© 2005 Society of Systematic Biologists
Phylogenetic Relationships, Divergence Time Estimation, and Global Biogeographic Patterns of Calopterygoid Damselflies (Odonata, Zygoptera) Inferred from Ribosomal DNA Sequences
Edited by Karl Kjer: Associate Editor
1 Department of Biology, Ghent University, Ledeganckstraat 35, B-9000 Ghent, Belgium E-mail: HenriDumont{at}UGent.be (H.J.D.)
2 Protozoology Laboratory, Scientific Institute of Public Health—Louis Pasteur J. Wytsmanstraat 14, B-1050 Brussels, Belgium
 |
Abstract |
|---|
 TopNotes Abstract Materials and MethodsResultsDiscussionConclusionsReferences |
|---|
The calopterygoid superfamily (Calopterygidae + Hetaerinidae) is composed of more than twenty genera in two families: the Calopterygidae (at least 17) and the Hetaerinidae (at least 4). Here, 62 calopterygoid (ingroup) taxa representing 18 genera and 15 outgroup taxa are subjected to phylogenetic analysis using the ribosomal 18S and 5.8S genes and internal transcribed spacers (ITS1, ITS2). The five other families of calopterid affinity (Polythoridae, Dicteriadidae, Amphipterygidae, Euphaeidae, and Chlorocyphidae) are included in the outgroup. For phylogenetic inference, we applied maximum parsimony, maximum likelihood, and the Bayesian inference methods. A molecular phylogeny combined with a geographic analysis produced a well-supported phylogenetic hypothesis that partly confirms the traditional taxonomy and describes distributional patterns. A monophyletic origin of the calopterygoids emerges, revealing the Hetaerinid clade as sister group to the Calopterygidae sensu strictu. Within Calopterygidae, seven clades of subfamily rank are recognized. Phylogenetic dating was performed with semiparametric rate smoothing by penalized likelihood, using seven reference fossils for calibration. Divergence time based on the ribosomal genes and spacers and fossil constraints indicate that Calopteryginae (10 genera, approximately 50% of all Calopterygid taxa studied here), Vestalinae (1 genus), and Hetaerinidae (1 genus out of 4 studied here) started radiating around 65 Mya (K/T boundary). The South American Iridictyon (without distinctive morphology except for wing venation) and Southeast Asian Noguchiphaea (with distinctive morphology) are older (about 86 My) and may be survivors of old clades with a Gondwanian range that went extinct at the K/T boundary. The same reasoning (and an even older age, ca. 150 My) applies to the amphipterygids Rimanella and Pentaphlebia (South America–Africa). The extant Calopterygidae show particular species and genus richness between west China and Japan, with genera originating between the early Oligocene and Pleistocene. Much of that richness probably extended much wider in preglacial times. The Holarctic Calopteryx, of Miocene age, was deeply affected by the climatic cooling of the Pliocene and by the Pleistocene glaciations. Its North American and Japanese representatives are of Miocene and Pliocene age, respectively, but its impoverished Euro-Siberian taxa are late Pliocene-Pleistocene, showing reinvasion, speciation, and introgression events. The five other calopterid families combine with the Calopterygidae and Hetaerinidae to form the monophyletic cohort Caloptera, with Polythoridae, Dicteriadidae, and Amphipterygidae sister group to Calopterygoidea. The crown node age of the latter three families has an age of about 157 My, but the Dicteriadidae and Polythoridae themselves are of Eocene age, and the same is true for the Euphaeidae and Chlorocyphidae. The cohort Caloptera itself, with about 197 My of age, goes back to the early Jurassic.
Keywords: Biogeography; Calopteridae; dating; divergence times; damselflies; internal transcribed spacers; odonata; phylogeny; phylogeography; 18S and 5.8S ribosomal DNA
Received March 28, 2002; Revised August 9, 2002; Accepted December 12, 2004
Calopterygoid damselflies sensu strictu (= the families Calopterygidae and Hetaerinidae) are cosmopolitan excluding only Australia, New Zealand, and the polar zones. Across this vast territory, they are a remarkably uniform in habitat choice (running waters, which they rarely leave) and in morphology, but the males present a wide array of precopulatory displays (Buchholtz, 1955; Heymer, 1972). Calopterygids can be distinguished by size, body, and wing color (almost invariably showing bright metallic sheens) and behavior, but rarely by morphology (Asahina, 1976). The phylogenetic relationships among genera and species are therefore unclear. Morphology is not a robust criterion for classifying taxa in which reproductive isolation is achieved by color-flashing and premating displays rather than by mechanical (lock-antilock) mechanisms (Dumont et al., 1987). In search of new technical approaches to taxonomy, electrophoretic analysis of allele frequencies was attempted (Maibach, 1985), but this approach is restricted to water-soluble enzymes and requires fresh specimens. In Calopteryx spp., quantitative wingspot analysis has been used for studying the taxonomy and range of species (Dumont et al., 1993). This technique is limited to taxa that have a wingspot, which is the case in less than half of the genera.
The internal transcribed spacers, in which substitutions are more frequent than in the structural rRNAs, are useful to distinguish between related species that otherwise show little genetic divergence (Porter and Collins, 1991; Fritz et al., 1994; Tang et al., 1996). To date, a handful of studies have addressed the phylogenetic relationship of the Odonata at the molecular level (Artiss et al., 2001; Chippendale et al., 1999; Kambhampati and Charlton, 1999; Misof et al., 2000, 2001; Weekers et al., 2001). Misof et al. (2000) sequenced fragments of the mitochondrial 16S rDNA to provide a partial phylogeny of the Calopterygids, with special emphasis on species groups within Calopteryx sensu strictu. Weekers et al. (2001) used the combined ITS1 and ITS2 to resolve the western and central European taxa of Calopteryx at the species-group, species, and, where possible, subspecies level. They confirmed most of Misof's conclusions and found evidence for incipient speciation in west Mediterranean taxa. Lindeboom (1996) used a combination of wing venation characters and male ligula structure to derive a hypothetical tree for 13 genera. Both Misof et al. (2000) and Lindeboom's (1996) phylogeny are partial and require further testing. Recently, Rehn (2003) published a morphology-based phylogeny of the order Odonata, which was particularly detailed for the suborder Zygoptera, placing the Calopterygoidea near the base of the Zygoptera.
