© 2002 by The Society for Integrative and Comparative Biology
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
Methods in Comparative Phylogeography, and Their Application to Studying Evolution in the North American Aridlands1
1 J. F. Bell Museum, 100 Ecology Building, University of Minnesota, St. Paul, Minnesota 55108
 |
SYNOPSIS |
|---|
TOP
 SYNOPSIS INTRODUCTIONCONCLUSIONSReferences |
|---|
Historical biogeography and comparative phylogeography have much in common. Both seek to discover common historical patterns in the elements of biotas, although typically at different tiers of evolutionary history. Comparative phylogeography is based on phylogeographic analyses of multiple taxa, usually widespread species. By comparing the phylogeographic structures of numerous widespread sympatric species, one can infer whether the current fauna has been historically stable, as evidenced by the relative frequency of geographically congruent reciprocally monophyletic groups. Alternatively, if species distributions are ephemeral over evolutionary time, a mixture of phylogeographic structures is expected. Coalescence analyses contribute information about history irrespective of whether haplotype phylogenies are structured or not. In the aridlands of North America, several isolating events are evident in the phylogeographic patterns of birds, mammals and herps. A mid-peninsular seaway in Baja California, dated at ca. one million years before present, had a pervasive effect, with 13 of 16 assayed species showing a concordant split. Hence, this community appears to have been a stable assemblage of species over the past one million years. In contrast, the avifauna of the Sonoran-Chihuahuan deserts consists of two species with a concordant split and three other species that are undifferentiated across both deserts. Hence, the species in this area have had different histories. The Baja biota appears to resemble its ancestral configuration to a greater degree than the Sonoran-Chihuahuan one. A deeper evolutionary event separated taxa in Baja California from the eastern deserts, showing that the aridlands fauna was affected by events at different times resulting in overlain tiers of history.
|
INTRODUCTION |
|---|
|
TOP
SYNOPSISINTRODUCTIONCONCLUSIONSReferences |
|---|
Diversification at all taxonomic levels has a spatial and temporal component. Historical biogeography and phylogeography represent analyses at different points on the continuum of evolutionary divergence—differing tiers of history. Historical biogeography deals with phylogenetic patterns among species and higher lineages attributable to relatively ancient events in earth history. Phylogeography, which some consider synonymous with biogeography (Arbogast and Kenagy, 2001
), deals with the origins and relationships of taxa at the lowest level of taxonomic organization (individuals within and among populations). I believe there is a distinction between phylogeography and biogeography. Phylogeography can involve historical inference even if phylogenetically structured trees are lacking, whereas biogeography typically requires resolved trees. However, the distinction between biogeography and phylogeography blurs when one discovers that a "species" includes multiple reciprocally monophyletic groups related in a hierarchical way. Not surprisingly, there are parallels in the theory and methods employed (Zink, 1996).
Phylogeography and biogeography are founded on phylogenetic analyses. Both are based on analyses of single taxa (clades in biogeography, "species" in phylogeography), but they are most powerful when comparisons of multiple lineages and species are involved. Historical (vicariance) biogeography seeks to find congruent phylogenetic patterns in multiple clades found in a common region. Congruence is interpreted to mean that a widespread ancestral biota was fragmented by a series of successive vicariant events (Avise, 1992). Comparative phylogeography (Bermingham and Avise, 1986; Zink, 1996) seeks to find concordant splits within contemporaneous, co-distributed species. The degree of phylogeographic concordance is a measure of the historical stability of the current species assemblage. Both historical biogeography and comparative phylogeography reveal the historical stability of communities (assemblages of clades, groups of species) at different temporal scales.
Congruent phylogenetic or phylogeographic patterns provide relatively straightforward inferences. Biogeography and comparative phylogeography differ in their potential to explain incongruent patterns, owing to the disparate time scales. Dispersal, extinctions, and overlapping events in earth history can obscure biogeographic history (Cracraft, 1988). Reasons why clades have incongruent phylogenetic histories are difficult to ascertain because of the age of the events—the trace grows colder with time. In contrast, comparative phylogeography deals with relatively recent events, and genetic signatures often reveal species histories irrespective of the geographic pattern of haplotype relationships. Thus, it is possible to decipher the history of each species whether the haplotype tree is structured or not (Templeton, 1998). In this paper, I explore methods in comparative phylogeography, and how the distribution of phylogeographic structures reveals the historical stability of community species composition, using examples from the aridlands fauna of North America.
