© 2007 Society of Systematic Biologists
Adding More Ecology into Species Delimitation: Ecological Niche Models and Phylogeography Help Define Cryptic Species in the Black Salamander (Aneides flavipunctatus)
Edited by John Weins: Guest Associate Editor
Department of Biological Sciences Box 870345 MHB Hall University of Alabama Tuscaloosa AL, 35487, USA E-mail: rissler{at}bama.ua.edu (L.J.R.)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
Being able to efficiently and accurately delimit species is one of the most basic and important aspects of systematics because species are the fundamental unit of analysis in biogeography, ecology, and conservation. We present a rationale and approach for combining ecological niche modeling, spatially explicit analyses of environmental data, and phylogenetics in species delimitation, and we use our methodology in an empirical example focusing on Aneides flavipunctatus, the black salamander (Caudata: Plethodontidae), in California. We assess the relationships between genetic, environmental, and geographic distance among populations. We use 11 climatic variables and point locality data from public databases to create ecological niche models. The suitability of potential contact zones between parapatric lineages is also assessed using the data from ecological niche modeling. Phylogenetic analyses of portions of the mitochondrial genome reveal morphologically cryptic mitochondrial lineages in this species. In addition, we find that patterns of genetic divergence are strongly associated with divergence in the ecological niche. Our work demonstrates the ease and utility of using spatial analyses of environmental data and phylogenetics in species delimitation, especially for groups displaying fine-scaled endemism and cryptic species.
Keywords: Aneides; California; Maxent; niche modeling; salamanders; species delimitation
Received April 25, 2007; Revised July 2, 2007; Accepted September 2, 2007
Integrating genetic and ecological approaches in the study of mechanisms driving geographic distributions of organisms is becoming more common (e.g., Hugall et al., 2002; Johnson and Cicero, 2002; Graham et al., 2004a; Lapointe and Rissler, 2005). This is partly due to the ever-increasing amount and accuracy of natural history collections' (NHCs) point-locality data, accessibility of public databases housing fine-scaled climate data, and new modeling techniques (Graham et al., 2004b; Hijmans et al., 2005; Elith et al., 2006). The use of such data in environmental niche modeling (Soberón and Peterson, 2005) is proving to be a powerful approach to understanding how abiotic factors (e.g., temperature, precipitation, and seasonality) impact the geographic limits of lineages and species (Graham et al., 2004b; Wiens and Graham, 2005).
New spatially explicit methods that combine bioclimatic information and presence-absence locality information from natural history databases are useful for understanding the ecological processes driving biodiversity patterns (Graham et al., 2004b; Elith et al., 2006). For example, environmental niche modeling uses georeferenced data from specimen records in combination with environmental data layers, usually in a GIS framework, using diverse algorithms (see Elith et al., 2006, for details on alternative algorithms), to identify areas of predicted presence on a map (Soberón and Peterson, 2005). This correlative approach identifies areas that are ecologically similar to regions where the point locality information was used to build the models. These predicted regions of occurrence represent the ecological niche, and they are presumed to contain the suite of environmental conditions necessary to maintain a viable population (Grinnell, 1917, 1924; Hutchinson, 1957; Graham et al., 2004b). Whether this predicted region is a true spatial representation of the fundamental or realized niche has been a point of controversy (Kearney and Porter, 2004; Soberón and Peterson, 2005; Araújo and Guisan, 2006), but in general, the "fundamental niche" has been defined as the environmental space where fitness is greater than or equal to one, in the absence of range-limiting biotic interactions and dispersal barriers (Hutchinson, 1957; see Chase and Leibold, 2003, for a detailed review of the niche concept). Because correlative models take presence data from natural history databases, and thus implicitly incorporate biotic interactions that are dependent on the abiotic variables, researchers have argued that ecological niche modeling more correctly identifies the "realized niche" (Kearney and Porter, 2004; Soberón and Peterson, 2005; Araújo and Guisan, 2006; Kearney, 2006).
Results from this type of correlative ecological niche modeling have provided insight into a variety of questions relevant to conservation and evolutionary biology. These include (1) the importance of niche conservatism to speciation (e.g., Peterson et al., 1999; Kozak and Wiens, 2006); (2) the geographic spread of invasive species (e.g., Peterson, 2003; Peterson and Robins, 2003); (3) distributions of undiscovered species (Raxworthy et al., 2003); (4) pinpointing regions of high endemicity and conservation value for genetic lineages (Rissler et al., 2006); (5) inferences of historical and future distributions (Hugall et al., 2002; Peterson et al., 2002; Thomas et al., 2004; Araújo et al., 2006; Hijmans and Graham, 2006); and (6) historical biogeography (Wiens et al., 2006). One area where niche modeling has not been used extensively is phylogeography (but see Hugall et al., 2002), although there is potential for niche modeling to provide insight into the abiotic factors affecting the geographic limits of genetic lineages and to shed light on whether speciation or genetic variation is associated with divergence in the ecological niche (Graham et al., 2004b; Kidd and Ritchie, 2006).
Historical biogeography and comparative phylogeography, in part, seek to explain patterns of geographic congruence in phylogenetic breaks across multiple taxa. The processes driving lineage divergence, speciation, and the buildup of biodiversity are many and include geographical, historical, and environmental factors that favor isolation and the cessation of gene flow between populations (Wiens, 2004). Recent work on modes of speciation has examined geographic, genetic, and climatic patterns across sister species in the hope of inferring the mechanisms driving differentiation (e.g., Kozak and Wiens, 2006). This approach has been somewhat controversial because it assumes that present-day distributions can provide information on the ancestral species at the time of speciation. Perhaps more importantly, and more relevant for the work in this paper, combining niche modeling and genetic analyses can provide some insight into the role of ecological divergence in speciation, regardless of the particular geographic mode of speciation (i.e., parapatric, sympatric, allopatric).
In addition, an understanding of the mechanisms driving divergence can also help in species delimitation. By combining information on geographic distributions, ecological niche models based on environmental data, and genetic information from multiple loci, researchers can make strong inferences when diagnosing species (Wiens and Graham, 2005). For example, two genetically unresolved lineages, with unique ecological niches, that are geographically distributed either parapatrically or allopatrically warrant further study to determine if the intervening region at the contact zone contains suitable habitat. If there is no suitable habitat in the contact zone, then gene flow between the lineages may be impeded, and this would support the conclusion that the two lineages are distinct, even with limited genetic divergence. However, if those two lineages had similar ecological niches with no biogeographic barrier separating their distributions and high-quality habitat for one or both at the contact zone, then it is less likely that the two lineages are distinct with no gene flow. Of course, the extent of gene flow in a suitable contact zone is still dependent on the level of reproductive isolation among lineages, but this information is often difficult to obtain. Therefore, an understanding of geographic distributions, environmental niche dimensions, and genetic patterns may help in species delimitation.