Here, we sequence and analyze the rDNA genes (18S, 5.8S) and internal transcribed spacers (ITS1, ITS2) of 62 Calopterygidae (all extant genera) and Hetaerinidae, and 15 outgroup taxa. The latter include the families Polythoridae, Dicteriadidae, Amphipterygidae, Euphaeidae, and Chlorocyphidae, supposed to form a clade of higher order, the cohort Caloptera. Other outgroup families included in the study were Megapodagrionidae, Protoneuridae, Platycnemidae, and Diphlebiidae. We combined molecular, geographical, and fossil data to investigate phylogenetic relationships, to relate these to their geographic ranges, and to link them to geological events.
|
Materials and Methods |
|---|
|
TopNotesAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
DNA Extraction, PCR Amplification, and Sequencing Reactions
The origins of the samples used in this study are listed in Table 1. Single samples of taxa that were expected to be well demarcated were analyzed, whereas several populations were sampled in the case of taxa that were a priori expected to be closely related. Muscular tissue was isolated from the thorax and total DNAs were prepared according to the protocol of the Puregene DNA isolation kit type D-5000A (Gentra Systems, Inc., BIOzym, Landgraaf, The Netherlands). The complete region of the ribosomal spacers (ITS1 and ITS2) and the ribosomal 18S, 5.8S, and part of the 28S genes was amplified using the polymerase chain reaction (PCR) with Qiagen DNA polymerase (Westburg, Leusden, The Netherlands). Eukaryote-specific external primers complementary to the 5'-terminus of the 18S rDNA gene (5'-TYCCTGGTTGATYYTGCCAG-3') and the 5'-terminus of the 28S rDNA gene (5'-TCCTCCGCTTABTDATATGCTTAA-3') were used to amplify the entire 18S-ITS1-5.8S-ITS2 and part of the 28S region. PCR amplifications, purification of the PCR products, and DNA sequencing was done according to standard procedures (Samraoui et al., 2003). External (see above) and internal primers in conserved regions of the 18S and 5.8S rDNA were used for sequencing (Weekers et al., 1994; Samraoui et al., 2003).
|
Sequence Alignment and the Construction of Datasets
Because we have nucleotide sequences from different genes (18S, 5.8S) and internal transcribed spacers (ITS1, ITS2) that evolve at different rates, several methods can be applied for treatment of partitioned data: the combined data, separate analysis, and conditional combination approaches (Huelsenbeck et al., 1996).
The DNA sequences covering the complete 18S-ITS1-5.8S-ITS2-28S (partial) region were aligned with CLUSTALW 1.64b (Thompson et al., 1994) using default settings, resulting in an initial data set. A second data set was created by fine-tuning the alignment of the initial data set based on secondary structural information, using DCSE 3.4 (Dedicated Comparative Sequence Editor program; De Rijk and De Wachter, 1993) or GeneDoc 2.6.002 (Nicholas et al., 1997). The alignment of the 18S gene region was manually optimized with published 18S rDNA sequences based on the conservation of both primary sequence data and inferred secondary structural features (the rRNA WWW Server: http://www-rrna.uia.ac.be/ssu/index.html) (The Ribosomal Database Project: http://rdp.cme.msu.edu/ download/SSU_rRNA/alignments/). The small and highly conserved 5.8S gene region and the small portion of the 28S gene were easy to align and were used to position the highly variable ITS1 and ITS2 regions. The boundaries of the ITS1 and ITS2 were determined by comparison of the aligned data set with previously published ITS sequences of calopterygids (Weekers et al., 2001), and other hexapod taxa available in the EMBL databank (e.g., Pseudococcus sp., Beuning et al., unpublished data; Luehdorfia sp., Makita et al., unpublished data; Anopheles sp., Beebe et al., 1999; Adalia sp., Schulenburg et al., 2001). The ITS regions were manually optimized based on conservation of both primary sequence data and inferred secondary structural features. The secondary structures of the ITS1 and ITS2 regions were predicted using the Mfold webserver for nucleic acid folding and hybridization prediction (Zuker, 2003) (http://www.bioinfo.rpi.edu/applications/mfold) and were compared with published data (Fritz et al., 1994; May and Coleman, 1997; Morgan and Blair, 1998; Gottschling et al., 2001).
Sequence and Phylogenetic Analyses
First, the likelihood-ratio test (LRT) and Akaike Information Criteria (AIC) in ModelTest 3.06 (Possada and Crandall, 1998) were used to select an appropriate substitution model of DNA evolution. The data set was analyzed with the Bayesian inference algorithm (MrBayes, version 3.0b4; Huelsenbeck and Ronquist, 2001), and the maximum parsimony (MP) and the maximum likelihood (ML) algorithms in PAUP* 4.0b10 (Swofford, 2003) to resolve the phylogenetic relationships. The Bayesian estimates of posterior probability and bootstrap analyses were included to assess support. The model with corresponding nucleotide frequencies, substitution rates and types, and Ti/Tv ratios was selected by ModelTest 3.06 (Posada and Crandall, 1998) and was used for MP and ML algorithms in PAUP*, and in the Bayesian analysis.
Pairwise sequence divergence data between taxa were computed for the complete 18S-ITS1-5.8S-ITS2-28S (partial) region. Absolute distance values and distances based on a maximum-likelihood distance matrix (PAUP*), with appropriate parameters for the DNA evolution model (ModelTest), were calculated.
