Species and Speciation
Richard G. Harrison
OUTLINE
1. Species concepts and definitions
2. Speciation as the evolution of intrinsic barriers to gene exchange
3. Classifying barriers to gene exchange
4. Studying speciation
Species are the fundamental units of biodiversity, but the definition of a species remains a subject of debate within evolutionary biology. One resolution of this debate views alternative species definitions as different stages in the process of speciation, in which conspecific populations diverge, accumulate intrinsic barriers to gene exchange, and ultimately become exclusive or reciprocally monophyletic groups. Most studies of speciation have focused on the evolution of reproductive isolation or intrinsic barriers to gene exchange. Such barriers may result from a variety of trait differences, some of which are simply a by-product of divergence in allopatry. Barriers may prevent individuals from meeting or mating, they may compromise gamete interactions, or they may reduce the viability and/or fertility of hybrid offspring.
GLOSSARY
Allopatric. Occupying different geographic regions; geographically isolated.
Gene Flow. Movement or incorporation of alleles from one population into one or more different populations.
Monophyletic. A group of organisms (taxa) that all share a most recent common ancestor not shared by any other organisms (taxa).
Sympatric. Occupying the same geographic area, with the opportunity for gene flow.
Zygote. The (usually diploid) cell formed by the union of two (usually haploid) gametes (e.g., sperm and egg).
The diversity of life comprises relatively discrete entities we call species. Like cells or individual organisms, species are widely viewed as fundamental units of biological organization. However, the defining qualities of species, the nature of the boundary between species, and even the reality of species remain matters of dispute. It is ironic that the concept or definition of species, so central to the studies of evolution, ecology, and conservation biology, has engendered so much confusion and debate. In contrast, the rules for naming species (for animals embodied in the International Code of Zoological Nomenclature) are very clearly described and widely accepted.
The last two decades have witnessed gradual acceptance of the view that there is not one “right” species concept or definition. One approach to resolving past disagreements is to recognize the difference between a species concept and a species definition. The former is what is meant by the word species; the latter involves defining the criteria used to delimit species. K. de Queiroz has suggested that most would agree that the defining property for a species is “a separately evolving metapopulation lineage,” in which a lineage is an ancestor-descendant series. Because species are defined over time as well as space, and because speciation is a process not an event, differences among species definitions may then reflect different landmarks along the path from conspecific populations to separate species.
1. SPECIES CONCEPTS AND DEFINITIONS
The evolutionary biology literature presents a bewildering array of different species concepts or definitions (as many as 24 have been identified). Some of the concepts are subtle variations on basic themes, but at least seven major concepts can be differentiated along a number of axes, including whether they are retrospective (species as products of history) or prospective (species as lineages extending into the future), whether they are relational or nonrelational, whether they are based on pattern or process, and the extent to which they are operational, including whether they can be applied to allopatric (geographically isolated) populations and to asexual lineages. The major concepts are summarized in table 1.
Table 1. Major species concepts or definitions
|
Biological Species Concept Recognition Species Concept Isolation Species Concept |
Character-Based Phylogenetic Species Concept Genealogical Species Concept Evolutionary Species Concept Genotypic Cluster Definition |
Character-Based Concepts
Perhaps the simplest and most intuitive concepts of species are those that are “character based.” Accordingly, the fundamental criterion for defining species is “diagnosability”: species are groups of organisms/populations that are diagnosably distinct, that is, groups exhibit fixed differences in states for at least one (but often more than one) character. Such a definition is unambiguous and easy to apply; it relies only on the ability to define and compare sets of characters (morphological, DNA sequence, etc.) in groups of organisms. Of course, care must be taken to ensure that phenotypically distinct groups do not simply represent differences between males and females or differences among life history stages.
Although easy to apply, character-based definitions view speciation as equivalent to divergence. If isolated populations diverged as a result of natural selection that led to local adaptation, or if allele frequencies drifted to fixation for different alleles at a gene locus, these populations would be viewed as distinct species. Strict application of such character-based concepts might result in a tremendous proliferation of species and the elevation of many current subspecies or races to species status.
