© 2004 Society of Systematic Biologists
Clade-Limited Colonization in Brood Parasitic Finches (Vidua spp.)
Edited by Kevin Johnson: Associate Editor
1 Department of Biology, Boston University Boston Massachusetts 02215, USA; E-mail: msoren{at}bu.edu (M.D.S.), cbala{at}bu.edu (C.N.B.)
2 Museum of Zoology, University of Michigan Ann Arbor Michigan 48109–1079, USA; E-mail: rbpayne{at}umich.edu (R.B.P.)
3 Department of Ecology and Evolutionary Biology, University of Michigan Ann Arbor Michigan 48109–1079, USA
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
The African brood parasitic finches (Vidua spp.) are host specialists that mimic the songs and nestling mouth markings of their finch hosts (family Estrildidae). Although recent molecular analyses suggest rapid speciation associated with host switches in some members of this group, the association of different Vidua lineages with particular host genera suggests the possibility of cospeciation at higher levels in the host and parasite phylogenies. We compared a phylogeny of all Vidua species with a phylogeny of their estrildid finch hosts and compared divergence time estimates for the two groups. Basal divergences among extant members of the Vidulidae and among Vidua species are more recent than those among host genera and species, respectively, allowing a model of cospeciation to be rejected at most or all levels of the Vidua phylogeny. Nonetheless, some tests for cospeciation indicated significant congruence between host and parasite tree topologies. This result may be an artifact of clade-limited colonization. Host switches in parasitic finches have most often involved new hosts in the same or a closely related genus, an effect that increases the apparent congruence of host and parasites trees.
Keywords: Brood parasitism; coevolution; cophylogeny; cospeciation; Estrildidae; Viduidae
Accepted September 12, 2003
Analyses of host–parasite cophylogeny often consider hosts and parasites that belong to different phyla with different life histories and rates of molecular evolution (e.g., Hafner et al., 1994; Moran et al., 1995; Page et al., 1998; Peek et al., 1998). In these cases, congruent phylogenies can provide evidence of cospeciation, but direct tests of similar divergence times, a second key prediction of cospeciation, are more difficult. Identical divergence times are assumed when completely congruent branching patterns are found, and this assumption provides a basis for comparing evolutionary rates (e.g., Moran et al., 1993). Where rates of sequence evolution differ between hosts and parasites, correlated branch lengths in congruent portions of host and parasite phylogenies provide additional evidence of cospeciation (Page, 1996). Systems in which hosts and parasites are more closely related allow a somewhat different approach. When both hosts and parasites can be included in a single phylogenetic analysis rooted with a more distant outgroup, rates of evolution and relative divergence times can be compared directly regardless of whether host and parasite phylogenies are congruent.
Recent studies addressing systems in which hosts and parasites are close relatives have tested Emery's Rule, which suggests that each parasitic species is most closely related to the host species it parasitizes (Emery, 1909; LeMasne, 1956; cited by Hölldobler and Wilson, 1990). Polyphyletic origins of parasitism and sister relationships between hosts and parasites as predicted by a strict interpretation of Emery's hypothesis were found in red algae (Goff et al., 1997), whereas a monophyletic origin of parasitism and subsequent colonization of closely related hosts is a more common result (Choudhary et al., 1994; Ronquist, 1994; Pedersen, 1996; Parker and Rissing, 2002). In all of the above examples, however, parasitic taxa fall within the host clade, suggesting their derivation from one of the current host lineages. A single origin of parasitism followed by subsequent diversification and cospeciation might also produce host and parasite clades that are sister groups with congruent phylogenies. A possible example are the brood parasitic finches, which are the sister group to the clade that includes most of their host species (Sorenson and Payne, 2001).
The African brood parasitic finches (family Viduidae) comprise 19 species of indigobirds and whydahs in the genus Vidua plus the cuckoo finch Anomalospiza imberbis (Friedmann, 1960; Payne, 1996, 1997b). The Vidua finches exhibit a greater degree of host specificity than do other parasitic birds, most Vidua species being associated with a single estrildid finch host species (Table 1). Young estrildids have prominent and sometimes elaborate mouth markings, including black spots and/or lines on the palate and papillae of different sizes, shapes, and colors at the gape. The palate, buccal cavity and gape area also vary in color among species. Although information is lacking for a few species, nestlings of most Vidua species exhibit remarkably precise mimicry of the mouth markings of their respective hosts (Nicolai, 1964; Payne, 1973, 1982). Mimetic mouth markings improve the success of Vidua nestlings (Payne et al., 2001) and apparently play a role in the unusual begging behavior shared by Vidua and their estrildid hosts. Like estrildids, young Vidua beg by twisting their heads nearly upside down and waving both head and tongue from side to side rather than by stretching upward (Nicolai, 1964; Payne et al., 2001; Payne and Payne, 2002). Other attributes of host young, such as colors of the skin, natal down, and juvenile plumage, also are mimicked by different Vidua species (Nicolai, 1964; Payne, 1997a).
|
Strong host specificity led Nicolai (1964) to suggest a history of cospeciation for the brood parasitic Vidua and their hosts. This model could explain both host–parasite associations and the evolution of mimetic mouth markings without requiring the parasites to have colonized new hosts with different markings. Recent analyses based on mitochondrial DNA (mtDNA), however, lead to two contrasting inferences about the history of parasitic finches and their hosts. First, a sister relationship between Anomalospiza and Vidua indicates a single ancient origin of obligate brood parasitism in African finches (Sorenson and Payne, 2001) dating back to before the common ancestor of all extant estrildids, thus suggesting that ancestral parasitic finches were closely related to their hosts and that parasitic finches and estrildids share a long coevolutionary history. In contrast, molecular data for indigobirds (Klein and Payne, 1998; Sorenson and Payne, 2002; Sorenson et al., 2003b) and paradise whydahs (Klein and Payne, 1998), two separate clades within Vidua, indicate a recent origin of extant species and a history of host colonization rather than cospeciation.
Rapid speciation following host colonization in indigobirds can be attributed to behavioral imprinting mechanisms that result in assortative mating of parasites reared by the same host species. Male indigobirds mimic the songs of their respective hosts (Nicolai, 1964; Payne, 1973, 1982, 1998; Payne et al., 1993; Payne and Payne, 1994, 1995), and recent experiments with captive birds show that this behavior is learned by young indigobirds (Payne et al., 1998). Female indigobirds also imprint on host songs and use song cues to choose both their mates and the nests they parasitize (Payne et al., 2000). Similar mechanisms likely account for speciation in paradise whydahs and perhaps other Vidua. An ancient origin of parasitism, however, and the association of most Vidua clades with particular host genera suggest the possibility of congruent branching patterns deeper in the phylogenies of Vidua and their hosts and longer term associations between particular host and parasite clades.