The Bayesian analysis was performed with MrBayes version 3.0b4 (Huelsenbeck and Ronquist, 2001), specifying the appropriate model structure for each partition (18S, ITS1, 5.8S, ITS2) in the data set. The appropriate mixed models for the heterogeneous data were specified, thus applying site-specific models of rate variation for each partition. In a first analysis, the length of the Bayesian run was tested in order to be certain of convergence. The Markov chain Monte Carlo (MCMC) process was set so that four chains ran simultaneously for 5,000,000 generations, with trees being sampled every 100 generations for a total of 50,000 trees in the initial sample. For the final analysis, five independent Bayesian runs were performed in order to confirm that there was adequate convergence and mixing. Each MCMC process started from random starting points, and was set so that four chains ran simultaneously for 1,000,000 generations, with trees sampled every 100 generations for a total of 10,000 trees in the initial sample. Variation in the ML scores in the samples was examined by inspecting the MrBayes log file, and the position where the ML scores stopped improving was determined. The portion of the trees before the position (tree number) where the ML score stopped improving dramatically and only fluctuated around a plateau was discarded. The posterior probability of the phylogeny and its branches was determined for all those trees in the plateau phase with near the best ML scores.
Equally weighted MP analyses were performed with PAUP*. Heuristic search settings were stepwise taxon addition, tree bisection-reconnection branch swapping, multiple trees retained, no steepest descent, rearrangements limited to 10,000, and accelerated transformation. Treating gaps as characters as in Swofford (1996) or Lutzoni et al. (2001) would have provided more information from these sites, but we treated gaps as missing data so that the MP analysis could be directly compared to the ML analyses. The nonparametric bootstrap analysis used 1000 replicates to assess the reliability of individual branches in the phylogenetic trees obtained by heuristic search with stepwise sequence addition (Felsenstein, 1985).
For ML analysis, the substitution model of DNA evolution with corresponding parameters that best fitted the data was determined by the LRT and the AIC, using ModelTest 3.06 (Posada and Crandall, 1998). Heuristic search settings were stepwise taxon addition, TBR branch swapping, MulTrees option in effect, no steepest descent, and rearrangements limited to 10,000. The nonparametric bootstrap analysis with 100 replicates was used to assess the reliability of individual branches in the phylogenetic trees obtained by heuristic search with stepwise sequence addition (Felsenstein, 1985). Trees were displayed with TREEVIEW 1.6.6 (Page, 1996).
Divergence Time Estimation
To date, many methods are available for phylogenetic dating (e.g., Britton et al., 2002; Thorne and Kishino, 2002; Yang and Yoder, 2003). Here, dating was done with the r8s program (Sanderson, 2002, 2003) by semiparametric rate smoothing using a penalized likelihood approach applied to the distances inferred from the phylogenetic trees (ML, MrB). It combines a model-based likelihood approach with a roughness penalty that prevents too much rate variation across the tree. The size of the roughness penalty is specified by a smoothing parameter obtained by a cross-validation procedure. Cross validation was performed for the trees with branch lengths, obtained by multiplying the per site values as reported by PAUP* with the number of sites. This procedure provides an objective method for model selection and choice of optimal smoothing value (Sanderson, 2002, 2003). Comparison of six independent dating analyses (ML trees with smoothing factor 31, 100, 316; MrB trees with smoothing factor 100, 316, 1000), using optimal (= lowest) and suboptimal smoothing factors, was used as a measure of confidence; average node date and confidence intervals were calculated.
We used seven reference fossils (Table 2) to estimate divergence times. All fossil calibration points were used simultaneously, using the fixage command of r8s. Alternatively, we tested dating each with one fossil as calibration point using the fixage command, and the other six as constraints specifying minimum ages using the constraints command.
|
|
Results |
|---|
|
TopNotesAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Sequence Analysis and Alignments
We obtained the complete, unambiguous DNA sequence for the 18S, 5.8S, partial 28S and internal transcribed spacers (ITS1, ITS2) of 62 calopterygoid (ingroup) and 15 outgroup taxa (Table 1). The length of the ribosomal genes (18S, 5.8S) and internal transcribed spacers (ITS1, ITS2) and their GC content are listed in Table 3.
|
Genetic Distances
Pairwise sequence comparison of 18S-ITS15.8S-ITS2-28S (partial) region using distance measurements by ML with settings corresponding to the GTR+G+I model showed remarkable differences in interspecific sequence diversity (Table 4; simple data matrix, data matrix for 77 taxa not shown). The tropical and subtropical taxa displayed higher interspecific sequence diversity than those of temperate Eurasia: viz. 0 to 5.5 and 0 to 0.8 substitutions per site. The holarctic genus Calopteryx showed little genetic variation (0 to 2.04 substitutions per site) and divides into North American, Euro-Siberian, and east Asian taxa. The Euro-Siberian taxa showed least genetic variation (< 0.08 substitutions per site) (Table 4; simple data matrix, data matrix for 77 taxa not shown).
|
Phylogenetic Analysis
The alignment contained 2745 aligned positions; 737 sites were variable, 603 of which were parsimony informative. Maximum parsimony, maximum likelihood, minimum evolution LogDet, and Bayesian inference analyses were used to construct phylogenetic trees covering all Calopterid taxa and to compare support among phylogenetic trees. These analyses resulted in well-resolved, strongly supported trees, showing consistent and corroborated topologies among phylogenetic methods (Figs. 1 to 3).