Phylogeography
Avise et al. (1987) coined the term phylogeography for the process of inferring the phylogenetic relationships among individual sequences (haplotypes) and superimposing the resulting haplotype phylogeny over the geographic locations of the samples to reveal population history. They recognized four types of phylogeographic structures: Type I—deep haplotype trees that are geographically structured; Type II—deep haplotype trees that are unstructured; Type III—shallow haplotype trees showing geographic structure; and Type IV—shallow unstructured trees. Each type is consistent with a particular history. Geographically structured groups of haplotypes (Types I, III) in an area typically result from isolation due to environmental or ecological barriers. In contrast a Type IV pattern is consistent with recent population expansion and a lack of isolating barriers. Thus, one sometimes obtains haplotype trees with "lackluster" resolution; however, assuming sufficient variation exists to diagnose haplotypes, such unstructured trees provide meaningful evidence of recent population history. No amount of sequence would change such a tree to a structured one—population growth and gene flow result in a bush-like topology in trees of individual haplotypes (Harpending et al., 1998). It is often instructive for systematists working at higher levels, who are uncomfortable with unresolved trees, to note that lack of resolution results because variation consists mostly of autapomorphies, and not conflicting synapomorphies.
Despite the seeming objectivity of the four types (Kuchta and Meyer 2001), in reality only one objective division exists. Either the haplotype tree exhibits a geographic pattern of reciprocal monophyly or not. Reciprocally monophyletic groups are the currency of phylogeography. Whether two reciprocally monophyletic sister groups differ by 0.5% (e.g., Type III) or 5% (Type I), reciprocal monophyly allows one to falsify the hypothesis that they are exchanging genes. This does not mean, however, that only reciprocally monophyletic groups are interesting. It is important to document all phylogeographic structures. For species without structure, further inferences about recent population history can be made by application of coalescence methods (Crandall and Templeton, 1996; Posada and Crandall, 2001). It is just as important to know that a group's history is one of population and range expansion as it is to document a history of isolation.
For example, in a phylogeographic study of the Greenfinch (Carduelis chloris), Merila et al. (1997) obtained a geographically unstructured haplotype tree (Type IV), leading to the inference that segments of the species have not undergone significant isolation. The authors discovered a significant northward decrease in nucleotide diversity from Spain to northern Finland. Declines in genetic diversity, presumably owing to sequential bottlenecks during recolonization, are signatures of both the magnitude and direction of population expansion into recently deglaciated areas (Hewitt, 2000). This phenomenon is termed "leading edge expansion" and results from successive sampling of haplotypes from the leading edge of the range expansion. Thus, the shallow, unstructured haplotype tree resulted from recent northward population expansion. The combination of phylogenetic analysis of haplotypes and coalescence analyses provides powerful tools for investigating recent population history.
Comparative phylogeography
Comparison of phylogeographic patterns among broadly sympatric species shows whether species have responded in parallel to recent isolating events. My aim is to introduce a framework (Fig. 1) for comparative phylogeography. The first step is to ascertain if reciprocally monophyletic groups exist within a sample of taxa (typically species) in a widespread community. The next step is to assess if the groups are sister groups and whether the reciprocally monophyletic groups are geographically concordant. If so, one can determine how long the groups have been isolated, either in a relative sense, or by invoking a molecular clock (Klicka and Zink, 1999). Next, assuming there is congruence indicating a historical barrier to at least some taxa, one examines reasons why other species in the current fauna do not show reciprocal monophyly across the same putative barrier. Either species were widespread and simply did not respond to the barrier that caused differentiation in other species, or they recently expanded their range across the barrier. Assuming that the barrier is not too old, genetic signatures can distinguish these two alternatives (Moritz, 1996; Crandall and Templeton, 1996). All things being equal, if the species has been widespread for a long time, nucleotide diversity should be equivalent on both sides of the barrier, the mismatch distribution should be ragged, and both interior and terminal haplotypes should occur on both sides of the barrier. If the species recently expanded its range, genetic signatures on either side of the barrier should differ, with the recently colonized area being less variable and exhibiting a Poisson-like mismatch distribution.