In addition to indigobirds and paradise whydahs, Vidua includes four other long-tailed whydah species (Table 1). The paradise whydahs are associated with melba finches Pytilia spp., V. fischeri and V. regia parasitize two different Granatina species, and V. hypocherina and the generalist V. macroura are associated with waxbills (Estrilda and Coccopygia); whereas most indigobirds parasitize firefinches Lagonosticta spp., although recent field studies have revealed hosts in five additional genera (Payne et al., 1992; Payne and Payne, 1994, 1995). We address here higher level relationships within Vidua and compare phylogenies of Vidua and their estrildid hosts to test for congruent branching patterns and similar divergence times in host and parasite clades. The close relationship of parasitic and estrildid finches allowed us to directly compare relative divergence times after correcting for differences in evolutionary rate between lineages.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
Taxa and Characters
Most of our analyses were based on an mtDNA data set comprising
1,650 base pairs (bp) of sequence data per taxon. We included in our analyses representatives of all Vidua species and the cuckoo finch Anomalospiza imberbis, which parasitizes more distantly related Old World warblers. We also included all known estrildid host species except V. charmosyna and at least one representative of all other estrildid genera except Paludipasser, Oreostruthus, and Zonaeginthus. Our analyses included 74 estrildids, 21 parasitic finches, and nine ploceid finches (Ploceidae) as the outgroup (see Sorenson and Payne, 2001).
The principal data set included two mtDNA regions. The first comprised 1,100 bp/taxon and included almost all of NADH dehydrogenase subunit 6 (ND6), glutamic acid tRNA, and the 5' half of the control region. The second comprised 549 bp/taxon and included the 3' half of ND2 and a small portion of tryptophan tRNA. ND2 was not sequenced for 8 of the 11 indigobird samples included in our analyses. Primers and laboratory protocols were as described previously (Sorenson et al., 1999; Sorenson and Payne, 2001; Payne et al., 2002). DNA was extracted from feathers or muscle tissue rather than blood to avoid problems associated with nuclear pseudogenes (see Sorenson and Fleischer, 1996; Sorenson and Quinn, 1998).
In an attempt to better resolve relationships among the major clades within Vidua, we compiled a larger data set for 14 taxa, including Anomalospiza imberbis, two estrildids (Chloebia gouldiae and Lagonosticta sanguinodorsalis), and 11 representative Vidua. Mitochondrial DNA regions included those mentioned above plus complete sequences of ND1, ND2, the small subunit ribosomal RNA (12S), and ATP synthase subunit 8 (ATP8), partial sequences of 16S, cytochrome oxidase subunit II (COII), ATP6, and ND5, plus complete or partial sequences of seven tRNAs. This larger mtDNA data set comprised 6,575 aligned nucleotide positions. We also sequenced six nuclear introns from five different genes: introns 3 and 9 of phosphenolpyruvate carboxykinase, intron 5 of transforming growth factor β 2, intron 5 of laminin receptor precursor/p40, intron 11 of glyceraldehyde-3-phosphate dehydrogenase, and intron 8 of 
-enolase. Our choice of introns followed previous studies (Friesen et al., 1999; Pacheco et al., 2002; Primmer et al., 2002; Sorenson et al., 2003a), but we designed our own EPIC (exon priming, intron crossing; Palumbi and Baker, 1994) primers based on GenBank sequences for Gallus gallus, Homo sapiens, and other species when available. The nuclear data set included 2,908 aligned nucleotide positions.
Information on voucher specimens and GenBank accession numbers is provided in Appendix 1. Primer sequences not reported in previous studies are provided in Appendix 2.
Phylogenetic Analyses
Because a comparison of divergence time estimates was a principal goal of this study, we limited most of our analyses to regions of unambiguous alignment and excluded all positions with a gap character in one or more taxa, leaving a set of 1,527 homologous characters present in all taxa. Tree searches under equal-weights parsimony were completed in PAUP* 4.0b10 (Swofford, 2002). To conduct a model-based analysis for the large number of taxa in our data set, we used Bayesian inference as implemented in MrBayes 3.04b (Huelsenbeck and Ronquist, 2001). Preliminary analyses using Modeltest (Posada and Crandall, 1998) indicated that the general time reversible model with unequal nucleotide frequencies, a proportion of invariant sites (I), and 
-distributed rate variation among sites provided the best fit to the data. In each of four replicate analyses for each data set, we ran four Markov chains for 2.2 million generations each and sampled trees every 500 generations. Excluding the first 0.5 million generations, we obtained the majority rule consensus of the sampled trees and used it for subsequent comparisons of host and parasite phylogenies and divergence times. Identical consensus trees with nearly identical posterior branch probabilities (±5%) were obtained in the replicate analyses.
Parsimony analyses of the larger mtDNA plus nuclear data set for representative Vidua included all 9,483 alignment positions. Gaps within the mtDNA alignment, most of which were clearly the result of single nucleotide insertion or deletion events, were treated as a fifth character state. Multiple base deletions in intron sequences were recoded as single characters. A set of 6,488 mtDNA characters present in all taxa was used in Bayesian analyses. Given weak support for basal nodes in the Vidua phylogeny, we explored the effect of outgroup by excluding either Anomalospiza or the two estrildid finches.
Analyses of Host–Parasite Associations
Prior to comparing host and parasite phylogenies and estimating divergence times, the Bayesian consensus tree from the initial analysis of 103 taxa was modified in the following ways. Estrildid species that are not regular hosts of Vidua parasites were pruned from the tree. Basal relationships within Vidua were adjusted to reflect results of the 14-taxon analyses, which were based on four to six times as many characters. So that all Vidua species could be included in the analysis, V. togoensis was added to the tree as the sister taxon of V. interjecta. A short mtDNA control region sequence obtained from a museum specimen of this species was identical to that of V. interjecta (M.D.S. and R.B.P., unpubl. data). Indigobird species are recently derived and not reciprocally monophyletic in mtDNA haplotypes (Sorenson et al., 2003b). A set of closely related haplotypes is shared among the species in each region, such that the single representatives included in this study form polytomies with internal branches of zero length. Because some analyses required fully resolved trees, we used estimates of population level differentiation (Sorenson et al., 2003b) to resolve a clade of six West African indigobird species and a clade of four southern species. Species showing greater differentiation in mtDNA haplotype and microsatellite allele frequencies were placed in basal positions. Variations on the above adjustments had little or no effect on inferences derived from our analyses.