Species delimitation can be especially difficult for organisms displaying cryptic morphological characters and fine-scaled endemism patterns. Most species have been diagnosed using morphological data, but the use of genetic data has led to the discovery of many unique, morphologically cryptic lineages (Sites and Marshall, 2003, 2004). Recent analyses have demonstrated high concordance in patterns of global terrestrial endemism across taxa (Brooks et al., 2006), but it is unclear whether or not conservation measures based on classically defined species adequately reflect evolutionary lineages (Agapow et al., 2004; Rissler et al., 2006). In addition, true biodiversity hotspots can be missed if taxonomic lists do not adequately reflect nature (Köhler et al., 2005). Therefore, it is important to develop methods that can be used efficiently and easily to aid in species delimitation.
In this paper we use the salamander species Aneides flavipuncatatus to demonstrate a general approach for combining ecological niche modeling and phylogenetics in species delimitation. Additional goals are to (1) examine phylogeographic patterns in A. flavipunctatus, a species for which there have been no previous mitochondrial analyses; (2) develop ecological niche models using georeferenced museum locality data and environmental data layers; and (3) assess whether genetic variation across lineages is associated with divergence in the ecological niche.
General Approach
Our view of species follows the evolutionary species concept (ESC) and general lineage species concept (GLC) (Wiley, 1978; de Queiroz, 1998), and our goal is to recognize historically distinct evolutionary lineages that are likely to remain distinct (e.g., Wake, 2006). We present a general approach to combining spatially-explicit environmental data with phylogenetic information in Figure 1 for aiding species delimitation. In general, one needs to (1) define the potential phylogenetic lineages that may deserve species' status; (2) determine the extent of ecological divergence for these lineages; and (3) assess whether the contact zones between lineages contain suitable habitat that may impede or facilitate gene flow.
|
Combining independent sets of data (ecological and genetic) can provide a more robust view of the independence of evolutionary lineages. Our approach focuses on adding information on ecological divergence because this is often an important step in the process of speciation (Mayr, 1947; Van Valen, 1976; Andersson, 1990; Streelman and Danley, 2003; Funk et al., 2006), and the information is available from public databases (Graham et al., 2004b; Hijmans et al., 2005; Elith et al., 2006). Information on the ecological niche can be especially important when other data (e.g., multiple genetic loci or behavioral) are lacking or insufficient to determine whether the lineages in question are truly distinct. Therefore, we consider those phylogenetic lineages that are also separated in ecological space as distinct evolutionary lineages that are likely to remain distinct; i.e., species.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
Species of Interest
Aneides flavipunctatus is endemic to the Californian Floristic Province and is distributed north of San Francisco Bay along the California coast and northeast into Shasta County, with an isolated and southern disjunct population around Santa Cruz, north of Monterey Bay. The range also extends northward into the Applegate Valley of extreme southern Oregon (Blackburn et al., 2001).
Tissue samples from frozen (–80°C) collections at the Museum of Vertebrate Zoology supplemented live specimen collections. We analyzed 42 samples from 18 localities of A. flavipunctatus, and almost all individuals were geocoded (latitude and longitude) or had detailed locality information associated with the specimen (Table 1). Sampling covered the majority of the species' range.
|
DNA Extraction and Sequencing
Extraction of DNA was done using the DNeasy tissue kit protocol (Qiagen, Valencia, CA). We sequenced portions of the ND4 and 12S regions in the mitochondrial genome. The valine transfer RNA region of the 12S gene was amplified using primers valB and valG (Rissler and Taylor, 2003). The ND4 region was amplified with primers designed for A. flavipunctatus: NADH-f (flav) (5'-GGGTATGGTATTATTCGAATT-3') and NADH-r (flav) (5'-TGGGGCAGATAATTAGCAGT-3'). All PCR reactions were carried out in a volume of 25 µ L containing 5 to 50 ng of genomic DNA, 2.5 µ L PCR buffer, 0.5 µ L of dNTPs (10 mM of each dNTP), 2.5 µ L of 100 mM primer, 0.2 µ L Taq, and water. Double-stranded PCR products were cleaned with a QIAquick PCR purification kit (Qiagen) or ExoSAP-IT (USB, Cleveland, OH) and labeled with fluorescent-dye labels through a cycle-sequencing reaction following standard protocols (Applied Biosystems, Perkin-Elmer). Cycle-sequencing products were cleaned with Sephadex columns and sequenced using an ABI Prism 377 automated sequencer (Applied Biosystems). All samples were sequenced in both directions and then combined in Sequencher (version 4.5). We used ClustalW in MacVector (version 8.0) for sequence alignments. Alignments were further corrected manually, and there was no difficulty in aligning sequences in hypothesized loop regions of the 12S gene. The two mitochondrial genes were appended and treated as a single locus.
Genetic Analyses
We used maximum likelihood (ML; PAUP*4.0b10 [Swofford, 2000] and Garli v.0.951 (Zwickl, 2006]) and Bayesian inference (MrBayes 3.01 [Huelsenbeck and Ronquist, 2001]) to infer the phylogeny. The model of evolution was chosen by ModelTest 3.04 (Posada and Crandall, 1998) using the AIC values, and model parameters were used in ML analyses. ML analyses were conducted using a heuristic search with the chosen model, TBR branch swapping, and 10 random-addition replicates. Nodal support was assessed using bootstrap analysis with 100 replicates for ML and Bayesian posterior probabilities (PP) for Bayesian analysis. For the Bayesian analyses, we used a flat Dirichlet probability density for the GTR rate matrix. Four incrementally heated Markov chains were run for 5.0 million generations, and trees were sampled every 5000 generations. The log-likelihood values of the chains fluctuated within a stable range of 0.008, and we plotted the ln-likelihood values against generation time to assess stationarity (data not shown). We allowed a burn-in of 2.5 million generations, and the remaining trees were used for measuring posterior probabilities of nodes. Data files have been uploaded to TreeBASE.org as accession number SN3580.
We defined lineages as those groups of populations that were monophyletic with good support values (> 95% ML bootstrap and 1.0 Bayesian PP) and were distributed in distinct geographic space; other methods could have been used (e.g., Templeton, 2001; Wiens and Penkrot, 2002; Pons et al., 2006). Some potential lineages were not considered as potential species in our analyses because of low sample sizes and/or support (e.g., AF22-27, AF28-31; Table 1), but with additional sampling, other lineages/species within our "Central" lineage could become evident.