Based on the results of ModelTest LRT and AIC evaluations, ML analysis was performed with the GTR+G+I model (for parameters see Fig. 1) and resulted in a tree (Ln likelihood = –18454.19013) with high ML bootstrap support for most of the clades. The MP analysis with heuristic search resulted in 94 most parsimonious trees (MPTs) of 3071 steps; the bootstrap 50% majority-rule consensus tree (Fig. 2) shows a similar topology to the ML tree with only a few minor topological changes. The Bayesian inference analysis ran for 5,000,000 generations and showed that a stable likelihood value was reached after 15,000 generations. Therefore, further multiple independent Bayesian analyses were confidently run with a maximum of 1,000,000 generations. All five independent Bayesian analyses resulted in identical topologies, showing a similar topology as the ML tree (Fig. 3). We also tested for variation in nucleotide bias among taxa by running an ME LogDet+I analysis in PAUP* with the appropriate model, in order to see if the nucleotide bias altered the phylogeny. LogDet distance analyses (ME) resulted in a tree topology similar to the ML, MP, and MrB results, clustering the same groups in the same way. The only difference was that for some nodes the bootstrap values increased or decreased a little (data not shown). The monophyly of cohort Caloptera is strongly supported (ML = 90, MP = 89, MrB = 100), and most major and minor clades within it are well supported, too (> 70). The branch point separating Euphaeidae and Chlorocyphidae from the other Caloptera is also well supported (ML = 92, MP = 90, MrB = 100). The next branch points separating Calopterygidae, Hetaerinidae, Polythoridae, Dicteriadidae, and Amphipterygidae are not consistently supported. The Calopterygoidea (clusters 1 to 7) are supported by ML (ML = 72) and Bayesian analysis (MrB = 99), but not by MP analysis (MP = 44). The monophyly of the true Calopterygidae (clusters 1 to 6) with the Hetaerinidae (cluster 7) as sister group is supported by all analyses (ML = 99, MP = 88, MrB = 100). A majority of the major clades in Calopterygidae (clusters 1 to 6) are monophyletic with high support. One of them, grouping clusters 1 to 3, is considered to include those genera that together form the Calopteryginae (ML = 90, MP = 91, MrB = 99). The crown of the tree is composed of two major monophyletic clades; one for the Neurobasis-Matrona group (ML = 97, MP = 93, MrB = 100) and one for the Calopteryx group (ML = 98, MP = 98, MrB = 100). Within each major clade, the topology of most taxa is well resolved, confirmed by high support, clustering the genera. Only the topology within the Calopteryx clade suffers from poor resolution, with only a few species-groups well resolved, with high support.
|
Dating Analysis
Application of the seven fossil calibration points in the penalized likelihood procedure applied to the ML and MrB trees provided us with a range of data. Initial results were obtained with the default settings for dating analysis in the r8s program, with cross validation function enforced. The rate smoothing parameters with optimal (= lowest) and suboptimal cross-validation scores were selected, and the dating procedure was then repeated. The result of the time divergence estimation is shown in Figure 4, and age estimates for all internal nodes are shown in Table 5. The analyses using alternative tree topologies (ML, MrB) and different smoothing factors, resulting from optimal and suboptimal cross-validation scores, yielded small deviations in age estimates. The use of several reference fossils is expected to reduce variation due to error. The age constraints imposed by the seven fossils in different parts of the tree are likely to restrict variation caused by a variety of other factors. The ages found vary from 238 My (deviation of Caloptera from outgroup) to virtually zero.
|
|
Discussion |
|---|
|
TopNotesAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Sequence Variation in Genes and Spacers
The length of the 18S (1860 to 1863 bp) and 5.8S genes (163 to 170 bp) of the Caloptera under examination is in the range of previously described odonate taxa (Samraoui et al., 2002, 2003) and other insect taxa (Genbank/EMBL). The internal transcribed spacers (ITS1, 156 to 233 bp; ITS2, 195 to 234 bp) are among the shortest known among eukaryotes. The GC content of the ribosomal genes and internal transcribed spacers is within normal range among eukaryotes (data from Genbank/EMBL). However, the GC content in the ITS2 is much higher than in the ITS1 and differs from other ITS sequences described in literature (Torres et al., 1990). A balanced GC content between ITS1 and ITS2 (Torres et al., 1990) is thus not found in Caloptera.
Genetic Distances
It is interesting to compare genetic distances within and among taxa to determine whether these particular sequences for a given group of Caloptera have diverged, on average, more or less than others. The minimum and maximum divergence values among Calopterygoidea (Calopterygidae + Hetaerinidae) vary greatly (Table 4). The highest genetic distance is found between Phaon, Noguchiphaea, Iridictyon, and Hetaerina. All these, plus Caliphaea and Vestalis, have at times been given subfamily rank in the literature (e.g., Bechly, 1996). This is here accepted, as is the family status of the Hetaerinidae. Of course, the genetic distance with the five "outgroup" families is even higher. A point that deserves analysis is the fact that tropical and subtropical taxa (clusters 2 to 7) display higher interspecific sequence diversity than those of the temperate region (cluster 1): respectively, 0 to 5.5 and 0 to 2.0 substitutions per site. The lowest variation in genetic distance is between Calopteryx taxa (cluster 1 in Fig. 1), even though these can be subdivided into a North American, an Euro-Siberian and an East Asian group. Especially the Euro-Siberian taxa reveal little genetic variation (< 0.8 substitutions per site) (Table 4; simple data matrix, data matrix for 77 taxa not shown).
Phylogenetic Relationships and Biogeographic Patterns
All tree topologies suggest that cohort Caloptera is monophyletic (ML = 90, MP = 89, MrB = 100) (Fig. 1 to 3) and about 200 My old (Fig. 4). Rehn's (2003) topologies differ from ours in two main respects: he places Amphipterygidae outside of Caloptera and shows an ambiguous topology for the Chlorocyphidae, Dicteriadidae, and Euphaeidae. In addition, he considers the hetaerinids inside Calopterygidae, with Caliphaea and Vestalis as basal sister groups. Here, we show the monophyly of the Calopterygidae sensu strictu (clusters 1 to 6), supported by all analyses (ML = 99, MP = 88, MrB = 100) (Figs. 1 to 3), with Hetaerinidae as basal sister group to Calopterygidae. However, like in Rehn's analyses (2003), the monophyly of the Calopterygoidea (clusters 1 to 7) has ML bootstrap support (72) and Bayesian probability support (100) (Figs. 1, and 3), although no MP bootstrap support (44) (Fig. 2).
|
|
|
The branching order among the 18 genera of Calopterygoidea (1 newly created, 2 reinstated from synonymy) shows a consistent tree topology. The superfamily arose some 175 Mya; from it, the Calopterygidae and Hetaerinidae started diversifying 151 Mya. Our taxon sampling of the Hetaerinidae is much less complete than that of the Calopteryigidae. The speciose genus Hetaerina radiated 61 Mya, slighty after the K/T boundary (65 Mya), but some of its three other genera (Garrison, 2005) may be as old as the oldest Calopterygidae.Iridictyon (surviving only in Venezuela), Noguchiphaea (presently only in North Thailand), and the amphipterygids Rimanella (only in Venezuela and Guyana) and Pentaphlebia (limited to the Cameroon rainforest) distinctly predate the K/T boundary (about 86 and 150 My, respectively). They probably belong to clades that were common and geographically widespread until the K/T events eliminated most of their representatives. These Gondwanian relicts consequently qualify as living fossils. Aside from these basic clades, it is of interest to define the status of some other within-calopterygid clades.
