ThorpeDiversificación Anolis Caribe

Evolutionary Diversication of Caribbean Anolis Lizards. In: Adaptive Speciation, eds.
Dieckmann U, Doebeli M, Metz JAJ & Tautz D, pp. 322344. Cambridge University Press 2004. c International Institute for Applied Systems Analysis
16
Evolutionary Diversication of Caribbean Anolis Lizards
16.1 Introduction
Jonathan B. Losos and Roger S. Thorpe The diversication of the lizard genus Anolis on Caribbean islands surely represents one of the best-studied cases of adaptive radiation in evolutionary biology. Over the course of the past four decades, researchers have studied almost every aspect of anole evolutionary ecology. These include systematics; community, physiological, and behavioral ecology; functional morphology; ethology; and demography. Studies have been conducted in the laboratory and in the eld, and have included basic natural history, geographic and temporal comparisons of populations, and a wide variety of experimental approaches to the study of phenotypic plasticity, ethology, ecology, and evolution [recent reviews include Losos (1994) and Roughgarden (1995)]. The result is an unusually broad and detailed understanding of the factors that promote and sustain evolutionary diversication and species coexistence. Speciation and adaptation in anoles Two conclusions from the current body of work are obvious. First, the genus Anolis has experienced extensive speciation. With more than 400 described species, and more being described every year, Anolis is the largest amniote genus, exceeded among tetrapods only by the potentially para- or polyphyletic frog genus Eleutherodactylus. The nearly 150 Caribbean species are descendants from as few as two initial colonizing species from the mainland (Jackman et al. 1997). Hence, the diversity of Caribbean species results almost entirely from speciation, rather than from repeated colonization. Second, adaptive diversication has been rampant. Within assemblages of anoles, species are clearly specialized to occupy different niches. Physiological and functional studies have revealed evidence for adaptation to particular microclimatic and structural habitats. Moreover, intraspecic comparisons indicate that populations adapt to their particular environmental conditions (reviewed in Malhotra and Thorpe 2000). Given the extensive adaptation and speciation exhibited by Caribbean anoles, one might wonder whether the two processes are linked. Can adaptive speciation the topic of this volume and the theory that as a population diverges under the pressure of disruptive selection, speciation ultimately ensues explain the
322
16 Evolutionary Diversication of Caribbean Anolis Lizards
323
evolutionary radiation of anoles? Alternatively, are adaptation and speciation in anoles largely unrelated, perhaps inuencing each other to some extent (speciation making adaptation more likely, or vice versa), but not necessarily linked? The anole radiations Anoles are small, insectivorous, and primarily arboreal lizards found throughout the Caribbean, Central America, northern South America, and the southeastern United States. Two morphological features characterize anoles. First is the possession of expanded subdigital toe pads. These toe pads are composed of laterally expanded scales, termed lamellae. Each lamella is covered with millions of microscopic hair-like structures, called setae. These setae allow anoles to adhere to smooth surfaces by the forces that form between electrons on the setae and electrons on the surface (Irschick et al. 1996 and references therein; see also Autumn et al. 2000). Very similar structures have evolved independently in two other groups of lizards, the geckos and prasinohaemid skinks (Williams and Peterson 1982). The second morphological structure that characterizes anoles is the presence of a gular throat fan, termed a dewlap. This structure, possessed by males of almost all species and females of many, is deployed in a variety of different contexts, including encounters with territorial rivals, potential mates, and predators. The color and pattern of the dewlap, as well as the specic pattern of head movements, is species specic and is used in species recognition (Rand and Williams 1970; Jenssen 1978; Losos 1985). The closest relative of Anolis is the genus Polychrus, which contains ve species in Central and South America (Frost and Etheridge 1989). Based on estimates from immunological studies, Anolis evolved at least 40 million years ago (Shochat and Dessauer 1981). Although the fossil record is sparse, several amber specimens from the Dominican Republic date to the Oligocene or Miocene (de Queiroz et al. 1998). Phylogenetic studies (reviewed in Jackman et al. 1997) indicate that Anolis originated in Central or South America and invaded the Caribbean twice. One lineage gave rise to the roquet group, which occupies the southern Lesser Antilles, whereas the second lineage gave rise to all other Caribbean anoles. In turn, the extensive radiation of the beta, or Norops, group on the mainland, composed of more than 150 species, appears to descend from a single colonist from the Caribbean back to the mainland. As many as 11 species of anoles can occur sympatrically in the Caribbean; assemblages nearly as large are known from the mainland. Comparisons of sympatric species indicate that species are almost invariably differentiated from sympatric congeners in some aspect of habitat use and either morphology or physiology, or both. On Caribbean islands, interspecic competition appears to be a potent force in the regulation of assemblage structure (reviewed in Losos 1994); on the mainland, both competition and predation may be important (Andrews 1976, 1979; Guyer 1988).
324
C Patterns of Speciation
The following two sections focus in turn on anole radiation in the Lesser Antilles (Section 16.2) and in the Greater Antilles (Section 16.3).
16.2
Adaptation and Speciation in Lesser Antillean Anoles
Roger S. Thorpe, Anita Malhotra, Andrew Stenson, and James T. Reardon The Lesser Antilles archipelago is composed of an older outer arc and a younger inner arc, with some islands (e.g., Martinique) being composed of elements of both (Figure 16.1). Younger islands tend to have a high elevation (e.g., Dominica), that in turn results in pronounced ecological (both biotic and physical) altitudinal and longitudinal zonation. These islands are occupied by two nonoverlapping series of anoles. The northern islands, down to and including Dominica, are occupied by anoles from the bimaculatus series and the southern islands, up to and including Martinique, are occupied by anoles from the roquet series (Figure 16.1; Underwood 1959). No major island is unoccupied and, naturally, an island has either one or two species. The far northern and southern islands tend to have two species, even if the islands are small and ecologically homogeneous, while central, large, ecologically heterogeneous islands have only a single species (Figure 16.1). As these islands may have been colonized from the northern and southern extremities (Gorman and Atkins 1969), the number of species that occupy an island may be related to the temporal opportunity for two successful colonizations rather than to ecological complexity of the habitat. In any event, these communities are characterized by less congeneric competition than the Greater Antillean anole communities, in which many species may live in sympatry (Section 16.3). While the number of sympatric species within a community is low, there are numerous allopatric species in this system. Here we attempt to elucidate the relative importance of adaptation in this system, the frequency and pattern of speciation, and the evidence for adaptive speciation. Evidence for adaptation by natural selection Evidence for the relative importance of adaptation in this system comes from three classes of study: Correlations between phenotype and habitat over space; Common-garden experiments; Natural selection experiments. Correlational evidence. Geographic variations in the phenotype of anoles from mountainous, single-species islands (Montserrat, Basse Terre, Dominica, Martinique, and St Lucia) have been investigated (Malhotra and Thorpe 1991a, 1997a, 1997b; Thorpe and Malhotra 1996). These islands tend to have rather similar ecological zonation with windswept littoral woodland on at least part of the Atlantic coast, xeric woodland on the Caribbean coast, and rainforest (giving way to eln woodland) at higher elevations. The anoles tend to adapt to these
325
Sombrero Anguilla Saba
A tl an ti c O
St Martin
St Barth
ce an
St Eustasius St Kitts Nevis

Redonda Montserrat Basse Terre Saintes
Anolis bimaculatus series Anolis roquet series
Barbuda Antigua
Grande Terre Desirade Petite Terre Marie Galante Dominica
Martinique
C ar ib b ea n Se a
St Lucia
St Vincent
Barbados
Saba
One spp. island
Grenadines Grenada
Nevis Two spp. island

Young chain Old chain
100 km
Figure 16.1 Distribution of the anole series in the Lesser Antilles, showing sympatry and simplied island age. The older chain arose primarily in the mid Eocene to mid Oligocene [about 4525 million years before present (MYBP)] with some activity in the late Oligocene to late Miocene (ca. 355 MYBP). The younger chain showed some activity in the Miocene, but arose primarily in the Pliocene (ca. 5 to 2 MYBP).
habitats (Figure 16.2) in their scalation, body dimensions, and color pattern (including human-invisible ultraviolet markings), which results in conspecic intergrading ecotypes (i.e., ecological races of the same species adapted to different, spatially segregated habitats, as distinct from ecomorph species see Section 16.3; Malhotra and Thorpe 1993b). The extent of this intraspecic geographic variation may be greater than the difference between allopatric species in some features. For example, there is a greater difference in body size between Caribbean lowland males and montane males within both Basse Terre and Dominica than there is between these island species (allowing for ecotype). Nevertheless, mitochondrial DNA (mtDNA) phylogenies for Dominica (Malhotra and Thorpe 2000) and Martinique (Thorpe and Stenson 2003) indicate that these ecotypes are not distinct lineages, although distinct mitochondrial lineages exist on these islands and partly correlate with color patterns.
326
Mon tane ecoty p
Canopy height (m)

ti c O
30
A tl an
North Caribbean form
300
J F MAM J J A S O ND
Rainfall
ce an
Canopy height (m)
30
Caribbean ecotypes
C ar
Atlantic ecotype Canopy height (m) 30