We used DnaSP 3.52 (Rozas and Rozas, 1999) to examine nucleotide diversity and other population genetic parameters within and across lineages. PAUP*4.0b10 was used to calculate maximum-likelihood corrected distances for comparisons of percent-corrected divergence across lineages.
Bioclimatic Modeling
Ecological niche models for A. flavipunctatus and its phylogeographic lineages were created using Maxent version 2.0 (Phillips et al., 2006). Maxent creates species distributional models by combining presence only data with ecological layers using a statistical approach known as maximum entropy. The maximum entropy approach estimates a species' environmental niche by finding a probability distribution that is based on a distribution of maximum entropy (with reference to a set of environmental variables). The principle of maximum entropy is useful for analyzing the available information to determine a unique probability distribution that assumes the least biased distribution encoding the given information is that which maximizes the information entropy. This method is equivalent to finding the maximum-likelihood distribution of a species (Phillips et al., 2004). Although Maxent is a new approach in species distributional modeling, it seems to perform better than other established methods like BIOCLIM, GARP, or DOMAIN (Phillips et al., 2004, 2006; Elith et al., 2006).
Maxent was run using point locality information from specimens in natural history museum collections summarized in two data portals, HerpNET (www.herpnet.org) and GBIF (www.gbif.org), combined with climatic layers downloaded from the WorldClim database (Hijmans et al., 2005). The WorldClim bioclimatic variables are biologically relevant temperature and precipitation layers (Hijmans et al., 2005). The WorldClim climate layers were created by interpolating observed climate from climate stations around the world, using a thin-plate smoothing spline set to a resolution of approximately 1 km, over the 30-year period from 1960 to 1990 (Hijmans et al., 2005).
To ensure that we did not overparameterize our niche models with redundant climatic information, we conducted a series of correlation tests intended to remove redundant variables. We extracted the environmental information from 50,000 randomly generated points from across the United States. Correlation matrices were then generated for all 19 variables within each of the two general climatic categories: temperature and precipitation. We used a Pearson correlation coefficient of 0.75 to identify highly correlated variables (Rissler et al., 2006). For pairs that were highly correlated we chose the variable that we considered more biologically meaningful and easier to interpret. Eleven variables were chosen and used in all subsequent analyses. These included BIO1 = annual mean temperature; BIO2 = mean diurnal range in temperature; BIO3 = isothermality (monthly/annual temperature range); BIO7 = annual range in temperature; BIO8 = mean temperature of wettest quarter of the year; BIO9 = mean temperature of driest quarter of the year; BIO15 = precipitation of seasonality (coefficient of variation); BIO16 = precipitation of wettest quarter of the year; BIO17 = precipitation of driest quarter of the year; BIO18= precipitation of warmest quarter of the year; BIO19 = precipitation of coldest quarter of the year. Finally, each climate layer was entered into Maxent as an ASCII raster grid.
Models for A. flavipunctatus were created using a total of 502 specimen localities, and the unsampled populations were split into four classes as defined by the phylogeographic analyses to create lineage models. Similar methods were used in Rissler et al. (2006). All runs were set with a convergence threshold of 1.0E–5 with 1000 iterations, and the regularization value was set to 1.0E–4. The resultant ASCII file was placed into raster format using ArcGIS 9.1 and viewed. The predicted potential distribution (ecological niche) of each lineage was displayed in four distinct categories using the natural breaks method of classification (Jenks' method) in ArcGIS 9.1. This method assigns the data to classes using an iterative algorithm that minimizes the variance within classes and maximizes variance between classes.
Spatial Statistical Analyses
We conducted principal components analyses (PCA) using each lineage with the values extracted for each climate layer to examine the overall levels of divergence in the ecological niche in STATISTICA v.6.0. We then used multivariate analyses of variance (MANOVA) with PCA axis scores as dependent variables and lineage as the fixed factors to determine whether separation in the ecological niche was statistically significant. Post hoc tests (e.g., Tukey's HSD) were used to understand which lineages differed, and we plotted the least-squares means (ls-means) to visually compare the differences in environmental variables across lineages.
It is also important to establish whether the associations are strictly due to spatial autocorrelation, and to do this, the geographic location of specimens must be statistically controlled for in multivariate analyses. We used partial Mantel tests to examine the relationships between genetic and environmental distances while controlling for geographic distance. Significant, positive partial correlations support the conclusion that genetic divergence is congruent with climatic gradients (e.g., Kozak and Wiens, 2006). Statistical significance was assessed by 999 permutations in the R-package v. 4.0 (Casgrain and Legendre, 2001). In these tests, the genetic distance matrices were based on the maximum-likelihood distances, and the environmental distances were based on Euclidean distances of the factor scores from the PCA of the 11 climatic variables at each locality. Only individuals that had genetic, geographic, and environmental information were used in the partial Mantel tests (n = 39).
Contact Zone Analyses
To determine whether the areas of potential geographic contact between parapatric lineages (contact zones) were broadly suitable for one or both lineages, we compared the probability of occurrence for each lineage in their respective ranges to those within the contact zone. If the contact zones are broadly suitable for both lineages, this suggests that environmental boundaries are not strong and gene flow could occur in the region, depending on the extent of reproductive isolation. On the other hand, if the two lineages are found in unique environmental space and separated by an inhospitable contact zone, then dispersal across this unsuitable region into an environmental space that may be outside of the physiological tolerance limits of the organisms may be unlikely. In fact, Kozak and Wiens (2006) suggest that even if lineages occupy the same environmental niche space, an unsuitable environment in the contact zone would still prevent gene flow.
We compared the contact zones of the following parapatric lineages: Northwest-Shasta and Northwest-Central. The ecological niche modeling technique (Maxent) provides probability of occurrence values for each grid cell (Phillips et al., 2006). Therefore, we extracted those probabilities for each historic collection locality from their respective ecological niche models. To assess contact zone suitability, we created 1,000 random points within a minimum convex polygon that spanned the contact zone (Fig. 2). Maxent probability values were then extracted from each lineage's ecological niche model within the contact zone polygon for those random points. Probability values for each lineage and contact zone were compared using ANOVAs in STATISTICA v.6.0. Low probability of occurrence values for one or both lineages in their contact zone is considered evidence that gene flow is unlikely between the parapatric lineages.