Among Calopterygidae, clusters 6a and 6b accommodate the South American Iridictyon, the Asian Noguchiphaea, and African Phaon. Iridictyon has conserved a typical calopterygid habitus and structure, even though its common ancestor with extant calopterygids was of Cretaceous age. Its present location is a small area of tropical forest situated in the ancient Serra Pacaraima, northeast South America. This, and its relatedness to Noguchiphaea, can only be understood in terms of continental drift and both are likely living fossils, relicts of a much richer fauna that is presently almost extinct. Typical calopterygid traits are also found in the African Phaon (support = 100) but the Thai endemic Noguchiphaea displays several unusual traits: it has petiolated wings, distinctive male appendages, and peculiar outgrowths on the pronotum. The common ancestor it shared with Phaon and Iridictyon around 152 Mya originated in the late Jurassic, before the separation of India from Africa and South America.
The Phaoninae, with a stem node age of 125 My, are old, and therefore, the presence of Phaon on Madagascar, where it is the only calopterygid on record, could qualify the Malagasy taxon as a Gondwanian relict. However, it speciated only some 32 Mya, and is of Oligocene age. About 95% of the zygopteran odonates of Madagascar are endemic, but the island has two types of endemics, "young" and "old" (Dijkstra and Clausnitzer, 2004). Furthermore, the aquatic fauna of the island lacks rheophilic species, as well among dragonflies (including the Calopteran families) as among fish (Kiener and Richar-Vindard, 1972). It is, in general, impoverished (its total number of dragonfly species is only 175, about half of what would be expected from its surface area). This suggests a phase of extinction, followed by a recolonization from the African mainland at a time when the distance to Madagascar was still bridgeable (Briggs, 2003). Although Phaon is rheophilic and not migratory, it covers a much wider range in Africa than the strict forest-dwelling Saphoinae, and hence may have some dispersal abilities.
Caliphaea (cluster 5a) and Vestalis (cluster 5b) from tropical Asia are separated and well supported, although the topology varies between the three analyses (Figs. 1 to 3). The geographic ranges of both taxa overlap, and the condition "stalked wings" is seen to have arisen in the ancestor to Caliphaea independently from Noguchiphaea, where stalked wings also occur and which have a similar stem node age (89 versus 87 My, respectively). Caliphaea with a stem node age of 89 My branched off around the same time as the common ancestor of Iridictyon and Noguchiphaea and is considered as a subfamily on account of its stalked wings. We here accept this position, meaning that claims of a subfamily status for taxa deeper than Caliphaeinae, viz. Vestalinae, Phaoninae, Iridictyoninae, and Noguchiphaeinae (e.g., Bechly, 1996) are also accepted, whether they have a distinctive wing venation and morphology (as in Caliphaea and Noguchiphaea), or mainly a distinctive wing venation (as in Vestalis, Iridictyon, and Phaon). The Saphoinae, forest-dwelling African calopterygids, with the morphologically well-defined genera Sapho and Umma (cluster 4), form a monophyletic clade (support = 100), split off from Calopteryginae some 81 Mya (cluster 4). Both genera have an age of about 26 My, corroborated by the finding, in France, of well-preserved wings of Umma and Sapho of Oligo-Miocene age (Nel and Paicheler, 1993). Thus, in the subtropical climate that characterized most of Eurasia during the Cenozoic, at least these calopterine genera already coexisted in Europe, far from their present geographic ranges. Their current confinement to the forest zone of tropical Africa reflects the cooling of the planet since the Cenozoic.
The Calopteryginae (clusters 1 to 3), grouping 10 genera, includes more than half of all extant calopterygids, East Palaearctic, but extending into the Oriental with the largest number of genera and species found between west China and Japan. Only the Hetaerinidae rivals Calopteryginae in species richness. Like the former, they are seen to have originated soon after the K/T boundary, and represent another example of radiation following the K/T events 65 Mya. Presumably, many dragonflies, calopteran and noncalopteran in nature, went extinct at that time, and left the old world open to the Calopteryginae.
Tropical and subtropical calopterygids (clusters 2 to 7) display higher interspecific sequence diversity than those of the temperate region (Table 4), revealing a dichotomy of the Calopterygoids in taxa with short branches (presumably young speciation events) and taxa with long branches (presumably old speciation events). The "young" group almost unequivocally corresponds to Holarctic taxa that populate North Africa, Europe, continental Asia as far as 60°N, and North America. Old taxa are predominantly tropical or straddle the temperate-tropical transition zone of East Asia, from the pacific coast of Russia ("Primoriye") to south China and Indochina-India.
Young genera, like Mnais, and the complex Matrona-Atrocalopteryx-Archineura are typically 2.7 to 10.0 My old, whereas old genera have stem node ages from 95 to 89 My old (Vestalis, Caliphaea). Mnais is of late Miocene age (ca. 10 My) and has a number of species in Japan that are notoriously "difficult" to distinguish, reminiscent of the C. splendens group. In fact, it is only marginally older than the entire cluster 1a, grouping all Eurasian Calopteryx. Like Calopteryx, Mnais is apparently still speciating. It is also noteworthy that the distinctive Chinese Archineura forms a subcluster with Atrocalopteryx and Matrona that is hardly 2.7 My old. Archineura thus appears to be a Pleistocene taxon, unrelated to the Laotian "Leucopteryx" hetaerinoides (estimated age 46 My), which is sometimes cited as the "second Archineura."