ea n
ib b
0 Rainfall
J F MAM J J A S O N D
Se a
0 Rainfall
J F MAM J J A S O N D South Caribbean form Seasonal Caribbean xeric woodland Evergreen Atlantic littoral woodland Transition between Evergreen rainforest and
10 km
Figure 16.2 Ecotypes of A. oculatus on Dominica. Open circles indicate seasonal Caribbean xeric woodland, lled circles indicate evergreen Atlantic littoral woodland, gray circles indicate a transition between these former vegetation zones, and squares indicate evergreen rainforest. The 300 m contour is indicated by the dashed curve. For the main vegetation zones, the canopy height (in meters) is illustrated diagrammatically, and relative monthly rainfall is shown by a histogram (the maximum monthly rainfall depicted is about 850 mm). Some aspects of the size, shape, and color pattern are indicated for the ecotypes associated with these vegetation zones. The southern and northern Caribbean forms live in a broadly similar habitat, but have some differences in appearance, perhaps for historical reasons (see Malhotra and Thorpe 2000).
The pattern of geographic variation may be tested against several putative causes, taking into account phylogenetic relationships (when available) and geographic proximity, using partial matrix correspondence (Mantel) tests (Box 16.1). When this is done, the hypothesis that the geographic variation in the phenotype is determined by adaptation to the physical and biotic conditions in these habitats cannot be rejected; in these adaptations the scalation tends to be adapted to the rainfall and/or moisture levels, the body size and shape tend to be adapted to altitude, and the color pattern tends to be adapted to vegetation (Table 16.1). It appears that natural selection has resulted in a close association between phenotype and habitat with the same trends being paralleled on each independent island.
327
Box 16.1 Testing causal hypotheses for patterns of geographic variation Matrix correspondence tests can be used to compare an observed pattern of variation (represented as a matrix of dissimilarities among localities) with a pattern derived from a causal hypothesis (also expressed as a dissimilarity matrix). If the measure of correspondence (e.g., regression or correlation) is not greater than, say, 95% of the correspondence values obtained when one of the matrices is randomized (e.g., 10 000 times), the null hypotheses of no correspondence may be accepted and the causal hypothesis rejected (Manly 1986a; Thorpe 1991). One complication is that alternative hypotheses may predict similar (correlated) patterns. With multiple hypotheses, this problem may be overcome by testing the hypotheses simultaneously using partial correlations (Smouse et al. 1986), or partial regressions (with the observed pattern as the dependent variable), as the measure of correspondence (Manly 1986b; Thorpe and Baez 1993; Thorpe and Malhotra 1996; Thorpe et al. 1996; Malhotra and Thorpe 2000). The correspondence between patterns of geographic variation in morphology with alternative patterns of climate and vegetational zonation may be tested to investigate the role of natural selection (Malhotra and Thorpe 1991a, 1997a, 1997b, 2000; Thorpe and Baez 1993; Thorpe and Malhotra 1996), and a molecular phylogeny (the matrix could represent the patristic distances or clade membership) may also be tested against patterns derived from alternative vicariance and/or phylogeographic hypotheses (Thorpe et al. 1995, 1996; Daltry et al. 1996; Thorpe 1996; Malhotra and Thorpe 2000). Moreover, factors such as spatial proximity and phylogenetic relatedness could also be included as independent variables to allow for them when assessing alternative causal hypotheses, such as the adaptation of morphology to ecological conditions (Thorpe et al. 1996; Malhotra and Thorpe 2000).
Table 16.1 Partial matrix correspondence (Mantel) tests show statistically signicant association between multivariate-generalized character systems in males and aspects of the environment (Box 16.1). Note the multiple parallels between independent mountainous island systems, in which scalation tends to adapt to rainfall and/or humidity, body dimensions tend to adapt to altitude, and color pattern tends to adapt to vegetation type. In both Dominica and Martinique, phylogeny is allowed for. Scalation Dominica Montserrat Basse Terre Martinique St Lucia Rainfall, vegetation Humidity Rainfall Rainfall Temperature Body dimensions Altitude, vegetation Altitude Altitude Vegetation Color pattern Vegetation Vegetation Vegetation
Common-garden experiments. However unlikely, it could be that the phenotype tracks habitat type through space entirely because of phenotypic plasticity, rather than natural selection (Section 16.3). To determine if the geographic pattern has a genetic component a common-garden experiment can be run in which individuals
328
Box 16.2 Common-garden experiments In studies of geographic variation, knowing the percent heritability of a phenotypic character is of little value, as it is a within-population measure and does not predict whether the geographic pattern is genetically determined (unless heritability is absolutely zero or 100%, which it never is). To test if the pattern of geographic variation in morphology of the Dominican anole results entirely from phenotypic plasticity, rather than from natural selection acting on characters whose variation is inuenced by genotype, Reardon and Thorpe (unpublished) ran a common-garden experiment. Pregnant females were sampled from 12 localities across the island and the resultant offspring were reared in a common garden situated in transitional woodland between northern Caribbean xeric woodland and montane forest. Insufcient offspring of both sexes were reared from three of the localities, but scalation and body dimension characters were recorded for the young adult offspring of the remaining nine localities. Their morphological characteristics were compared [using analysis of variance (ANOVA) and/or analysis of covariance (ANCOVA)]: Among localities, within offspring; Between offspring and adults previously sampled from the locality of origin. The multivariate divergence (Mahalanobis DM ) between common-garden-reared offspring from a given locality and a sample that occured naturally at that locality was compared. Finally, a canonical analysis scatter diagram that summarized the afnities of the geographic localities based on offspring was similar to a scatter diagram for specimens from the original localities. These all reject the hypothesis that the geographic variation arises from phenotypic plasticity (see text).
from different localities are hatched and raised under the same natural conditions in the eld (Garland and Adolph 1991). However, apart from the husbandry being excessively time consuming, these experiments can be problematic, because the more critically important natural selection is, the more likely it is that mortality will be high in individuals from habitats very different from the habitat at the site of the common garden. For example, an anole adapted by natural selection for a damp cool rainforest may fail to ourish and so die in a common garden sited on a hot dry coast. This results in low sample sizes for offspring for some localities, a reduced number of localities, and a reduced character suite when some characters (e.g., sexual characteristics of mature males) take a long time to develop. In spite of these difculties, a successful common-garden experiment was run in Dominica on A. oculatus, using body dimensions and scalation (Box 16.2), and established that: No signicant morphological differences occurred between experimentally reared offspring and adults from the original source locality. Signicant morphological differences occurred between offspring from different localities.
329
Box 16.3 Natural selection experiments The enclosures were constructed as in Malhotra and Thorpe (1991b, 1993a) and assiduously emptied of anoles. Specimens collected from control and test localities were anesthetized prior to individual labeling (toe clipping) and a multivariate prole was recorded from the scalation, body dimensions, and color pattern (the last via standardized high-quality Kodachrome macrophotographs). They were then released into the enclosures at the original biomass levels. At regular intervals, a census of the enclosures was taken to record those still alive. Selection intensity was estimated by the multivariate distance (Mahalanobis DM ) between surviving and nonsurviving groups within an enclosure (i.e., of the same ecotype). Selection coefcients were estimated by partial regression coefcients, in which the morphological characters were the independent variables and tness (number of census time periods survived) was the dependent variable (Brodie et al. 1995; Thorpe and Malhotra 1996).
The extent of morphological divergence between experimental offspring and anoles from the same locality did not depend on the extent of difference between the habitat at the common-garden site and the locality of origin. A multivariate scatter diagram that summarized geographic variation in the morphology of offspring showed the same pattern as one based on anoles from the original localities. This evidence strongly supports the proposal that the pattern of geographic variation is largely the product of natural selection working on characters that are, at least partially, genetically determined, rather than the pattern being a product of phenotypic plasticity. Natural selection experiments. Components of natural selection may be measured directly in eld experiments (Endler 1986; Brodie et al. 1995). To this end, four large-scale natural enclosures were established on the west coast of Dominica to investigate natural selection in A. oculatus in a series of chronologically and synchronously replicated translocation experiments. Following the methodology outlined in Box 16.3, an initial experiment was set up to test for a signicant selection intensity in montane, littoral woodland, south Caribbean, and north Caribbean ecotypes, with the last acting as the experimental control. After 60 days, a multiple analysis of variance (MANOVA) indicated signicant selection within ecotypes, and it was shown that, although the controls had no signicant selection acting on them, the translocated montane animals were subject to a signicant selection intensity (Malhotra and Thorpe 1991b; Thorpe and Malhotra 1992, 1996). Moreover, the intensity of the selection correlated signicantly to the extent of ecological difference between the locality of origin and the enclosure site. After more time had passed, signicant selection (as shown by selection intensity) was shown to be acting on the littoral woodland ecotype.
330
This was supported by a partial reciprocal experiment with enclosures in the littoral woodland, which showed a signicant selection intensity for the north xeric woodland ecotype, but no signicant selection for the littoral woodland control (Thorpe and Malhotra 1996). To investigate this further, critical aspects of the experiment were replicated chronologically by a pair of experiments with their own (internal) synchronous replication. This pair of experiments concentrated on montane translocates and north Caribbean controls, with each type in two enclosures. One experiment was run in the dry season and another in the wet season. A complete-block design MANOVA with interaction showed that, for both males and females, there was a signicant selection effect (a signicant difference between survivors and nonsurvivors) and that the intensity varied among ecotypes (a signicant interaction between survivornonsurvivor and ecotype). For both the dry season and the wet season experiments, and for both males and females, the accumulated selection intensity tended to be high and signicant for the translocated montanes and low and insignicant for xeric woodland controls. The targets of selection were also investigated by partial correlation (Box 16.3). Several characters were the targets of selection in the montane group and those characters with signicant selection coefcients were shown to distributed among all three character types (scalation, color pattern, and body size and shape). Finally, an experiment was run on 12 samples (replicated in the four enclosures) from a lowland-to-montane ecological gradient (using just females to maximize sample size). This investigated the relationship between the intensity of selection and the intensity of the difference between the ecological conditions at the site of the enclosures and the site of origin. This study showed a strong correlation between the two, irrespective of the time at which selection intensity was measured. This supports the contention that the lizards are adapting, by natural selection, to the spatially various ecological conditions. This series of replicated experiments consistently indicates that natural selection is acting on A. oculatus and that the geographic variation in this anole is likely to result from adaptation to the geographic variation in physical and biotic conditions within Dominica. The correlations between habitat and morphology (irrespective of phylogenetic history), the observed parallels in these correlations on comparable islands, the common-garden experiments, and the series of natural selection experiments, taken together, provide overwhelming evidence that adaptation is a very inuential and important evolutionary factor in this system. Evidence for speciation In a sympatry-dominated system, such as the Greater Antillean anoles, species recognition is quite straightforward, but it is arbitrary when forms are allopatric, and interbreeding is untestable. Consequently, it has to be acknowledged that in an allopatry-dominated system, such as the Lesser Antillean anoles, to assess the extent of speciation will have an arbitrary element. This is illustrated by a series of
331
Box 16.4 Assessing gene-ow patterns using microsatellites A series of ve polymorphic microsatellites (Stenson et al. 2000) were screened for up to 50 specimens from each of 33 localities (as in Figure 16.2) across Dominica (Malhotra and Thorpe 2000). The gene ow between pairs of adjacent sites (adjusted for differential distance among sites by regression) was plotted on a minimum-contact, minimum-distance connectivity network that joined adjacent sites (Stenson et al. 2002). This is visualized by contouring these values to give a nonprobabilistic representation of the geographic pattern of reduction in gene ow (Figure 16.4). This shows that along the Caribbean coast a sharp reduction in gene ow is found between northern and southern types.
taxonomic changes. The species status of the two northern groups (one comprises A. wattsi, pogus, foressti, and schwartzi, and the other comprises A. bimaculatus, gingivinus, and leachi), as well as the southern group comprising A. griseus and richardi, have all changed, with more allopatric forms recognized as separate species (Schwartz and Henderson 1991; Malhotra and Thorpe 1999). Patterns of between-island speciation. Molecular phylogenetic analysis based on mtDNA and supported by nuclear microsatellites established the relationships within the A. bimaculatus series and were used to hypothesize the colonization sequence and spaciotemporal pattern of speciation, following the procedures in Thorpe et al. (1994). Early on in the history of the group this shows a step-wise pattern of colonization of adjacent islands (or island banks) in a general northto-south, old-to-young progression (Figure 16.3). The one interesting exception to this is the very recent colonization of Saba from southwest Basse Terre in a long-range, southeast-to-northwest track shown by many hurricanes. Similar molecular phylogenies are available for the southern A. roquet series (Giannasi et al. 2000) and are compatible with the concept of speciation by early founder colonization between major islands and island banks. There is therefore a substantial amount of between-island speciation in this model. Within-island speciation. In Dominica, the geographic variation in the morphology of A. oculatus generally shows all the indications that would result from adaptation by natural selection, such as incongruence in the patterns of geographic variation of different characters, gradual character change correlated to changes in environmental variables, and little association with the molecular phylogenetic history (Malhotra and Thorpe 2000). However, one exception to this is the change between the north and south Caribbean coastal forms (Malhotra and Thorpe 1994). Here the changes are very sharp and congruent, between both various types of morphological character and mtDNA lineage boundaries. Moreover, the sharp transition is not closely associated with sharp environmental changes, or extrinsic barriers to gene ow. Is this an example of partial reproductive isolation within an island? To test the extent of reproductive isolation, gene ow among populations
332
Anguilla
St Martin
St Barth Saba
Barbuda St Eustasius St Kitts Nevis