Darwin clearly recognized that evolution (including the origin of species) is a process and that sampling diversity at any one point in time should reveal populations at all stages in the process. Indeed, Darwin saw the continuum of differences between populations as direct evidence for the evolutionary process. He wrote:
Certainly no clear line of demarcation has as yet been drawn between species and sub-species. …; or again between sub-species and well-marked varieties, or between lesser varieties and individual differences. These differences blend into each other in an insensible series; and a series impresses the mind with the idea of an actual passage.
Thus, species boundaries may be fuzzy and develop gradually, with few rules about how much difference needs to accumulate before lineages should be recognized as different species. However, for allopatric populations, reliance on amount (or quality) of difference seems to be the only possibility. Darwin commented that in cases “in which intermediate links [between populations] have not been found … naturalists are compelled to come to a determination by the amount of difference between them.” When differentiated populations co-occur, a direct test of character-based definitions is whether the populations remain distinct in sympatry. The approach has been formalized in an updated version of Darwin’s views, in which species are recognized when there are few/no intermediates in a zone of overlap. This genotypic cluster definition is also character based, but by examining character differences in sympatry, it attempts to assess whether the existing differences persist.
The Biological Species Concept (or Isolation Concept)
For many years, the prevailing species concept, promoted by two famous twentieth-century evolutionary biologists, Ernst Mayr and Theodosius Dobzhansky, has been the biological species concept (BSC). In contrast with character-based definitions, the BSC defines species in terms of gene flow or gene exchange between populations. Species are characterized by interbreeding within a population and reproductive isolation among groups of populations. Reproductive isolation is a consequence of intrinsic barriers to gene exchange, barriers due to the properties of the organisms themselves and not simply geographic separation.
The great strength of the BSC is that it is clearly based on an appreciation for the evolutionary process; it attempts to define species using a criterion (gene exchange) that has important implications for the future of the lineages and is not simply a product of history. The presence of intrinsic barriers to gene exchange suggests that two species will persist, although ecological differentiation may also be a prerequisite for coexistence.
The BSC has a number of obvious limitations, many of which are shared by other definitions of species. First, it is difficult to estimate amount of gene flow in natural populations; therefore, absence of gene flow is often inferred from patterns of differentiation for genotypic or phenotypic markers. Second, the BSC cannot be applied to allopatric populations; if it is, amounts of gene flow cannot be estimated. Mayr’s definition introduces the notion of “actually or potentially interbreeding” populations; use of “potentially interbreeding” presumably implies that we should be able to infer whether populations would interbreed should they come into contact. A third limitation is that the BSC is relevant only for organisms that reproduce sexually; interbreeding clearly has no meaning for obligately asexual lineages. Finally, the BSC must confront the issue of the “fuzzy” species boundary. Patterns of gene exchange vary not only in space and time but also across the genome. In some regions of the genome there may be no gene exchange, whereas in other regions, hybridization results in the flow of alleles between the two populations/species. In the extreme, the BSC may need to be defined for individual genes or gene regions. The bright side is that observed patterns of differential gene flow (introgression) can be used to gain insight into the genetic architecture of speciation.
Phylogenetic or Genealogical Concepts
Descent with modification produces a nested hierarchy of traits and taxa. The structure of this nested hierarchy can be revealed by phylogenetic analysis, which documents the pattern of branching events (forward in time) or coalescent events (backward in time) that define extant individuals, populations, or species. A phylogenetic or genealogical perspective suggests that species should be considered to be monophyletic or exclusive groups. An exclusive group is one whose members are all more closely related to one another than to any individual outside the group. The genealogical species concept of Baum and Shaw argues that exclusivity is an important criterion for species status. It seems quite reasonable that all members of a species should be closely related, but because of ancestral polymorphism and incomplete lineage sorting and/or ongoing hybridization, individuals within a species may, in fact, be more closely related to individuals of other species. Relationships depend on which gene or gene region is sampled, and exclusivity may characterize some genome regions and not others. It is not clear what proportion of the genome must be “exclusive” before a group is considered a genealogical species. According to a strict definition, many entities now viewed as independent evolutionary lineages would be considered conspecific, because genealogical speciation requires very long periods of time.