The east Palaearctic Atrocalopteryx (2.7 My old) accommodates the former "Calopteryx" atrata, and possibly a few other little-known Chinese "Calopteryx." For C. atrata, Selys-Longchamps and Hagen (1854) had created a special "section" with the following characters: wings compatively long and narrow, without a pterostigma in both sexes; postocular tubercles very small to absent; the wing veins R2 and IR3 detach together form R4+5 (in Calopteryx, R2 detaches from R+M). We here rename this "section" Atrocalopteryx n.g., with atrata as the type species.
The genus Calopteryx arose around 35 Mya and started to speciate approximately 21 ± 3 Mya. It is shared between Eurasia and North America, and may have extended across the Beringian and Thulean routes. The North Atlantic (Thulean) route was available to thermophilic organisms like swallowtail butterflies until about 35 Mya (Zakharov et al., 2004), but because of progressive cooling, and the transformation of the Thulean route into a chain of islands, younger and thermally less demanding organisms like Calopteryx (Tiffney, 1985) may have taken advantage of the Beringian route. North American Calopteryx started radiating less than 16 Mya; the Eurasian group began its radiation only around 6.2 Mya. The first product of this radiation, around 5.3 Mya, was the C. virgo-group (including the Japanese C. cornelia); the C. splendens-group s.l., with only some 3.7 My, is even younger. At that time, corresponding to the onset of the Pliocene, the climatic deterioration that culminated in the Pleistocene glaciations seems to have been sufficiently advanced to start causing significant extinctions. This condition, which has been repeatedly addressed in the literature, and was recently reviewed by Hewitt (2000): the regions of the globe that were directly hit by the Pleistocene glaciations were impacted by huge genetic losses, and were later repopulated by a limited number of taxa with reduced genetic variation. Japan, the Pacific fringe, China, and North America, less or not affected by mass extinctions and easier to repopulate than Europe, show this effect to a lesser extent, and have conserved identified species-groups and even genera (like Mnais). In Europe, the west Mediterranean refugium allowed C. xanthostoma-exul and C. haemorrhoidalis to survive the glaciations. Both are between 2.2 and 2.4 My old, and thus originated just before the onset of the major climatic cooling. The origin of C. splendens s.s. is hard to identify, and its numerous subspecific tax are all of very recent origin, and hybridize freely. Even C. xanthostoma is currently suffering from introgression by an advancing edge of C. splendens coming from the east, and its range is both shrinking and breaking up into disjunct fragments (Weekers et al., 2001).
|
Conclusions |
|---|
|
TopNotesAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Combined analyses of the ribosomal genes (18S, 5.8S) and their internal transcribed spacers (ITS1, ITS2) produced a fairly satisfactory reconstruction of the phylogeny of a major group of the Zygoptera, the cohort Caloptera, with particular emphasis on the family Calopterygidae. The group is found to be of early Mesozoic age. The Calopterygidae (Old World plus the ancient Serra Pacaraima upland in South America) form a clade, whereas a second clade contains the neotropical Hetaerinidae. We found several Gondwanaland disjunctions in taxa with relict distributions such as Iridictyon, Noguchiphaea, and the Amphipterygid genera Rimanella and Pentaphlebia. Our results also provide support to the concept that taxa in the temperate zone, having been affected by a series of glaciations because the Pleistocene have not recovered from that climatic vagary. If sequence variation can be assumed to be relatively steady, then all speciation events in the Calopteryx splendens group and in Mnais are recent. In Europe, particularly, introgression between previously disjunct taxa is continuing. The Tropics, in contrast, have conserved the old taxa cited above, some of which occur in such extremely relict geographic ranges that they qualify as living fossils.
|
Acknowledgements |
|---|
The following individuals kindly provided specimens: V. Alekseev (St. Peterburg, Russia), A. Brancelj (Ljubljana, Slovenia), A. Budieri (Amman, Jordan), G. Chiambeng (Limbe, Cameroon), V. Clausnitzer (Marburg, Germany), A. Cordero-Rivera (Pontevedra, Spain), J. Demarmels (Caracas, Venezuela), J. Dommanget/V. Bosc (Bois d'Arcy, France), S. W. Dunkle (Plano, TX, USA), G. Fleck (Strassbourg, France), M. Hämäläinen (Helsinki, Finland), H. Heidari (Teheran, Iran), K. Inoue and I. Wakana (Osaka, Japan), A. Mitra (Calcutta, India), L. Nacelli-Flores (Palermo, Sicily), L. Nagorskaja (Minsk, Russia), M. Papazian (Marseille, France), M. J. Parr (Somerset, United Kingdom), M. Pavesi (Milan, Italy), S. Rong (Hohhot, China), B. Samraoui (Annaba, Algeria), W. Schneider (Darmstadt, Germany), K. Van Damme (Ghent, Belgium), K. Watanabe (via Yeh) (Taipei, Taiwan), K.O.P. Wilson (Hong Kong, China), W.C. Yeh (Taipei, Taiwan), and Y. Zaitsev (Odessa, Ukraine) (for full addresses see: http://www.bechly.de/ododir.htm). We also wish to thank Matti Hämäläinen, Michael May, Karl Kjer, Chris Simon, and an anonymous reviewer for their critical reading and excellent suggestions. Funding for this research was provided by Belgian National Foundation for Scientific Research to HJD and JFDJ (FWO grant G.0066.96) and JVF (FWO grant G.0292.00). PHHW was funded in 1996 by a visiting postdoctoral fellowship from the Belgian National Foundation for Scientific Research (FWO) and in 1997 by a grant from Ghent University. All experimental procedures and phylogenetic and DTE analyses were performed by PHHW.
|
Notes |
|---|
|
TopNotesAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
3 Present address: Research Unit for Tropical Diseases, Christian de Duve Institute of Cellular Pathology, Avenue Hippocrate 74–75, B-1200 Brussels, Belgium
|
References |
|---|
|
TopNotesAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
-
Artiss T., Schultz T. R., Polhemus D. A., Simon C. Molecular phylogenetic analysis of the dragonfly genera Libellula, Ladona and Plathemis (Odonata: Libellulidae) based on mitochondrial cyrochrome oxidase I and 16S rRNA sequence data. Mol. Phylogen. Evol. (2001) 18:348–361.[CrossRef][Web of Science][Medline]
Asahina S. Description of one new genus and two new species of Caliphaeaidae (Odonata, Calopterygidae) from Thailand, with taxonomic notes on the subfamily Kontyu (Tokyo). (1976) 44:387–402.
