What increases the likelihood of allopatric speciation taking place more quickly?

Publisher Summary

Allopatric speciation, also known as geographic speciation, is speciation that occurs when biological populations of the same species become isolated due to geographical changes such as mountain building or social changes such as emigration. In the neo-Darwinian era, the species was universally recognized as the fundamental taxonomic unit of evolution. One of the most widely accepted concepts of species is the biological species concept (BSC). According to the BSC, allopatrically formed species are postzygotically isolated, i.e., even when they secondarily come in contact and can interbreed, they are incapable of producing fertile hybrids. This chapter describes the various theories for verification of allopatric models of evolution, such as the basic allopatric model, the reinforcement model, and divergence-with-gene-flow model. However, evidence has shown that many well-established vertebrate and invertebrate species are reproductively not isolated. Also, from the neo-Darwinian perspective, formation of new species and higher taxa was seen as nothing more than an extension, or an unavoidable finalization, of the process of gradual accumulation of microevolutionary (mutational) changes in genes. Darwinian evolutionists distinguished between two main forms of speciation, allopatric speciation, taking place under conditions of spatial isolation of populations, and sympatric speciation, occurring within populations sharing the same habitat. In this light, the study illustrates peripatric evolution using models such as the founder effect model and the bottleneck model. Finally, it sheds light on the phenomenon of ecological separation.

A ALLOPATRIC SPECIATION

Allopatric speciation is widely believed to be the most likely mechanism of speciation in most animal taxa (Mayr, 1963, 1977). Several recent studies of cryptic species favor allopatric explanations of speciation (sticklebacks, Schluter and McPhail, 1992; mouse-eared bats, Arlettaz, 1995). Allopatric speciation involves geographical isolation of a population from other populations of the parental species, and acquisition of characters that promote or ensure reproductive isolation once sympatry is reestablished.

A plausible scenario for pipistrelle speciation may be as follows. Imagine that bats in the parent population echolocate at 45 kHz. A small population becomes isolated by, for example, mountain barriers. Perhaps a glaciation event pushed this isolated population into a refuge that ensured its isolation. Bats in the isolated population changed echolocation call frequency to 55 kHz, social call structure altered, but morphological conformity with the parent population was maintained. Perhaps echolocation call frequency changed to exploit a new insect resource encountered by the isolates. As conditions became warmer, barriers between the populations broke down; however, the changes that occurred during isolation caused the previous parent and isolated populations to remain reproductively isolated. Both nascent species then spread over a wide geographic range and avoided competition because they used different call frequencies. This scenario may be testable if the date of divergence of the species could be ascertained by application of a molecular clock. Conditions at the time of divergence could be explored, to determine if geographic separation of populations would have been facilitated by climatic conditions at that time.

Glossary

Allopatric speciation

The evolutionary development of new species in the presence of a geographical barrier which reduces gene flow and promotes genetic divergence.

A biogeographic region that includes the New World tropics, extending from southern Mexico through the Southern Cone of South America to Tierra del Fuego. Many different ecosystems are found here, including tropical rainforest.

This has two connotations. In a broad sense, it simply refers to the number of species of a particular taxonomic group living within a given area and is used synonymously with species richness. In a narrow sense, it refers to the number of species within a given area while simultaneously taking into account their relative abundances. In this article, species diversity is used in a broad sense and is used interchangeably with species richness.

The number of species of a particular taxonomic group living within a given area.

A forest that occurs below 1000 m in elevation, experiences high, relatively constant temperatures, and receives at least 200 mm of rainfall per year.

Hypothesis 2: Divergent Mating Factors

Allopatric speciation events can generate two related taxa, physically separated, which can become secondarily sympatric due to migration or changes in the extent of suitable habitat (Section 22.2.1). If during the period of separation substantial genetic drift occurs in the sequences of genes determining mating compatibility, or in other genes such that hybrid offspring are unlikely to be viable, then the species barrier will remain intact, and recombination between the two forms will not occur. As argued above, the most likely reason for the evolution of two forms of ovale malaria parasite is that at least two independent host transitions into ancestors of modern humans occurred, separated by a lengthy period of time. This scenario is of course analogous to allopatry, in that the two lineages would have been “physically” separated by occupying different hosts. Thus the lack of recombination between twenty-first century human-dwelling populations of P. ovale curtisi and P. ovale wallikeri may be due to substantial changes in key genes encoding molecules essential for mate recognition, fertilization, or meiosis. It is also possible that gross chromosomal rearrangements have occurred in one or both lineages, thus rendering meiotic chromosome pairing impossible, and hybrid zygotes unviable.