Montserrat Basse Terre
Antigua
Grande Terre Desirade

Petite Terre
Saintes
Marie Galante Dominica
50 km
Figure 16.3 Hypothesized colonization pathways of the large-bodied anoles from the A. bimaculatus series. Ovals enclose island banks (or lineages within which colonization direction is equivocal) and arrows indicate the main colonization events associated with speciation. For anoles of the bimaculatus series, around 1200 base pairs from cytochrome b and cytochrome oxidase I were sequenced and a gene tree reconstructed using maximum parsimony and maximum likelihood. Several key aspects of the mitochondrial tree were corroborated by nuclear genes (microsatellites). The small-bodied A. wattsi, pogus, foressti, and schwartzi group was a separate lineage to the large-bodied A. bimaculatus, gingivinus, and leachi group. Taking the phylogeny of this latter group, it can be converted into a colonization sequence following the branch-length procedure of Thorpe et al. (1994), whereby a node is allocated a geographic locality based on the shortest patristic distance to the terminal node (geographic population). Nodes are then joined in phylogenetic sequence. This procedure is sensitive to the branch lengths of the tree and so only colonization steps common to both maximum parsimony and maximum likelihood are shown here. This procedure suggests a geographic sequential colonization in this anole series, with the speciation events associated with relatively early colonization of major islands or island banks, while later colonizations of islands within the same bank (e.g., within the St Martin, St Kitts, and Antigua banks) or colonization of adjacent islets (e.g., Petite Terre) are not associated with speciation. The important exception to this is the late, long-range colonization of Saba (from southwest Basse Terre) and speciation of A. sabanus after the colonization of much of the remaining Guadeloupe archipelago (Desirade, Petite Terre, Grande Terre, and Basse Terre) without speciation. The Saintes and Dominica have separate species, but the colonization sequence is equivocal, as is the colonization of Redonda.
333
10 km
40 + 20 to 40 0 to 20 20 to 0 40 to 20 Below 40
Figure 16.4 Geographic variation in gene ow between adjacent localities in Dominica based on A. oculatus microsatellites. Residual values from the regression of linearized FST (between adjacent sites) against log( z ), where z is the geographic distance between adjacent sites, were multiplied by 1000. The darker the area, the lower the gene ow. Based on 33 localities, as in Figure 16.2.
was measured using microsatellites and contoured across space (Box 16.4; Stenson et al. 2002). This shows, that while there is a very sharp reduction in gene ow at this point and some reproductive isolation between these forms along the coast, genes are exchanged via the inland populations at much the same rate as between contiguous population across the rest of the island (Figure 16.4). Consequently, this example does not provide convincing evidence of complete within-island speciation. One possible scenario is that volcanic activity has eradicated populations from some areas temporarily and that this coastal transition zone marks a point of secondary contact (Malhotra and Thorpe 2000). Factors may yet operate that will eventually result in speciation (e.g., there may be assortative mating at this coastal transition, and this warrants further investigation). In Martinique, the molecular (mtDNA) phylogeny of Anolis roquet indicates that major phylogenetic divisions (about 11% uncorrected base pair differences) occur within the current island (Thorpe and Stenson 2003). Matrix correspondence tests and geological information indicate that this pattern most likely arises from several peripheral precursor islands (southern, southwestern, and northwestern) that formed Martinique when a central precursor island emerged and joined with each of the peripheral precursors (Thorpe and Stenson 2003). The phylogeographic pattern suggests secondary contact of the populations on these precursor islands, where at least the mtDNA has diverged in allopatry. This has resulted in three major secondary contact zones within the island marked by the distribution of the distinct mtDNA lineages. Moreover, there is distinct habitat zonation on the island, and the multivariate morphology (Giannasi 1997), color pattern, and hue relate largely to this zonation (Thorpe and Stenson 2003). Some of these forms are so divergent that there is an indication of multivariate bimodality at the point
334
where at least one pair of parapatric forms meet (Thorpe and Stenson 2003), and a signicant reduction in microsatellite gene ow between several pairs of parapatric forms (Ogden and Thorpe 2002). Moreover, the Martinique form is parapatric, as another full species, A. extremus from Barbados, is nested inside the Martinique complex (Thorpe and Stenson 2003). Hence, there is evidence of incipient speciation and some parapatric forms may warrant recognition as full species. However, given the occurrence of several potentially important factors (ecological zonation and past vicariance), and the difference between a signicant reduction in gene ow and complete speciation, more research needs to be undertaken before it can be suggested that full parapatric speciation has taken place in situ within this island. The evidence is overwhelming for numerous speciation events generally associated with early colonization. Anoles appear to speciate readily in geographic isolation. While there is clear evidence of strong morphological and molecular divergence within Lesser Antillean islands and some evidence of parapatric forms being incipent species, there is no adequate evidence for complete speciation in situ within islands at this stage. These studies on the differentiation between incipient speciation and speciation within islands are ongoing. Has adaptive speciation occurred? Clearly, both widespread adaptation and speciation occur within this Lesser Antillean anole system, but does adaptive speciation? This is not an easy question to answer, as it can be difcult to know the rate of some factors, irrespective of which model of speciation is suspected. For example, interpretation of the molecular phylogeny can suggest the timing and geographic origin of the Saba species, and its morphological differentiation can be measured, but what the relative roles of chance and adaptation were in this speciation event is not known. This is because the process is unique, as are so many of the facets of speciation events. An area in which repetition may allow the cause to be tested is in the evolution of body size. It has long been known that on islands with only one lizard the body size tends to be intermediate, whereas on islands with two species, one species tends to be large and the other small (Schoener 1969, 1970; Plate 9). This may result from size assortment (i.e., the tendency for only species of dissimilar sizes to colonize and coexist on the same island because of competitive exclusion or interbreeding), or character displacement in which, after colonization of an island by precursors of similar size, they evolve into sympatric species of different sizes by character displacement to minimize the competition for resources (Losos 1990a; Giannasi et al. 2000). Ancestral character reconstruction methods (Maddison and Maddison 1992) allied to molecular phylogenies allow these two hypotheses to be tested in the southern A. roquet series. These tests show that an element of character displacement is likely, with size assortment playing a role (Giannasi et al. 2000).
335
To what extent does this interisland study of size offer evidence for the role of adaptation in speciation? In the north, the larger species are of the bimaculatus group and the smaller belong to the wattsi group. These are distinct lineages (Stenson 2000), so within a given island the two species will not be sister taxa. Hence there is no question of a common ancestor split by adaptive speciation. In the south, the phylogenetic relationships of the roquet group are resolved only partially (see Thorpe and Stenson 2003 and references therein), so it is not always possible to know if a pair of species on an island (Plate 9) are sister taxa that could have speciated in situ by adaptive speciation. Work on the phylogeny and colonization sequence of this group is ongoing. Within islands there is evidence of parapatric forms with restricted gene ow among them on both Dominica (Stenson et al. 2002) and Martinique (Ogden and Thorpe 2002). The latter case was interpreted as evidence in favor of the process of ecological speciation, as the forms are dened by patterns of current habitat rather than past allopatry. Adaptive speciation may occur on an ecological gradient (Doebeli and Dieckmann 2003; Chapter 7), but as pointed out in Chapter 1, such a parapatric situation may, or may not, involve adaptive speciation. Consequently, whether the Martinique (Ogden and Thorpe 2002), and other (Thorpe and Richard 2001) lizard cases represent examples of adaptive speciation (Tautz 2003) remains to be investigated fully.
16.3
Adaptation and Speciation in Greater Antillean Anoles
Jonathan B. Losos This section rst reviews the evidence that both speciation and adaptation played important roles in anole diversication, focusing primarily on the anoles of the Greater Antilles (Cuba, Hispaniola, Jamaica, and Puerto Rico). It then addresses the extent to which the two processes are intimately linked. The theory of adaptive speciation presents one mechanism by which the two processes might occur in an integrated fashion, but there are other possibilities. In part this requires a discussion of what constitutes a species of Anolis, so that it is possible to recognize when two lineages have diverged to the level of different species. Finally, this section addresses some exciting new developments that affect our understanding of the processes important in anole adaptive radiation. Evidence for speciation The extent of speciation in Caribbean anoles can be viewed at several levels. Most generally, the number of Caribbean anole species indicates the high rates of speciation that have occurred. Anolis is the most species-rich genus of vertebrate in the Caribbean and phylogenetic studies indicate that Caribbean anoles are the descendants of only two colonizations from mainland taxa (Jackman et al. 1999). Hence, the Caribbean anole radiation has resulted almost entirely from in situ speciation, rather than from repeated colonization. Examination of anole phylogeny (Figure 16.5), however, can tell us considerably more about patterns of speciation.
336
(a)
(b)
Jamaica (6)* Mainland Cuba (15) Hispaniola (1) Cuba (4) Cuba (4)* Cuba(4) Cuba (14) Hispaniola (8) Puerto Rico (8)* Hispaniola (6) NLA Hispaniola (7) Puerto Rico (1)* Cuba (5) Cuba (2) Hispaniola (1) Hispaniola (6) Hispaniola (7) Cuba (2) Cuba (6) SLA & mainland Puerto Rico (1) Mainland
A tl
Bahamas
500 km
ti c O ce an
an
Cuba
C ar ib b ea n Se a
Hispaniola Jamaica Puerto Rico
Lesser Antilles
Figure 16.5 (a) Map of the Caribbean. (b) Phylogeny of Anolis based on mtDNA (Jackman et al. 1999). Relatively few interisland movements are required to account for the more than 100 species of anoles on the Greater Antilles. Numbers in parentheses indicate the number of species within each terminal clade. Monophyly of these terminal clades is based on traditional morphological data. Molecular phylogenetic analyses using DNA indicated that the placement of species into species groups based on morphology is nearly always correct (e.g., Jackman et al. 1999; Schneider et al. 2001; Jackman et al. 2002). Species from small islands in the Caribbean (e.g., A. acutus on St Croix, A. conspersus on Grand Cayman) occur within terminal clades (designated with an asterisk) and are not indicated. In all cases, these species are apparently derived from clades on one of the Greater Antilles [e.g., A. conspersus from Jamaica, A. acutus from Puerto Rico (Jackman et al. 2002)].