Bechli G. Morphologische Untersuchungen am Flügelgeäder der rezenten Libellen und deren Stammgruppenvertreter (Insecta; Pterygota; Odonata); unter besonderer Berücksichtigung der Phylogenetischen Systematik und des Grundplanes de *Odonaten. Petalura, spec.-vol. (1996) 2:1–402. ISSN 0944-0453.
Beebe N. W., Ellis J. T., Cooper R. D., Saul A. DNA sequence analysis of the ribosomal DNA ITS2 region for the Anopheles punctulatus group of mosquitoes. Insect. Mol. Biol. (1999) 8:381–390.[CrossRef][Web of Science][Medline]
Bernardi G., Mouchiroud D., Goutier C. Compositional patterns in vertebrate genomes: conservation and change in evolution. J. Mol. Evol. (1988) 28:7–18.[CrossRef][Web of Science][Medline]
Briggs J. C. The biogeography and tectonic history of India. J. Biogeography (2003) 30:381–388.
Britton T., Oxelman B., Vinnersten A., Bremer K. Phylogenetic dating with confidence intervals using mean path-lengths. Mol. Phylogen. Evol. (2002) 24:58–65.[CrossRef][Web of Science][Medline]
Buchholtz C. Eine vergleichende Ethologie der orientalischen Calopterygiden (Odonata) als Beitrag zu ihrer systematischen Deutung. Z. Tierpsychol. (1955) 12:363–386.
Chippendale P. T., Davé V. K., Whitemore D. H., Robinson J. V. Phylogenetic relationships of North American damselflies of the genus Ischnura (Odonata: Zygoptera: Coenagrionidae) based on sequences of three mitochondrial genes. Mol. Phyl. Evol. (1999) 11:110–121.[CrossRef][Web of Science][Medline]
De Rijk P., De Wachter R. DCSE v2.54, an interactive tool for sequence alignment and secondary structure research. Comput. Appl. Biosci. (1993) 9:735–740.
Dijkstra K.-D. B., Clausnitzer V. Critical species of Odonata in Madagascar. Int. J. Odonatol. (2004) 7:219–228.
Dumont H. J., Demirsoy A., Verschuren D. Breaking the Calopteryx-bottleneck: Taxonomy and range of Calopteryx splendens waterstonii Schneider, 1984 and C. splendens tschaldirica Bartenef, 1909 (Zygoptera; Calopterygidae). Odonatologica (1987) 16:239–247.
Dumont H. J., Mertens J., De Coster W. The Calopteryx splendens cline in southwestern France, analysed by quantitative wingspot analysis (Zygoptera: Calopterygidea). Odonatologica (1993) 22:345–351.
Felsenstein J. Confidence limits on phylogenies: An approach using the bootstrap. Evolution (1985) 39:783–791.[CrossRef][Web of Science]
Fritz G. N., Conn J., Cockburn A., Seawright J. Sequence analysis of the ribosomal DNA internal transcribed spacer 2 from populations of Anophelesnuneztovari (Diptera: Culicidae)??. Mol. Biol. Evol. (1994) 11:406–416.[Abstract]
Garrison R. A synopsis of the genera Mnesarete Cowley, Bryoplathanon gen. Nov. and Ormenophlebia gen. Nov. (Odonata: Calopterygidae). Science (2005) In press.
Gottschling M., Hilger H. H., Wolf M., Diane N. Secondary structure of the ITS1 transcript and its application in a reconstruction of a phylogeny of Boraginales. Pl. Biol. (2001) 3:629–636.[CrossRef]
Hewitt G. The genetic legacy of the Quaternary ice ages. Nature (2000) 405:907–913.[CrossRef]
Heymer A. Comportement social et territorial des Calopterygidae (Odonata, Zygoptera). Ann. Soc. Ent. Fr. (N.S.) (1972) 8:3–53.
Huelsenbeck J. P., Bull J. J., Cunningham C. W. Combining data in phylogenetic analysis. TREE (1996) 11:152–158.
Huelsenbeck J. P., Ronquist F. R. MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics (2001) 17:754–755.
Kambhampati S., Charlton R. E. Phylogenetic relationships among Libellula, Ladona and Plathemis (Odonata: Libellulidae) based on DNA sequence of mitochondrial 16S rRNA gene. Syst. Entomol. (1999) 24:37–49.[CrossRef]
Kiener A., Richard-Vindard G. Fishes of the continental waters of Madagascar. In: Biogeography and ecology of madagascar—Battistini R., Richard-Vindard G., eds. (1996) Junk: The Hague. Pages 477–501.
Lindeboom M. Fortpflanzungsbiologie der Gebänderten Prachtlibelle Calopteryx splendens (Calopterygidae, Odonata) (1996) Germany: University of Freiburg. PhD. Thesis.
Lutzoni F., Pagel M., Reeb V. Major fungal lineages are derived from lichen symbiotic ancestors. Nature (2001) 411:937–940.[CrossRef][Medline]
Maibach A. Révision systématique du genre Calopteryx Leach (Odonata, Zygoptera) pour l'Europe Occidentale. I. Analyses biochemique. Mitt. Schweiz Ent. Ges. (1985) 58:477–492.