|
Comparative phylogeography reveals the historical stability of the current community. If the current species pool was historically coherent, widespread, and bisected by an isolating barrier, a preponderance of geographically congruent, reciprocally monophyletic patterns should be recovered. A preponderance of Type IV patterns suggests that species composition of the community has been fluid, with ranges shifting over time scales that preclude the evolution of reciprocally monophyletic groups (e.g., Taberlet et al., 1998
). Species with genetic signatures of recent expansion might be relatively recent arrivals in the current community. Alternatively, they might have been widespread and gone through a bottleneck followed by expansion. I would argue that if a large number of species show a consistent pattern, an inference can be drawn. For example, if many species in an area showed signatures of recent population expansion and shallow haplotype trees, it would not be parsimonious to conclude that they had all simultaneously recovered from bottlenecks. Thus, the power of comparative phylogeography derives from analysis of many co-distributed species. By tabulating the distribution of phylogeographic types one gains an understanding of a fauna's recent historical stability, and by comparing distributions among communities, one can infer which current communities have changed the least in species composition over the recent past (Fig. 2). Thus comparative phylogeography (Figs. 1, 2) offers a means to diagnose the history of biotic stability.
|
Phylogeographic patterns in the Aridlands of North America
The North American deserts (Fig. 3) provide a useful context in which to explore methods in comparative phylogeography. The deserts apparently act as areas of endemism, resulting in some clear patterns of diversification. A relatively large number of bird and mammal species have been studied (Riddle et al., 2000a,
b, c; Zink et al., 2001), which allow some generalizations about the history of faunas at different spatial scales. Below I consider three apparent isolating barriers and their effects on the constituent communities.
|
Mid-peninsular seaway
In Baja California, 13 of 16 taxa (Table 1) show reciprocally monophyletic groups that meet today at 28–30° N (Fig. 3, areas A and B). A mid-peninsular seaway was presumed to exist 1 MYBP, isolating the ancestral community in southern and central Baja from northern and eastern populations (Riddle et al., 2000a
; but see Grismer, 2002). Genetic divergence across this region is relatively high for all 13 taxa, but there is also considerable variation in levels of divergence. Because different genes were surveyed for many taxa, there is little reason to explore reasons for the variation. Suffice it to say that a significant number of taxa of mammals, herps and birds show a concordant phylogeographic division in Baja California.
|
Three taxa show no phylogeographic divisions: Polioptila californica (California Gnatcatcher), Callipepla californica (California Quail), and Bufo punctatus (Red-spotted Toad). Using Figure 1, it is possible to ascertain why these taxa appear genetically homogeneous across the region where many taxa were apparently sundered by the mid-peninsular seaway. Polioptila californica was examined in detail by Zink et al. (2000
, 2001). High southern nucleotide diversity, basal southern haplotypes, and the shape of mismatch distributions, all led to the inference that upon closure of the seaway, the gnatcatcher dispersed northwards unimpeded by conspecific individuals. Thus, it appears that the gnatcatcher does not show the division common to many other taxa because its ancestral population did not exist on both sides of this seaway. Because it is a relatively recent addition to the community in the northern part of its range, one might make different predictions about this species north and south of the seaway. For example, ecological interactions with competing species have existed longer in the south than in the north.
The other avian species that apparently shows no break is C. californica. However, the only data for this species are allozymes, which in birds are less likely to reveal geographic variation than mtDNA (Zink, 1997). Examination of the data in Zink et al. (1987), however, suggests that the southern populations are more variable than those to the north. Hence, the quail and the gnatcatcher might both be relatively recent arrivals in the north.
Data for Bufo punctatus are currently unavailable (only the phylogenetic tree of haplotypes was published by Riddle et al. [2000a]), and inspection of the pattern of variability is needed to infer the history of this interesting species.
Thus, comparisons of species in Baja California show a preponderance of species with geographically concordant reciprocally monophyletic groups, suggesting that an ancestral community was fragmented by a common isolating event. The current species composition was probably stable for a million years. This conflicts with the notion (Brown 1995) that species distributions are too ephemeral to yield congruent phylogeographic splits. Furthermore, it appears that species lacking the split have recently expanded northwards from a southern Baja refugium.