Parasites and hosts in our analysis are sister groups that can be included in a single phylogenetic analysis. This combined analysis facilitates the comparison of divergence times by allowing a direct comparison of rates of sequence evolution in parasites and hosts. We conducted relative rates tests using the method of Steel et al. (1996) to compare evolutionary rates among Anomalospiza, Vidua, and estrildids. Consistent with the findings of Sorenson and Payne (2001), these tests indicated significantly different rates of sequence evolution in all three clades, with the highest rate in Anomalospiza, an intermediate rate in Vidua, and a lower rate in estrildids.
To convert branch length information into comparable estimates of divergence time for hosts and parasites, we used the Langley–Fitch method as implemented in the program r8s (Sanderson, 2003) to obtain ML estimates of divergence times under a local molecular clock model with different rates of sequence evolution for Anomalospiza, Vidua, and estrildids, respectively. This method converts branch length estimates into calibrated divergence time estimates for each node under the assumption that rates of sequence evolution are constant within each of the specified clades, an assumption consistent with results of relative rates tests within each clade. In the present example, this approach yields parasite divergence times that are more recent than host divergence times for a given level of sequence divergence, as is appropriate given a faster rate in the parasitic lineage. Results using the penalized likelihood method (Sanderson, 2002) with the smoothing parameter set to one were qualitatively and quantitatively similar. Maximum likelihood (ML) branch lengths on the modified consensus tree were estimated in PAUP* using the following parameters: base frequencies: A = 0.3589, C = 0.3634, G = 0.0993, T = 0.1785. relative transformation rates: A 
C = 0.3672, A G = 7.2795, A T = 0.4334, C G = 0.2875, C T = 5.6874, G T = 1.0000; proportion of invariant sites: I = 0.3873; shape parameter for the distribution: = 0.7861. We fixed the age of the most recent common ancestor of parasitic and estrildid finches at 20 million years based on the rough estimate of Sorenson and Payne (2001). Although we have little confidence in the accuracy of this estimate, error in this calibration point has no effect on the comparison of relative divergence times between Vidua and estrildids because we estimated times for both groups in the same analysis.
We compared host and parasite phylogenies and examined associations between them in three different ways. First, we used TreeMap 1.0b (Page, 1995) to compare phylogenies for the estrildid host species and Vidua parasites, respectively. Host and parasite trees were reconciled without host switching, and the number of reconstructed cospeciation events was compared with a null distribution derived by randomizing the parasite tree (Page, 1990). The maximum number of cospeciation events when host switches are allowed was also determined using an exact search in TreeMap (Page, 1994a, 1994b). We constructed jungles using TreeMap 2.0.2β (Charleston and Page, 2002) to find reconstructions of host–parasite history with minimal cost; costs of duplication, sorting, and host-switching events were set at one, whereas cospeciation had a cost of zero. This last analysis was completed with a reduced data set in which each parasite was associated with a single primary host species.
The program ParaFit (Legendre et al., 2002) implements a general test of coevolution between host and parasite clades that accommodates both multiple hosts per parasitic lineage and uncertainty in tree topologies. The null hypothesis that the evolution of host and parasites has been independent is tested by combining information from genetic distance matrices for hosts and parasites, respectively, with a matrix of host–parasite associations. We obtained ML estimates of pairwise genetic distances using the model parameters noted above and converted the distance matrices to matrices of principal coordinates using the program DistPCoA (Legendre and Anderson, 1998). ParaFit combines these matrices with a matrix of host associations into a single global test statistic that is compared with a null distribution generated by randomizing host–parasite associations. The significance of individual associations between hosts and parasites is also tested by sequentially deleting each host–parasite pair from the analysis. We completed this analysis with all of the host associations listed in Table 1 and with a single principal host per parasitic species.
Because our analyses indicated much shorter divergence times for parasites than for hosts, we reasoned that most or all host–parasite associations in this system are the result of host switching rather than codivergence. We therefore tested for nonrandom association of parasites with host genera by treating host genus as a character on the parasite tree. The number of steps in this character was compared with a null distribution obtained by randomizing character states (host genera) across parasitic taxa (see Kelley and Farrell, 1998).
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
Estrildid Phylogeny and Distribution of Hosts
A genus-level phylogeny of estrildid finches is shown in Figure 1. The family is comprised of two clades, one primarily African but including two Amandava species in Asia and one primarily Australasian that includes three African genera and Lemuresthes in Madagascar. The African clade and groups within it are supported by high posterior probabilities from the Bayesian analysis. In the most-parsimonious (MP) tree, the genera Amadina, Ortygospiza, and Amandava are paraphyletic and basal to all other estrildids, but this result is sensitive to the outgroups included. Excluding parasitic finches, which have a faster rate of sequence evolution, and using only plocieds as the outgroup, the MP tree is consistent with the Bayesian analysis in placing this group of three genera at the base of the African clade.
|
Most African genera that nest in open habitats are regularly parasitized by Vidua. Exceptions include the three cordon-bleus (Uraeginthus spp.) and two Amadina species. Cordon-bleus are abundant and broadly distributed and are behaviorally appropriate hosts for indigobirds (Payne et al., 2001). Although anecdotal cases of parasitism of cordon-bleus by V. regia, V. fischeri, V. hypocherina, and V. macroura are known (Payne, 1996, 2004), the lack of a Vidua species regularly associated with Uraeginthus is puzzling. Estrildids nesting in forests or other more heavily wooded habitats (F in Fig. 1) apparently are unavailable to Vidua parasites, which live in open woodlands and shrubby grassland habitats. African genera falling within the primarily Australasian clade also are not regularly parasitized by Vidua, although the bronze manikin Spermestes cucullatus is parasitized by the generalist V. macroura in some areas (Friedmann, 1960; MacDonald, 1980).
Parasite Phylogeny
The results of all the analyses conducted in this study were consistent with those of a previous study (Sorenson and Payne, 2001) indicating monophyly of the parasitic finches (Vidua plus Anomalospiza) and a sister relationship between parasitic and estrildid finches (Fig. 1). None of our analyses, however, provided clear resolution of basal relationships within the genus Vidua. The genus includes four main clades: (1) steel-blue whydah V. hypocherina, (2) pin-tailed whydah V. macroura, (3) a group of five paradise whydah species, and (4) a clade comprising 10 closely related indigobird species plus straw-tailed whydah V. fischeri and shaft-tailed whydah V. regia.