Many islands in the Lesser Antilles and elsewhere in the Caribbean, such as the Bahamas and the Virgin Islands, contain more than one anole species. However, these co-occurring species are almost invariably distantly related, which suggests that the ancestors of these species independently colonized the island, and are not the result of a speciation event that occurred on that island. By contrast, in the Greater Antilles, the situation is quite different. Many monophyletic clades are found only on a single island, which indicates that substantial speciation has occurred within an island. On Puerto Rico, for example, eight of the ten species are members of the cristatellus group, whereas in Jamaica, six of the seven species belong to the grahami group. Similarly, the alutaceus group of grass anoles, found only on Cuba, contains 14 species. In each case, the most parsimonious explanation is that the ancestor of the group occurred on the island in question and then, by a series of divergent speciation events, the single ancestral species gave rise to numerous descendants. Indeed, anole phylogeny suggests that as few as 11 interisland colonization events are necessary to explain the distribution of the 110 or more Greater Antillean species (Jackman et al. 1997); thus, all remaining species are the result of within-island diversication (Losos and Schluter 2000).
337
Evidence for adaptation Several lines of evidence indicate the great extent of anole adaptive diversity. First, convergent habitat specialization is widespread in the Greater Antilles. Williams (1972, 1983) noted, for anole assemblages within the Greater Antilles, that species exhibit specializations to using different parts of the structural habitat. Thus, for example, twig species are streamlined with extremely short limbs and tails and trunkground species (which perch low on the trunk and frequently descend to the ground) are stocky with extremely long hind limbs and poorly developed toe pads. The similarities even extend to patterns of sexual size dimorphism: trunkground anoles, for example, are always highly dimorphic, whereas twig anoles exhibit little dimorphism (Butler et al. 2000). In all, six types of habitat specialists have been recognized. Such a pattern of assemblage similarity could result either because each specialist evolved once only and made its way to each of the islands, or because highly similar assemblages evolved independently on each island. Phylogenetic studies make clear that the latter explanation is correct (Losos et al. 1998). Habitat specialists on different islands are almost never closely related; hence, these similar assemblages are the result of convergent evolution. Patterns of divergent evolution also provide evidence of adaptation. For example, correlations exist between habitat use and morphology within both Lesser (Section 16.2) and Greater Antillean anoles (Lister 1976; Losos et al. 1994). Among populations of A. sagrei, for instance, a correlation exists between mean perch height and mean number of toe-pad lamellae (Lister 1976). Similar examples are presented from Lesser Antillean taxa (Section 16.2) Convergent evolution and correlations between morphology and environment have long been taken as evidence of adaptation (reviewed in Harvey and Pagel 1991; Losos and Miles 1994). Nonetheless, to understand fully the adaptive basis of character evolution, we need to understand why particular traits evolved in particular environments. Doing so requires an understanding of the functional consequences of character variation, as well as detailed information of what organisms actually do in nature (Lauder 1981; Arnold 1983; Greene 1986; Wainwright 1988; Losos 1990b; Arnold 1994). In the case of anoles, laboratory functional studies have revealed how variations in limb length and toe-pad dimensions affect the ability to run, jump, and cling (Irschick et al. 1996; Irschick and Losos 1999, and references therein). As might be expected, species with longer hind limbs have greater sprinting and jumping abilities, whereas toe-pad dimensions are related to clinging abilities. Moreover, functional capabilities are not independent of the environment. Long-legged lizards, for example, can run faster than short-legged species on broad surfaces, but not on narrow ones (Losos and Sinervo 1989; Losos and Irschick 1996). In turn, eld studies have revealed some of the contexts in which maximal performance is important. Most species run at maximal speed to escape predators, but not to capture prey (Irschick and Losos 1998). Moreover, species that run quickly
338
tend primarily to use habitats in which they can run at top speed (i.e., broad surfaces), whereas slow species, which rely more on crypsis and slow locomotion, are less constrained in their habitat use (Irschick and Losos 1999). In a similar vein, physiological studies have established how species adapted to different thermal environments by altering their physiological capabilities. Thus, species that use cooler, forest habitats thermoregulate at lower temperatures, whereas species in hot, open habitats maintain higher temperatures. Further, research on Caribbean and mainland species has established that peak functional capabilities correspond to preferred temperatures (e.g., Hertz 1979, 1992; Hertz et al. 1979; Huey 1983; Van Berkum 1986). Thus, although many questions remain unanswered, we do have a solid understanding of the functional consequences of observed variation in traits. This understanding supports the conclusion, based on convergence and correlation, that the extensive variation in traits such as limb length, toe-pad dimensions, and thermal physiology observed among anole species represents adaptation to living in different parts of the environment. Has adaptive speciation occurred? The observations that sympatric anoles almost always differ in habitat use, with attendant morphological or physiological adaptations, and that most speciation has occurred within islands in the Greater Antilles suggest that adaptive speciation has played an important role in anole adaptive radiation (Dieckmann and Doebeli 1999). Consider Jamaica, for example. Five species are widespread throughout the island, and often occur sympatrically. All ve descended from a common ancestor on Jamaica, and each has adapted to its own specic niche. One ready explanation is that this diversity arose via adaptive speciation, as disruptive selection led rst to polymorphism within a single ancestral species and then, as reproductive isolation evolved, to speciation and the production of an assemblage of species adapted to different parts of the environment. Indeed, given that the species now occur island wide, that it is not clear that the island was ever divided into separate parts by climate change or sea-level rise, and that there are few peripherally isolated offshore islands (Lazell 1996), a scenario of allopatric speciation is in some ways distinctly less parsimonious. Although opportunities for allopatric speciation are greater on the other Greater Antillean islands, which are larger and more varied topographically, the argument for adaptive, sympatric speciation still remains. However, before the process of speciation can be discussed meaningfully, an understanding of what constitutes a species is necessary. Thus, before the evidence relative to modes of speciation in anoles is examined, rst a digression is needed to discuss what denes an anole species and how one recognizes whether different taxa belong to the same species. Concepts of anole species and speciation. Recent years have seen much criticism of the biological species concept and a proliferation of other ideas about what a species is [e.g., Howard and Berlocher (1998), and references therein]. Nonetheless, the vast majority of sympatric anoles coexist without interbreeding. Indeed, despite 40 years of extensive eldwork in the Caribbean, hybridization has only
339
Table 16.2 Reported cases of hybridization in Caribbean Anolis. Species A. aeneus A. trinitatis A. allisoni A. porcatus A. brevirostris A. distichus Island Trinidad Cuba Notes References
Infertile; species introduced to Gorman and Atkins 1968; island, not naturally sympatric Gorman et al. 1971 Hybridization suggested by morphological intermediacy Ruibal and Williams 1961 Webster 1977
Hispaniola Abnormal meiosis, thus probably infertile; dewlaps very similar; hybridization possibly frequent Hispaniola Will intermate, but no documented hybrids; dewlaps different color
A. caudalis A. websteri
Jenssen 1996
A. chlorocyanus Hispaniola Hybridization suggested by the Garcia et al. 1994 A. coelestinus presence of morphologically intermediate forms and interspecic courtship in staged trials A. grahami A. lineatopus Jamaica One individual; probably infertile because of chromosomal irregularities Jenssen 1977
been suggested between six pairs of Caribbean species, most at a single locality only (Table 16.2). Further, the evidence in several of these cases is far from compelling. Thus, as a general rule, Caribbean anoles are characterized by premating reproductive isolation. Fortunately, we have a good idea of what factors promote reproductive isolation in anoles. Anoles are visually oriented animals and communicate by visual displays that involve vertical movements of the head, termed head bobs, and by extending their colorful dewlaps. Several lines of evidence indicate that anoles are able to distinguish conspecics from nonconspecics by the form of their bobbing display and by the appearance of the dewlap. With regard to the head-bobbing patterns, each species has its own stereotyped pattern of movements, which differ, among species, in rhythm and amplitude (Jenssen 1978). A video playback experiment with one Central American species indicated that female A. nebulosus can discriminate males that perform their normal, unaltered displays from males whose displays have been experimentally altered (Jenssen 1970; see also Macedonia et al. 1993). With regard to dewlaps, it is notable that sympatric species always differ in the size, color, or patterning of their dewlaps (Plate 10; Rand and Williams 1970; Williams and Rand 1977; Losos and Chu 1998). In itself, this decidedly nonrandom pattern suggests that the dewlap is used as a species-recognition signal. Moreover, an experimental study with a pair of sympatric species indicates that
340
