IN THIS CHAPTER WE PAY homage to Ed Wilson as Naturalist. His influence on our research on speciation has been much greater than this chapter will reveal, so we begin by making one explicit connection. In the Theory of Island Biogeography, MacArthur and Wilson (1967) came close to discussing speciation in chapter 7 when referring to the prevailing view, associated with Mayr (1963), that given enough time isolated populations will diverge genetically to the point at which they are incapable of exchanging genes when finally they encounter each other. They made the insightful point that if islands could be reached once they could be reached again; therefore repeated immigration (and breeding) would retard divergence and a balance would be struck between these opposing processes, rather like the immigration-extinction balance they so successfully modeled. Since then the dynamics of gene flow and selection have been thoroughly investigated (Slatkin 1975, Barton and Slatkin 1986), and they form the core of divergence-with-gene-flow ideas about how speciation occurs (e.g., Rice and Hostert 1993, Smith et al. 1997, Price 2008).
The last forty years of research on bird speciation on islands has yielded different pictures or models of the speciation process (Grant 2001, Price 2008, Grant and Grant 2008a). One elaborates the views on allopatric speciation described above. Divergence takes place in allopatry, and barriers to interbreeding arise there as a result of selection, with or without gene flow from parent to daughter population (model I; see also Clegg, this volume). Founder effects may contribute at the beginning. Speciation is both initiated and completed in allopatry. In the next two models speciation begins in allopatry and is completed in sympatry. The second (model II) lays stress on accelerated divergence at the time of secondary contact through selective reinforcement of reproductive and/or ecological trait differences that initially evolved in allopatry. A third (model III) emphasizes an exchange of genes at the sympatric stage through episodic introgressive hybridization. The exchange does not simply destroy the differences, but through selective backcrossing creates new combinations of genes that enhance responsiveness to selection. The result is speeded up divergence along existing trajectories or change to new trajectories. Fission tendencies alternate with fusion. A fourth view (model IV) holds that all changes occur sympatrically; there is no allopatric phase, and hence no secondary contact. For sympatric speciation to occur there must be assortative mating among members of two groups formed from one by disruptive selection.
The four models differ in biogeographic features, and in how selection is supposed to occur. They combine elements of ecological and non-ecological speciation (Schluter 1996, 2001, Price 2008). Three are variations on the allopatric speciation theme. All involve a secondary sympatric phase through immigration, and therefore can be accommodated by the theory of island biogeography fairly simply. The fourth, sympatric speciation, is fundamentally different because it proposes in situ, within-island, origination of new species without immigration. It enhances diversity over and above the effects of immigration, and for that reason we focus on it.
One fruitful approach to the problem of understanding speciation is to study directly the processes hypothesized to be important. This complements the more often used, indirect, comparative method for inferring evolutionary history. The critical processes that need to be demonstrated to discriminate among these four models are effects of intraspecific gene flow from island to island, introgressive hybridization within islands, enhancement of differences between populations soon after secondary contact is made, mate choice, and selection, be it disruptive or directional. All these are amenable to direct study by observation, measurement and experimentation.
Species in statu nascendi are especially suitable for direct study of dynamical interactions in sympatry and for extrapolation to the unobserved history of species that are now completely reproductively isolated. This chapter discusses what has been learned recently about speciation through field study of two groups of such species; buntings in the Tristan da Cunha archipelago in the south Atlantic and ground finches in the Galápagos archipelago in the eastern tropical Pacific. In their isolated locations, one can be confident the species evolved where they are now found. In contrast, species in many continental regions and on less isolated islands like the Baltic islands of Gotland and Öland (Tegelström and Gelter 1990) and Britain (Newton 2003) may have evolved in one place and now, postglacially, occupy another.
Sympatric speciation (model IV) has been invoked in both of the cases we review. Serious investigation of sympatric speciation began with a theoretical analysis by Maynard Smith (1966), coincidentally at about the same time as the first synthesis of island biogeography theory (MacArthur and Wilson (1963, 1967). The theories have had largely independent lives since then. In the Discussion we explore some connections between them.
Solitary islands provide the strongest evidence of sympatric speciation. One species is likely to have given rise to two, sympatrically, if they occupy a single and solitary island, too small to allow for spatial segregation, and they are more related to each other than either is to a third. For example, two species of palms apparently evolved on the single, Australian, Lord Howe Island (Savolainen, Anstett et al. 2006, Savolainen, Lexer et al. 2006, Stuessy 2006, Gavrilets and Vose 2007). This example is similar to fish that have apparently undergone diversification and speciation in single bodies of water where opportunities for spatial segregation are minimal (Schliewen et al. 1994, 2006, Barluenga et al. 2006a,b, Gavrilets et al. 2007). These are essentially insular environments, solitary islands in effect. Coyne and Price (2000) surveyed the relevant bird literature and could find no such examples. Where they might have found examples they didn’t. For example, the Cocos finch has been present on the well-isolated Cocos Island long enough to have given rise to other species, and its environment is varied enough to support a variety of feeding types in the population (Werner and Sherry 1987), and yet it has remained a single species under conditions suitable for sympatric, but not allopatric, speciation.
However, three recent studies have suggested that birds may indeed undergo sympatric speciation on islands. One investigated Nesospiza buntings on islands in the South Atlantic Tristan da Cunha archipelago (Ryan et al. 2007), and another studied a population of Geospiza finches in the Galápagos archipelago (Huber et al. 2007). A third one, suggesting that Oceanodroma petrels have speciated sympatrically as a result of breeding in the same location at different times (Friesen et al. 2007), was published after this chapter was written and is briefly mentioned in the Discussion.
Two species of Nesospiza occur together on Inaccessible and Nightingale, two out of the three islands in the Tristan da Cunha group. One species is large (N.w. wilkinsi, Nightingale; N.w. dunnei, Inaccessible) and one is small (N.a. questi, Nightingale; N.a. acunhae, Inaccessible). They are ecologically separated on each island by bill-related food size. N. wilkinsi exploit Phylica fruits and N. acunhae eat grass (Spartina) and sedge seeds which are much smaller (Hagen 1952, Elliott 1957, Ryan et al. 2007). Reproductively they are separated by their song, plumage, and size differences (Ryan et al. 2007).
Arrival of buntings in the archipelago can be dated at ~3.3 mya on the basis of a 6.7% difference in cytochrome b sequences between Nesospiza and the presumptive sister species, Rowettia goughensis, on the solitary Gough Island 350km to the south (Ryan et al. 2007). How did Nesospiza speciation then take place? To answer this question Ryan et al. (2007) analyzed mtDNA and microsatellite variation, and found almost complete lineage sorting by island (figure 12.1). This is consistent with in situ splitting of a single population into two species, on each of the two islands. Sympatric speciation is the hypothesis favored by Peter Ryan and colleagues. They support it with observations of assortative mating by size, and evidence of ecotypic variation in the smaller species on Inaccessible that is suggestive of disruptive selection and incipient speciation.