Evolutionary Species Concept
For most of evolutionary history, the only data we have about species and speciation come from the fossil record. Thus, the data are purely phenotypic (usually morphological), and neither the nature nor the quantity of the data allow direct assessment of gene exchange or exclusivity. Fixed differences can be defined, and character-based species definitions apply.
A very different and more general view that has been applied to fossil data is the evolutionary species concept, originating with the paleontologist G. G. Simpson and updated and modified by others. This concept defines species as populations through time (which can be followed in the fossil record) that maintain their separate identity (in the presence of other lineages) and exhibit an independent evolutionary trajectory.
Cohesion and Recognition as the Basis for Defining Species
All the species concepts discussed thus far rely on comparison of two (or more) lineages—they are relational. Fixed differences, reproductive isolation, exclusivity, and separate identities are all patterns or characteristics that must be defined in terms of differences between individuals and populations. A number of evolutionary biologists have argued that species concepts should be nonrelational, defined in terms of what is shared, rather than what is different.
The recognition concept defines species as populations of sexual organisms that share a common fertilization system or specific mate recognition system. The emphasis is on defining an interbreeding unit, a “field for recombination,” a group of organisms held together by “genetic cohesion.” In many ways the recognition concept is a reaction to the mention of “isolation” and “isolating mechanisms” in versions of the BSC. In practice, defining cohesion mechanisms is virtually equivalent to defining reproductive isolation. In either case, a group of individuals has to be partitioned into two or more subgroups, each of which shares fertilization and mate recognition systems within the subgroup but differs in these respects from other subgroups.
Most species concepts focus on genetic cohesion or isolation. But if two species are to persist in sympatry, the competitive exclusion principle from ecology says that they must be ecologically distinct. Ecology has not played much of a role in the development of species concepts. One exception is the cohesion species concept, which defines species as “the most inclusive population of individuals having the potential for phenotypic cohesion,” which is mediated by both genetic and demographic exchangeability. The former is essentially equivalent to interbreeding or gene flow. However, demographic exchangeability emphasizes ecological interactions. Groups of organisms that are demographically exchangeable are ecological equivalents. This concept is of particular use in sympatric asexual lineages, in which even in the absence of any gene exchange, demographic exchangeability implies that lineages belong to the same species.
2. SPECIATION AS THE EVOLUTION OF INTRINSIC BARRIERS TO GENE EXCHANGE
In the simplest model of speciation, allopatric populations diverge in the absence of gene flow as a result of the fixation (due to natural selection or genetic drift) of new mutations and/or different ancestral alleles. It is useful to think of the process of divergence as the life history of a species. Fixation of alternative alleles or phenotypes in the two populations results in their being considered character-based phylogenetic species. Eventually, some of the fixed differences affect the ecology, behavior, physiology, or reproductive biology of the diverging lineages. As a by-product of this divergence, the populations accumulate differences that affect the probability of their interbreeding or the success of their progeny should an individual mate with a member of the “other” population. At this point, the populations become biological (or isolation) species. However, these populations are exclusive groups only at some (perhaps relatively few) regions of the genome. Over time, an increasing proportion of the genome diverges, and each of the diverging populations becomes an exclusive group across the genome. At this point, the populations are genealogical species.
The critical event in this life history is the evolution of reproductive isolation, the appearance of intrinsic barriers to gene exchange. This transition alters the outcome if or when secondary contact occurs between the diverging populations. Should the populations become sympatric, intrinsic barriers will prevent gene flow and the erosion of genetic differences. Similarly, in models of sympatric speciation, the evolution of reproductive barriers leads to the cessation of gene flow and enables further divergence of two subpopulations.