Relationships among these four main lineages were identical in Bayesian and parsimony analyses based on 6,488 bp of mtDNA sequence for 11 Vidua and three outgroup taxa (Fig. 2). The same topology was found in one of three MP trees for the nuclear intron data and in a combined parsimony analysis of the nuclear and mtDNA data. Relationships among the four Vidua clades were not strongly supported, however, and the topology was sensitive to the outgroups included. Although Anomalospiza and Vidua are sister taxa, sequence divergence between them is greater than that between Vidua and the estrildids because of a faster rate of evolution in parasitic finches (Sorenson and Payne, 2001), particularly on the branch leading to Anomalospiza (Fig. 3). With Anomalospiza excluded, alternative arrangements among the four main Vidua clades were found in Bayesian and parsimony analyses, respectively. The 1,527-bp mtDNA data set with a much larger number of outgroup taxa yielded yet another topology in the Bayesian analysis (Fig. 1). With Anomalospiza excluded, the MP tree for the 1,527 bp data set included a sister relationship between V. macroura and V. hypocherina, a result not seen in other analyses. Both of these taxa parasitize waxbills (genus Estrilda).
|
|
In spite of relatively high posterior branch probabilities in some analyses, sensitivity of results to the outgroups included suggest the presence of systematic biases in the mtDNA data set that are not accounted for by the model of sequence evolution used in our analyses. In contrast, poor resolution in analyses based on nuclear sequences resulted primarily from a lack of informative variation. Clearly, short internal nodes at the base of the Vidua phylogeny make resolution of the branching order among the major Vidua clades difficult. Given more recent divergence times in Vidua than in their hosts (see below), uncertainty in the Vidua phylogeny is not particularly problematic with respect to the questions addressed here.
For the following comparisons of host and parasite phylogenies and divergence times, we combined the host tree summarized in Figure 1 with the parasite tree in Figure 2 and included several additional Vidua species. This consensus phylogeny is shown in Figure 3, in which both a faster rate of sequence evolution in parasitic finches (long basal branches) as well as many recent speciation events within Vidua (short terminal branches) are evident.
Cophylogenetic Analyses
The association of different parasite clades with particular host genera produces at least the appearance of congruence between phylogenies for Vidua and their hosts (Fig. 4). Paradise whydahs are associated with melba finches (genus Pytilia), whereas most indigobirds are associated with firefinches (genus Lagonosticta), but relationships within each group are incongruent with the host tree. Considering only deeper level relationships within Vidua and the principal genus with which each group of parasites is associated, the host tree predicts the following topology: ((V. hypocherina, V. macroura) ((V. fischeri, V. regia)(paradise-whydahs, indigobirds))). Of the five sister relationships predicted, our analyses consistently supported only V. fischeri + V. regia (see also Klein and Payne, 1998), although a sister relationship between V. hypocherina and V. macroura was found in one analysis. The association of V. regia and V. fischeri with two Granatina species may be the best candidate for an example of cospeciation in parasitic finches and their hosts.
|
|
Including all of the host–parasite associations listed in Table 1, tree reconciliation with no host switching suggests eight cospeciation events between Vidua and their hosts, a number not significantly greater than expected based on a null distribution obtained by randomizing the parasite tree (P = 0.15). Considering only the primary host for each parasite (see Table 1), nine cospeciation events are reconstructed, significantly more than expected by chance (P = 0.027). This reconstruction, however, requires 55 sorting events (parasite extinctions) and nine duplication events (parasite speciation without host switch) and posits as many as seven parasitic lineages simultaneously associated with some ancestral hosts. Given the importance of imprinting mechanisms in mate choice and reproductive isolation in Vidua (Payne et al., 1998, 2000), both multiple parasites per host and duplication events seem unlikely. Among extant species, there are no examples of two different Vidua species regularly parasitizing the same host species in the same region.
Jungle reconstructions in TreeMap 2.0.2β allowing an increasing number of host switches resulted in increasingly lower cost solutions (because of smaller numbers of sorting and duplication events), suggesting that a history involving host switching provides a simpler explanation for the associations of Vidua and their hosts. Computational limitations prevented us from exploring reconstructions with more than three host switches, but these reconstructions had the lowest cost, suggesting that lower cost reconstructions with a greater number of host switches would be possible (see Johnson et al., 2003). In reconstructions with three host switches, the program identified switches from ancestral firefinch Lagonosticta lineages to three of the other genera parasitized by indigobirds (Amandava, Ortygospiza, and Hypargos). Incorporating these switches to relatively distant host genera apparently resulted in the greatest reduction in the number of sorting and duplication events needed to explain current host–parasite associations and also reduced from seven to four the maximum number of parasite lineages associated with a single ancestral host.
Comparison of host and parasite phylogenies using ParaFit suggests that phylogenies of Vidua and their hosts are significantly nonindependent, whether we consider all host–parasite associations listed in Table 1 (P < 0.001) or just one primary host for each parasitic species (P = 0.036). This result suggests that pairs of closely related hosts are more often parasitized by pairs of closely related parasites than expected by chance, a result suggesting that cospeciation may have been an important process in the history of these birds. Results of significance tests for individual host–parasite links, however, were less than intuitive. With all host–parasite associations included, almost all of the individual host–parasite links contributed significantly to the overall test statistic. Nonsignificant links included those between V. camerunensis and its hosts Clytospiza and Euschistospiza and between V. codringtoni and Hypargos, cases in which indigobirds have colonized more distantly related hosts. Other nonsignificant links, however, included those between V. fischeri and V. regia and their respective hosts. In contrast, links involving V. raricola and V. nigeriae, two other indigobird species associated with distantly related genera, were significant in both analyses and were among the few individual links that were significant in the test with one primary host per parasite. Thus, the method did not consistently identify as nonsignificant the links most obviously involving host switching (cf. Legendre et al., 2002).