The data are consistent with alternative hypotheses. According to one, ancestral Nesospiza buntings colonized the archipelago not once but twice from South America. Sequential invasions of the same lineage have been repeatedly hypothesized to explain the occurrence of two related species on some islands yet only one in the mainland source region (Grant 1968, 2001, Coyne and Price 2000). For example, two species of Sephanoides hummingbirds occur on the Juan Fernandez Islands off the coast of Chile, whereas there is only one on the mainland. If the island had been invaded once and the two island species had evolved sympatrically they should be sister species. Phylogenetic reconstruction by Roy et al. (1998) shows they are not. Instead, it supports the double-invasion hypothesis by showing that the mainland species is more closely related to one, presumably a relatively recent colonist, than to the other (Grant 2001).
In the case of Nesospiza buntings one species could have colonized the archipelago twice, or two species could have colonized once. Comparisons with continental species and phylogenetic reconstruction performed so far suggest that Tristan da Cunha was colonized only once, and all Nesospiza evolution took place within the archipelago (Ryan et al. 2007).
Figure 12.1. Diagram of the relationships among Tristan buntings. Inaccessible (I) and Nightingale (N) Islands are each occupied by a small species (Nesospiza acunhae) and a large one (N. wilkinsi). The unrooted dendrogram of microsatellite DNA differences placed between the islands shows each sympatric pair to be most similar to each other genetically. In contrast to this, phenotypic similarities are strongest between allopatric pairs. Adapted from Ryan et al. (2007).
According to allopatric models of speciation, birds dispersing either from South America or from Gough Island ~3.3 mya colonized Nightingale, the oldest island in the Tristan archipelago. The population gave rise through dispersal to another on Inaccessible, and the two populations diverged, thereby beginning the process of speciation. Sympatry was subsequently established through further dispersal of members of each population to the island occupied by the other: within-archipelago double invasions after differentiation, a Darwin’s finch radiation in miniature (Lowe 1923, Lack 1947). If this actually happened, why is it not reflected in the pattern of phylogenetic relationships? The answer is a well-known problem in island speciation inferred from molecular phylogenies (Clarke et al. 1998; see also Chan and Levin 2005): one sympatric lineage has “captured” another through introgressive hybridization, and the phylogenetic signal has become obliterated. Hybridization is now occurring on the younger Inaccessible (3 my), but apparently not on Nightingale (>18 my). It may have occurred on Nightingale earlier, gradually diminishing through time. If so it might be detected with coalescent methods (e.g. Peters et al. 2007).
A prediction of the allopatric hypothesis is that sympatry on Nightingale is no older than 3 my, the age of the younger island. If it is older than 3 my, the allopatric model would have to be abandoned and the sympatric alternative would be upheld. Mitochondrial data do not support such an ancient split: they yield an estimate of 0.3–0.4 my for the separation of Nightingale and Inaccessible buntings (0.7% cytochrome b sequence difference). Therefore the allopatric model cannot be abandoned. On the question of whether the earliest split is between species on different islands, as expected from the allopatric hypothesis, or between populations on the same island, as expected from the sympatric hypothesis, the data are equivocal. There are no mitochondrial differences between populations on the same island. This is not expected under a sympatric speciation model. One explanation among others (Ryan et al. 2007) is introgressive hybridization after initial divergence in allopatry. On the other hand, the species on Nightingale differ more in microsatellite profiles, marginally, than either does from buntings on Inaccessible (Ryan et al. 2007). This is consistent with the sympatric hypothesis.
A curious feature of Tristan buntings is that for the first 80–90% of their history on Nightingale only one species existed, to judge from the cytochrome b data considered at face value. Even allowing for imprecision in age estimates and the biasing effects of lineage sorting, the magnitude of the delay is remarkable. There is no comparable long delay in finch speciation in two other volcanic archipelagoes, Hawaii (Fleischer and McIntosh 2001) and Galápagos (Grant and Grant 2008a). In the first 80–90% of Darwin’s finch history (2–3 my), for example, more than half of the species evolved. Galápagos differs from Tristan in that a minimum of five (volcanic) islands was always present during finch history. This may have allowed species to persist and accumulate in Galápagos even when individual populations became extinct. Note that 14 species of Darwin’s finches evolved in a shorter time (2–3 my) than was available to Tristan buntings (3–4 my).
Such a long “waiting time” to speciation (Bolnick 2004) is not expected under the sympatric speciation model except under a set of restricted (genetic) conditions governing mate choice. Neither is it expected under an allopatric model, because for most of that time Inaccessible was present. It takes no more than 0.2 my for a new island to be colonized (see below). The long delay in speciation could be explained ecologically. Plants that constitute one of the niches may have arrived recently, perhaps in the last 0.5 mya. In principle this could be tested with a phylogeny of the food plants (Phylica trees, Spartina grasses, and sedges). A testable expectation under the allopatric model is that volcanic activity on Inaccessible rendered the island uninhabitable for all or part of its early history, but at the same time was less drastic on the older Nightingale. Volcanic activity occurred in the last 0.5 my on both islands, and therefore probably earlier. It may have extirpated populations on Inaccessible, and possibly also on Nightingale, thereby obscuring the history of the survivor(s).
For a better understanding of the evolutionary history of these buntings it would help to include molecular data from a population of the smaller species (N. acunhae) on a third island, Tristan (0.2 my old), because birds from this island may have contributed to the mixture on Inaccessible. The Tristan population is now extinct, owing to human activity; Spartina tussocks were destroyed (Hagen 1952) and predatory feral cats, rats, and mice were introduced (Elliott 1957). Unfortunately, it appears that only one specimen of bunting from Tristan exists in museum collections (Lowe 1923, Elliott 1957).
Second, it would help to root the tree. This might permit identification of the oldest species, thereby allowing a more precise framing of the food-niche test described above. N. wilkinsi on Nightingale is the best candidate, as it is genetically the most distinctive from the rest. Further, if a root is established with mainland species (e.g., Sicalis or Melanodera spp.) it might be possible to distinguish between two colonization hypotheses: separate colonizations of Tristan da Cunha and Gough Island from South America, or colonization of one followed by the other. The first hypothesis was suggested by Lowe (1923) and developed by Rand (1955). It is preferred by Ryan et al. (2007) because, among other reasons, population sizes are larger on the mainland than on the islands. Assuming Rowettia and Nesospiza are truly sister genera, we consider a single, sequential, colonization to be at least as likely as two separate ones, because the South American mainland is 3,000 km away, whereas Gough is little more than a tenth of this distance from the Tristan archipelago.
In summary, ecological, morphological, and genetic patterns among Tristan buntings display elements of all models outlined at the beginning except for one with reinforcement (II). Consistency with the model of sympatric speciation is noteworthy in view of the rarity of evidence for this mode of speciation in birds (e.g. Sorenson et al. 2003, Price 2008). We cannot draw a stronger conclusion because the issue of sympatric speciation is unresolved, and perhaps unresolvable in the light of introgressive hybridization and possible extinctions. The next example provides more evidence of sympatric speciation, of a different kind.
Darwin’s finches are a classical example of a young adaptive radiation (Grant 1986, Grant and Grant 2008a). In recent and ongoing radiations the distinction between species is often blurred because there has been insufficient time for complete discreteness to evolve and speciation is incomplete (Grant and Grant 2005). A taxonomist’s nightmare is an evolutionary biologist’s treasure. Incomplete speciation provides opportunities to study the process. There is no more confusing, and at the same time potentially more rewarding, situation than on Santa Cruz Island. The remainder of this chapter discusses what has been learned from field studies of finches on this and the neighboring island of Daphne Major.