Evolutionary biologists who study speciation therefore focus on the nature of intrinsic barriers to gene exchange: what they are, when they act in the life cycle of the organism, to what extent they reduce gene flow, and when in the history of divergence they arose. Answers to these questions emerge from comparisons of closely related (recently diverged) species and “incipient” species.” Such comparisons are particularly informative when the diverging populations occur in sympatry, and important contributions have come from the study of hybrid zones where individuals from distinct lineages meet and mate, producing some offspring of mixed ancestry (see chapter VI.6).
3. CLASSIFYING BARRIERS TO GENE EXCHANGE
Many phenotypic differences between diverging lineages can result in barriers to gene exchange. The traditional approach to classifying such barriers is to organize them with respect to whether they act before mating and/or zygote formation or whether they are a consequence of the reduced fitness (viability, fertility) of hybrid offspring. Most classifications recognize three distinct sorts of barriers: (1) premating, (2) postmating but prezygotic, and (3) postzygotic.
Premating Barriers
Premating barriers are those that result from trait differences that prevent hybridization between distinct species. Barriers to gene exchange will exist if potential mates do not meet (temporal and habitat or ecogeographic isolation), if potential mates meet but do not mate (behavioral isolation), or if attempted copulation does not result in sperm transfer (mechanical isolation). Premating barriers to gene exchange have been studied in many different animal and plant systems, revealing a host of different mechanisms whereby gene exchange is limited or prevented.
Temporal isolation reflects seasonal or diurnal differences in the times at which adults are present or sexually active. Thus, flowering time differences in plants, and major life cycle differences in animals (e.g., different overwintering life stages or different rates of development in insects) lead to partial or complete seasonal isolation. In many marine invertebrates, mating (in the narrow sense) does not occur, and eggs and sperm are simply broadcast in the water column. Spawning times can be determined by lunar cycles, so that in some corals, eggs and sperm from closely related species are unlikely to encounter each other in the water column. Similar patterns are seen in some moths, in which sexual activity is limited to a relatively narrow window in the diurnal cycle and may be displaced from the corresponding window for a sympatric close relative.
Habitat or resource isolation results from the association of particular populations or lineages with specific habitats or resources. Observed associations can be the result of differential adaptation to habitats or differential preference for habitats. Many examples come from the insect–host plant literature, in which insect lineages have apparently diversified by adapting to new host plants, resulting in reproductive isolation between the derived forms. There are also numerous examples of plants with different habitat needs or requirements; some of the best studied involve adaptation to different soils (e.g., serpentine soils or soils contaminated by heavy metals). Geographic isolation can result from local adaptation if habitats or resources are geographically separate; this phenomenon has been termed “ecogeographic isolation.” This type of isolation is a special form of ecological or habitat isolation, because unlike most barriers, which are studied in sympatry (where they prevent individuals from meeting or mating), ecogeographic isolation is an intrinsic barrier that characterizes allopatric taxa. A related concept is that of “immigrant inviability,” which refers to the “reduced survival of immigrants on reaching foreign habitats that are ecologically divergent from their native habitat” (see chapter VI.4).
If ecological factors do not prevent individuals from meeting, then behavioral differences may well prevent them from mating. Many examples have been documented of species-specific communication systems that function in mate finding and mate recognition. These include visual communication (e.g., plumage coloration in birds, color patterns in fish, flashing patterns of “fireflies,” attraction of pollinators to flowers), acoustical communication (e.g., songs of [mostly male] birds, frogs, and insects and the corresponding preference functions of females), and chemical communication (e.g., sex pheromones in insects). Behavioral barriers have been well studied in a diversity of animal systems, and sexual selection is often invoked to explain patterns of divergence. A primary focus has been on sexual selection by female preference. Depending on the nature of female preferences, the outcome of sexual selection may be “arbitrary”—that is, it will not have any “adaptive value” or relationship to environment (e.g., in runaway sexual selection). Prezygotic barriers can then arise as a result of different outcomes of sexual selection in isolated populations. Differences in female preferences (and in corresponding male traits) will act as barriers to gene exchange should populations come into secondary contact (see chapter VI.5).