|
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
Gene Genealogies
Sequence lengths for the valine transfer region of the 12S gene averaged around 600 bp, and the ND4 region averaged around 730 bp. All sequences have been deposited in Genbank (Table 1). There were 296 parsimony-informative characters. The chosen model of evolution was the K81+
with base frequencies A = 0.3332, C = 0.2179, G = 0.1379, and T = 0.3110; substitution model A-C = 1.0000, A-G = 3.4425, A-T = 1.2944, C-G = 1.2944, C-T = 3.4425, and G-T = 1.000; and gamma distribution shape parameter equal to 0.6363. All methods of phylogenetic inference identified the same lineages within A. flavipunctatus; therefore, we only present the ML topology (Fig. 3). This tree suggests that A. flavipuncatus can be divided into four lineages (Southern Disjunct, Shasta, Northwest, and Central). The ML tree was rooted with the Southern Disjunct lineage because it clearly is a monophyletic and geographically bounded unit that is the sister taxon of other lineages within A. flavipunctatus based on unpublished data (Rissler, unpublished) that includes multiple outgroups (including other species within the genus Aneides and other species within the families Plethodontidae, Ambystomatidae, and Salamandridae). The Shasta and Northwest California lineages are also distinct (Figs. 3, 4). The remaining samples include individuals from the central coastal region of California (Central lineage, Fig. 3, 4). High genetic diversity that is not geographically structured characterizes this lineage, yet there is a deeply diverged lineage found within Mendocino County (AF28–31; Fig. 3).
|
The four main lineages within A. flavipuncatus are quite distinct (Tables 2, 3). For example, there is an average of 66 nucleotide differences between the Southern Disjunct lineage and the Shasta lineage (Table 3). Even within a lineage, there is high haplotype diversity (Table 2). The Northwest lineage (Siskiyou, Humboldt, and Trinity counties) also has high haplotype diversity and number of nucleotide differences, despite a much smaller sample size. The Shasta lineage has the lowest number of nucleotide differences between individuals (Table 2), but it differs from each of the other lineages by an average of more than 29 nucleotide sites (Table 3). The Southern Disjunct lineage has both high haplotype and nucleotide diversity despite a very isolated geographic distribution.
|
|
Bioclimatic Modeling and Spatial Statistical Analyses
For the lineages within A. flavipunctatus, the ecological niche models had little overprediction, meaning the models did not predict broadly outside of the modeled populations, except for some common overprediction into the Sierra Nevada Mountains (Fig. 5). However, the Southern Disjunct lineage that was sampled only around Santa Clara County (Fig. 4, A on map), predicted an optimal range that extended north into the Bay Area. The PCA indicated that the phylogenetic lineages were predominately distributed in unique environmental space, with the Central and Northwest lineages showing the most overlap (Fig. 6). The Southern Disjunct lineage showed no overlap with any lineage, and the Shasta lineage had very little variation across the first two factor scores. The PCA scores differed significantly between the lineages (Wilks' lambda = 0.204; F6, 960= 194.350; P < 0.000), and post hoc tests revealed both factor 1 and factor 2 were significant. Six of the 11 climatic variables differed significantly across all lineages at P < 0.000 (Fig. 7).
|
|
|
Northwest; • Shasta; 
Southern Disjunct; 
Central lineage.
|
The partial Mantel test that examined the relationship of environmental distance and genetic distance, while controlling for geographic distance, was significant (r = 0.331; P = 0.002). To determine if the Southern Disjunct lineage was driving that significance, we tested the relationship using only the other three lineages and found that the significant relationship held (r = 0.155; P = 0.005). However, when only the Shasta and Northwest lineages were included in the analysis, the relationship between environmental distance and genetic distance was not significant (r = –0.008; P = 0.509).
Contact Zone Analyses
For the first major contact zone (Northwest and Shasta), Maxent probabilities of all NHC locality points under the Northwest ecological niche model (Fig. 5) averaged around 78% for the Northwest points, less than 50% for the Shasta points, and about 34% for the 1000 random points in the contact zone (Figs. 2, 8a). In this comparison the contact zone was even less suitable for the Northwest lineage than the parapatric sister lineage's range. The overall ANOVA was significant (F2, 269= 125.202; P < 0.001), and all pairwise comparisons were significantly different from each other. The same comparison but using the Shasta ecological niche model (Fig. 5) averaged around 88% for the Shasta points, about 20% for the Northwest points, and about 38% for the random points in the contact zone (Figs. 2, 8b). The overall ANOVA was significant (F2, 269= 274.130; P < 0.001), and all pairwise comparisons were significant. Therefore, for the Shasta lineage, the contact zone was not suitable, nor was the range of the parapatric sister lineage.
For the second major contact zone (Central versus Northwest), Maxent probabilities of all NHC locality points under the Central ecological niche model (Fig. 5) averaged around 70%, about 25% for the Northwest points, and about 45% for the random points in the contact zone (Figs. 2, 8c). The contact zone in this case is relatively unsuitable at 45% for the Central lineage, and the range of the Northwest lineage is not suitable for the Central lineage. The overall ANOVA was significant (F2, 458= 188.860; P < 0.001), and all pairwise comparisons were significant. Under the Northwest ecological niche model (Fig. 5), Northwest points averaged around 78%, Central points less than 25%, but points in the contact zone averaged around 72%. Although the overall ANOVA was significant (F2, 458= 205.514, P < 0.001), the important comparison of the contact zone points and the Northwest points did not differ (Figs. 2, 8d; P = 0.3245). Therefore, the contact zone is suitable for the Northwest lineage.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
Implications for Species Delimitation in Aneides flavipunctatus
Our analyses of the spatial data (PCA, MANOVA, partial Mantel) all support the conclusion that genetic diversity is associated with significant divergence in the ecological niche of A. flavipunctatus. The extent of the divergence across lineages varies and is a result of isolation that can be geographic or environmental. For example, in the case of the Shasta lineage, the population is not necessarily geographically isolated (as is the Southern Disjunct lineage) but it is environmentally isolated. The potential contact zone region between the Northwest and Shasta lineages is unlikely to support viable populations of either lineage, because the Maxent probability values are well under 40% in that region under both lineage niche models. This was found despite the nonsignificant association between environmental and genetic distances (controlling for geographic distance) in the partial Mantel test, suggesting that the two environments for the Shasta and Northwest lineages are not that different but these sister lineages are isolated by intervening, unsuitable habitat. However, to confirm this, more thorough sampling should be conducted in this potential contact zone. We acknowledge that there could be very limited intervening suitable habitat in this region, but to date (2007), no records of A. flavipuntatus are known from this potential contact zone.
The Central lineage is genetically quite distinct from the Shasta, Northwest, and Southern Disjunct lineages. However, the contact zone between the Northwest lineage and Central lineage is broadly suitable for the Northwest lineage, but, interestingly, not for the Central lineage. Despite suitability, the extent of gene flow in this region is entirely dependent on the level of reproductive isolation between these lineages, and we have no evidence of reproductive isolation at the current time. More thorough sampling should be conducted across the Central lineage's range to determine whether additional lineages exist and to refine the location of potential contact zones.