The population of medium ground finches (Geospiza fortis) on this island displays an unusual feature: beak sizes are bimodally distributed (figure 12.2) at some localities and at some times (Hendry et al. 2006). The bimodality is not accounted for by average size differences between males and females or between young and old birds. Phenotypic variances are unusually large, and this fact, combined with bimodality, raises the possibility of disruptive selection as a cause of the origin as well as the maintenance of the bimodality (Ford et al. 1973). The hypothesis of current disruptive selection has yet to be tested by quantifying survival and breeding success of individuals in relation to beak sizes, diets and food availability. This is difficult to do in a local area embedded within a larger region because of uncontrolled movement of birds in and out of the study area, and for that reason analysis needs to be restricted to known residents. The best evidence for disruptive selection is the nonrandom persistence of adults from one year to the next in the El Garrapatero study area (Hendry et al. 2009).
Figure 12.2. Bimodal distribution of beak depth in a sample of male G. fortis from the El Garrapatero locality, southern Santa Cruz Island, Galápagos, in 2004. From Hendry et al. (2006), fig. 3.
Figure 12.3. Assortative pairing of medium ground finches (G. fortis) at El Garrapatero on Santa Cruz Island in (A) 2004–5 (dry conditions), (B) late 2005 (very wet), and (C) 2006 (moderately wet). From Huber et al. (2007).
Another factor maintaining the bimodality is a strong tendency for birds to pair assortatively (Huber et al. 2007). The pattern of morphological variation among pairs (figure 12.3) suggests that large birds mate preferentially with large birds and small birds mate preferentially with small birds. There is no assortative mating within size groups; it is manifest only when size groups are combined. Characteristics of song vary with body and beak size (Huber and Podos 2006), so the cues used in mate choice could be provided by song, by morphology, or by both (Grant and Grant 2008a). Experiments with other populations of Geospiza species have demonstrated discrimination on the basis of each set of cues independent of the other (Ratcliffe and Grant 1983, 1985).
As with Tristan buntings, the origin of this interesting situation is unknown. Divergence could have originated sympatrically or allopatrically.
A bimodal beak size frequency distribution coupled with assortative pairing on the basis of beak size is consistent with the idea that the population is in the process of splitting into two, sympatrically, through disruptive selection (model IV). The split has reached the point at which large and small members of the population differ in microsatellite allele frequencies and rarely breed with each other (Huber et al. 2007).
If sympatric divergence is a correct interpretation of their origin, the process has been occurring for a century or more. Specimens of medium ground finches collected on Santa Cruz island at the beginning of the nineteenth century for museums show exactly the same positively skewed frequency distributions with bimodal tendencies as do modern samples, at both northern and southern localities, and in early (<1906) and later (>1924) samples (figures 12.4 and 12.5). Two species of ground finches that are sympatric with G. fortis, the small ground finch (G. fuliginosa) and the cactus finch (G. scandens), show standard normal distributions and no such skew (figure 12.6). They are a kind of “control” for the ongoing “experiment” with medium ground finches (G. fortis). The sample of measurements of a fourth species, the large ground finch (G. magnirostris), is too small for analysis.
The morphological and mating patterns are also consistent with allopatric model III, under which the population we call G. fortis is actually two populations. The bimodality could be the result of unusually large medium ground finches immigrating from another island where average size is large, such as San Cristóbal or Floreana to the south, and breeding with residents on Santa Cruz to some, but apparently incomplete, extent. If so, fission and fusion tendencies have yet to be resolved one way or the other. Nothing is known about current immigration to Santa Cruz. In the absence of other factors it would have to be persistent to account for the persistent bimodality and skew.
Yet another possibility is that skew and bimodality are produced by hybridization with G. magnirostris; either residents on Santa Cruz or immigrants from another island. The hypothesis of interbreeding on Santa Cruz is supported by one observation of a mixed pair (Huber et al. 2007), by the genetic (microsatellite) similarity of these species compared with allopatric pairs of the same species (Grant et al. 2005), and by the similarity in songs of G. magnirostris and large members of G. fortis (Bowman 1983, Grant and Grant 1995, 2008a, Huber and Podos 2006).
Figure 12.4. Frequency distributions of beak size of medium ground finches (G. fortis) collected for museums in the north (1868–1939) and south (1868–1968) of Santa Cruz Island. g1 is a measure of skewness, N is sample size, and two asterisks indicate a significant departure from normality at P<0.01 (Snedecor and Cochran 1989). Data originally analyzed in Grant et al. (1985).
Thus there is not one but three explanations for the unusual frequency distributions of finch morphology (Grant 1986, Huber et al. 2007), and few data available to discriminate among them. An expanded array of molecular markers is needed to detect and identify immigrants, F1 hybrids, and backcrosses. Therefore, for a better understanding of the dynamics of immigration and hybridization, we turn to a long-term study of ground finches on the neighboring small island of Daphne Major (0.34 ha), 8 km north of Santa Cruz. We then apply the findings from Daphne to the question of G. fortis evolution on Santa Cruz.
Figure 12.5. Frequency distributions of beak size of medium ground finches (G. fortis) collected for museums on Santa Cruz Island, early (1868–1904) and late (1924–1968). Symbols as in figure 12.3.
For the immigration hypothesis to be supported it needs to be shown that immigrants from a morphologically differentiated population breed with residents. For the hybridization hypothesis to be supported it needs to be shown that introgressive hybridization results in a skewed distribution.
Figure 12.6. Frequency distributions of beak size of medium ground finches (G. fortis), small ground finches (G. fuliginosa) and cactus finches (G. scandens) collected for museums on Santa Cruz Island. Symbols as in figure 12.3.
Figure 12.7. Frequency distributions of beak size of live medium ground finches (G. fortis) trapped and measured on Daphne Major Island. Gray bars indicate individuals, mainly immigrants and offspring, which are statistically responsible for the skew. Symbols as in figure 12.3.
Medium ground finches immigrate to Daphne. Their detection is made difficult by the large overlap in frequency distributions of beak and body traits between Daphne resident G. fortis and G. fortis from other islands. Moreover the 13 populations are not differentiated enough genetically (Grant et al. 2004) to enable us to identify island of origin of individuals by using assignment tests (e.g., Pritchard et al. 2000). Nevertheless, some immigrants can be detected by their phenotype. Daphne residents are smaller on average than all other conspecific populations. Therefore large birds beyond the size range of Daphne residents and within the upper size range of birds on other islands are recognizable as immigrants. They cause the frequency of beak sizes to be positively skewed. They (and their offspring) can be identified as the minimum number of individuals that must be serially deleted from the upper end of a frequency distribution to eliminate the skewness (figure 12.7).
Identified by this means, immigration of large birds is rare and intermittent. The total is 30 out of 3245 (1.0%), and they arrived at only four times. Six arrived in 1977, the year following a long breeding season in the archipelago, one arrived in 1981, 22 arrived sometime after the end of the 1983 El Niño and were captured in 1983–5, and the remaining two arrived in 2000–1. Twenty-five were never seen after their year of capture, and one was seen two years after capture. All these were in immature plumage. Therefore immigration usually ends with the disappearance of the immigrants (death or emigration). We know there were none in 1991 and 1992 because all birds on the island were banded at that time.