Finally, if mating is attempted, successful transfer of sperm may not occur. Genitalic mismatch (the “lock-and-key hypothesis”) is an oft-cited reason for this failure, but the role of mismatch as a barrier to gene flow is not entirely clear. One well-documented case of mechanical isolation involves snails that show a dimorphism for coiling: some individuals exhibit dextral coiling, and others, sinistral coiling. Mating within coiling types presents no problems, but in matings between types the genital openings fail to match up.
Postmating, Prezygotic Barriers
Mating (or spawning) and zygote formation are often separated in time, and therefore it is important to recognize postmating but prezygotic barriers. These include sperm or pollen competition, cryptic female choice, and gametic incompatibility. Sperm and pollen competition demonstrate that male-male competition can continue after mating: when females are multiply inseminated, production of offspring sired only by conspecific males may occur even when sperm are successfully transferred from both conspecific and heterospecific males, and when heterospecific pollen or sperm are known to be able to combine with eggs to form viable zygotes. This phenomenon, documented in insects, vertebrates, and plants, is known as conspecific sperm and pollen precedence. The outcome of competition may be mediated by the female, in which case it is referred to as cryptic female choice.
Postmating, prezygotic barriers can also result from reduced gamete compatibility. Sperm-egg (pollen-ovule) interactions are often mediated by specific proteins, and changes in either of the gamete recognition proteins may affect the rate and/or ultimate success of the interactions. In abalone, the sperm protein lysin interacts with an egg receptor, VERL. Both these proteins are rapidly evolving and subject to directional selection, and differences in lysin-VERL interactions in closely related species may be responsible for barriers to gene exchange.
Postzygotic Barriers
The reduced viability and fertility of “hybrids” (often meaning the F1 offspring of a cross between two distinct parental types, but also more broadly used to refer to any individual of mixed ancestry) have been documented in a wide variety of taxa. Traditionally, postzygotic isolation has referred to developmental defects in hybrids that lead to full or partial inviability and/or infertility. The origin of such barriers presents a challenge (noted by Darwin), because if isolation is due to heterozygote disadvantage at a single locus, then the origin of such a barrier requires passing through a less fit intermediate state (and therefore would be opposed by selection). An alternative model imagines that incompatibilities arise because of fixation of mutations at two different (but interacting) loci, with each mutation arising uniquely in one allopatric population. If the two gene products do not “work well” together, then secondary contact between the divergent populations will lead to less fit hybrids (because they will carry both mutations). This model (the Dobzhansky-Muller model) is now widely accepted, and this scenario has now been supported in a number of different systems (see chapter VI.8).
There is an extensive literature on the genetics of postzygotic barriers, particularly in Drosophila. Part of that literature has been motivated by Haldane’s rule, the observation that “when in the offspring of two different animal races one sex is absent, rare, or sterile, that sex is the heterozygous sex.” By “heterozygous” Haldane meant heterogametic, that is, having two different sex chromosomes (individuals that are XY or ZW). Haldane’s rule appears to apply in diverse animal taxa, including species in which males are the heterogametic sex (flies and mammals) and species in which females are the heterogametic sex (birds and Lepidoptera) (see chapter VI.8).
The postzygotic barriers discussed are “intrinsic” barriers, in the sense that they result from developmental problems, apparently independent of the environment. More recently, it has been recognized that postzygotic barriers can be “extrinsic.” In these cases, the reduced fitness of hybrids results because they “fall between the niches” (are less fit than either parental type in the environments in which the species live) or because they are less successful at obtaining mates. For example, two forms of lake sticklebacks (benthic and limnetic) are distinct in morphology and feeding ecology. Hybrids are intermediate in morphology and are less successful than either of the parental types in the habitats in which the parents feed (see chapter VI.4).