Although some of the above tests suggest significant nonindependence of host and parasite phylogenies, comparison of divergence times indicates that few of the current associations between parasitic finches and their hosts are the result of cospeciation. Branch lengths in Figure 4 are based on divergence time estimates from the program r8s, in which rate variation among lineages was accommodated using local molecular clocks for estrildids, Vidua, and Anomalospiza, respectively. Estimated rates of sequence evolution for these taxa were 1.1%, 1.4%, and 2.3% per lineage per million years (MY) (assuming 20 MY for the common ancestor of estrildid and parasitic finches). The most recent common ancestor (MRCA) of the host species was estimated to be 11.3 MY, whereas the MRCA of Vidua was only 5.0 MY. In contrast, the MRCA of Vidua and Anomalospiza was 13.6 MY, representing a minimum estimate for the origin of parasitism that is older than the MRCA of all the estrildids included in Figure 1. The penalized likelihood method (Sanderson, 2002), in which rate variation is allowed throughout the tree, yielded similar results, except for a somewhat lower rate of evolution in Anomalospiza (1.65%) and an older estimate for the MRCA of Vidua and Anomalospiza (14.9 MY).
We compared selected nodes in the host and parasite trees that represent possible cospeciation events if only tree topology is considered (see Fig. 4). In most cases, divergence times for the parasites are substantially more recent than those for hosts (Fig. 5). Indigobirds and paradise whydahs in particular are much more recently derived than their hosts (see also Klein and Payne, 1998; Sorenson and Payne, 2002; Sorenson et al., 2003b). In contrast, the divergence time of southern and western V. macroura is relatively old and similar to the MRCA of five Estrilda species that are its principal hosts. This finding suggests a long-term association between V. macroura and this host clade and diversification of host species without speciation in the parasite. The divergence between V. fischeri and V. regia is about half the divergence between their respective hosts in the genus Granatina, also suggesting long-term associations between these hosts and parasites even if the associations were established by colonization. Allowing for some host switching and uncertainty in basal Vidua relationships, splits between any of the four main Vidua lineages also might represent cospeciation events. The MRCA of Vidua, however, is substantially more recent than the base of the host tree or any of the intergeneric divergences in the host tree. In one case, however, host and parasite divergence times are nearly equal: the divergence between V. macroura and V. hypocherina is similar to the divergence between their primary hosts in the genus Estrilda. Vidua macroura, however, has additional hosts in other more distantly related genera, including Coccopygia, Amandava in South Africa (Mines, 1999), and Spermestes, which falls within the primarily Australasian clade and is the most divergent of all Vidua hosts.
Although most host associations in parasitic finches are the result of colonization rather than cospeciation, switches between host genera have been infrequent. When host genus is treated as a character and is reconstructed on the parasite tree (an approach that assumes all host associations are due to colonization events), nine switches between host genera are required, a number significantly smaller than expected if host associations were randomized on the parasite tree (P < 0.0001). A similar result is obtained when considering a single primary host for each parasitic species (P < 0.0001). The largest number of switches involve indigobirds, which have colonized hosts in five genera in addition to their primary hosts, the firefinches (genus Lagonosticta).
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
Current associations between brood parasitic finches and their hosts have developed primarily through host colonization rather than cospeciation. This colonization process, however, has been limited such that host switches have most often involved new hosts in the same or a closely related genus, a conclusion reached in several studies on other taxa (e.g., Choudhary et al., 1994; Ronquist and Liljeblad, 2001; Charleston and Robertson, 2002). This clade-limited colonization results in host and parasite trees that are significantly nonindependent, as would be expected if cospeciation were important in the history of these species. Comparison of relative divergence times, however, allows cospeciation to be rejected at most levels of the Vidua phylogeny.
As noted by Charleston and Robertson (2002), cophylogenetic analyses may identify cospeciation events even when none have occurred if host shifts occur primarily between closely related hosts. Clade-limited host switching also may result in correlated genetic distances between pairs of hosts and their associated parasites, another prediction of cospeciation that can be tested even when rates of evolution differ between parasites and hosts (e.g., Page, 1996; Paterson and Banks, 2001). This effect likely explains the significant results we obtained using ParaFit, which distills matrices of genetic distances among hosts and parasites, respectively, and a matrix of host–parasite associations into a single global test statistic. Other processes such as similar responses of hosts and parasites to shared biogeographic history also may produce tree topologies consistent with cospeciation, making a comparison of absolute divergence times an important additional test of cospeciation (Page, 2003). The close relationship of parasitic finches and their estrildid finch hosts allowed us to account for differences in evolutionary rates and to obtain comparable estimates of divergence times for both groups in the context of a single phylogenetic analysis.
Clade-limited colonization is a general expectation for parasites that switch hosts in that new hosts closely related to the original may share characteristics that influence parasite fitness (e.g., Reed and Hafner, 1997; Tompkins and Clayton, 1999; Perlman and Jaenike, 2003). In the case of brood parasitic finches, successful host colonization may be influenced by the degree of mismatch between the mouth markings of parasitic chicks and those of the young of a novel host. Mouth patterns in estrildid finches vary substantially more among genera than among species within genera (R.B.P., unpubl. data). Thus, the predominance of firefinches (Lagonosticta spp.) among indigobird hosts, for example, may be due to greater success in colonizing host species with mouth markings that are more similar to those of the ancestral host, which (based on begging calls mimicked by indigobirds) may be L. rubricata or L. rhodopareia (Payne and Payne, 2002). In an experiment using captive birds, village indigobirds V. chalybeata (with mouths that mimic red-billed firefinches L. senegala) were more likely to survive in firefinch nests than in nests of Uraeginthus or Amandava, although inappropriate mouth markings were not an absolute barrier to being reared by another species (Payne et al., 2001). Behavioral mechanisms of host choice by female parasites (Payne et al., 2000) also may favor the colonization of closely related hosts that share similar songs, nest characteristics, and habitats.
The colonization of new hosts by indigobirds is facilitated by behavioral imprinting mechanisms that result in reproductive isolation of parasites associated with different host species (Payne et al., 1998, 2000). Recent speciation following host switches accounts for the limited but significant genetic differentiation among extant indigobirds (Sorenson et al., 2003b). A similar level of mtDNA differentiation among paradise whydah species (Fig. 4; Klein and Payne, 1998) may reflect a similar history of recent host colonization and speciation. The history of host–parasite associations for other Vidua, however, may not follow the same pattern. The estimated divergence time of V. fischeri and V. regia, for example, is approximately 1.4 MY, 39% of the divergence time of their respective hosts (Fig. 4), suggesting a fairly long period of continuous association for each parasite with its host. Both hosts and parasites are allopatric sister taxa, suggesting a geographic component in the speciation process in both host and parasite lineages. In addition, our estimates of divergence time did not account for the potential discrepancy between the divergence time of two populations and their respective mtDNA lineages (Edwards and Beerli, 2000; Arbogast et al., 2002). Given generally smaller population sizes in parasitic birds than in their hosts, divergence times for the two Granatina species and V. regia + V. fischeri may be more similar than indicated above. Factoring in stochasticity in the lineage sorting process, it is at least possible that these associations are the result of cospeciation.