Fig. 12.8. The pedigree of large immigrants on Daphne Island below the female of the F1 generation (from Grant and Grant 2008b). Genealogical relationships were inferred from genetic (microsatellite) data and from observations. Solid symbols are genotyped birds (circles females, squares males, diamond sex unknown). The unfilled symbol refers to an individual that was known but not genotyped. Gray symbols refer to two birds whose genetic relationships are hypothesized from their phenotypes (see text). Double lines connect the breeding of close relatives. Photo by G. B. Estes.
Only five large immigrants are known to have stayed to breed; one of unknown sex arrived in the early 1970s, two males arrived at different times in the 1980s, and a male and a female arrived in 2000–1. When single birds arrived they bred successfully with residents (Grant and Grant 1996). When the male and female arrived at approximately the same time they bred with each other. Thus, as shown by this pair, some degree of reproductive isolation occurs between large immigrants and residents. This makes plausible the hypothesis of immigration as a source of bimodality, skew, and assortative mating in the Santa Cruz population of G. fortis.
This breeding pair is remarkable. It provides a rare example of the crucial step in the allopatric model of the establishment of sympatry. Observations and genotypes allow us to reconstruct the pattern of events and relationships among the participants (figure 12.8). The original male was first seen in 2000 in immature plumage. It had probably hatched in 1998. It set up a territory, built a nest, and sang, but probably did not breed. The female was first seen the following year, a year of little or no breeding. They bred in 2002, and died in 2003 or early in 2004. Two offspring hatched in 2002 and bred with each other for the first time in 2005, producing at least five offspring (figure 12.8).
The original mother was captured, measured, and genotyped. Her offspring matched her genotype at all 15 microsatellite loci, and matched no other individual’s complete genotype. This allowed us to exclude as the mother all G. fortis known or suspected to be resident in 2002–5, as well as G. magnirostris and G. scandens. The genotype of the missing father can be deduced at 12 of the loci; both alleles can be identified by default at nine of them. This enabled us to exclude all G. magnirostris and all G. scandens as possible fathers as well as all resident G. fortis. Altogether 263 G. fortis, 60 G. magnirostris and 100 G. scandens were excluded as parents. Moreover phenotypic data are also inconsistent with a hypothesis of cryptic, that is unobserved, hybridization. The large birds are not intermediate in beak proportions between those of G. fortis and G. magnirostris as they should have been if they were F1 and F2 hybrids (figure 12.9). Thus both the original mother and father must have immigrated. The source island is unknown. On geographical grounds Santa Cruz is the most likely candidate. Parents, offspring, and grand-offspring are above average for Santa Cruz G. fortis, spanning the 60th to 90th percentile range in bill characters.
Notice in figure 12.7 how few immigrants can create skewness. The degree of skewness in the frequency distribution of G. fortis beak sizes on Daphne in the combined samples from 2002 to 2007 (g1 = 0.938, N = 332, t = 7.22, P<0.0001) is greatly influenced by the measured immigrant in 2000–01 and the six offspring and grand-offspring. When they are deleted from the sample the skewness is more than halved (g1 = 0.421, N = 325, t = 3.24, P<0.0005). Statistical significance can be eliminated altogether just by deleting the next three largest birds (g1 = 0.248, N = 322, t = 1.91, P >0.05).
The medium ground finch hybridizes with the small ground finch (G. fuliginosa) and the cactus finch (G. scandens). Hybridization is rare but persistent, carries no fitness disadvantage under favorable environmental (feeding) conditions that we have been able to discover, and, in the years following the exceptionally strong El Niño event of 1982–3 when favorable feeding conditions persisted, it resulted in a genetic and morphological convergence of the medium ground and cactus finches (figure 12.10). Introgression has the effect of increasing both variance and skewness of the recipient population (figure 12.11; Grant and Grant 2002a). Therefore skewness in the Santa Cruz frequency distributions can be plausibly explained by introgressive hybridization with large ground finches.
Figure 12.9. Medium ground finches (G. fortis) and large ground finches (G. magnirostris) on Daphne Island 2002–7.
With the known facts about immigration and hybridization on Daphne, we should expect a blurring of the morphological distinction between sympatric species on Santa Cruz. As expected, there is no clear distinction between G. fortis and G. magnirostris when large samples are analyzed (figure 12.12). Neither we, nor our colleagues, have been able to establish explicit criteria for characterizing each species and distinguishing between them. As a result, individuals between two peaks in the frequency distribution of beak sizes could be G. fortis, G. magnirostris, F1 hybrids, or backcrosses.
Figure 12.10. Introgressive hybridization after 1983 blurred the morphological distinction between medium ground finches (G. fortis) and cactus finches (G. scandens) on Daphne Island. Polygons enclose males that sang the species-specific songs and their mates.
However, a fortuitous circumstance enables us to identify G. magnirostris individuals objectively. G. magnirostris and G. fortis occur on Daphne Major without interbreeding. A breeding population of G. magnirostris was established on the island at the beginning of the El Niño event in 1982–83 (Gibbs and Grant 1987, Grant et al. 2001), when three female and four male immigrants stayed to breed. Numbers increased gradually as a result of breeding and local recruitment, augmented by additional immigration. Over the following 25 years, when G. fortis was hybridizing with G. scandens, large ground finches did not hybridize with G. fortis, probably because the morphological difference between them here is unusually large, as a result of the small average size of the G. fortis (figure 12.10). G. magnirostris on Daphne can therefore be used to identify G. magnirostris on neighboring Santa Cruz, on the assumption that distributions of beak sizes of G. magnirostris on the two islands are the same. This may not be exactly correct in view of evidence that some G. magnirostris immigrate to Daphne from Santiago (Grant et al. 2001). However, any bias arising from inclusion of birds from Santiago is conservative, in that large ground finches on Santiago are slightly larger on average than those on Santa Cruz (Lack 1947, Grant et al. 1985).
Figure 12.11. Increase in the variance (above) and skewness (below) in the frequency distribution of cactus finch (G. scandens) beak shape as a result of interbreeding with medium ground finches (G. fortis). From Grant and Grant (2002a).
First, we combined measurements of live G. fortis and G. magnirostris on Santa Cruz and Daphne and performed a principal-components analysis of three beak dimensions (length, depth, and width). We used PC 1 as an index of size because it accounts for most of the variance (97.2%), and loadings of all three beak dimensions were high (0.977–0.992). We then used the lower boundary of the Daphne G. magnirostris distribution as a criterion for identifying G. magnirostris on Santa Cruz. No adjustment for skewness was needed; there was none (g1=0.024). In the final step we ranked the Santa Cruz birds in order of decreasing size, serially deleted birds from beyond the apparent upper end of a normal distribution, and stopped when skewness was at a minimum (g1 = 0.035). Birds lying below this boundary are G. fortis, while birds above this boundary but below the lower G. magnirostris boundary belong to neither species and are therefore identified as hybrids and backcrosses (figure 12.12).
Figure 12.12. Combined frequency distributions of beak size of medium ground finches (G. fortis) and large ground finches (G. magnirostris) on Santa Cruz and Daphne Major Islands. The right hand broken line shows the lower limit of G. magnirostris sizes, calculated from figure 12.9. The left-hand broken line shows the upper limit of the G. fortis sizes on Santa Cruz, calculated by serially deleting large individuals from the Santa Cruz sample until skewness dis appeared. Individuals between the lines are presumed to be F1 hybrids and backcrosses.