4. STUDYING SPECIATION
If speciation is defined as the origin of barriers to gene exchange between diverging lineages, then evolutionary biologists interested in revealing the details of the process must confront the following questions: (1) What are the barriers that are currently operating between species, and to what extent does each of these barriers currently reduce gene flow? (2) When did each of these barriers arise in the history of the diverging lineages, and therefore which of the barriers was initially responsible for a reduction in gene flow? (3) In what geographic context did the barriers arise (and in particular, did divergence occur in the face of some gene flow)? (4) What roles have natural selection, sexual selection, and genetic drift played in driving the divergence? (5) What is the genetic architecture of reproductive isolation? Genetic architecture refers to the number, effect (major or minor), and chromosomal distribution (clustered or distributed across the genome) of the genes that determine speciation phenotypes.
The first question has been examined in a wide variety of animal and plant systems, although comprehensive analyses of barriers and their effects are still relatively rare. A focus on the order in which barriers act (during the life history of the organism) emphasizes the importance of early-acting barriers (those that prevent potential mates from meeting, e.g., ecogeographic barriers, differences in habitat or timing of reproduction) and distinguishes between the absolute strength of the barrier and its proportional contribution to total isolation. Because each barrier can reduce gene flow that persists only after earlier-acting barriers have had their effect, later-acting barriers inevitably make a smaller proportional contribution. However, it is not appropriate to extrapolate from the current contribution of individual barriers to their role in the speciation process, because it is the order in which barriers arose during the history of the lineages, rather than their proportional contribution now, that is most relevant to understanding the speciation process.
Debate about the geographic context in which speciation occurs has been a central issue in the speciation literature for many years (see chapter VI.3) and, as with so many issues in evolutionary biology, can be traced to Darwin. The central debate has been whether sympatric speciation (speciation in the face of substantial gene flow) can occur. Theoretical models support the notion that sympatric speciation is possible, and a few model systems can be cited in evidence, but the consensus remains that sympatric speciation is relatively rare. However, there is also increasing evidence that selection can drive sympatric divergence in the face of some gene flow.
The relative importance of drift and selection in causing divergence that leads to speciation has also been a central theme in the speciation literature. Genetic drift can alter allele frequencies in small populations, and founder events may therefore be an important route to speciation. In certain situations (colonization of oceanic islands, fixation of different chromosome rearrangements), drift may well determine outcomes, but a wider role for genetic drift in speciation is not well supported.
In contrast, selection can play many possible roles in the speciation process, including the following: (1) Populations diverge in allopatry as a result of different selection pressures in two regions or habitats, and barriers arise as a simple by-product of the allopatric divergence. (2) In zones of secondary contact, hybridization results in less fit progeny, and selection favors those individuals that do not or cannot hybridize. (3) Disruptive ecological selection (adaptation to two local habitats or resources) within a single population leads directly to sympatric species. (4) Sexual selection follows different trajectories in allopatric populations, leading to divergence in mate recognition and/or fertilization systems. In many examples, probably two or more of these scenarios may apply.
Finally, because the genomes of sexually reproducing organisms are mosaics of different histories, the pattern of the mosaic can reveal the recent history of selection across the genome. Furthermore, when diverging populations are in contact, genome regions that include genes that contribute to reproductive barriers tend to remain distinct, whereas gene flow (introgression) at neutral loci may erase patterns of differentiation. Recent genome or transcriptome scans have provided insights into genomic patterns of differentiation, which in turn are revealing candidate gene regions worthy of further investigation. These regions are candidates for harboring “barrier genes”—genes that contribute to reproductive isolation.
FURTHER READING
Coyne, J. A., and H. A. Orr. 2004. Speciation. Sunderland, MA: Sinauer.
de Queiroz, K. 2007. Species concepts and species delimitation. Systematic Biology 56: 879–886.
Harrison, R. G. 1998. Linking evolutionary pattern and process. In D. J. Howard and S. H. Berlocher, eds., Endless Forms. New York: Oxford University Press 19–31.
Mayr, E. 1963. Animal Species and Evolution. Cambridge, MA: Harvard University Press.
Schluter, D. 2001. Ecology and the origin of species. Trends in Ecology & Evolution 16: 372–380.