The divergence between V. macroura and V. hypocherina also is similar to the divergence time of their principal hosts (point 7, Fig. 5). Whether or not this finding represents an old cospeciation event, both species have likely had long and continuous associations with their current hosts (Fig. 4). The estimated divergence time of southern and West African V. macroura is comparable to the time of the MRCA of the five Estrilda species that are this species' principal hosts (point 6, Fig. 5). In contrast to other Vidua, V. macroura are not known to mimic host song (Payne, 1997b). Behavioral isolating mechanisms found in other Vidua therefore may be absent in V. macroura, resulting in continuing gene flow among individuals associated with different hosts and a failure to speciate in response to host speciation (see Johnson et al., 2003).
A remaining puzzle is the contrast between an ancient origin of brood parasitism in finches (Sorenson and Payne, 2001), likely dating back to before the MRCA of extant estrildids, and a relatively recent origin of extant Vidua. What were the host associations of ancestral Vidua during those time periods represented by long branches in the parasite tree? Ancestral paradise whydahs, for example, were probably associated with melba finches Pytilia during the long branch leading to node 5 in Figure 4. Ancestral parasites along the branch leading to node 2 might have been associated either with firefinches Lagonosticta or with grenadiers Granatina. If they were associated with grenadiers, recent colonization of firefinches may have led to the radiation of extant indigobirds. Alternatively, if the ancestor of V. fischeri and V. regia was derived from a parasitic lineage associated with firefinches, which firefinch species was associated with the ancestral parasite lineage and why are there no divergent lineages among the extant parasites of firefinches? Similar questions apply to the long branch leading to the MRCA of extant Vidua (node 4 in Fig. 5), representing a time period of perhaps 8 MY. Extant Vidua may represent only the most recent products of a recurring process of host colonization, speciation, and extinction in the parasitic lineage. The association of a particular parasitic lineage with a given host clade may be older than any of the extant parasitic species and may result in clade-level coevolution between hosts and parasites that reinforces a pattern of clade-limited colonization. This scenario seems likely for paradise whydahs and melba finches and certainly applies more generally to the association of Vidua and estrildid finches.
|
Appendix 1 |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
|
|
Appendix 2 |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
|
|
Acknowledgments |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
Field permits and collecting permits were provided by the governments of Cameroon, Gambia, Malawi, Mali, Nigeria, Zambia, and Zimbabwe. Several museums contributed loans of specimens and/or genetic material for the study: American Museum of Natural History; Durban Museum; Field Museum of Natural History; National Museum of Zimbabwe; Museum National d'Histoire Naturelle, Paris; Forschungsinstitut Senckenberg, Frankfurt; University of Michigan Museum of Zoology; U.S. National Museum of Natural History; University of Washington Burke Museum; Museum Alexander Koenig, Bonn; Universitets Zoologiske Museum, Copenhagen; and Zoologische Staatssammlung München, München. For institutional support we thank the Zoological Research Institute of Cameroon; the University of Jos, Nigeria; the National Museums of Malawi; the Department of National Parks and Wildlife, Zambia; and the National Museum of Zimbabwe. For help in the field we thank Clive Barlow, Bob Dowsett, Mark Hopkins, Kit Hustler, Nedra Klein, Kevin Njabo, Martin Nhlane, Laura Payne, and Bob Stjernstedt. Ornithologists Luis Baptista, Clive Barlow, Les Christidis, Adrian Craig, Phil Hall, John Hook, Kit Hustler, Jürgen Nicolai, Terry Root, Tom Smith, and Bob Stjernstedt and aviculturists Bob Adams, Dave Armer, Paul Boulden, Julie Duimstra, Emma Greig, Steve Hopman, P. J. Maijer, Sigie Meyer, Steve Payne, David Porter, Peter Rindon, Levin Tilghman, Pieter van den Hooven, and John Wilson provided feathers from wild and captive finches. Some of the molecular analyses were completed in David Mindell's laboratory at the University of Michigan. Janet Hinshaw helped organize the genetic materials. The research was supported by the National Science Foundation, Boston University, and the University of Michigan Museum of Zoology.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionAppendix 1Appendix 2AcknowledgmentsReferences |
|---|
-
Arbogast B. S., Edwards S. V., Wakeley J., Beerli P., Slowinski J. B. Estimating divergence times from molecular data on phylogenetic and population genetic timescales. Annu. Rev. Ecol. Syst. (2002) 33:707–740.[CrossRef][Web of Science]
Charleston M. A., Page R. D. M. TreeMap version 2.0.2β. (2002) Available at http://taxonomy.zoology.gla.ac.uk.
Charleston M. A., Robertson D. L. Preferential host switching by primate lentiviruses can account for phylogenetic similarity with the primate phylogeny. Syst. Biol. (2002) 51:528–535.
Choudhary M., Strassmann J. E., Queller D. C., Turillazzi S., Cervo R. Social parasites in polistine wasps are monophyletic: Implications for sympatric speciation. Proc. R. Soc. Lond. B (1994) 257:31–35.
Desjardins P., Morais R. Sequence and gene organization of the chicken mitochondrial genome. J. Mol. Biol. (1990) 212:599–634.[CrossRef][Web of Science][Medline]
Edwards S. V., Beerli P. Perspective: Gene divergence, population divergence, and the variance in coalescence time in phylogeographic studies. Evolution (2000) 54:1839–1854.[CrossRef][Web of Science][Medline]
Emery C. Über den Ursprung der dulotischen, parasitischen und myrmekophilen Ameiksen. Biol. Centralbl. (1909) 29:352–362.
Friedmann H. The parasitic weaverbirds. U.S. Natl. Mus. Bull. (1960) 223:1–196.
Friesen V. L., Congdon B. C., Kidd M. G., Birt T. P. Polymerase chain reaction (PCR) primers for the amplification of five nuclear introns in vertebrates. Mol. Ecol. (1999) 8:2147–2149.[CrossRef][Medline]
Goff L. J., Ashen J., Moon D. The evolution of parasites from their hosts: A case study in the parasitic red algae. Evolution (1997) 51:1068–1078.[CrossRef][Web of Science]
Hafner M. S., Sudman P. D., Villablanca F. X., Spradling T. A., Demastes J. W., Nadler S. A. Disparate rates of molecular evolution in cospeciating hosts and parasites. Science (1994) 265:1087–1090.