The results were as follows. Nine Santa Cruz individuals considered by us on capture to be G. fortis were identified as G. magnirostris. All were from Academy Bay in late 1973. An additional 17 were identified as hybrids. The total is 26 out of 278, or approximately 10%.
Hendry et al. (2006) plotted beak depth against beak length of G. fortis, measured in the same way as we did, at three localities on Santa Cruz. The results are all positively skewed. Using the classification developed for our own specimens, we estimate that 5% of the Borrero Bay sample shown as G. fortis are in fact G. magnirostris and/or hybrids and backcrosses, and at the other two localities (Academy Bay and El Garrapatero), at least 25% are (figure 12.13). These numbers are approximate and could be somewhat in error; A. P. Hendry (personal communication) considers them to be too high (see Foster et al. 2008). Nevertheless there are clearly G. magnirostris in these samples, and probably hybrids and backcrosses too. For example, at El Garrapatero three individuals exceed 14 mm in beak depth, and four exceed 14 mm in beak length.
Figure 12.13. Beak sizes of medium ground finches (G. fortis) on Santa Cruz Island, from Hendry et al. (2006). The broken line, calculated from estimates in figure 12.12, separates large ground finches (G. magnirostris) and presumed hybrids (above and to the right) from G. fortis (below and to the left).
To summarize, large and small members of the G. fortis population differ in microsatellite allele frequencies, have different song characteristics on average, and rarely breed with each other (Huber and Podos 2006, Huber et al. 2007). These are characteristics of sympatric species, which suggests they could be cryptic species; a minifortis and a megafortis. The group of large G. fortis, the megafortis, is heterogeneous; it comprises G. fortis, some individuals indistinguishable from G. magnirostris, and probably F1 hybrids and backcrosses. The group also appears to be reproductively isolated from larger members of the G. magnirostris population (Huber et al. 2007). The group of large G. fortis may owe its origin not to a splitting of a single population into two through disruptive selection as envisaged in models of sympatric speciation but to a pooling of genes of two species. In other words, it could be a rare example of hybrid (homoploid) speciation in birds. Ongoing studies of this population (A. P. Hendry and S. Huber, personal communication) are designed to clarify the roles of selection, competition for food, mating structure, and relationships with the small (G. fuliginosa) and large ground finches (G. magnirostris).
As originally formulated by MacArthur and Wilson (1963, 1967), the theory of island biology was ecological and not evolutionary. Whittaker et al. (this volume) summarize efforts to extend the theory by incorporating speciation in archipelagoes (see also Gillespie and Baldwin, this volume). A biogeographically important distinction is to be made between modes of speciation. Allopatric speciation increases the number of species on an island through intra-archipelago immigration, whereas sympatric speciation increases the number on an island without immigration. Sympatric speciation is dependent on environmental heterogeneity (opportunity) within an island persisting for a long time under conditions of low rates of immigration. Logically, therefore, it is to be expected more in the middle of a radiation than early or late (Rosenzweig 1995), on large rather than small islands (Grant and Grant 1989a), and on distant rather than near islands. If sympatric speciation is common, island biogeography theory needs to be modified to allow for an increase in island diversity without immigration (Heaney 2000, Losos and Schluter 2000, Gillespie 2004, Gillespie and Baldwin, this volume). But how likely is this form of speciation for birds on islands?
Despite numerous theoretical investigations into how it might occur (Doebeli 1996, Kawecki 1997, Dieckmann and Doebeli 1999, Kondrashov and Kondrashov 1999, Doebeli and Dieckmann 2000, Dieckmann et al. 2004, Van Doorn et al. 2004, Bürger and Schneider 2006, Bürger et al. 2006, Bolnick and Fitzgerald 2007, Gavrilets et al. 2007, Gavrilets and Vose 2007) sympatric speciation is believed by many to be a rare process in nature, requiring special conditions and circumstances (Coyne and Price 2000, Coyne and Orr 2004, Gavrilets 2004, Bolnick and Fitzgerald 2007). An example of special conditions and circumstances is provided by seabirds. Like some insects (Tauber and Tauber 1989), a few have a relatively unvarying food supply, and this enables them to breed in discretely different seasons at the same place (Bourne 1957, Harris 1969a, Friesen et al. 2007). Coupled with this, some of them (petrels: procellarids) are incapable of relaying for several months if an egg is destroyed (Harris 1969b), and as a result failed breeders, after molting, are likely to return to breed out of synchrony with most of the population (see also Ashmole 1965).
Land birds lack this unusual combination of ecological opportunity and relaying constraint. Hence it is especially noteworthy that two possible cases of sympatric speciation in island land birds have been reported recently. The population of medium ground finches (G. fortis) on Santa Cruz Island in the Galápagos shows morphological signs of splitting into two through disruptive selection (Hendry et al. 2006), and nonrandom mating (Huber et al. 2007). The Tristan Nesospiza buntings display the molecular signature expected of a species that has already split into two, sympatrically, on two islands (Ryan et al. 2007).
As we have discussed in this chapter, all of the observations interpreted as evidence for sympatric speciation (model IV) can be explained alternatively in terms of an allopatric phase of divergence, followed by a sympatric phase with a reversal of divergence caused by introgressive hybridization (model III). Therefore the question arises, how can one choose between them? For the Darwin’s finch example, we consider the allopatric alternative to be more parsimonious because it is more strongly supported by observations of evolutionary processes. We suggest that on Santa Cruz Island there are essentially three and a half niches for granivorous finches; hybrids and backcrosses occupy the half, and this situation has persisted for at least a century and probably much more.
Where direct observation of processes is lacking, appeals to parsimony do not provide a clear answer. For example, after the Tristan da Cunha archipelago was colonized by buntings, there was either one additional island colonization and two speciations (sympatric model) or three island colonizations and one speciation (allopatric model). Since the probabilities of colonization and speciation are not known they cannot strictly be compared. Nevertheless, colonization (an event) seems to us to have a much higher likelihood of occurring than speciation (a long process), and on that basis alone the allopatric model has the stronger support.
Sympatric divergence due to selection and ecological and reproductive interactions may be identical under the two models. The crucial distinction between the models lies in the initiation of speciation. The allopatric model specifies geographical separation as the condition under which a population begins to split into two. It can be falsified with molecular data, although the scope for doing so is restricted. We attempted to falsify an allopatric speciation hypothesis for the evolution of Tristan buntings. Molecular data, showing that two islands were present when the initial split occurred, failed to reject it.
In contrast, the sympatric speciation model is difficult if not impossible to test and reject, as far as we can judge, even though some observations are not easily explained by it, e.g., the long waiting time to speciation on Nightingale. It therefore becomes a default model if there are grounds for rejecting an allopatric alternative. Where allopatric speciation cannot be rejected, we believe it is simpler than sympatric speciation because it does not have to confront the following difficulty. Disruptive selection has to be very strong to produce two morphological groups that are ecologically different enough to coexist. This can happen only if some degree of reproductive isolation allows their independent evolution. And yet mate choice of many passerine bird species is based on the learning of signals, and these must be different enough to isolate two groups reproductively. How the groups get to that point of “sufficient” difference is not clear because disruptive selection is not effective without some degree of reproductive isolation. This is the sympatric speciation dilemma.