Sonoran-Chihuahuan deserts
Five putative biological species of birds, whose current distributions span the Sonoran and Chihuahuan deserts (Table 2; areas C–D in Fig. 3), were studied by Zink et al. (2001). Toxostoma curvirostre (Curve-billed Thrasher) and Pipilo fuscus (Canyon Towhee) show congruent splits, which were apparently contemporaneous given the similarity in levels of sequence divergence (ca. 2% in Cytochrome b). Three species, Polioptila melanura (Black-tailed Gnatcatcher), Auriparus flaviceps (Verdin), and Campylorhynchus brunneicapillus (Cactus Wren), show no differentiation across this region. Genetic signatures can be examined to infer why there are no splits in these species (Fig. 1). The Chihuahuan Desert population of A. flaviceps is genetically depauperate, with only two haplotypes, and a nucleotide diversity of 0.0002. In contrast, the 18 western individuals represent 13 haplotypes, which do not sort phylogenetically by locality, but exhibit a nucleotide diversity of 0.0062. The large difference in nucleotide diversity suggests that the eastern Verdin populations are relatively recent (or recently underwent a bottleneck). That is, they likely only recently expanded into the Chihuahuan desert across whatever barrier isolated ancestral populations of T. curvirostre and P. fuscus. The western populations appear to be growing, whereas for the Chihuahuan groups, 
0 and 1 are of similar magnitude, suggesting that the population has not experienced recent growth.
|
In Polioptila melanura, the eastern and western groups are of similar nucleotide diversity (Table 2), and both groups appear to be significantly growing (
1 >> 0). A haplotype network (Fig. 4) shows that both interior and terminal haplotypes are intermingled in both deserts, as one would expect if there has been interchange for a considerable period. Populations of C. brunneicapillus differ by a factor of two in nucleotide diversity, with the Chihuahuan desert group most variable, although large standard errors (not shown) preclude definitive conclusions. The similarity of the theta values for both Chihuahuan and Sonoran populations of wrens suggests that neither experienced recent growth (Zink et al., 2001 pooled these samples, which suggested population growth). However, the mismatch distributions (Fig. 5) in each desert area are quite different, with that for the Sonoran samples being unimodal, suggesting growth, but for the Chihuahuan samples it is flatter, which is consistent with stable population size. Thus, it is likely that populations on either side of the barrier had different histories.
|
|
The two species that show reciprocally monophyletic groups can also be examined to determine if the reciprocally monophyletic groups in each desert have had similar recent histories. Populations of T. curvirostre have similar levels of nucleotide diversity and modest population growth, suggesting no major differences in recent population histories on either side of the barrier (the coalescence point within each taxon is more recent than that between them, meaning that some early history is lost in the gene tree). Pipilo fuscus appears to have much lower nucleotide diversity in the Chihuahuan Desert, although the small sample size precludes further analysis.
The five bird species show a mixture of phylogeographic types and population histories, with two congruent splits and three species lacking structure. Some taxa appear to have undergone recent growth whereas others appear relatively stable. This community therefore consists of species with differing recent histories, that is, the current configuration of species is likely a recent one.
Relatively few nonavian species have been surveyed (Table 2), but two mammals and one herp show differentiation across this region. Perhaps the avian community will show a variety of responses to isolating events, whereas other less vagile taxa will be consistently divided.
Sonoran desert—Baja California
The third major isolating event was discussed in detail by Zink et al. (2000). Six of six avian lineages have sister species in the Sonoran Desert and to the west in Baja California/California (areas C and D versus A and B) suggesting a common vicariant event. The degrees of divergence range from under 2% to over 6%, which might be attributable to differences in rates (use of different genes precludes inference of rates versus differing isolating barriers). Several of the sister species have ranges that span the two isolating events discussed above. Pipilo fuscus and P. crissalis each appear to be differentiated themselves, although at a level roughly equal to half of the sequence divergence between them. Alternatively, Polioptila melanura and P. californica differ by nearly 5%, suggesting a long period of independent evolution, but neither exhibits phylogeographic structure. Possibly, these species are less likely to respond to isolating events because of their dispersal capabilities. In general, the divergences between the taxa that span this barrier appear older than the previous two, showing different tiers of diversification in the aridlands of North America.
|
CONCLUSIONS |
|---|
|
TOP
SYNOPSISINTRODUCTIONCONCLUSIONSReferences |
|---|
The aridland fauna shows evidence of at least three major isolation events. The mid-peninsular seaway, and the Sonoran-Baja events appear pervasive. The split between Sonoran and Chihuahuan populations is less extensive, appearing in two of five species of birds, and potentially higher percentages of mammals and herps. Interestingly, some species, such as C. brunneicapillus and A. flaviceps, show the mid-peninsular split, but not the Sonoran-Chihuahuan split. Some species, like T. curvirostre, does not occur in all regions, and its nearest congener in Baja is several phylogenetic nodes away (Zink et al., 1999
). No taxon shows evidence of all three splits, showing considerable flux in species ranges over a longer period. Thus, different tiers of history are evident in the aridlands fauna.