Currently, A. flavipunctatus is considered a distinct species, and sometimes two subspecies are recognized: A. f. flavipunctatus (speckled black salamander) and A. f. niger (Santa Cruz black salamander; Myers and Maslin, 1948). Petranka (1998) follows Lynch's (1981) recommendations to not recognize subspecies based on the conclusion that morphological and color variation is predominately clinal and ontogenetic (Lynch, 1981). Aneides f. niger corresponds to our Southern Disjunct lineage, and it is completely geographically isolated. However, other lineages within the species are not geographically isolated, and ecological niche modeling can aid in our assessment of species status, especially when morphological delineations are not clear-cut. Although morphological variation is somewhat high in this species (i.e., in degree of brassy pigmentation and diameter of white dorsal iridophores), there is no evidence that this is genetically or racially based (Lynch, 1974, 1981). Allozyme analyses have also indicated that both the Shasta lineage and Southern Disjunct lineage are distinct (Larson, 1980). Highton (2000) reanalyzed Larson's (1980) data and found that there were three groups within A. flavipunctatus (Southern Disjunct, Shasta, and remaining populations) that differed from each other by Nei D > 0.15. However, both Larson (1980) and Lynch (1974, 1981) concluded that there was insufficient evidence to elevate these populations to species status, although Highton (2000) disagreed. Interestingly, early ecological and morphological studies by Lowe (1950) in a University of California Los Angeles unpublished doctoral thesis separated A. flavipunctatus into five subspecies that roughly correspond to the genetic lineages we found. Names are available for three of these: flavipunctatus Strauch 1870 (most similar to our Central lineage), niger Myers and Maslin 1948 (our Southern Disjunct lineage), and iëcanus Cope 1883 (our Shasta lineage). Should subsequent work validate species status for all five clusters, new names will be required for the Northwest and AF28–31 lineages. Lowe's evidence was almost completely ecological and based on the separation of humid, coniferous forest habitat along the coastal populations and the drier, more openly vegetated interior populations; populations along the coast had reduced levels of white spotting, although later analyses have shown morphological variation in A. flavipuncatatus to be largely due to ontogenetic variation (Lynch, 1981).
Based on our mtDNA analyses and ecological modeling, along with the corroboration of allozyme data, we believe that the Shasta and Southern Disjunct lineages should clearly be elevated to species status, and the southern extent of the Northwest lineage should be determined before that lineage is elevated to species status. We have taxonomic publications in progress, and we suggest additional geographic and genetic sampling should be conducted to refine the limits of populations within the Central lineage, especially around Mendocino County (e.g., AF28–31, Fig. 3). There appears to be little morphological divergence across these species, and in general we suggest that genetic analyses combined with ecological data can be far stronger evidence for delimiting species than morphological data due to the ubiquitous nature of morphological homoplasy from convergence, parallelism, and reversals (e.g., Wake, 1991; Mueller et al., 2004).
|
Adding Information on the Ecological Niche to Studies of Speciation and Species Delimitation
We believe that natural selection through abiotic pressures can be a major driver of divergence and ultimately speciation, and the analyses of ecological divergence in relation to phylogenetic diversity can shed insight into biodiversity patterns and the potential processes driving those patterns. A recent analysis of over 500 comparisons across plant, invertebrate, and vertebrate taxa found a highly consistent and significant association between ecological divergence (including habitat divergence) and reproductive isolation (Funk et al., 2006). Funk et al. (2006) concluded that speciation is an "... inherently ecological process," and this has been supported by theoretical and experimental work (Mayr, 1947; Dobzhansky, 1951; Rice and Salt, 1990; Rice and Hostert, 1993; Filchak et al., 2000; Schluter, 2000; Rundle and Nosil, 2005; Dettman et al., 2007). Another recent review (Streelman and Danley, 2003) focused on the stages of vertebrate evolutionary radiation and concluded that habitat was the initial axis of divergence in many adaptive radiations, followed by divergence in morphology and communication. Therefore, we suggest that quantifying divergence in the ecological niche should be an important component of current phylogeographic studies and could be useful for species delimitation (e.g., Wiens and Graham, 2005).
In species delimitation it is important to distinguish between the primary species concept (entities believed to be species) and the secondary species concept (operational methods for the discovery of those entities; Sites and Marshall, 2003, 2004). As mentioned previously, our view of species follows the evolutionary species concept (ESC) and general lineage species concept (GLC; Wiley, 1978; de Queiroz, 1998) with the goal of recognizing historically distinct evolutionary lineages that are likely to remain distinct (e.g., Wake, 2006). The methodologies used to distinguish those lineages are likely to vary across systems because of the plethora of evolutionary processes operating within and among populations at varying spatiotemporal scales (Harrison, 1998; reviewed in Sites and Marshall, 2003, 2004). Methods are generally grouped into tree-based and non–tree-based methods, but none of the currently recognized operational criteria for delimiting species considers spatially explicit environmental (climatic) information on the niche (e.g., Sites and Marshall, 2003, 2004; Wiens and Graham, 2005; Pons et al., 2006). We argue that data on the ecological niche can be highly informative, especially in combination with more traditional phylogenetic data (e.g., mtDNA), because it can provide a surrogate for physiological adaptations as well as a direct view of the abiotic variables likely impacting divergence. Our approach, therefore, focuses on the distinctness of lineages in historical (genetic) and ecological space, rather than giving primacy to gene flow.
|
Summary |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
Amphibians, as well as many other taxa, commonly exhibit conservative morphological evolution that masks significant genetic diversity (e.g., Highton, 1995; Bickford et al., 2007). This hampers our understanding of the patterns of biodiversity and the processes that drive those patterns, and because amphibians are one of the most threatened groups of organisms on the planet (Stuart et al., 2004), a limited understanding of their biodiversity patterns can significantly affect conservation assessments.
The California Floristic Province is one of 34 world hotspots of biological diversity (Mittermeier et al., 2005), and it is a region with high endemism and concordant phylogeographic breaks that are often congruent with environmental gradients (Calsbeek et al., 2003; Lapointe and Rissler, 2005; Rissler et al., 2006). Temperature and precipitation regimes are especially likely to impact the distributions of amphibians because of their permeable skin and general physiology (Duellman and Trueb, 1986; Buckley and Jetz, 2007). When populations become isolated geographically, unique environmental conditions can drive divergence (e.g., Dobzhansky, 1940; Orr and Smith, 1998; Schluter, 2001; Funk et al., 2006). Within the range of A. flavipunctatus, environmental gradients are steeper in an east-west direction (Rissler et al., 2006), and morphological and genetic data support this (Lynch, 1974, 1981; Larson, 1980). These strong gradients in combination with the generalized phenomenon of niche conservatism (see Wiens, 2004) likely provided multiple opportunities for isolation and subsequent divergence in the early evolutionary history of A. flavipunctatus. Whether or not reproductive isolation is complete across lineages/species in A. flavipunctatus is unknown at the current time, but plethodontid salamanders can show reproductive isolation even in the absence of observable morphological or ecological changes (Highton, 1979).