Hölldobler B., Wilson E. O. The ants (1990) Cambridge, Massachusetts: Belknap Press.
Huelsenbeck J. P., Ronquist F. MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics (2001) 17:754–755.
Johnson K. P., Adams R. J., Page R. D. M., Clayton D. H. When do parasites fail to speciate in response to host speciation? Syst. Biol. (2003) 52:37–47.
Kelley S. T., Farrell B. D. Is specialization a dead end? The phylogeny of host use in Dendroctonus bark beetles (Scolytidae). Evolution (1998) 52:1731–1743.[CrossRef][Web of Science]
Klein N. K., Payne R. B. Evolutionary associations of brood parasitic finches (Vidua) and their host species: Analyses of mitochondrial DNA restriction sites. Evolution (1998) 52:566–582.[CrossRef][Web of Science]
Legendre P., Anderson M. J. Program DistPCoA (1998) de Montréal, Montréal: Départment de Sciences Biologiques, Université.
Legendre P., Desdevises Y., Bazin E. A statistical test for host–parasite coevolution. Syst. Biol. (2002) 51:217–234.
LeMasne G. Recherches sur les formes parasites. C. R. Hebd. Scéances Acad. Sci. (1956) 243:673–675.
MacDonald M. A. Observations on Wilson's widowfinch and the pin-tailed whydah in southern Ghana, with notes on their hosts. Ostrich (1980) 51:21–24.[Web of Science]
Mines B. E. Hosts of the pin-tailed whydah (Vidua macroura) in southern Africa. Bird Numbers (1999) 8(1):6.
Moran N. A., Munson M. A., Baumann P., Ishikawa H. A molecular clock in endosymbiotic bacteria is calibrated using the insect hosts. Proc. R. Soc. Lond. B (1993) 253:167–171.
Moran N. A., von Dohlen C. D., Baumann P. Faster evolutionary rates in endosymbiotic bacteria than in cospeciating insect hosts. J. Mol. Evol. (1995) 41:727–731.[Web of Science]
Nicolai J. Der Brutparasitismus der Viduinae als ethologisches Problem. Z. Tierpsychol. (1964) 21:129–204.
Pacheco N. M., Congdon B. C., Friesen V. L. The utility of nuclear introns for investigating hybridization and genetic introgression: A case study involving Brachyramphus murrelets. Conserv. Genet. (2002) 3:175–182.[CrossRef]
Page R. D. M. Temporal congruence and cladistic analysis of biogeography and cospeciation. Syst. Zool. (1990) 39:205–226.
Page R. D. M. Maps between trees and cladistic analysis of historical associations among genes, organisms, and areas. Syst. Biol. (1994a) 43:58–77.
Page R. D. M. Parallel phylogenies: Reconstructing the history of host–parasite assemblages. Cladistics (1994b) 10:155–173.[CrossRef][Web of Science]
Page R. D. M. TreeMap, version 1.0b. (1995) Available at http:// taxonomy.zoology.gla.ac.uk.
Page R. D. M. Temporal congruence revisited: Comparison of mitochondrial DNA sequence divergence in cospeciating pocket gophers and their chewing lice. Syst. Biol. (1996) 45:151–167.
Page R. D. M. Introduction. In: Tangled trees: Phylogeny, cospeciation, and coevolution—Page R. D. M., ed. (2003) Chicago: The University of Chicago Press. 1–32. Pages.
Page R. D. M., Lee P. L. M., Becher S. A., Grif fiths R., Clayton D. H. A different tempo of mitochondrial DNA evolution in birds and their parasitic lice. Mol. Phylogenet. Evol. (1998) 9:276–293.[CrossRef][Web of Science][Medline]
Palumbi S. R., Baker C. S. Contrasting population structure from nuclear intron sequences and mtDNA of humpback whales. Mol. Biol. Evol. (1994) 11:426–435.[Abstract]
Parker J. D., Rissing S. W. Molecular evidence for the origin of workerless social parasites in the ant genus Pogonomyrmex. Evolution (2002) 56:2017–2028.[Web of Science][Medline]
Paterson A. M., Banks J. Analytical approaches to measuring cospeciation of host and parasites: Through a glass, darkly. Int. J. Parasitol. (2001) 31:1012–1022.[CrossRef][Web of Science][Medline]
Payne R. B. Behavior, mimetic songs and song dialects, and relationships of the parasitic indigobirds (Vidua) of Africa. Ornithol. Monog. (1973) 11:1–333.
Payne R. B. Species limits in the indigobirds (Ploceidae, Vidua) of West Africa: Mouth mimicry, song mimicry, and description of new species. Misc. Publ. Univ. Mich. Mus. Zool. (1982) 162:1–96.
Payne R. B. Field identification of the indigobirds. Bull. Afr. Bird Club (1996) 3:14–25.
Payne R. B. Avian brood parasitism. In: Host–parasite evolution: General principles and avian models—Clayton D. H., Moore J., eds. (1997a) Oxford, U.K: Oxford Univ. Press. 338–369. Pages.
Payne R. B. Field identification of the brood-parasitic whydahs Vidua and cuckoo finch Anomalospiza. Bull. Afr. Bird Club (1997b) 4:18–28.
Payne R. B. A new species of firefinch Lagonosticta from northern Nigeria and its association with the Jos Plateau indigobird Vidua maryae. Ibis (1998) 140:369–381.[CrossRef]
Payne R. B. Family Viduidae. In: The birds of Africa, Volume 7—Fry C. H., Keith S., eds. (2004) London: Elsevier.
Payne R. B., Hustler K., Stjernstedt R., Sefc K. M., Sorenson M. D. Behavioural and genetic evidence of a recent population switch to a novel host species in brood-parasitic indigobirds Vidua chalybeata. Ibis (2002) 144:373–383.[CrossRef]
Payne R. B., Payne L. L. Song mimicry and species associations of West African indigobirds Vidua with quail-finch Ortygospiza atricollis, goldbreast Amandava subflava and brown twinspot Clytospiza monteiri. Ibis (1994) 136:291–304.[CrossRef]
Payne R. B., Payne L. L. Song mimicry and association of brood-parasitic indigobirds (Vidua) with Dybowski's twinspot (Euschistospiza dybowskii). Auk (1995) 112:649–658.[Web of Science]
Payne R. B., Payne L. L. Begging for parental care from another species: Behavioural specialization and generalization in brood-parasitic finches. In: The evolution of begging: Competition, cooperation and communication—Wright J., Leonard M. L., eds. (2002) Dordrecht, The Netherlands: Kluwer. 429–450. Pages.