Darwin’s finches on the island of Genovesa illustrate the dilemma. In 1978 two groups of large cactus finches (G. conirostris), recognizable by their different songs, differed in average beak size and diets. They appeared to be undergoing a split, sympatrically, into two feeding and breeding groups (Grant and Grant 1979). However, females in this population and related ones (Grant and Grant 2002b) learn both (or all) song types sung by males. In the next generation mating was random with respect to song. Incipient ecological and morphological divergence collapsed as a result of random mating (Grant and Grant 1989b).
We know of only one example of an escape from the sympatric speciation dilemma, and it involves a discrete, rather than a graded, shift in reproductive niche. African viduine finches parasitize the nest of other finch species. As nestlings the parasites learn the characteristics of the hosts and nests. As adults they use the songs learned from their hosts to court and mate at the nests of the hosts. When they switch hosts, as they have done in the past several times, they switch mating signals as well as locations, and in so doing become reproductively isolated from the rest of the population from which they were derived, at one stroke (Payne et al. 2002, Sorenson et al 2003, Price 2008). The success of the new population may depend on overcoming deleterious effects of close inbreeding.
In contrast to the field studies of two putative cases of sympatric speciation on islands, recent observations of medium ground finches on Daphne Major Island (figure 12.8) have been made close to the time of origin of nonrandom mating. An unusually large male and a large female immigrated at approximately the same time and bred with each other, as did their offspring, and their grand-offspring. The pairing pattern of the large immigrants and their offspring reflects a degree of reproductive isolation from the rest of the population unmatched, in our 35-year experience, by any other finch family: pairing among residents on Daphne is almost always random with respect to size traits (Grant and Grant 2008b). Two members of the pedigree have not been characterized genetically, and therefore the degree of reproductive isolation is uncertain. It could be complete.
Even if the reproductive isolation is transitory, it offers two insights into the important stage in speciation when two, differentiated, populations establish sympatry. First, reproductive isolation was apparently fostered by morphological divergence in allopatry. Size, especially beak size, undergoes evolutionary change through natural selection when feeding conditions change (Grant and Grant 2002a, 2008a); hence size-based reproductive isolation is a by-product of ecological divergence under natural selection (Dobzhansky 1937, Schluter 2000, Grant and Grant 2002b). An alternative possibility, that the immigrants and offspring bred with each other and not with the residents because they differed in song, can be ruled out. The immigrant G. fortis sang one of the song types prevalent (but rare) among Daphne residents (Grant and Grant 2008b).
Second, small numbers of colonists imply close inbreeding in the initial stages of the sympatric phase of speciation. Colonization of Daphne by two, assortatively mating, G. fortis individuals parallels the establishment of a breeding population of large ground finches (G. magnirostris) on the same island through immigration of five individuals in the 1980s (Grant et al. 2001). In both cases the population was started with a small number of founders and underwent close inbreeding in the next two generations.
We conclude that sympatric speciation in island birds is likely to be rare, dwarfed in importance by the allopatric alternative. Nevertheless, there are two additional forms of within-island speciation to consider. Speciation might occur allopatrically on a single island, but only if it is very large, like Madagascar or New Zealand (Diamond 1977). Alternatively it might occur parapatrically, that is, with partial spatial isolation; this is sometimes referred to as contiguous allopatry.
Parapatric speciation can be justified theoretically (Doebeli and Dieckmann 2003, Gavrilets 2004), and models have been developed to capture the essence of well-studied field examples of speciation in palms (Gavrilets and Vose 2007) and fish (Gavrilets et al. 2007). A requirement of the models is a small number of genetic loci with large effects on mate preferences. This makes them inapplicable to the numerous bird species whose mate choice is based on sexual imprinting and not on genetic variation (but see Saether et al. 2007). To be applicable to birds, cultural, nongenetic, influences on mate preferences need to be modeled (e.g., Laland 1994, Boyd and Richerson 2002, Ihara et al. 2003).
For island birds the starting condition could be partially isolated populations along an altitudinal, ecologically varying, gradient connected by limited dispersal and gene flow (Endler 1977, Gavrilets 2004). Spatial segregation (parapatry) could allow the evolution of small, site-specific, differences in ecology and morphology, and divergence in mate preferences based on sexual imprinting. Research is needed to determine if these small morphological differences could then be subsequently magnified, perhaps as a result of divergent coevolutionary dynamics with their foods (seeds, fruits), leading to reproductive isolation of the groups when they later invade each other’s ranges and become spatially intermingled. On Inaccessible Island, observations of altitudinal differentiation of N. acunhae bunting morphology in relation to variation in the habitat (Ryan et al. 1994, 2007) fit the parapatric speciation alternative. On Galápagos, a similar example has been found with the small ground finch (G. fuliginosa) on Santa Cruz Island (Kleindorfer et al. 2006). Consistent with incipient speciation, G. fortis on this island can discriminate between local songs and songs sung by birds only 11 km away (Podos 2007).
Nonetheless the question remains: is geographical differentiation within islands an evolutionary end point, or a stage toward completion of speciation marked by coexistence with little or no interbreeding? If within-island speciation initiated parapatrically on moderately large islands proves to be more than just feasible, but likely to occur, the fundamental relation in island biogeography will require a minor modification: at equilbrium, I (immigration) + W (within-island speciation) = E (extinction). If extinction is stochastic, species arising within an island should be just as likely to become extinct as those originating on another island, in which case the equilibrium should not be much affected by how it is reached. On the other hand, within-island speciation might be expected to affect (enhance) the rate of approach to the equilibrium. When calibrated by a measure of time, nonequilibrial communities should have more species than predicted from geography alone (see also Gillespie and Baldwin, this volume).
We thank Andrew Hendry, Sarah Huber, Jonathan Losos, Trevor Price, Bob Ricklefs, Peter Ryan, and an anonymous reviewer for their comments, advice, and suggestions.
Ashmole, N. P. 1965. Adaptive variation in the breeding regime of a tropical sea bird. Proceedings of the National Academy of Sciences U.S.A. 53:311–18.
Barluenga, M., K. N. Stölting, W. Salzburger, M. Muschick, and A. Meyer. 2006a. Sympatric speciation in crater lake cichlid fish. Nature 439:719–23.
Barluenga, M., K. N. Stölting, W. Salzburger, M. Muschick and A. Meyer. 2006b. Reply: Evidence for sympatric speciation? Nature 444:E13.
Barton, N. H., and M. Slatkin. 1986. A quasi-equilibrium theory of the distribution of rare alleles in a subdivided population. Heredity 56:409–15.
Bolnick, D. 2004. Waiting for sympatric speciation. Evolution 58:895–99.
Bolnick, D. I., and B. M. Fitzpatrick. 2007. Sympatric speciation: models and empirical evidence. Annual Reviews of Ecology, Evolution, and Systematics 38:459–87.
Bourne, W.R.P. 1957. Additional notes on the birds of the Cape Verde Islands, with particular reference to Bulweria mollis and Fregata magnificens. Ibis 99:182–90.
Bowman, R. I. 1983. The evolution of song in Darwin’s finches. In Patterns of Evolution in Galápagos Organisms, ed. R. I. Bowman, M. Berson, and A. E. Leviton, 237–537. San Francisco: American Association for the Advancement of Science, Pacific Division.