Happily, genomic sequencing of a 1977 isolate of P. ovale curtisi is well advanced (Section 22.1.4; Sutherland et al., 2010), and we have now prepared two patient isolates each of P. ovale curtisi and P. ovale wallikeri, collected in 2009–2010, for next-generation highly parallel sequencing on the Solexa platform at WTSI. When these data are compared with the capillary genome sequence of the 1977 P. ovale curtisi, it should be possible to address directly the question of whether the two ovale parasite species have accumulated either crucial mutations (particularly in mating, fertilization, or meiotic function genes), or chromosomal structural changes sufficient to prevent viable hybridization. However, these data can also answer some very important questions about genome-wide inter-species polymorphisms: which loci have diverged the most between P. ovale curtisi and P. ovale wallikeri? Are erythrocyte invasion molecules prominent among them? Finally, we will also be able to compare contemporary twenty-first century isolates of P. ovale curtisi with the capillary sequenced isolate collected in Nigeria over 30 years previously, and thus also gain some interesting new insights into intra-species polymorphism, in both time and space, at the genome level.

Neurocognitive Sympatric Speciation

Although allopatric speciation is theoretically possible and seems to have occurred, contrary to the conventional wisdom, most of the scientifically proven cases of speciation are those related to the reproductive isolation occurring in sympatry. Premating isolation in sympatry seems to be the most frequent of demonstrated factors in speciation in nature. In most of the described cases, the sympatric premating isolation is behavior dependent: incipient species avoid the interbreeding, while being capable to interbreed and produce fertile hybrid offspring, under natural conditions. They use specific (e.g., visual, olfactory, and auditory) cues to discriminate against breeding nonself-like individuals. This sexual avoidance behavior is determined by the activity of neural circuits involved in the discrimination of self-like from nonself-like individuals.

Thus premating reproductive isolation, as the first step in the process of speciation, is a new and exclusive property of eumetazoans, related to the evolution of the nervous system in the kingdom Animalia. Changes in mating behavior and mating sensory signals represent the most frequently documented mechanism of reproductive isolation in sympatry.

Neo-Darwinians have considered sympatric speciation to be unlikely or have been skeptical of whether it can ever occur. However, recently two theoretical models have been developed (Dieckmann and Doebeli, 1999; Kondrashov and Kondrashov, 1999) in order to eliminate the discrepancy between the neo-Darwinian theoretical prediction of the impossibility of sympatric speciation under conditions of gene flow, on the one hand, and the facts of occurrence of sympatric speciation, on the other. Both models are based on conditions that are not demonstrated to exist in nature. The models are criticized of requiring “unlikely conditions” (Drès and Mallet, 2002), but construction of models based on assumptions that are not likely to exist nature seems to heuristically be of little use.

Vast observational and experimental evidence suggests that neither of the neo-Darwinian requirements of speciation (physical separation prevention of gene flow or preliminary accumulation of genetic differences between populations) are necessary for reproductive isolation to take place. Populations of a species which are genetically similar (sharing a common gene pool) may create, in sympatry, a separate fertilization system and enter the process of speciation by rapidly evolving nongenetic changes in mate preferences or courtship behavior.

2.4 Hybridization

During allopatric or ecological speciation, divergent selection and gradual evolution of genic incompatibilities occur over many generations. However, the formation of new species can also occur instantaneously as is the case of hybridization. Hybridization is the successful mating between individuals of otherwise genetically isolated lineages. If the parental species are similar in genome content and sequence, the hybridization can occur through recombination between homologous chromosomes resulting in progenies with the same chromosome number as their parents (homoploid hybridization). Alternatively, hybridization can also result in polyploid or aneuploid progenies when chromosome numbers are distinct as documented in many hybrid plant species (Rieseberg & Willis, 2007).