anoles use the color of the dewlap as a cue to determine how to respond to another male (Losos 1985; see also Macedonia and Stamps 1994). This understanding of reproductive isolating mechanisms [or specic materecognition systems (Paterson 1982)] is important in two regards. First, it provides a mechanistic understanding of what goes on during speciation; before two species can coexist in sympatry, they apparently must evolve differences in these systems. Second, it permits an objective means of evaluating whether allopatric populations have differentiated to the extent that they would interact as distinct species were they to become sympatric: if the populations differ in dewlap color or head-bobbing pattern, they constitute distinct species. Of course, the converse is not always true. It is possible that allopatric populations may evolve postzygotic reproductive isolation and thus be isolated reproductively, even in the absence of premating isolating mechanisms. Thus, this method for determining specic status may, in some cases, fail to recognize populations that already constitute different biological species, a situation that may exist in some Lesser Antillean taxa (Section 16.2). Evidence for adaptive speciation. Given this understanding of what constitutes speciation in anoles, under what conditions has speciation occurred? The adaptive speciation hypothesis suggests that speciation was sympatric and that speciation and adaptation are related intimately, as discussed elsewhere in this volume (Chapter 1). Despite years of work on the evolutionary ecology of Caribbean anoles, little direct evidence exists to determine the geographic context in which speciation has occurred. However, data do exist from which the hypothesis that speciation and adaptation are related can be evaluated. Although far from conclusive, these data indicate that no necessary link exists between these two processes. This argument has two lines of evidence. First, many examples indicate that adaptation can occur in the absence of speciation in Caribbean anoles. As discussed above, comparisons indicate that populations of the same species adapt to different environments in both the Lesser and Greater Antilles. Although detailed phylogeographic studies have been conducted for only a few of these species, such studies conrm that populations of several of these species are part of a single species, rather than distinct but unrecognized species (see Section 16.2). Thus, adaptive evolution can occur in the absence of speciation. Conversely, speciation can occur, but it produces species that are adaptively undifferentiated. Evidence for this claim comes from complexes of allopatrically distributed species on the islands of Cuba and Hispaniola. On Cuba, for example, the monophyletic alutaceus group of grass anoles contains 14 species. Two of these species are found island wide, but the other 12 have narrow distributions, mostly centered on different mountain ranges. For the most part, these species are little differentiated, both ecologically and morphologically. However, differences in their dewlaps suggest that the species are isolated reproductively and, thus, valid species. Similar complexes of mostly allopatric species are seen in the sagrei group in Cuba (14 species) and the cybotes group in Hispaniola (eight species), as well as in a number of smaller complexes. This phenomenon reveals
341
both the efcacy of allopatric speciation and the possibility that speciation can occur without substantial adaptive divergence (it could, of course, be suggested that these populations initially arose in sympatry and only later attained allopatry, but given the small and disjunct ranges often centered on mountain ranges of many of these species, such an explanation seems decidedly unlikely). Adaptation can occur without speciation, and speciation can occur without adaptation. Does this indicate that adaptive, sympatric speciation does not occur? Of course not. No one would claim that adaptive speciation must account for all of the speciation and adaptation evident in the anole radiation. Moreover, although adaptive differentiation does occur intraspecically, the amount of divergence is relatively limited and does not approach the extent of differentiation between different habitat specialists, such as the difference between twig and trunkground specialists. All that can be said is that no evidence exists to support the occurrence of adaptive speciation, and alternative mechanisms have been demonstrated that can produce adaptation and speciation in other ways. An alternative hypothesis is that species arose in allopatry and perhaps differentiated adaptively to some extent. Then, when they came back into sympatry, interspecic interactions most likely competition lead to habitat partitioning and subsequent adaptation and specialization to different habitats. Such a scenario of character displacement is a standard view of the manner in which adaptive radiation proceeds (Grant 1986; Schluter 2000). Much experimental and observational data indicate that anole species do alter their habitat use in the presence of congeners and some data suggest that this leads to morphological change (reviewed in Losos 1994). However, the extent of habitat shifts and morphological change are minor relative to the differences between different habitat specialists. Thus, further investigation is required of both adaptive speciation and character displacement hypotheses for the adaptive radiation of anoles. An alternative link between adaptation and speciation. Adaptation and speciation may be linked directly in a manner different from that envisioned in the adaptive speciation hypothesis. As discussed above, speciation in anoles results when changes occur in the species-recognition systems, the dewlap and head-bobbing patterns. These systems may evolve, in turn, when populations adapt to new habitats. The reason is that these signals, being visual, must be seen to be effective. Consequently, if a population for some reason occupies a new habitat in which the visual environment is different, natural selection may favor evolutionary change in the signal to maximize the effectiveness of communication, both intra- and interspecically [this is the sensory drive hypothesis of Endler (1992); see Fleishman (1992, 2000)]. For example, the most visible color for a dewlap varies depending on the environment. In closed forests, light levels are reduced and the light that does penetrate is mostly in the yellow and green parts of the spectrum. As a result, white or yellow dewlaps are optimal because they reect the little light that is available. By contrast, in open areas there are few limits in the light availability and the most effective dewlaps are nonreective and dark, producing a darkness contrast to the
342
bright background (Fleishman 1992, 2000). Indeed, in a survey of Caribbean anole dewlap colors, Fleishman (1992, 2000) found that most closed-forest anole species have yellow or white dewlaps, whereas the majority of open-habitat species have orange, red, black, or blue dewlaps. Thus, if an open-habitat species for some reason moved into a closed habitat [or perhaps the open area itself is transformed and the populations stay put, akin to Vanzolini and Williams (1981) vanishing refugium hypothesis], selection would favor change in dewlap color to maximize the ability to communicate intraspecically. An indirect by-product of this adaptation, however, may be that the population becomes reproductively isolated from other populations remaining in the original habitat. Thus, adaptation to a new visual environment may lead to speciation. This scenario could operate whether the populations are initially sympatric or allopatric; it could easily result as allopatric populations differentiate in different habitats, but it also might be the result of adaptive speciation in sympatry. Similar arguments may apply to head-bobbing patterns. To be detected, the movements of a displaying lizard must differ from movements of the background vegetation. In habitats in which there is not much background vegetation, or in which the vegetation tends not to move much, small amplitude displays may be conspicuous. By contrast, when background vegetation moves greatly, more pronounced displays may be needed to catch the attention of other lizards (Fleishman 1992). As with dewlap color, adaptive change in head-bobbing patterns thus could produce speciation. In addition, new habitats may select for other features as well as those used in communication, because the structural and thermal environments may differ also. Consequently, adaptations in a variety of different traits may all be linked to speciation because of the relationship between communication, light environments, and specic mate-recognition systems. This hypothesis is not easy to test, but important advances in our understanding of anole visual ecology and physiology are being made by Fleishman and colleagues (Fleishman et al. 1997; Persons et al. 1999; Leal and Fleishman 2002). Phenotypic plasticity and anole adaptive radiation. Recent work has suggested an intriguing new hypothesis: phenotypic plasticity might play an important role in anole adaptation and adaptive radiation. The background for this work was a pair of studies on populations of A. sagrei in the Bahamas. A comparison of natural populations revealed a correlation between the mean diameter of perches used and mean relative limb length (Losos et al. 1994). Further, examination of populations experimentally established on tiny Bahamian cays 1015 years previously (Schoener and Schoener 1983) revealed a similar trend (Losos et al. 1997). These correlations paralleled at a lesser scale the trend seen among anole habitatspecialists, exemplied by the extremely long-legged trunkground anoles that inhabit tree trunks and the short-legged anoles that inhabit twigs. Thus, one interpretation is that microevolutionary change over relatively short periods produced the same pattern as evident over macroevolutionary time; by extrapolation, the latter was simply an extension of the former over longer periods of time.
343
However, an alternative hypothesis is that limb length is a phenotypically plastic trait. Perhaps young A. sagrei that grow up using narrower supports develop shorter limbs than individuals that grow up using broader supports. Surprisingly, a laboratory study found just that: hind-limb length is a phenotypically plastic trait affected by perch diameter (Losos et al. 2000). This study has two implications. More narrowly, it suggests that the differences observed between both experimental and natural populations in the Bahamas may be the result of plasticity, rather than genetic differentiation. Studies in which individuals from different islands are raised in the same environment are needed to test this hypothesis more directly. More generally, these ndings suggest the intriguing possibility that plasticity may play an important role in adaptive evolution. By permitting lizards to occupy a new habitat in which they otherwise might not be able to survive, plasticity may allow anoles to occupy new habitats. Once in these habitats, lizard behavior might change and, as new mutations arise, these could be selected for and thus greatly accentuate the initial, relatively minor, changes in limb length. In this way, plasticity might represent the rst stage in major adaptive shifts. If nothing else, it is striking that phenotypic plasticity produces the same morphologyenvironment correlation as observed among habitat specialists [note that the differences in limb length among habitat specialists are vastly greater than those produced in the plasticity experiment and surely represent genetic differences (Losos et al. 2000)]. The hypothesis that plasticity might be important in adaptive differentiation was originally put forth 50 years ago by Schmalhausen (1949), Waddington (1953a, 1953b), and others. Long ignored, the idea has recently been revived (WestEberhard 1989; Schlichting and Pigliucci 1998). Anoles may represent a good system in which to further explore these ideas.
16.4
Concluding Comments