Bürger, R., and K. A. Schneider. 2006. Intraspecific competitive divergence and convergence under assortative mating. American Naturalist 167:190–205.
Bürger, R., K. A. Schneider, and M. Willensdorfer. 2006. The conditions for speciation through intraspecific competition. Evolution 60:2185–206.
Chan, K.M.A., and S. A. Levin. 2005. Prezygotic isolation and porous genomes: rapid introgression and maternally inherited DNA. Evolution 59:720–29.
Clarke, B., M. S. Johnson, and J. Murray. 1998. How ‘molecular leakage’ can mislead about island speciation. In Evolution on Islands, ed. P. R. Grant, 181–95. Oxford: Oxford University Press.
Coyne, J. A., and H. A. Orr. 2004. Speciation. Sunderland, MA: Sinauer Associates.
Coyne, J. A., and T. D. Price. 2000. Little evidence for sympatric speciation in island birds. Evolution 54:2166–71.
Diamond, J. M. 1977. Continental and insular speciation in Pacific island birds. Systematic Zoology 26:263–68.
Dieckmann, U., and M. Doebeli. 1999. On the origin of species by sympatric speciation. Nature 400:354–57.
Dieckmann, U., M. Doebeli, J. A. J. Metz, and D. Tautz. 2004. Adaptive Speciation. Cambridge: Cambridge University Press.
Dobzhansky, T. 1937. Genetics and the Origin of Species. New York: Columbia University Press.
Doebeli, M. 1996. A quantitative genetic competition model for sympatric speciation. Journal of Evolutionary Biology 9:893–909.
Doebeli, M., and U. Dieckmann. 2000. Evolutionary branching and sympatric speciation caused by different types of ecological interaction. American Naturalist 156 (suppl.):S77–S101.
———. 2003. Speciation along environmental gradients. Nature 421:259–64.
Elliott, H.F.I. 1957. A contribution to the ornithology of the Tristan da Cunha group. Ibis 99:545–86.
Fleischer, R. C., and C. E. McIntosh. 2001. Molecular systematics and biogeography of the Hawaiian avifauna. In Evolution, Ecology, Conservation, and Management of Hawaiian Birds: A Vanishing Avifauna, Studies in Avian Biology no. 22, ed. J. M. Scott, S. Conant, and C. van Riper III, 51–60. Lawrence, KS: Allen Press.
Ford, H. A., D. T. Parkin, and A. W. Ewing. 1973. Divergence and evolution in Darwin’s finches. Biological Journal of the Linnean Society 5:289–95.
Foster, D. J., J. Podos, and A. P. Hendry. 2008. A geometric morphometric appraisal of beak shape in Darwin’s finches. Journal of Evolutionary Biology 21:263–75.
Friesen, V. L., A. L. Smith, E. Gómez-Díaz, M. Bolton, R. W. Furness, J. González-Solis, and L. R. Monteiro. 2007. Sympatric speciation by allochrony in a seabird. Proceedings of the National Academy of Sciences U.S.A. 104:18589–94.
Gavrilets, S. 2004. Fitness Landscapes and the Origin of Species. Princeton, NJ: Princeton University Press.
Gibbs, H. L., and P. R. Grant. 1987. Ecological consequences of an exceptionally strong El Niño event on Darwin’s finches. Ecology 39:1735–41.
Gillespie, R. G. 2004. Community assembly through adaptive radiation in Hawaiian spiders. Science 303:356–59.
Grant, B. R., and P. R. Grant. 1979. Darwin’s finches: population variation and sympatric speciation. Proceedings of the National Academy of Sciences U.S.A. 76:2359–63.
———. 1989a. Evolutionary Dynamics of a Natural Population: The Large Cactus Finch of the Galápagos. Chicago: University of Chicago Press.
———. 1989b. Sympatric speciation in Darwin’s finches. In Speciation and Its Consequences, ed. D. Otte and J. A. Endler, 343–60. Sunderland, MA: Sinauer.
———. 2002b. Simulating secondary contact in allopatric speciation: an empirical test of premating isolation. Biological Journal of the Linnean Society 76:545–56.
Grant, P. R. 1968. Bill size, body size and the ecological adaptations of bird species to competitive situations on islands. Systematic Zoology 17:319–33.
———. 1986. Ecology and Evolution of Darwin’s Finches. Princeton, NJ: Princeton University Press.
———. 2001. Reconstructing the evolution of birds on islands: 100 years of research. Oikos 92:385–403.
Grant, P. R., I. Abbott, D. Schluter, R. L. Curry, and L. K. Abbott. 1985. Variation in the size and shape of Darwin’s finches. Biological Journal of the Linnean Society 25:1–39.
Grant, P. R., and B. R. Grant 1995. The founding of a new population of Darwin’s finches. Evolution 49:229–40.
———. 1996. Finch communities in a fluctuating environment. In Long-Term Studies of Vertebrate Communities, ed. M. L. Cody and J. A. Smallwood, 343–90. New York: Academic Press,.
———. 2002a. Unpredictable evolution in a 30-year study of Darwin’s finches. Science 296:707–11.
———. 2005. Species before speciation is complete. Annals of the Missouri Botanical Garden 93:94–102.
———. 2008a. How and Why Species Multiply. The Radiation of Darwin’s Finches. Princeton, NJ: Princeton University Press.
———. 2008b. Pedigrees, assortative mating and speciation in Darwin’s finches. Proceedings of the Royal Society of London, Series B 275:661–68.
Grant, P. R., B. R. Grant, J. A. Markert, L. F. Keller, and K. Petren. 2004. Convergent evolution of Darwin’s finches caused by introgressive hybridization and selection. Evolution 58:1588–99.
Grant, P. R., B. R. Grant, and K. Petren. 2000. The allopatric phase of speciation: the sharp-beaked ground finch (Geospiza difficilis) on the Galápagos islands. Biological Journal of the Linnean Society 69:287–317.
———. 2001. A population founded by a single pair of individuals: Establishment, expansion, and evolution. Genetica 112/113:359–82.
———. 2005. Hybridization in the recent past. American Naturalist 166:56–67.
Hagen, Y. 1952. Birds of Tristan da Cunha. Results of the Norwegian Scientific Expedition to Tristan da Cunha 1937–38. 20:1–248.
Harris, M. P. 1969a. The breeding seasons of sea-birds in the Galápagos Islands. Journal of Zoology, London 159:145–65.
———. 1969b. The biology of storm petrels in the Galápagos Islands. Proceedings of the California Academy of Sciences 37:95–166.
Heaney, L. R. 2000. Dynamic disequilibrium: A long-term, large-scale perspective on the equilibrium model of island biogeography. Global Ecology and Biogeography 9:59–74.
Hendry, A. P., P. R. Grant, B. R. Grant, H. A. Brewer, M. J. Brewer, and J. Podos. 2006. Possible human impacts on adaptive radiation: beak size bimodality in Darwin’s finches. Proceedings of the Royal Society of London, Ser. B 273:1887–94.
Hendry, A. P., S. K. Huber, L. De León, A. Herrel and J. Podos. 2009. Disruptive selection in a bimodal population of Darwin’s finches. Proceedings of the Royal Society of London, Ser. B 276:753–59.
Huber, S. K., L. F. De León, A. P. Hendry, E. Bermingham, and J. Podos. 2007. Reproductive isolation of sympatric morphs in a population of Darwin’s finches. Proceedings of the Royal Society of London, Series B 274:1709–14.