We suggest that future phylogeographic and biogeographic analyses will be more explicitly concerned with the spatial and geographic components of genetic variation and speciation than is currently evident (Kidd and Richie, 2006). We believe ecological niche modeling of climatic variables has shown promise in helping to elucidate the abiotic processes impacting evolutionary lineages, but the technique is not a panacea for inadequate sampling, nor can one definitively conclude that the climatic variables that statistically differentiate two lineages are the reason for their genetic divergence. We do, however, suggest that because amphibians are physiologically constrained to particular environmental conditions (Feder and Burggren, 1992), climate is likely to play an especially important role in driving global and local biodiversity patterns. In fact, Buckley and Jetz (2007) analyzed the distributions of 5,634 of the known 6,200 amphibian species (www.amphibiaweb.org; accessed 30 August 2007) and concluded that while history plays an important role in broad scale richness patterns, environment is by far the stronger predictor of species richness within geographic realms. Future research is important to elucidate the causal relationships between environmental variables, ecological divergence in phenotypic traits, and speciation. For example, for species showing divergence in the ecological niche, further physiological and behavioral comparisons should be made across lineages. These types of analyses will become increasingly important for predictions of range distributions and extinctions during climate change, especially for species that are physiologically constrained to environmental conditions (e.g., Pounds et al., 1999; Lips et al., 2003; Buckley and Jetz, 2007).
In summary, we have demonstrated the ease and utility of combining ecological niche models, spatially explicit analyses of environmental data, and phylogeographic information to help define cryptic species. Especially for organisms displaying fine-scaled endemism and cryptic diversity, these methods should provide an objective and rigorous tool to aid in investigations of biogeography and species delimitation.
|
Acknowledgments |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
We are grateful to the Museum of Vertebrate Zoology, University of California at Berkeley, and especially Carla Cicero for tissues. David Wake and Craig Moritz provided guidance and support. Saori Haigo and Lindsey Smith were instrumental in assisting with genetic analyses; Robert Hijmans and Catherine Graham kindly assisted with discussions of bioclimatic modeling. We also thank J. Wiens, K. Kozak, and an anonymous reviewer for many helpful comments on this manuscript. Funding was provided by the National Science Foundation (DEB 0414033 and DEB 0074509) and Museum of Vertebrate Zoology (Kellog Award) to LJR.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionSummaryAcknowledgmentsReferences |
|---|
-
Agapow P. M., Bininda-Edmonds O. R. P., Crandall K. A., Gittleman J. L., Mace G. M., Marshall J. C., Purvis A. The impact of species concept on biodiversity studies. Q. Rev. Biol. (2004) 79:161–179.[CrossRef][Medline]
Andersson L. The driving force: Species concepts and ecology. Taxon (1990) 39:375–382.[CrossRef][Web of Science]
Araújo M. B., Guisan A. Five (or so) challenges for species distribution modeling. J. Biogeogr. (2006) 33:1677–1688.[CrossRef]
Araújo M. B., Thuiller W., Pearson R. G. Climate warming and the decline of amphibians and reptiles in Europe. J. Biogeogr. (2006) 33:1712–1728.[CrossRef]
Bickford D., Lohman D. J., Sodhi N. S. P. K. L., Ng Meier R., Winke K., Ingram K. K., Das I. Cryptic species as a window on diversity and conservation. Trends Ecol. Evol. (2006) 22:148–155.[CrossRef]
Blackburn L., Nanjappa P., Lannoo M. J. U.S. amphibian distribution maps (2001) http://home.bsu.edu/home/00mjlannoo/.
Brooks T. M., Mittermeier R. A., Fonsenca G., Gerlach J., Hoffman M., Lamoreux J., Mittermeier C., Pilgrim J., Rodrigues A. Global biodiversity conservation priorities. Science (2006) 313:58–61.
Buckley L. B., Jetz W. Environmental and historical constraints on global patterns of amphibian richness. Proc. R. Soc. Lond. B (2007) 274:1167–1173.[Medline]
Calsbeek R., Thompson J., Richardson J. Patterns of molecular evolution and diversification in a biodiversity hotspot: the California Floristic Province. Mol. Ecol. (2003) 12:1021–1029.[CrossRef][Medline]
Casgrain P., Legendre P. The R-Package for multivariate and spatial analysis, version 4.0 (2001) http://www.fas.umontreal.ca/biol/casgrain/en/labo/R/index.html.
Chase J. M., Leibold M. A. Ecological niches: Linking classical and contemporary approaches (2003) Chicago, Illinois: University of Chicago Press.
de Queiroz K. The general lineage concept of species, species criteria, and the process of speciation. In: Endless forms: Species and speciation—Howard D. J., Berlocher S. H., eds. (1998) Oxford, UK: Oxford University Press. 57–75.
Dettman J. R., Sirjusingh C., Kohn L. M., Anderson J. B. Incipient speciation by divergent adaptation and antagonistic epistasis in yeast. Nature (2007) 447:585–588.[CrossRef][Web of Science][Medline]
Dobzhansky T. Speciation as a stage in evolutionary divergence. Am. Nat. (1940) 74:312–321.[CrossRef][Web of Science]
Dobzhansky T. H. Genetics and the origin of species (1951) 3rd edition. New York: Columbia University Press.
Duellman W., Trueb L. Biology of amphibians (1986) Baltimore, Maryland: John Hopkins University Press.
Elith J., Graham C., Anderson R., Dudík M., Ferrier S., Guisan A., Hijmans R. J., Huettman F., Leathwick J., Lehmann A., Li J., Lohmann L., Loiselle B., Manion G., Moritz C., Nakamura M., Nakazawa Y., Overton J., Peterson A., Phillips S., Richardson K., Scachetti-Pereira R., Schapire R., Soberón J., Williams S., Wisz M., Zimmermann N. Novel methods improve prediction of species' distributions from occurrence data. Ecography (2006) 29:129–151.
Feder M., Burggren W. Environmental physiology of the amphibians (1992) Chicago, Illinois: University of Chicago Press.
Filchak K. E., Roethele J. B., Feder J. L. Natural selection and sympatric divergence in the apple maggot, Rhagoletis pomonella. Nature (2000) 407:739–742.[CrossRef][Medline]
Funk D. J., Nosil P., Etges W. J. Ecological divergence exhibits consistently positive associations with reproductive isolation across disparate taxa. Proc. Nat. Acad. Sci. USA (2006) 103:3209–3213.