Roger S. Thorpe and Jonathan B. Losos The previous two sections on one of the most speciose genera of amniote vertebrates emphasize the Lesser Antilles (Section 16.2) and the Greater Antilles (Section 16.3) systems. These two systems have many basic differences. The former is dominated by solitary anole species (or at most two natural species in sympatry); no, or relatively low, congeneric competition; numerous colonization events between island banks associated with speciation; and somewhat arbitrary allopatric species. The latter is dominated by multiple-species sympatry, competition, few colonization events, speciation within islands, and a relatively high degree of condence in what constitutes a species. Despite these basic differences, the conclusions suggested by the two sections are very similar. In both the Lesser and Greater Antillean models, there is substantial speciation and substantial evidence of adaptation. In the former, evidence of adaptation comes from intraspecic, within-island adaptation to different habitat types (ecotypes) supported by correlational evidence, parallels, translocation
344
experiments, and common-garden experiments. In the latter it comes from convergent habitat specialization among species (ecomorphs) supported by functional and physiological studies. Even though both speciation and adaptation are substantial in both systems, currently neither system shows hard evidence of adaptive speciation. Perhaps we have not been able to reveal any clear-cut cases of adaptive speciation because it does not occur in this group. However, it may occur and our failure to reveal it may be because either the criteria used are too demanding or the appropriate studies have not been carried out. If the denition of adaptive speciation in Chapter 1 is used, an important subset of cases involve speciation in sympatry. In this subset of cases, anoles are unlikely models for adaptive speciation as there are no well-supported cases of sympatric speciation in squamates with normal sexual reproduction and without chromosomal changes. However, if adaptive speciation could occur where adaptation plays a key role in the speciation of populations in ecological (parapatric) contact, as indicated in Chapter 7, then adaptive speciation may play a role in the speciation of anoles and other squamates. For the Greater Antilles, Losos (Section 16.3) suggests the possibility that adaptation to new habitats could trigger speciation indirectly by leading to changes in the behavioral or morphological facets of sexual signals, but this remains to be demonstrated. So in the Greater Antilles, perhaps, parapatric adaptive speciation may have played a part that future studies will reveal. However, the dewlap, which is thought to play a key role in sexual signaling, may show relatively little difference among habitat types within Lesser Antillean islands (e.g., A. oculatus in Malhotra 1992), an observation that does not provide support for the above proposal. Yet, more recent studies using spectrometric analysis of dewlap hue, in the roquet group, show distinct variation in relation to habitat type (Thorpe 2002; Thorpe and Stenson 2003). With regard to the Lesser Antilles model, there is evidence (Ogden and Thorpe 2002) of a reduction in gene ow between parapatric habitat forms (incipient speciation), but this is not complete speciation, and even if these parapatric forms have become partially isolated in situ, this is not necessarily adaptive speciation (Chapter 1). With other contact zones within Martinique, which may warrant full species recognition, further work is required to exclude condently a role for divergence in allopatry.
Acknowledgments Roger S. Thorpe thanks the Natural Environment Research Council, the Linnean Society, and the Leverhulme Trust for support.
Color Plates
Plate 9 Sympatric anole species from the Lesser Antilles (Section 16.2).
Top: Anolis trinitatis. Bottom: A. griseus. A. trinitatis is the smaller of the two and both are widely sympatric on the island of St Vincent. Sympatric pairs always differ in size; it remains to be shown whether this is because of size assortment or character displacement.
Color Plates
Plate 10 Dewlap color and pattern diversity in Caribbean anole species (Section 16.3).
Top left: Anolis sagrei is a trunkground anole from the Bahamas, but occurs also elsewhere in the Caribbean, including Cuba. Top right: A. mestrei is a trunkground anole from Cuba. Bottom left: A. grahami is a trunkcrown anole from Jamaica. Bottom right: Chamaelinorops barbouri is a leaflitter dwelling species from Hispaniola with no ecological counterparts on other islands. C. barbouri falls phylogenetically within Anolis.
References
References in the book in which this chapter is published are integrated in a single list, which appears on pp. 395444. For the purpose of this reprint, references cited in the chapter have been assembled below.
Andrews RM (1976). Growth rate in island and mainland anoline lizards. Copeia 1976:477482 Andrews RM (1979). Evolution of life histories: A comparison of Anolis lizards from matched island and mainland habitats. Breviora 454:151 Arnold SJ (1983). Morphology, performance, and tness. American Zoologist 23:1347 1361 Arnold EN (1994). Investigating the origins of performance advantage: Adaptation, exaptation and lineage effects. In Phylogenetics and Ecology, eds. Eggleton P & Vane-Wright R, pp. 123168. London, UK: Academic Press Autumn K, Liang YA, Hsieh T, Fisher RN, Zesch W, Chan WP, Kenny TW, Fearing R & Full RJ (2000). Adhesive force of a single gecko foot-hair. Nature 405:681685 Brodie ED III, Moore AJ & Janzen FJ (1995). Visualising and quantifying natural selection. Trends in Ecology and Evolution 10:313318 Butler MA, Schoener TW & Losos JB (2000). The relationship between sexual size dimorphism and habitat use in Greater Antillean Anolis lizards. Evolution 54:259272 Daltry J, Wster W & Thorpe RS (1996). Diet and snake venom evolution. Nature 379:537 540 De Queiroz K, Chu LR & Losos JB (1998). A second Anolis lizard in Dominican amber and the systematics and ecological morphology of Dominican amber anoles. American Museum Novitates 3249:123 Dieckmann U & Doebeli M (1999). On the origin of species by sympatric speciation. Nature 400:354357 Doebeli M & Dieckmann U (2003). Speciation along environmental gradients. Nature 421:259264 Endler JA (1986). Natural Selection in the Wild. Monographs in Population Biology 21. Princeton, NJ, USA: Princeton University Press Endler JA (1992). Signals, signal conditions, and the direction of evolution. The American Naturalist 139:S125S153 Fleishman LJ (1992). The inuence of sensory system and the environment on motion patterns in the visual displays of anoline lizards and other vertebrates. The American Naturalist 139:S36S61 Fleishman LJ (2000). Signal function, signal efciency and the evolution of anoline lizard dewlap color. In Animal Signals: Signalling and Signal Design in Animal Communication, eds. Espmark Y, Amundsen T & Rosenqvist G, pp. 209236. Trondheim, Norway: Tapir Academic Press Fleishman LJ, Bowman M, Saunders D, Miller WE, Rury MJ & Loew ER (1997). The visual ecology of Puerto Rican anoline lizards: Habitat light and spectral sensitivity. Journal of Comparative Physiology 181:446460 Frost DR & Etheridge R (1989). A phylogenetic analysis and taxonomy of iguanian lizards (Reptilia: Squamata). Miscellaneous Publications, University of Kansas Museum of Natural History 81:165 Garcia R, Queral A, Powell R, Parmerlee JS Jr, Smith DD & Lathrop A (1994). Evidence of hybridization among green anoles (Lacertilia: Polychrotidae) from Hispaniola. Caribbean Journal of Science 30:279281
Garland T Jr & Adolph SC (1991). Physiological differentiation of vertebrate populations. Annual Review of Ecology and Systematics 22:193228 Giannasi N (1997). Morphological, Molecular and Behavioural Evolution of the Anolis roquet group. PhD thesis. Bangor, UK: University of Wales Giannasi N, Thorpe RS & Malhotra A (2000). A phylogenetic analysis of body size evolution in the Anolis roquet group (Sauria: Iguanidae): Character displacement or size assortment? Molecular Ecology 9:193202 Gorman GC & Atkins L (1968). Natural hybridization between two sibling species of Anolis lizards: Chromosome cytology. Science 159:13581360 Gorman GC & Atkins L (1969). The zoogeography of the Lesser Antilles Anolis lizards: An analysis based on chromosomes and lactic dehydrogenases. Bulletin of the Museum of Comparative Zoology 138:5380 Gorman GC, Licht P, Dessauer HC & Boos JO (1971). Reproductive failure among the hybridizing Anolis lizards of Trinidad. Systematic Zoology 20:112 Grant PR (1986). Ecology and Evolution of Darwins Finches. Princeton, NJ, USA: Princeton University Press Greene HW (1986). Diet and arboreality in the emerald monitor, Varanus prasinus, with comments on the study of adaptation. Fieldiana Zoology, New Series 31:112 Guyer C (1988). Food supplementation in a tropical mainland anole, Norops humilis: Effects on individuals. Ecology 69:362369 Harvey PH & Pagel MD (1991). The Comparative Method in Evolutionary Biology. Oxford, UK: Oxford University Press Hertz PE (1979). Sensitivity to high temperatures in three West Indian grass anoles (Sauria: Iguanidae), with a review of heat sensitivity in the genus Anolis. Journal of Comparative Biochemistry and Physiology 62A:217222 Hertz PE (1992). Temperature regulation in Puerto Rican Anolis lizards: A eld test using null hypotheses. Ecology 73:14051417 Hertz PE, Arce-Hernandez A, Ramirez-Vazquez J, Tirado-Rivera W & Vazquez-Vives L (1979). Geographical variation of heat sensitivity and water loss rates in the tropical lizard, Anolis gundlachi. Comparative Biochemistry and Physiology 62A:947953 Howard DJ & Berlocher SH (1998). Endless Forms: Species and Speciation. Oxford, UK: Oxford University Press Huey RB (1983). Natural variation in body temperature and physiological performance in a lizard (Anolis cristatellus). In Advances in Herpetology and Evolutionary Biology: Essays in Honor of Ernest E. Williams, eds. Rhodin AGJ & Miyata K, pp. 484490. Cambridge, MA, USA: Museum of Comparative Zoology, Harvard University Irschick DJ & Losos JB (1998). A comparative analysis of the ecological signicance of maximal locomotor performance in Caribbean Anolis lizards. Evolution 52:219226 Irschick DJ & Losos JB (1999). Do lizards only use habitats in which performance is maximal? The relationship between sprinting capabilities and structural habitat use in Caribbean anoles. The American Naturalist 154:293305 Irschick DJ, Austin CC, Petren K, Fisher RN, Losos JB & Ellers O (1996). A comparative analysis of clinging ability among pad-bearing lizards. Biological Journal of the Linnean Society 59:2135 Jackman T, Losos JB, Larson A & de Queiroz K (1997). Phylogenetic studies of convergent adaptive radiations in Caribbean Anolis lizards. In Molecular Evolution and Adaptive Radiation, eds. Givnish T & Systma K, pp. 535557. Cambridge, UK: Cambridge University Press