Ecologists have long been interested in ascertaining if the species composition of modern biotas has been stable (Brown, 1995). There is clear evidence that many plant species have undergone dramatic range shifts in the past 15,000 years in North America. In fact, some species that today do not occur sympatrically were once part of the same communities. If this were the rule, that community composition is ephemeral, then one would not expect many congruent reciprocally monophyletic patterns. If species "come and go" from communities over relatively short spans of time, then one would expect idiosyncratic patterns, or no patterns at all.
The Baja California and Sonoran-Chihuahuan communities apparently exhibit different histories. The Baja California fauna appears to have been stable and widespread for at least one million years. It exhibits a high proportion of geographically concordant, reciprocally monophyletic groups across the species in the community. In contrast, the birds today found across the Chihuahuan and Sonoran deserts exhibit different histories, including common vicariance (Toxostoma curvirostre, Pipilo fuscus), range expansion (A. flaviceps), and possibly long-term stable, interconnected populations (Polioptila melanura, C. brunneicapillus). These five species have likely not had a long history of co-association in both deserts; more data are needed to infer if a consistent colonization direction occurred. Thus, the Baja community has changed less over recent history than the Sonoran-Chihuahuan one (Fig. 2).
One future direction for comparative phylogeography is to compile the distributions of phylogeographic types for faunas. The relative percentages of phylogeographic histories provide a record of the historical stability of species ranges, and whether they responded in common to isolating events. Comparisons of distributions of phylogeographic types among faunas reveal which current ones most closely resemble historical configurations.
|
ACKNOWLEDGMENTS |
|---|
Many individuals are thanked in the papers that this analysis is based on; I repeat my thanks to them. B. Riddle, P. Brito, A. Jones, and A. Kessen provided comments that improved this manuscript. Funding was provided by the NSF (DEB-9317945).
|
FOOTNOTES |
|---|
1 From the Symposium Integrated Approaches to Biogeography: Patterns and Processes on Land and in the Sea presented at the Annual Meeting of the Society for Integrative and Comparative Biology, 2–6 January 2002, at Anaheim, California.
2 E-mail: rzink{at}biosci.umn.edu
|
References |
|---|
|
TOP
SYNOPSISINTRODUCTIONCONCLUSIONSReferences |
|---|
Arbogast, B. S., and G. J. Kenagy. 2001. Comparative phylogeography as an integrative approach to historical biogeography. J. Biogeography, 28:819-825.[CrossRef]
Avise, J. C. 1992. Molecular population structure and the biogeographic history of a regional fauna: A case history with lessons for conservation biology. Oikos, 63:62-76.[CrossRef][Web of Science]
Avise, J. C., J. Arnold, R. M. Ball, E. Bermingham, T. Lamb, J. E. Neigel, C. A. Reeb, and N. C. Saunders. 1987. Intraspecific phylogeography: The mitochondrial DNA bridge between population genetics and systematics. Ann. Rev. Ecol. Syst. , 18:489-522.[Web of Science]
Bermingham, E., and J. C. Avise. 1986. Molecular zoogeography of freshwater fishes in the southeastern United States. Genetics, 113:939-965.
Brown, J. H. 1995. Macroecology. Univ. Chicago Press, Chicago.
Clement, M., D. Posada, and K. A. Crandall. 2000. TCS: A computer program to estimate gene genealogies. Mol. Ecol, 9:1657-1660.[CrossRef][Medline]
Cracraft, J. 1988. Deep-history biogeography: Retrieving the historical pattern of evolving continental biotas. Syst. Zool, 37:221-236.