Huber, S. K., and J. Podos. 2006. Beak morphology and song features covary in a population of Darwin’s finches (Geospiza fortis). Biological Journal of the Linnean Society 88:489–98.
Ihara, Y., K. Aoki, and M. W. Feldman. 2003. Runaway sexual selection with paternal transmission of the male trait and gene-culture determination of the female preference. Theoretical Population Biology 63:53–62.
Kawecki, T. 1997. Sympatric speciation by habitat specialization driven by deleterious mutations. Evolution 51:1751–63.
Kleindorfer, S., T. W. Chapman, H. Winkler, and F. J. Sulloway. 2006. Adaptive divergence in contiguous populations of Darwin’s small ground finch (Geospiza fuliginosa). Evolutionary and Ecological Research 8:357–72.
Kondrashov, A. S., and F. A. Kondrashov. 1999. Interactions among quantitative traits in the course of sympatric speciation. Nature 400:351–54.
Lack, D. 1947. Darwin’s Finches. Cambridge: Cambridge University Press.
Laland, K. N. 1994. On the evolutionary consequences of sexual imprinting. Evolution 48:477–89.
Losos, J., and D. Schluter. 2000. Analysis of an evolutionary species-area relationship. Nature 408:84–50.
Lowe, P. R. 1923. Notes on some land birds of the Tristan da Cunha group collected by the Quest expedition. Ibis 5, series 11:519–28.
MacArthur, R. H., and E. O. Wilson. 1963. An equilibrium theory of island biogeography. Evolution 17:373–87.
———. 1967. The Theory of Island Biogeography. Princeton, NJ: Princeton University Press.
Maynard Smith, J. 1966. Sympatric speciation. American Naturalist 100:637–50.
Mayr, E. 1963. Animal species and evolution. Cambridge, MA: Belknap Press.
Mayr, E., and J. Diamond. 2001. The Birds of Melanesia. Cambridge, MA: Harvard University Press.
Newton, I. 2003. The Speciation and Biogeography of Birds. San Diego: Academic Press.
Payne, R. B., K. Hustler, R. Stjernstedt, K. M. Sefc, and M. D. Sorenson. 2002. Behavioural and genetic evidence of a recent population switch to a novel host species in brood-parasitic indigobirds Vidua chalybeata. Ibis 144:373–83.
Peters, J. L., Y. Zhuravlev, I. Fefelov, A. Logie, and K. E. Omland. 2007. Nuclear loci and coalescent methods support ancient hybridization as cause of mitochondrial paraphyly between gadwall and falcated duck (Anas spp.). Evolution 61:1992–2006.
Podos, J. 2007. Discrimination of geographical variants by Darwin’s Finches. Animal Behaviour 73:833–44.
Price, T. 2008. Speciation in Birds. Greenwood Village, CO: Roberts & Co.
Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945–59.
Rand, A. S. 1955. The origin of the land birds of Tristan da Cunha. Fieldiana Zoology 37:139–66.
Ratcliffe, L. M., and P. R. Grant 1983. Species recognition in Darwin’s Finches (Geospiza, Gould). I. Discrimination by morphological cues. Animal Behaviour 31:1139–43.
———. 1985. Species recognition in Darwin’s Finches (Geospiza, Gould). III. Male responses to playback of different song types, dialects aand heterospecific songs. Animal Behaviour 33:290–307.
Rice, W. R., and E. E. Hostert 1993. Laboratory experiments on speciation: What have we learned in 40 years? Evolution 47:1637–53.
Rosenzweig, M. L. 1995. Species Diversity in Space and Time. Cambridge: Cambridge University Press.
Roy, M. S., J. C. Torres-Mura, and F. Herel. 1998. Evolution and history of hummingbirds (Aves: Trochilidae) from the Juan Fernandez Islands, Chile. Ibis 140:265–73.
Ryan, P. G., P. Bloomer, C. L. Moloney, T. J. Grant, and W. Delport. 2007. Ecological speciation in south Atlantic island finches. Science 315:1420–23.
Ryan, P. G., C. L. Moloney, and J. Hudon. 1994. Color variation and hybridization among Nesospiza buntings on Inaccessible Island, Tristan da Cunha. Auk 111:314–27.
Saether, S.-A., G. P. Sætre, T. Borge, C. Wiley, N. Svedin, G. Andersson, T. Veen, J. Haavie, M. R. Servedio, S. Bures, M. Kral, M. B. Hjernquist, L. Gustafsson, J. Träff, and A. Qvarnström. 2007. Sex chromosome-linked species recognition and evolution of reproductive isolation in flycatchers. Science 318:95–97.
Savolainen, V., M.-C. Anstett, C. Lexer, I. Hutton, J. J. Clarkson, M. V. Borup, M. P. Powell, D. Springate, N. Salamin, and W. J. Baker. 2006. Sympatric speciation in palms on an oceanic island. Nature 441:210–13.
Savolainen, V., C. Lexer, M.-C. Anstett, I. Hutton, J. J. Clarkson, M. V. Borup, M. P. Powell, D. Springate, N. Salamin, and W. J. Baker. 2006. Replying to T. F. Steussy. Nature 443:E12–E13.
Schliewen, U., T. Kocher, K. R. McKaye, O. Seehausen, and D. Tautz. 2006. Evidence for sympatric speciation? Nature 444:E12–E13.
Schliewen, U. K., D. Tautz, and S. Päabo. 1994. Sympatric speciation suggested by monophyly of crater lake cichlids. Nature 368:629–32.
Schluter, D. 2000. The Ecology of Adaptive Radiation. Oxford: Oxford University Press.
———. 2001. Ecology and the origin of species. Trends in Ecology and Evolution 16:372–80.
Slatkin, M. 1975. Gene flow and the geographic structure of natural populations. Science 236:787–92.
Smith, T. B., R. K. Wayne, D. J. Girman, and M. W. Bruford. 1997. A role for ecotones in generating rainforest biodiversity. Science 276:1855–57.
Snedecor, G. W., and W. G. Cochran. 1989. Statistical Methods, 8th ed. Ames: Iowa State University Press.
Sorenson, M. D., K. M. Sefc, and R. B. Payne. 2003. Speciation by host switch in brood parasitic indigobirds. Nature 424:928–31.
Stuessy, T. T. 2006. Sympatric speciation in islands? Nature 443:E12.
Tauber, C. A., and M. J. Tauber. 1989. Sympatric speciation in insects: perception and perspective. In Speciation and its Consequences, ed. D. Otte and J. A. Endler, 307–44. Sunderland, MA: Sinauer Associates.
Tegelström, H., and H. P. Gelter 1990. Haldane’s rule and sex-biased gene flow between two hybridizing flycatcher species (Ficedula albicollis and F. hypoleuca, Aves: Muscicapidae). Evolution 44:2012–21.
Van Doorn, G. S., U. Dieckmann, and F. J. Weissing. 2004. Sympatric speciation by sexual selection: a critical reevaluation. American Naturalist 163:709–25.
Werner, T. K., and T. W. Sherry. 1987. Behavioral feeding specialization in Pinaroloxias inornata, the “Darwin’s Finch” of Cocos island, Costa Rica. Proceedings of the National Academy of Sciences U.S.A. 84:5506–10.