Jackman TR, Irschick DJ, de Queiroz K, Losos JB & Larson A (2002). Molecular phylogenetic perspective on evolution of lizards of the Anolis grahami series. Journal of Experimental Zoology: Molecular and Developmental Evolution 294:116 Jackman TR, Larson A, de Queiroz K & Losos JB (1999). Phylogenetic relationships and the tempo of early diversication in Anolis lizards. Systematic Biology 48:254285 Jenssen TA (1970). Female response to lmed displays of Anolis nebulosus (Sauria: Iguanidae). Animal Behaviour 18:640647 Jenssen TA (1977). Morphological, behavioral and electrophoretic evidence of hybridization between the lizards, Anolis grahami and Anolis lineatopus neckeri, on Jamaica. Copeia 1977:270276 Jenssen TA (1978). Display diversity in anoline lizards and problems of interpretation. In Behavior and Neurology of Lizards, eds. Greenberg N & MacLean PD, pp. 269285. Rockville, MD, USA: National Institute of Mental Health Jenssen TA (1996). A test of assortative mating between sibling lizard species, Anolis websteri and A. caudalis, in Haiti. In Contributions to West Indian Herpetology: A Tribute to Albert Schwartz, eds. Powell R & Henderson RW, pp. 303315. Ithaca, NY, USA: Society for the Study of Amphibians and Reptiles Lauder GV (1981). Form and function: Structural analysis in evolutionary morphology. Paleobiology 7:430442 Lazell J (1996). Careening Island and the Goat Islands: Evidence for the aridinsular invasion wave theory of dichopatric speciation in Jamaica. In Contributions to West Indian Herpetology: A Tribute to Albert Schwartz, eds. Powell R & Henderson RW, pp. 195 205. Ithaca, NY, USA: Society for the Study of Amphibians and Reptiles Leal M & Fleishman LJ (2002). Evidence for habitat partitioning based on adaptation to environmental light in a pair of sympatric lizard species. Proceedings of the Royal Society of London B 269:351359 Lister BC (1976). The nature of niche expansion in West Indian Anolis lizards: II. Evolutionary components. Evolution 30:677692 Losos JB (1985). An experimental demonstration of the species recognition role of Anolis dewlap color. Copeia 1985:905910 Losos JB (1990a). Ecomorphology, performance capability, and scaling of West Indian Anolis lizards: An evolutionary analysis. Ecological Monographs 60:369388 Losos JB (1990b). A phylogenetic analysis of character displacement in Caribbean Anolis lizards. Evolution 44:558569 Losos JB (1994). Integrative approaches to evolutionary ecology: Anolis lizards as model systems. Annual Review of Ecology and Systematics 25:467493 Losos JB & Chu LR (1998). Examination of factors potentially affecting dewlap size in Caribbean anoles. Copeia 1998:430438 Losos JB & Irschick DJ (1996). The effect of perch diameter on escape behaviour of Anolis lizards: Laboratory-based predictions and eld tests. Animal Behaviour 51:593602 Losos JB & Miles DB (1994). Adaptation, constraint, and the comparative method: Phylogenetic issues and methods. In Ecological Morphology: Integrative Organismal Biology, eds. Wainwright PC & Reilly SM, pp. 6098. Chicago, IL, USA: University of Chicago Press Losos JB & Schluter D (2000). Analysis of an evolutionary speciesarea relationship. Nature 408:847850 Losos JB & Sinervo B (1989). The effect of morphology and perch diameter on sprint performance of Anolis lizards. Journal of Experimental Biology 145:2330
Losos JB, Irschick DJ & Schoener TW (1994). Adaptation and constraint in the evolution of specialization of Bahamian Anolis lizards. Evolution 48:17861798 Losos JB, Warheit KI & Schoener TW (1997). Adaptive differentiation following experimental island colonization in Anolis lizards. Nature 387:7073 Losos JB, Jackman TR, Larson A, de Queiroz K & Rodrguez-Schettino L (1998). Historical contingency and determinism in replicated adaptive radiations of island lizards. Science 279:21152118 Losos JB, Creer DA, Glossip D, Goellner R, Hampton A, Roberts G, Haskell N, Taylor P & Etling J (2000). Evolutionary implications of phenotypic plasticity in the hindlimb of the lizard Anolis sagrei. Evolution 54:301305 Macedonia JM & Stamps JA (1994). Species recognition in Anolis grahami (Sauria: Iguanidae): Evidence from responses to video playbacks of conspecic and heterospecic displays. Ethology 98:246264 Macedonia JM, Evans CS & Losos JB (1993). Male Anolis lizards discriminate videorecorded conspecic and heterospecic displays, performance, and tness. Animal Behaviour 47:12201223 Maddison WP & Maddison DR (1992). MacClade, Version 3 Analysis of Phylogeny and Character Evolution. Sunderland, MA, USA: Sinauer Associates Inc. Malhotra A (1992). What Causes Geographic Variation: A Case Study of Anolis oculatus. PhD thesis. Aberdeen, UK: University of Aberdeen Malhotra A & Thorpe RS (1991a). Microgeographic variation in Anolis oculatus on the island of Dominica, West Indies. Journal of Evolutionary Biology 4:321335 Malhotra A & Thorpe RS (1991b). Experimental detection of rapid evolutionary response in natural lizard populations. Nature 353:347348 Malhotra A & Thorpe RS (1993a). An experimental eld study of an eurytopic anole, Anolis oculatus. Journal of Zoology 229:163170 Malhotra A & Thorpe RS (1993b). Anolis oculatus. Catalogue of American Amphibians and Reptiles, Society for the Study of Amphibians and Reptiles 540:14 Malhotra A & Thorpe RS (1994). Parallels between island lizards suggests selection on mitochondrial DNA and morphology. Proceedings of the Royal Society of London B 257:3742 Malhotra A & Thorpe RS (1997a). Size and shape variation in a Lesser Antillean anole, Anolis oculatus (Sauria: Iguanidae) in relation to habitat. Biological Journal of the Linnean Society 60:5372 Malhotra A & Thorpe RS (1997b). Microgeographic variation in scalation of Anolis oculatus (Dominica, West Indies): A multivariate analysis. Herpetologica 53:4962 Malhotra A & Thorpe RS (1999). Reptiles and Amphibians of the Eastern Caribbean. London, UK: MacMillan Press Malhotra A & Thorpe RS (2000). The dynamics of natural selection and vicariance in the Dominican anole: Patterns of within-island molecular and morphological divergence. Evolution 54:245258 Manly BJF (1986a). Multivariate Statistical Methods: A Primer. London, UK: Chapman & Hall Manly BJF (1986b). Randomization and regression methods for testing associations with geographical, environmental and biological distances between populations. Researches on Population Ecology 28:201218 Ogden R & Thorpe RS (2002). Molecular evidence for ecological speciation in tropical habitats. Proceedings of the National Academy of Sciences of the USA 99:1361213615
Paterson HEH (1982). Perspective on speciation by reinforcement. South African Journal of Science 78:5357 Persons MH, Fleishman LJ, Frye MA & Stimphil ME (1999). Sensory response patterns and the evolution of visual signal design in anoline lizards. Journal of Comparative Physiology 184:585607 Rand AS & Williams EE (1970). An estimation of redundancy and information content of anole dewlaps. The American Naturalist 104:99103 Reardon JT & Thorpe RS. Natural selection and common garden experiments on Anolis oculatus. Unpublished Roughgarden J (1995). Anolis Lizards of the Caribbean: Ecology, Evolution, and Plate Tectonics. New York, NY, USA: Oxford University Press Ruibal R & Williams EE (1961). Two sympatric Cuban anoles of the carolinensis group. Bulletin of the Museum of Comparative Zoology 125:183208 Schlichting CD & Pigliucci M (1998). Phenotypic Evolution: A Reaction Norm Perspective. Sunderland, MA, USA: Sinauer Associates Inc. Schluter D (2000). The Ecology of Adaptive Radiation. Oxford, UK: Oxford University Press Schmalhausen II (1949). Factors of Evolution. Philadelphia, PA, USA: Blakiston Schoener TW (1969). Size patterns in West Indian Anolis lizard: I. Size and species diversity. Systematic Zoology 18:386401 Schoener TW (1970). Size patterns in West Indian Anolis lizard: II. Correlations with the sizes of particular sympatric species displacement and convergence. The American Naturalist 104:155174 Schoener TW & Schoener A (1983). The time to extinction of a colonizing propagule of lizards increases with island area. Nature 302:332334 Schneider CJ, Losos JB & de Queiroz K (2001). Evolutionary relationships of the Anolis bimaculatus group from the northern Lesser Antilles. Journal of Herpetology 35:112 Schwartz A & Henderson RW (1991). Amphibians and Reptiles of the West Indies: Descriptions, Distributions, and Natural History. Gainsville, FL, USA: University of Florida Press Shochat D & Dessauer HC (1981). Comparative immunological study of albumins of Anolis lizards of the Caribbean islands. Comparative Biochemistry and Physiology 68A:6773 Smouse PE, Long J & Sokal RR (1986). Multiple regression and correlation extensions of the Mantel test of matrix correspondence. Systematic Zoology 35:627632 Stenson AG (2000). Use of Molecular Markers at Different Taxonomic Levels: Evolution of the Northern Lesser Antillean Anole Radiation. PhD thesis. Bangor, UK: University of Wales Stenson AG, Malhotra A & Thorpe RS (2000). Highly polymorphic microsatellite loci from the Dominican anole (Anolis oculatus) and their amplication in other bimaculatus series anoles. Molecular Ecology 9:16801681 Stenson AG, Malhotra A & Thorpe RS (2002). Population differentiation and gene ow in a species displaying pronounced geographic variation in morphology, the Dominican anole (Anolis oculatus). Molecular Ecology 11:16791688 Tautz D (2003). Splitting in space. Nature 421:225226 Thorpe RS (1991). Clines and cause: Microgeographic variation in the Tenerife gecko Tarentola delalandii. Systematic Zoology 40:172187 Thorpe RS (1996). The use of DNA divergence to help determine the correlates of evolution of morphological characters. Evolution 50:524531
Thorpe RS (2002). Analysis of color spectra in comparative evolutionary studies: Molecular phylogeny and habitat adaptation in the St Vincent Anole, Anolis trinitatis. Systematic Biology 51:554569 Thorpe RS & Baez M (1993). Geographic variation in scalation of the lizard Gallotia stehlini within the island of Gran Canaria. Biological Journal of the Linnean Society 48:7587 Thorpe RS & Malhotra A (1992). Are Anolis lizards evolving? Nature 355:506 Thorpe RS & Malhotra A (1996). Molecular and morphological evolution within small islands. Philosophical Transactions of the Royal Society of London B 351:815822 Thorpe RS & Richard M (2001). Evidence that ultraviolet markings are associated with patterns of molecular gene ow. Proceedings of the National Academy of Sciences of the USA 98:39293934 Thorpe RS & Stenson AG (2003). Phylogeny, paraphyly and ecological adaptation of the colour and pattern in the Anolis roquet complex on Martinique. Molecular Ecology 12:117132 Thorpe RS, McGregor DP, Cumming AM & Jordan WC (1994). DNA evolution and colonization sequence of island lizards in relation to geological history: mtDNA RFLP, cytochrome b, cytochrome oxidase, 12s rRNA sequence, and nuclear RAPD analysis. Evolution 48:230240 Thorpe RS, Malhotra A, Black H, Daltry JC & Wster W (1995). Relating geographic patterns to phylogenetic processes. Philosophical Transactions of the Royal Society of London B 349:6168 Thorpe RS, Black H & Malhotra A (1996). Matrix correspondence tests on the DNA phylogeny of the Tenerife lacertid elucidates both historical causes and morphological adaptation. Systematic Biology 45:335343 Underwood G (1959). The anoles of the eastern Caribbean. Part III. Revisionary notes. Bulletin of the Museum of Comparative Zoology, Harvard 121:191226 Van Berkum FH (1986). Evolutionary patterns of the thermal sensitivity of spring speed in Anolis lizards. Evolution 40:495504 Vanzolini PE & Williams EE (1981). The vanishing refuge: A mechanisms for ecogeographic speciation. Papeis Avulsos Zoologici 34:251255 Waddington CH (1953a). Genetic assimilation of an acquired character. Evolution 7:118 126 Waddington CH (1953b). Epigenetics and evolution. Symposium of the Society of Experimental Biology 7:186199 Wainwright PC (1988). Morphology and ecology: Functional basis of feeding constraints in Caribbean labrid shes. Ecology 69:635645 Webster TP (1977). Geographic variation in Anolis brevirostris. In The Third Anolis Newsletter, ed. Williams EA, pp. 153164. Cambridge, MA, USA: Museum of Comparative Zoology West-Eberhard MJ (1989). Phenotypic plasticity and the origins of diversity. Annual Review of Ecology and Systematics 20:249278 Williams EE (1972). The origin of faunas. Evolution of lizard congeners in a complex island fauna: A trial analysis. Evolutionary Biology 6:4789 Williams EE (1983). Ecomorphs, faunas, island size, and diverse end points in island radiations of Anolis. In Lizard Ecology: Studies of a Model Organism, eds. Huey RB, Pianka ER & Schoener TW, pp. 326370. Cambridge, MA, USA: Harvard University Press
Williams EE & Peterson JA (1982). Convergent and alternative designs in the digital adhesive pads of scincid lizards. Science 215:15091511 Williams EE & Rand AS (1977). Species recognition, dewlap function, and faunal size. American Zoologist 17:261270