Graham C. H., Ferrier S., Huettman F., Moritz C., Peterson A. New developments in museum-based informatics and applications in biodiversity analysis. Trends Ecol. Evol. (2004b) 19:497–503.[CrossRef][Medline]
Graham C. H., Ron S., Santos J., Schneider C., Moritz C. Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution (2004a) 58:1781–1793.[Web of Science][Medline]
Grinnell J. The niche-relationships of the California Thrasher. Auk (1917) 34:427–433.
Grinnell J. Geography and evolution. Ecology (1924) 5:225–229.[CrossRef][Web of Science]
Harrison R. G. Linking evolutionary pattern and process: The relevance of species concepts for the study of speciation. In: Endless forms: species and speciation—Howard D. J., Berlocher S. H., eds. (1998) Oxford, UK: Oxford University Press. 19–31.
Highton R. A new cryptic species of salamander of the genus Plethodon from the southeastern United States (Amphibia: Plethodontidae). Brimleyana (1979) 1:31–36.
Highton R. Speciation in eastern North American salamanders of the genus Plethodon. Ann. Rev. Ecol. Syst. (1995) 26:579–600.[CrossRef][Web of Science]
Highton R. Detecting cryptic species using allozyme data. In: The biology of plethodontid salamanders—Bruce R. C., Jaeger R. G., Houck L. D., eds. (2000) New York, New York: Kluwer Academic/Plenum Publishers. 215–241.
Hijmans R. J., Cameron S., Parra J., Jones P., Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int. J. Clim. (2005) 25:1965–1978.[CrossRef]
Hijmans R. J., Graham C. H. The ability of climate envelope models to predict the effect of climate change on species distributions. Global Change Biol. (2006) 12:1–10.[CrossRef]
Huelsenbeck J. P., Ronquist F. MrBayes: Bayesian inference of phylogeny. Bioinformatics (2001) 17:754–755.
Hugall A., Moritz C., Moussalli A., Stanisic J. Reconciling paleodistribution models and comparative phylogeography in the Wet Tropics rainforest land snail Gnarosophia bellendenkerensis (Brazier 1875). Proc. Nat. Acad. Sci. USA (2002) 99:6112–6117.
Hutchinson G. E. Concluding remarks. Cold Springs Harbor Symp. Quant. Biol. (1957) 22:415–427.
Johnson N. K., Cicero C. The role of ecologic diversification in sibling speciation of Empidonax flycatchers (Tyrannidae): Multi-gene evidence from mtDNA. Mol. Ecol. (2002) 11:2065–2081.[Medline]
Kearney M. Habitat, environment, and niche: What are we modeling? Oikos (2006) 115:186–191.[CrossRef][Web of Science]
Kearney M., Porter W. Mapping the fundamental niche: Physiology, climate and the distribution of nocturnal lizards across Australia. Ecology (2004) 85:3119–3131.[CrossRef][Web of Science]
Kidd D., Ritchie M. Phylogeographic information systems: putting the geography into phylogeography. J. Biogeogr. (2006) 33:1851–1865.[CrossRef]
Köhler J., Vieites D., Bonett R., Garcia F., Glaw F., Steinke D., Vences M. New amphibians and global conservation: A boost in species discoveries in a highly endangered vertebrate group. Bioscience (2005) 55:693–696.[CrossRef][Web of Science]
Kozak K. H., Wiens J. J. Does niche conservatism promote speciation? A case study in North American salamanders. Evolution (2006) 60:2604–2621.[Web of Science][Medline]
Lapointe F. J., Rissler L. J. Congruence, consensus, and the comparative phylogeography of co-distributed species in California. Am. Nat. (2005) 166:290–299.[CrossRef][Web of Science][Medline]
Larson A. Paedomorphosis in relation to rates of morphological and molecular evolution in the salamander Aneides flavipunctatus (Amphibia, Plethodontidae). Evolution (1980) 34:1–17.[CrossRef][Web of Science]
Lips K. R., Reeve J. D., Witters L. R. Ecological traits predicting amphibian population declines in Central America. Cons. Biol. (2003) 17:1078–1088.[CrossRef]
Lowe C. H. J. Speciation and ecology in salamanders of the genus Aneides (1950) Los Angeles: University of California. PhD dissertation.
Lynch J. F. Ontogenic and geographic variation in the morphology and ecology of the black salamander (Aneides flavipunctatus) (1974) University of California Berkeley. PhD dissertation.
Lynch J. F. Patterns of ontogenetic and geographic variation in the black salamander (Aneides flavipunctatus). Smiths. Contrib. Zool. (1981) 324:1–53.
Mayr E. Ecological factors in speciation. Evolution (1947) 1:263–288.[CrossRef][Web of Science]
Mittermeier R. A., Gil P. R., Hoffman M., Pilgrim J. D., Brooks T. M., Mittermeier C. G., Lamoreux J. F., da Fonseca D. A. Hotspots revisited: Earth's biologically richest and most endangered terrestrial eco-regions (2005) Washington, DC: Conservation International.
Mueller R. L., Macey J. R., Jaekel M., Wake D. B., Boore J. L. Morphological homoplasy, life history evolution, and historical biogeography of plethodontid salamanders inferred from complete mitochondrial genomes. Proc. Nat. Acad. Sci. USA (2004) 101:13820–13825.
Myers G. S., Maslin T. P. The California plethodont salamander Aneides flavipunctatus (Strauch) with description of a new subspecies and notes on other western Aneides. Proc. Biol. Soc. Washington (1948) 61:127–138.
Orr M. R., Smith T. B. Ecology and speciation. Trends Ecol. Evol. (1998) 13:502–506.[CrossRef]
Peterson A. T. Predicting the geography of species invasions via ecological niche modeling. Q. Rev. Biol. (2003) 78:419–433.[CrossRef][Medline]
Peterson A. T., Ortega-Huerta M. A., Bartley J., Sanchez-Cordero V., Soberón J., Buddemeier R. H., Stockwell D. R. B. Future projections for Mexican faunas under global climate change scenarios. Nature (2002) 416:626–629.[CrossRef][Medline]
Peterson A. T., Robins C. R. Using ecological-niche modeling to predict barred owl invasions with implications for spotted owl conservation. Cons. Biol. (2003) 17:1161–1165.[CrossRef]
Peterson T. A., Soberon J., Sanchez-Cordero V. Conservatism of ecological niches in evolutionary time. Science (1999) 285:1265–1267.
Petranka J. W. Salamanders of the United States and Canada (1998) Washington, DC: Smithsonian Institution Press.