Usually, the genetic forces that have evolved to maintain the genetic integrity of distinct species counteract the generation of hybrids. These are, as described in the preceding text, prezygotic and/or postzygotic reproductive barriers that either prevent interspecific mating or result in inviable or sterile hybrid progenies. The existence and occasional strong success of hybrid species however illustrate that reproductive barriers can be permissive and that interspecific crossing in some cases can result in the formation of viable and fertile progenies. Hybrid speciation requires successful mating between distinct species and self-fertile hybrids that can evolve reproductive barriers to become sufficiently isolated from the parental species. Evidently, hybrids must also be fit and able to explore and occupy a niche in their environment even in competition with their parents.

How can reproductive isolation between parental species be permissive? As mentioned in the succeeding text, reproductive barriers evolve under different conditions in allopatric and parapatric/sympatric species. In allopatric species, a genetic basis for reproductive isolation is not directly required because populations already are physically isolated. Reproductive barriers therefore evolve as a by-product of genetic divergence rather than as reinforcement of reproductive isolation. Consequently, these barriers may be less strong, and when allopatric species are brought into contact, they may be able to interbreed as shown in species such as Armillaria mellea and Neurospora crassa (Anderson et al., 1980; Dettman, Jacobson, Turner, Pringle, & Taylor, 2003). Consistently, these studies have revealed that reproductive barriers are stronger between parapatric and sympatric species than between allopatric species. Hybridization therefore likely occurs more readily between allopatric species. This idea should receive special consideration given the fact that the human-mediated dispersal of fungal pathogens greatly has enhanced that potential for allopatric species to encounter each other (Fisher et al., 2012).

Hybrids may be self-fertile and propagate through mating within the hybrid population, or propagation may occur by backcrossing to parental individuals. However, extensive backcrossing will over time dilute the hybrid genome with parental alleles, and ultimately, the hybridization event may only be visible as traces of past introgression. But if hybrids are able to occupy a distinct niche or even outcompete the parental species, they may be able to persist and establish genetic integrity as new species.

One paradoxical issue to consider in the study of hybrid formation is how genetic incompatibilities are overcome. A hybrid genome consists of genes, which have not coevolved in the same genetic background and therefore were not optimized in parallel by natural selection. Indeed, many examples of hybrid sterility and inferior fitness observed in plants, animals and fungi support the hypothesis that incompatibilities between individual genes with major effects, or larger number of genes with a cumulative effect, are responsible for the deleterious effects in hybrids (Orr, 1995).

In fungi, a high tolerance for structural changes and genomic plasticity may enable genic incompatibilities to be overcome more easily (Croll & Mcdonald, 2012; Raffaele & Kamoun, 2012). Genome comparisons have confirmed previous karyotyping studies demonstrating considerable amounts of within-species variability in chromosome structure and chromosome numbers (see review by Zolan, 1995). To what extent chromosome rearrangements and genome plasticity affect homologous recombination between individuals is not known.

Another paradoxical issue of hybrids is how they can become established in an environment where the parental species also exist. The parental species are expected to be considerably better adapted in their environment relative to hybrids of intermediate genotypes and phenotypes. Hybrids may however contain new gene combinations that contribute new phenotypic characteristics allowing them to exploit other environmental niches not occupied by the parents. In yeast, experimental evolution studies revealed that hybrid progenies under certain environmental conditions can exhibit an even higher fitness than their parents (Greig, Louis, Borts, & Travisano, 2002). In particular, under intermediate or fluctuating conditions, hybrids may have a fitness advantage relative to their parents.