Crandall, K. A., and A. R. Templeton. 1996. Applications of intraspecific phylogenetics. In P. H. Harvey, A. J. L. Brown, J. M. Smith, and S. Nee (eds.), New uses for new phylogenies, pp. 81–99. Oxford Univ Press, New York.
Fu, Y.-X. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics, 147:915-925.[Abstract]
Grismer, L. L. 2002. A re-evaluation of the evidence for a Mid-Pleistocene Mid-Peninsular seaway in Baja California: A reply to Riddle et al. Herpetological Review, 33:15-16.
Harpending, H. C., M. A. Batzer, M. Gurvey, L. B. Jorde, A. R. Rogers, and S. T. Sherry. 1998. Genetic traces of ancient demography. PNAS, 95:1961-1967.
Hewitt, G. M. 2000. The genetic legacy of the Quaternary ice ages. Nature, 405:907-913.[CrossRef]
Klicka, J., and R. M. Zink. 1999. Pleistocene effects on North American songbird evolution. Proc. R. Soc. London B, 266:695-700.
Kuchta, S. R., and D. Meyer. 2001. A genealogical view of geographical variation. Mol. Ecol, 10:2569-2576.[CrossRef]
Merila, J., M. Bjorklund, and A. J. Baker. 1997. Historical demography and present day population structure of the greenfinch, Carduelis chloris—an analysis of mtDNA control region sequences. Evolution, 51:946-56.
Moritz, C. 1996. Uses of molecular phylogenies for conservation. In P. H. Harvey, A. J. L. Brown, J. M. Smith, and S. Nee (eds.), New uses for new phylogenies, pp. 203–214. Oxford Univ Press, New York.
Posada, D., and K. A. Crandall. 2001. Intraspecific gene genealogies: Tres grafting into networks. Trends Ecology and Evolution, 16:37-45.
Riddle, B. R., D. J. Hafner, L. F. Alexander, and J. R. Jaeger. 2000a. Cryptic vicariance in the historical assembly of a Baja California peninsular desert biota. PNAS USA, 97:14438-14443.
Riddle, B. R., D. J. Hafner, and L. F. Alexander. 2000b. Comparative phylogeography of Baileys' Pocket Mouse (Chaetodipus baileyi) and the Peromyscus eremicus species group: Historical vicariance of the Baja California peninsular desert. Mol. Phyl. Evol, 17:161-172.[CrossRef][Web of Science][Medline]
Riddle, B. R., D. J. Hafner, and L. F. Alexander. 2000c. Phylogeography and systematics of the Peromyscus eremicus species group and the historical biogeography of North American warm regional deserts. Mol. Phyl. Evol, 17:145-160.[CrossRef][Web of Science][Medline]
Schneider, S., D. Roessli, and L. Excoffier. 2000. Arlequin ver. 2.00. A software for population genetics data analysis. http://anthro.unige.ch/arlequin.
Taberlet, P., L. Fumagalli, A.-G. Wust-Saucy, and J.-F. Cosson. 1998. Comparative phylogeography and postglacial colonization routes in Europe. Mole. Ecol, 7:453-464.
Templeton, A. R. 1998. Nested clade analyses of phylogeographic data: testing hypotheses about gene flow and population history. Mole. Ecol, 7:381-397.
Zink, R. M. 1996. Comparative phylogeography of North American birds. Evolution, 50:308-317.[CrossRef][Web of Science]
Zink, R. M. 1997. Phylogeographic studies of North American birds. In D. P. Mindell (ed.), Avian molecular evolution and systematics, pp. 297–320. Academic Press.
Zink, R. M., D. Anderson, and D. Lott. 1987. Genetic variation, population structure and evolution of California Quail. Condor, 89:395-405.
Zink, R. M., G. F. Barrowclough, R. C. Blackwell-Rago, and J. L. Atwood. 2000. Conservation genetics of the California Gnatcatcher. Conservation Biology, 14:1394-1405.[CrossRef]
Zink, R. M., D. L. Dittmann, J. Klicka, and R. C. Blackwell. 1999. Evolutionary patterns of morphometrics, allozymes and mitochondrial DNA in thrashers (Genus Toxostoma). Auk, 116:1021-1038.
Zink, R. M., A. E. Kessen, T. V. Line, and R. C. Blackwell-Rago. 2001. Comparative phylogeography of some aridland bird species. Condor, 103:1-10.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||