Phillips S. J., Anderson R. P., Schapire R. E. Maximum entropy modeling of species geographic distribution. Ecol. Mod. (2006) 190:231–259.[CrossRef][Web of Science]
Phillips S. J., Dudik M., Schapire R. E. A maximum entropy approach to species distribution modeling. Proceedings of the Twenty-First International Conference on Machine Learning (2004) 69:83–99.
Pons J., Barraclough T. G., Gomez-Zurita J., Cardoso A., Duran D. P., Hazell S., Kamoun S., Sumlin W. D., Vogler A. P. Sequence-based species delimitation for the DNA taxonomy of under described insects. Syst. Biol. (2006) 55:595–609.
Posada D., Crandall K. A. ModelTest: Testing the model of DNA substitution. Bioinformatics (1998) 14:817–818.
Pounds J. A., Fogden M. P. L., Campbell J. H. Biological response to climate change on a tropical mountain. Nature (1999) 398:611–615.[CrossRef][Web of Science]
Raxworthy C. J., Martinez-Meyer E., Horning N., Nussbaum R. A., Schneider G. E., Ortega-Huerta M. A., Peterson T. A. Predicting distributions of known and unknown reptile species in Madagascar. Nature (2003) 426:837–841.[CrossRef][Medline]
Rice W. R., Hostert E. E. Laboratory experiments on speciation: What have we learned in 40 years? Evolution (1993) 47:1637–1653.[CrossRef][Web of Science]
Rice W. R., Salt G. W. The evolution of reproductive isolation as a correlated character under sympatric conditions: Experimental evidence. Evolution (1990) 44:1140–1152.[CrossRef][Web of Science]
Rissler L. J., Hijmans R. J., Graham C. H., Moritz C., Wake D. B. Phylogeographic lineages and species comparisons in conservation analysis: a case study of California herpetofauna. Am. Nat. (2006) 167:655–666.[CrossRef][Web of Science][Medline]
Rissler L. J., Taylor D. R. The phylogenetics of desmognathine salamander populations across the southern Appalachians. Mol. Ecol. (2003) 27:197–211.
Rozas J., Rozas R. An integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics (1999) 15:174–75.
Rundle H., Nosil P. Ecological speciation. Ecol. Lett. (2005) 8:336–352.[CrossRef][Web of Science]
Schluter D. The ecology of adaptive radiation (2000) Oxford, UK: Oxford University Press.
Schluter D. Ecology and the origin of species. Trends Ecol. Evol. (2001) 16:372–380.[CrossRef][Medline]
Sites J. W. J., Marshall J. C. M. Delimiting species: A renaissance issue in systematic biology. Trends Ecol. Evol. (2003) 18:462–470.[CrossRef]
Sites J. W. J., Marshall J. L. Operational criteria for delimiting species. Ann. Rev. Ecol. Syst. (2004) 35:199–227.[CrossRef][Web of Science]
Soberón J., Peterson A. T. Interpretation of models of fundamental ecological niches and species' distributional areas. Biodiv. Infor. (2005) 2:1–10.
Streelman T. J., Danley P. D. The stages of vertebrate evolutionary radiation. Trends Ecol. Evol. (2003) 18:126–131.[CrossRef]
Stuart S. N., Chanson J. S., Cox N. A., Young B. E., Rodrigues A. S. L., Fischman D. L., Waller R. W. Status and trends of amphibian declines and extinctions worldwide. Science (2004) 306:1783–1786.
Swofford D. L. PAUP*: Phylogenetic analysis using parsimony (*and other methods) (2000) Sunderland, Massachusetts: Sinauer Associates.
Templeton A. R. Using phylogeographic analyses of gene trees to test species status and processes. Mol. Ecol. (2001) 10:779–791.[CrossRef][Medline]
Thomas C. D., Cameron A., Green R. E., Bakkenes M., Beaumont L. J., Collingham Y. C., Erasmus B. F. N., de Siqueira M. F., Grainger A., Hannah L., Hughes L., Huntley B., van Jaarsveld A. S., Midgley G. F., Miles L., Ortega-Huerta M. A., Peterson A. T., Phillips O. L., Williams S. E. Extinction risk from climate change. Nature (2004) 427:145–148.[CrossRef][Medline]
Van Valen L. Ecological species, multispecies, and oaks. Taxon (1976) 25:233–239.[CrossRef][Web of Science]
Wake D. B. Homoplasy—The result of natural selection, or evidence of design limitations. Am. Nat. (1991) 138:543–567.[CrossRef][Web of Science]
Wake D. B. Problems with species: Patterns and processes of species formation in salamanders. Ann. Miss. Bot. Gard. (2006) 93:8–23.[CrossRef]
Wiens J. J. What is speciation and how should we study it? Am. Nat. (2004) 163:914–923.[CrossRef][Web of Science][Medline]
Wiens J. J., Graham C. H. Niche conservatism: Integrating evolution, ecology, and conservation biology. Ann. Rev. Ecol. Evol. Syst. (2005) 36:519–539.[CrossRef]
Wiens J. J., Graham C. H., Moen D. S., Smith S. A., Reeder T. W. Evolutionary and ecological causes of the latitudinal diversity gradient in Hylid frogs: Treefrog trees unearth the roots of high tropical diversity. Am. Nat. (2006) 168:579–596.[CrossRef][Web of Science][Medline]
Wiens J. J., Penkrot T. A. Delimiting species using DNA and morphological variation and discordant species limits in spiny lizards (Sceloporus). Syst. Biol. (2002) 51:69–91.
Wiley E. O. The evolutionary species concept reconsidered. Syst. Zool. (1978) 27:17–26.
Zwickl D. J. Genetic algorithm for rapid likelihood inference. Program distributed by editor (2006) http://www.bio.utexas.edu/faculty/antisense/garli/Garli.html.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. H. Parks, M. Porter, S. Churcher, S. Wang, C. Blouin, J. Whalley, S. Brooks, and R. G. Beiko GenGIS: A geospatial information system for genomic data Genome Res., October 1, 2009; 19(10): 1896 - 1904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Leache, M. S. Koo, C. L. Spencer, T. J. Papenfuss, R. N. Fisher, and J. A. McGuire From the Cover: Quantifying ecological, morphological, and genetic variation to delimit species in the coast horned lizard species complex (Phrynosoma) PNAS, July 28, 2009; 106(30): 12418 - 12423. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Bond and A. K. Stockman An Integrative Method for Delimiting Cohesion Species: Finding the Population-Species Interface in a Group of Californian Trapdoor Spiders with Extreme Genetic Divergence and Geographic Structuring Syst Biol, August 1, 2008; 57(4): 628 - 646. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||