Theoretical support for allopatric speciation

Much of the theory associated with allopatric speciation focuses on the evolution of intrinsic postzygotic isolation (Dobzhansky, 1937, Muller and Pontecorvo, 1942). When emerging taxa are geographically isolated, two or more loci may fix for different alleles in each population. If these alleles produce negative epistatic interactions, hybrids between the taxa may be sterile or inviable (Orr, 2001). This form of isolation is often considered a hallmark of allopatric speciation because selection is expected to remove such negative interactions if they arose in the face of gene flow. While many other forms of isolation can also arise in allopatry, their evolution is often modeled indirectly. Indeed, theoretical treatment of allopatric speciation are simply models of sympatric speciation with the constraints imposed by gene flow removed. Felsenstein (1981) famously described the antagonism between selection and recombination when divergence occurs in the face of gene flow. His initial model considers a simplified case of two loci involved in divergent selection in alternate habitats. If mating is random, selection alone produces linkage disequilibrium between coadapted alleles; therefore, speciation cannot occur under these conditions unless selection is strong enough to remove every recombinant. If a locus conferring a pre-zygotic barrier is added in the form of assortative mating, speciation can occur; however, recombination severely limits favorable conditions by breaking up associations between assortative mating and adaptive loci. However, reductions in migration between alternative habitats facilitates the speciation process, with the lower extreme of allopatric speciation (m=0) completely free of the constraints imposed by recombination.

Peripatric or Founder Effect Speciation

Peripatric speciation represents a variation on allopatric speciation. In this case, a small population forms at the periphery of a larger population. This type of geographical isolation can happen anywhere, but is most easy to visualize in the situation of founders colonizing oceanic islands or, more generally, isolated pockets of habitat. Because the founding of peripheral populations can sometimes involve the movement of only a few individuals (or even a single gravid female), this type of speciation has also become known as founder effect speciation. The emphasis here is on the interaction between genetic drift and natural or sexual selection that occurs during the early stages of speciation.

If a new population is founded by a small number of individuals, just by chance the genetic composition of the founding population may differ significantly from that of the original source population because of genetic drift (see Genetic Drift). The population need not remain small for very long in order for this sampling effect to influence the future evolutionary trajectory of the population. In addition to this immediate genetic change, oftentimes small populations founded in peripheral habitats or on islands also experience changed environments (both the physical environment, including such things as the quality of the habitat, the distribution of resources, or the presence of competitors or predators, as well as the mating environment, represented by a shift in the distribution of available mating types and preferences), creating new selection regimes. Even in the complete absence of new selective environments, the shift in allele frequencies alone can potentially have a profound effect on how the population will respond to selection because of the changed internal genetic environment that results from drift. The combination of shifting gene frequencies by drift and the presence of potentially new selection regimes under this scenario has led some researchers to propose that this type of speciation can occur more rapidly than the more standard form of allopatric speciation which generally involves populations of larger size and less drastic changes in the physical and mating environment on isolation (for reviews, see Hollocher article in Grant, 1998; Templeton, 2008).

The theoretical framework used to justify the conclusion that speciation would be accelerated during founder effect speciation stems directly from Wright's model of an adaptive landscape (see Natural Selection and Genetic Drift). A major underlying genetic assumption that enters into the idea that the random sampling of alleles during the founder event can have a profound effect on the evolutionary trajectory of a population is that epistasis (where interactions between alleles at different loci produce phenotypic effects that are not predicted by the action of the individual allelic effects considered alone) and pleiotropy (where a single locus can directly influence more than one phenotypic trait) are quite common. It is under the assumptions of this type of genetic architecture that fitness peaks of varying heights will exist in the adaptive landscape and where random shifts in allele frequencies can have profound effects (e.g., see Gavrilets and Hastings, 1996; Gavrilets, 2004). If allelic effects governing traits important in speciation are more additive (where interactions between alleles at different loci are minimal), then allele frequency changes will not necessarily impact the trajectory of natural selection greatly.

Much of the debate surrounding the likelihood of founder effects accelerating the process of speciation has focused on the specific influence drift alone would have on the probability of shifting from one fitness peak to another (for a reviews, see the Barton and Hollocher articles in Grant, 1998; Coyne and Orr, 2004; Gavrilets, 2004). What has emerged from these theoretical studies has been the idea that the actual size of the founding population does not play as crucial a role in determining the probability of shifting from one fitness peak to another as does the underlying genetic architecture of fitness. On the basis of these theoretical results, researchers have begun to shift their focus to evaluating the genetic architecture underlying traits that change during speciation to see how often epistasis is an important component (see Genetic Patterns and Processes of Species Differentiation; see also Phillips, 2008). In addition to the genetic architecture influencing rates of change, the actual nature of the trait itself can affect the type of response that is expected under founder effect speciation. Reproductive isolation (both prezygotic and postzygotic) can be particularly susceptible to rapid change under this scenario because of the tight coevolution of male and female traits that normally occurs via sexual selection (see Sexual Selection). The random sampling of individuals during a founder event can easily move the population away from the stable equilibrium that characterizes the reproductive system in the original population. Reestablishment of a new equilibrium can often involve a radical shift in the mating system of the new population relative to the ancestral one. For sexually selected traits, random genetic drift coupled with sexual selection can act as a particularly powerful mechanism for driving speciation (Lande, 1981; Boake, 2005).