Payne R. B., Payne L. L., Nhlane M. E. D. Song mimicry and species status of the green widowfinch Vidua codringtoni. Ostrich (1992) 63:86–97.[Web of Science]
Payne R. B., Payne L. L., Nhlane M. E. D., Hustler K. Species status and distribution of the parasitic indigo-birds Vidua in east and southern Africa. Proc. Pan-Afr. Ornithol. Congr. (1993) 1992:40–52.
Payne R. B., Payne L. L., Woods J. L. Song learning in brood-parasitic indigobirds Vidua chalybeata: Song mimicry of the host species. Anim. Behav. (1998) 55:1537–1553.[CrossRef][Web of Science][Medline]
Payne R. B., Payne L. L., Woods J. L., Sorenson M. D. Imprinting and the origin of parasite–host species associations in brood-parasitic indigobirds, Vidua chalybeata. Anim. Behav. (2000) 59:69–81.[CrossRef][Web of Science][Medline]
Payne R. B., Woods J. L., Payne L. L. Parental care in estrildid finches: Experimental tests of a model of Vidua brood parasitism. Anim. Behav. (2001) 62:473–483.[CrossRef][Web of Science]
Pedersen B. V. A phylogenetic analysis of cuckoo bumblebees (Psithyrus, Lepeletier) and bumblebees (Bombus, Latreille) inferred from sequences of the mitochondrial gene cytochrome oxidase. Mol. Phylogenet. Evol. (1996) 5:289–297.[CrossRef][Web of Science][Medline]
Peek A. S., Feldman R. A., Lutz R. A., Vrijenhoek R. C. Cospeciation of chemoautotrophic bacteria and deep sea clams. Proc. Natl. Acad. Sci. USA (1998) 95:9962–9966.
Perlman S. J., Jaenike J. Infection success in novel hosts: An experimental and phylogenetic study of Drosophila parasitic nematodes. Evolution (2003) 57:544–557.[Web of Science][Medline]
Posada D., Crandall K. A. Modeltest: Testing the model of DNA substitution. Bioinformatics (1998) 14:817–818.
Primmer C. R., Borge T., Lindell J., Saetre G. P. Single-nucleotide polymorphism characterization in species with limited available sequence information: High nucleotide diversity revealed in the avian genome. Mol. Ecol. (2002) 11:603–612.[CrossRef][Medline]
Reed D. L., Hafner M. S. Host specificity of chewing lice on pocket gophers: A potential mechanism for cospeciation. J. Mammal. (1997) 78:655–660.[CrossRef][Web of Science]
Ronquist F. Evolution of parasitism among closely related species: Phylogenetic relationships and the origin of inquilinism in gall wasps (Hymenoptera, Cynipidae). Evolution (1994) 48:241–266.[CrossRef][Web of Science]
Ronquist F., Liljeblad J. Evolution of the gall wasp–host plant association. Evolution (2001) 55:2503–2522.[Web of Science][Medline]
Sanderson M. J. Estimating absolute rates of molecular evolution and divergence times: A penalized likelihood approach. Mol. Biol. Evol. (2002) 19:101–109.
Sanderson M. J. r8s, version 1.60. Available from the author, Univ (2003) California, Davis.
Sorenson M. D., Ast J. C., Dimcheff D. E., Yuri T., Mindell D. P. Primers for a PCR-based approach to mitochondrial genome sequencing in birds and other vertebrates. Mol. Phylogenet. Evol. (1999) 12:105–114.[CrossRef][Web of Science][Medline]
Sorenson M. D., Fleischer R. C. Multiple independent transpositions of mitochondrial DNA control region sequences to the nucleus. Proc. Natl. Acad. Sci. USA (1996) 93:15239–15243.
Sorenson M. D., Oneal E., García-Moreno J., Mindell D. P. More taxa, more characters: The hoatzin problem is still unresolved. Mol. Biol. Evol. (2003a) 20:1484–1499.
Sorenson M. D., Payne R. B. A single ancient origin of brood parasitism in African finches: Implications for host–parasite coevolution. Evolution (2001) 55:2550–2567.[CrossRef][Web of Science][Medline]
Sorenson M. D., Payne R. B. Molecular genetic perspectives on avian brood parasitism. Integr. Comp. Biol. (2002) 42:388–400.
Sorenson M. D., Quinn T. W. Numts: A challenge for avian systematics and population biology. Auk (1998) 115:214–221.[Web of Science]
Sorenson M. D., Sefc K. M., Payne R. B. Speciation by host switch in brood parasitic indigobirds. Nature (2003b) 424:928–931.[CrossRef][Medline]
Steel M. A., Cooper A. C., Penny D. Confidence intervals for the divergence time of two clades. Syst. Biol. (1996) 45:127–134.
Swofford D. L. PAUP*. Phylogenetic analysis using parsimony (*and other methods), version 4.0b10 (2002) Sunderland, Massachusetts: Sinauer.
Tompkins D. M., Clayton D. H. Host resources govern the specificity of swiftlet lice: Size matters. J. Anim. Ecol. (1999) 68:489–500.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. N. Balakrishnan and S. V. Edwards Nucleotide Variation, Linkage Disequilibrium and Founder-Facilitated Speciation in Wild Populations of the Zebra Finch (Taeniopygia guttata) Genetics, February 1, 2009; 181(2): 645 - 660. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. N Balakrishnan, K. M Sefc, and M. D Sorenson Incomplete reproductive isolation following host shift in brood parasitic indigobirds Proc R Soc B, January 22, 2009; 276(1655): 219 - 228. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Light and M. S. Hafner Codivergence in Heteromyid Rodents (Rodentia: Heteromyidae) and Their Sucking Lice of the Genus Fahrenholzia (Phthiraptera: Anoplura) Syst Biol, June 1, 2008; 57(3): 449 - 465. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Kruger Cuckoos, cowbirds and hosts: adaptations, trade-offs and constraints Phil Trans R Soc B, October 29, 2007; 362(1486): 1873 - 1886. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||