May J. C., Coleman A. W. The internal transcribed spacers2 exhibits a common secondary structure in green algae and flowering plants. J. Mol. Evol. (1997) 44:258–271.[CrossRef][Web of Science][Medline]
Misof B., Anderson C. L., Hadrys H. A Phylogeny of the damselfly genus Calopteryx (Odonata) using mitochondrial 16S rDNA markers. Mol. Phyl. Evol. (2000) 15:5–14.[CrossRef][Web of Science][Medline]
Misof B., Rickert A. M., Buckley T. R., Fleck G., Sauer K. P. Phylogenetic Signal and its Decay in Mitochondrial SSU and LSU rRNA Gene Fragments of Anisoptera. Mol. Biol. Evol. (2001) 18:27–37.
Morgan J. A. T., Blair D. Trematode and monogenean rRNA ITS2 secondary structures support a four-domain model. J. Mol. Evol. (1998) 47:406–419.[CrossRef][Web of Science][Medline]
Nel A., Brisac P. Découverte du premier Calopteryx fossile du Miocène de l'Ardèche (Odonata, Calopterygidae). Bull. Soc. Entom. Fr. (1994) 99:27–30.
Nel A., Paicheler J.-C. Les Odonates fossiles: État actuel des connaissances. Huitième Partie: Les Calopterygoidea fossiles (Odonata, Zygoptera). Bull. Soc. Ent. Fr. (1992) 97:381–396.
Nicholas K. B., Nicholas H. B. Jr., Deerfield D. W. GeneDoc: Analysis and visualization of genetic variation. EMBNEW.NEWS (1997) 4:14.
Page R. D. M. TREEVIEW: An application to display phylogenetic trees on personal computers. Comput. Appl. Biosci. (1996) 12:357–358.
Porter C. H., Collins F. H. Species-diagnostic differences in a ribosomal DNA internal transcribed spacer from sibling species Anopheles freeborniAnopheles hermsi (Diptera: Culicidae). Am. J. Trop. Med. Hyg. (1991) 45:271–279.
Posada D., Crandall K. A. ModelTest: Testing the model of DNA substitution. Bioinformatics (1998) 14:817–818.
Rehn A. C. Phylogenetic analysis of higher-level relationships of Odonata. Syst. Entomol. (2003) 28:181–239.[CrossRef]
Samraoui B., Weekers P. H. H., Dumont H. J. The Enallagma (Odonata, Zygoptera: Coenagrionidae) of the western and central Palaearctic. Odonatologica (2002) 31:371–381.[Web of Science]
Samraoui B., Weekers P. H. H., Dumont H. J. Two taxa within the North African Lestes virens complex (Odonata, Zygoptera). Odonatologica (2003) 32:131–142.[Web of Science]
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: Inferring absolute rates of molecular evolution and divergence Times in absence of a molecular clock. Bioinformatics (2003) 19:301–302.
Schulenburg H. J. G. V. D., Hancock J. M., Pagnamenta A., Sloggett J. J., Majerus M. E. N., Hurst G. D. D. Extreme length and length variation in the first ribosomal internal transcribed spacer of ladybird beetles (Coleoptera: Coccinellidae). Mol. Biol. Evol. (2001) 18:648–660.
Sélys-Longchamps E. de, Hagen H. A. Monographie des Caloptérygines. Mém. Soc. Sci. Liège (1854) 9.
Swofford D. L. PAUP*: Phylogenetic analysis using parsimony (*and other methods) (2003) Sunderland, Massachusetts: Sinauer Associates. Version 4.
Swofford D. L., Thorne J. L., Felsenstein J., Wiegmann B. M. The topology-dependent permutation test for monophyly does not test for monophyly. Syst. Bio. (1996) 45:575–579.[CrossRef]
Tang J., Toè L., Back C., Unnasch T. R. Intra-specific heterogeneity of their DNA internal transcribed spacer in the Simulium damnosum (Diptera: Simuliidae) complex. Mol. Biol. Evol. (1996) 13:244–252.[Abstract]
Tiffney B. The Eocene North-Atlantic land-bridge—its importance in tertiary and modern phytogeography of the northern hemisphere. 2. J. Arnold Arobor. (1985) 66:243–273.
Thompson J. D., Higgins D. G., Gibson T. J. CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Res. (1994) 22:4673–4680.
Thorne J. L., Kishino H. Divergence time and evolutionary rate estimation with multilocus data. Syst. Biol. (2002) 51:689–702.
Torres R. A., Ganal M., Hemleben V. GC balance in the internal transcribed spacers ITS1 and ITS2 of nuclear ribosomal DNA. J. Mol. Evol. (1990) 30:170–181.[CrossRef][Web of Science][Medline]
Weekers P. H. H., De Jonckheere J. F., Dumont H. J. Phylogenetic relationship, inferred from ribosomal ITS sequences, and detecting biogeographic patterns in the representatives of the genus Calopteryx (Insecta: Odonata) of the West Mediterranean and adjacent West European zone. Mol. Phylogenet. Evol. (2001) 20:89–99.[CrossRef][Web of Science][Medline]
Weekers P. H. H., Gast R. J., Fuerst P. A., Byers T. J. Sequence variations in small-subunit ribosomal RNAs of Hartmanella vermiformis and their phylogenetic implications. Mol. Biol. Evol. (1994) 11:684–690.[Abstract]
Yang Z., Yoder A. D. Comparison of likelihood and Bayesian methods for estimating divergence times using multiple gene loci and calibration points with application to a radiation of cute-looking mouse lemur species. Syst. Biol. (2003) 52:705–716.
Zakharov E. V., Caterino M. S., Sperling F. A. H. Molecular phylogeny, historical biogeography, and divercence time estimates for swallowtail butterflies of the Genus Papilio (Lepidoptera: Papilionidae). Syst. Biol. (2004) 53:193–215.
Zuker M. Mfold: Webserver for nucleic acid folding and hybridization prediction. Nucleic Acids Res. (2003) 31:3406–3415.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. R. Clark, R. H. Ree, M. E. Alfaro, M. G. King, W. L. Wagner, and E. H. Roalson A Comparative Study in Ancestral Range Reconstruction Methods: Retracing the Uncertain Histories of Insular Lineages Syst Biol, October 1, 2008; 57(5): 693 - 707. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||