3 Geo-dispersal, a paleontological contribution to biogeographic theory

Unfortunately, with the increased acceptance of the importance of allopatric speciation came a dogmatic insistence by some authors (e.g., Croizat et al., 1974; Nelson and Platnick, 1981; Patterson, 1983; Humphries and Parenti, 1986), most of whom studied the extant biota, that vicariance was the only biogeographic process that produced congruent responses in biotas. As already discussed, the emergence of geographic barriers can isolate populations of several different co-occurring species and promote divergence. However, this ignores the opposite side of the coin because just as geological and climatic change can sometimes cause barriers to form, at other times, they may cause barriers to fall, allowing many taxa to simultaneously expand their range. It also ignores a long history of paleontological research which shows that the fossil record is replete with numerous examples of congruent range expansion by independent clades, a pattern that has been termed geo-dispersal by Lieberman and Eldredge (1996) and Lieberman (1997, 2000). Some prominent examples of geo-dispersal in the paleontological literature include McKenna (1975, 1983), who documented numerous cases of wholesale movement by mammals between Europe and North America and North America and Asia throughout the Cenozoic related to tectonic events and climatic changes. Beard (1998) also recovered more evidence for geo-dispersal by mammals from Asia into North America in the late Paleocene and early Eocene driven by warming events. At a broad scale, Hallam (1992) described numerous cases of geo-dispersal spanning the Phanerozoic in marine invertebrate faunas related to rising and falling sea-level. Lieberman and Eldredge (1996) also documented movements by trilobites between different marine basins in eastern North America during the Middle Devonian related to sea-level rise which joined marine connections between formerly isolated epeiric seas. Finally, Sereno et al. (1996) and Sereno (1997, 1999) found evidence for geo-dispersal by dinosaurian faunas within and across continents during the Early Cretaceous (Fig. 1). Studies of the modern biota by Sanmartin et al. (2001) and Conti et al. (2002) have reiterated these results from fossil faunas by their strong support for the existence of geo-dispersal. Not only is geo-dispersal a valid process, but biogeographic methods that only look for vicariance and ignore geo-dispersal will miss important biogeographic patterns and will potentially be inaccurate.

The recognition of the existence of geo-dispersal has a long historical pedigree which extends at least back to paleontological information described by Lyell (1832), suggesting that the initial impetus for the recognition of this process came from the analysis of the fossil record. However, a growing number of biologists have come to recognize the potential biogeographic importance of geo-dispersal (e.g., Riddle, 1996; Ronquist, 1998; Bisconti et al., 2001; Brooks and McLennan, 2002; Conti et al., 2002), although not all of these authors used that term to describe the process. This increasing recognition by biologists of the existence of processes first documented in the fossil record is further evidence for the growing geobiological synthesis between paleobiogeographers and biogeographers.

KEY INNOVATIONS AND SPECIATION

Diversification of taxa involves speciation events and these speciation events can occur allopatrically or sympatrically. Allopatric speciation occurs in the absence of gene flow and sympatric speciation in the presence of gene flow. The two processes of speciation are very different and it is likely that the role of key innovations as a trigger of diversity differs as a consequence.