ThorpeDiversificación Anolis Caribe

Uploaded by

Copyright:

Available Formats

ThorpeDiversificación Anolis Caribe

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ThorpeDiversificación Anolis Caribe

Uploaded by

Copyright:

Available Formats

Evolutionary Diversication of Caribbean Anolis Lizards. In: Adaptive Speciation, eds.

16 Evolutionary Diversication of Caribbean Anolis Lizards

Adaptation and Speciation in Lesser Antillean Anoles

16 Evolutionary Diversication of Caribbean Anolis Lizards

Sombrero Anguilla Saba

St Eustasius St Kitts Nevis

Grande Terre Desirade Petite Terre Marie Galante Dominica

One spp. island

Nevis Two spp. island

Mon tane ecoty p

Canopy height (m)

North Caribbean form

Canopy height (m)

Atlantic ecotype Canopy height (m) 30

16 Evolutionary Diversication of Caribbean Anolis Lizards

16 Evolutionary Diversication of Caribbean Anolis Lizards

16 Evolutionary Diversication of Caribbean Anolis Lizards

Barbuda St Eustasius St Kitts Nevis

Grande Terre Desirade

Marie Galante Dominica

16 Evolutionary Diversication of Caribbean Anolis Lizards

16 Evolutionary Diversication of Caribbean Anolis Lizards

Adaptation and Speciation in Greater Antillean Anoles

Hispaniola Jamaica Puerto Rico

16 Evolutionary Diversication of Caribbean Anolis Lizards

16 Evolutionary Diversication of Caribbean Anolis Lizards

16 Evolutionary Diversication of Caribbean Anolis Lizards

16 Evolutionary Diversication of Caribbean Anolis Lizards

You might also like