contact as a result of geographic range shifts. Given the fluctuations of climate over geological and evolutionary time scales (e.g., glacial advances), shifts in geographic ranges have been common, and divergence that might have accumulated in geographic isolation will have been erased on secondary contact, unless protected by traits that contributed to reproductive isolation. Likewise, in recent history, humans have disturbed natural habitats and caused range shifts that have brought previously isolated lineages into geographic contact. If hybrids have fitness equal to that of their parental lineages, divergent alleles will flow between them, and they will cease to evolve independently. Complete loss of some divergent lineages is expected to be common, but difficult to observe directly. Well-studied examples of loss of divergence due to hybridization include fish species (e.g., sticklebacks, suckers, sculpins, trout, ciscoes, and cichlids) and a large number of flowering plants (including hybridization between crops and their wild relatives). For example, rainbow trout that were transplanted and introduced by humans hybridize with and threaten the persistence of cutthroat trout in many drainages in the western United States.

A second outcome of hybridization is the formation of hybrid zones, areas in which a population of hybrids persists adjacent to the parent species. Hybrid zones can occur at geographic range margins—where two species meet—or can be less spatially structured, occurring as a patchwork within the ranges of the parental species (mosaic hybrid zones). The dynamics of hybrid zones are affected by the rate of input of parental alleles through ongoing interspecific hybridization and inputs from crosses between hybrids themselves. Likewise, the composition and persistence of hybrid zones is affected by the dispersal of adults and gametes, spatial and temporal variation in ecological conditions, and the contribution of extrinsic and intrinsic factors to fitness variation. Consequently, hybrid zones are inherently dynamic: their composition can change over time, they can move across the landscape, or they can vanish when species cease hybridizing. Some of the best-studied hybridizing species illustrate some of this variation among hybrid zones. Two forms of house mice (Mus musculus and M. domesticus) come into contact from the east and west in Europe and form a long hybrid zone that stretches 2400 km from Denmark to the South Caucasus. Transects through the hybrid zone have indicated that the area of hybridization extends for more than 50 km, with many agricultural barns occupied by populations of hybrid mice. This extensive, highly geographically structured hybrid zone of mice (and many other similar examples) can be contrasted with many hybrid populations in which hybrids co-occur with both parental species in a single, relatively small area, or different forms occur in a more complex geographic mosaic (e.g., sunflowers, sticklebacks within a single lake, and riparian cottonwoods). In many cases, there are multiple hybrid zones between the same pair of species, so that the outcomes and dynamics of hybridization may be compared. Despite decades of study of hybrid zones, we are only beginning to understand key aspects of the nature of genetic exchange that occurs within them.

By virtue of the incomplete and variable reproductive isolation that can be found in hybrid zones, they have the potential to offer us key insights into the traits and genetics that underlie reproductive isolation between species, which is otherwise difficult to study between lineages that are completely reproductively isolated. Hybrids between divergent lineages will possess novel combinations of alleles (and in some cases novel ploidy levels). These genotype combinations in hybrids that are rare or missing in the parental lineages can lead to inviability or infertility and contribute to reproductive isolation (e.g., Muller-Dobzhansky incompatibilities; see chapter VI.8) or lead to novel phenotypes in hybrids that have high fitness in certain contexts.

Finally, hybrids can become reproductively isolated from both their parental species and form new species. Hybrid speciation comes in two forms. In homoploid hybrid speciation, the genomes of the parental species merge without an increase in the number of chromosomes. This means that the genome of the hybrid species is a mosaic of genetic material from each of the parental species. In allopolyploid speciation, which involves a doubling (or other multiple) of chromosome number (polyploidy), the genomes of both parental species are retained in the novel hybrid species, at least early in its evolutionary history. Over time, the original genome multiplication can become fractionated and evolve toward a diploid number of copies of each genomic region. Both homoploid and polyploid hybrid speciation are examples of speciation with gene flow, since they begin with a hybridization event, but once it is formed, the derived hybrid must itself avoid homogenizing gene flow with the parental lineages if it is to persist and become evolutionarily independent. Thus hybrid speciation is made more likely if the hybrid lineage is somewhat spatially or ecologically isolated from the parents. For example, if the genotypes of hybrids predispose them to breed at a different time than a parental species, or to occupy a habitat in which the parental species cannot survive, this increases the chances of their success as an independent species. This type of ecological shift and resulting isolation has occurred in homoploid hybrid species of sunflowers (Helianthus): each of the three known hybrid species occupies an extreme habitat relative to the parental species. Likewise, among butterflies, hybrid species of Lycaeides in the Sierra Nevada of California occur at higher altitude and utilize a novel host plant relative to the ancestral lineages. Homoploid hybrid speciation may be more common than once thought and is currently the subject of intense investigation in a variety of taxonomic groups, including Heliconius butterflies, sculpins, house sparrows, and pines.

In general, it is important to recognize that hybridization often results in a genetically and phenotypically diverse hybrid progeny. Given this variability and known ecological influences on the outcomes of hybridization, hybridization is likely to lead to a diversity of genetic and evolutionary outcomes. For example, hybridization is common between two annual sunflower species (Helianthus annuus and petiolaris) wherever their geographic ranges meet. As noted earlier, in at least three instances this hybridization has led to novel, homoploid species that differ from the parental species. More commonly, their hybridization simply leads to a mixed population of various hybrid sunflowers and the parental species. Similarly, hybridization is common among various members of the genus Senecio (like Helianthus, also plants in the family Asteraceae). Hybridization in this genus has also led to new species, including novel species with the same chromosome number or with double the number of chromosomes of the parental species, but also has resulted in hybrid zones. Understanding the causes of this variation in outcomes of hybridization will continue to be the focus of considerable research and will lead to a better understanding of how species are formed.

5. HOW TO THINK ABOUT SPECIES IN THE CONTEXT OF GENE FLOW AND HYBRIDIZATION

The processes that lead to speciation have been a subject of long-standing interest in evolutionary biology, and recent research has advanced knowledge of the complex interactions of gene flow, hybridization, and speciation. Recognition that hybridizing species can nevertheless evolve independently as a result of traits that contribute to their isolation has clarified the expectation that different portions of their genomes will differ in their degree of divergence and amount of genetic exchange as a result of variation across the genome in mutation rates, effective population size, natural selection, and recombination rates. This means that a pair of hybridizing lineages will appear to exhibit different levels of isolation and divergence depending on which portion of the genome is examined (see chapters VI.1 and VI.9). This is a much more dynamic and realistic view of species, in contrast with the previously held notion that reproductive isolation between species is a genome-wide property. From a practical standpoint, variability in degree of isolation across the genome makes the task of recognizing and naming species more difficult. This challenge is not unique to evolutionary biology but is also encountered in other domains of biology where levels of differentiation may vary depending on which components are measured, such as the transition between juvenile and adult, or the boundary between cell types of different tissues.

Ours is an era of evolutionary biology in which researchers are seeking substantial advances in our understanding of the origin of species and the biological means by which novel lineages escape gene flow from their ancestors. Many researchers are studying traits and genomes in recently diverged lineages and their hybrids, in both natural and experimental settings. It is unlikely that the fundamental concept of species as isolated, independent lineages will be changed by this research, but it is likely that many mechanisms and processes, and ultimately generalities, associated with speciation will be revealed and that these will provide a more complete understanding of how novel species arise.