In the sympatric scenario, polymorphic populations are kept divided by assortative mating habits (Bush, 1994). Disruptive selection then leads to divergence of the polymorphic populations. Sexual selection which acts on the mating system can be disruptive (Lande, 1982; Wu, 1985; Turner and Burrows, 1995; Payne and Krakauer, 1997). Natural selection on other traits (e.g., body size) with pleiotropic effects on mate preference can also have a disruptive effect (Thoday and Gibson, 1959; Maynard Smith, 1966; Rosenzweig, 1978; Kondrashov and Mina, 1986; Johnson et al., 1996). A large number of speciation events and mating barriers produced by sexual selection is not enough to maintain species diversity. Species diversity is determined by the balance of the numbers of species that originate and become extinct. When, after speciation events, the new species are indistinguishable ecologically, species will be lost in a process akin to random drift (Wright, 1931). However, the species will never be exactly similar ecologically, in which case the extinction process will be considerably more rapid, except when the differences lead to niche differentiation (MacArthur and Levins, 1967; Meszéna and Metz, in press). Thus, disruptive natural selection that avoids extinction because of limiting similarity is necessarily involved in diversification.

Diversification of the body plan occurs in both sympatrically and allopatrically evolved radiations. An important difference is the nature of the selection processes, disruptive followed by directional selection in sympatric speciation and directional in allopatric speciation. Evolution in a speciose allopatric clade requires a body plan that can be diversified, but the modifications can occur slowly since there is no competition between diverging incipient species. Evolution of a speciose sympatric clade is only possible when rapid disruptive selection can repeatedly occur within the evolving clade. Rapid disruptive selection poses high demands on the relative ease with which diversification of a body plan can occur. It is, thus, to be expected that several structural key innovations will have been involved in the history of such a radiation, each of which increased the number of degrees of freedom of the body plan.

The cichlid species flock of Lake Victoria in Africa provides one of the most spectacular examples of speciation and diversification (e.g., Fryer and Iles, 1972; Lowe-McConnel, 1987). Diversification was in this case facilitated by the availability of an empty habitat (relaxation of stabilizing selection), subsequent to the formation of the lake after the most recent ice age (Johnson, 1996). Speciation has apparently mainly been driven by sexual selection for strikingly coloured males (Seehausen et al., 1997; Seehausen and van Alphen, 1997), although allopatric speciation will certainly have also played a role given the size of the lake and the diversity of habitats. Cichlid fishes have indeed acquired several structural key innovations, at least two more structural key innovations relative to their presumed ancestors (see section “Decouplings and radiations,” Fig. 6). These key innovations have led to a flexible and versatile pharyngeal jaw apparatus (Galis and Drucker, 1996).

There are two reasons why the flexible and versatile pharyngeal jaw apparatus of cichlids promotes evolutionary diversification (Galis and Metz, 1996). First, it provides behavioral plasticity; second, it provides evolvability. Although cichlids usually act as specialists, occupying particular feeding niches, they can eat very diverse food items when necessary, albeit with lower efficiency. This is probably relevant right from the start of the speciation process, because if competition forces a polymorphic population toward diversification, this type of phenotypic plasticity immediately allows rapid shifts. The second reason why the versatile pharyngeal jaw apparatus is important for evolutionary diversification is that quite small evolutionary changes in morphology and behavior allow cichlids to specialize on different food items. The striking diversity of feeding niches that characterizes cichlids of Lake Victoria suggests that niche differentiation occurred by rapid specialization for different feeding niches. This implication is strengthened by the observation that sibling species are always pigeonholed by small differences in feeding behavior (Hoogerhoud et al., 1983; Bouton et al., 1997).

In the case of the haplochromine cichlids, two spectacular radiations have independently occurred, in Lake Victoria and in Lake Malawi. In rivers, no substantial radiation of haplochromine cichlids has taken place, which is probably because of the much lower structural diversity of a fast-flowing river. The lower structural diversity has both consequences for potential niche diversification and for drift in mate recognition systems. Tilapiine cichlids possess the same key structural innovations as the haplochromines and two of the three big genera (Oreochromis and Sarotherodon) also possess the same key behavioral innovation (mouth brooding), but they have not radiated (Trewavas, 1983). As tilapiines also occur in the same lakes where haplochromines have radiated it is not immediately apparent which ecological factors could be responsible for this striking difference. Sexual dichromatism, which presumably has played an important role in the diversification of haplochromines (Seehausen et al., 1997; Seehausen and van Alphen, 1997), is also common in Oreochromis (Trewavas, 1983). Possibly, genetic factors have constrained diversification of tilapiines.

The low diversity of haplochromines in rivers and of tilapiines in lakes provides a good example of how diversity is the result of many interacting factors of which key innovations are but a few.