Coleoptera: Curculionidae

IMB_354.fm Page 453 Wednesday, August 28, 2002 10:49 AM

Insect Molecular Biology (2002) 11(5), 453 – 465

Elongation Factor 1 α resolves the monophyly of the haplodiploid ambrosia beetles Xyleborini (Coleoptera: Curculionidae)

Blackwell Science, Ltd

B. H. Jordal Department of Zoology, University of Bergen, Allegt 41, N-5007 Bergen, Norway Abstract Elongation Factor 1-α was used to test the monophyly of the wood boring beetle tribe Xyleborini, where all species are haplodiploid and perform regular inbreeding by brother–sister mating. Due to their feeding requirements, being highly dependent on ophiostomatoid fungi which they cultivate in wood tunnels, monophyly may be expected due to nutritional constraints. During the course of analyses, two copies α were amplified in these beetles, differing in of EF-1α intron structure. The high similarity between paralogous amino acid sequences (93–94%) indicates a rather recent duplication in beetles, but phylogenetic analyses of different copies in insects rejected this hypothesis. Subsequent phylogenetic analyses of eighty orthologous sequences from Xyleborini and allied taxa, using the single-intron bearing copy, were greatly improved in resolution and node support by including the intron sequences (c. 60 bp). Most analyses resulted in a monophyletic Xyleborini, implying one origin of fungus feeding in this tribe. However, clear evidence for a polyphyletic Xyleborus and three more xyleborine genera calls for further revision of xyleborine classification. Keywords: haplodiploidy, sib-mating, gene duplication, intron sequences, weevils. Introduction One of the largest radiations in wood-boring phytophagous beetles coincided with the evolution of sib-mating, haplodiploidy and fungus feeding in scolytine weevils of the Received 12 February 2002; accepted after revision 10 June 2002. Correspondence and present address: School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, U.K. E-mail: [email protected]

© 2002 The Royal Entomological Society

pantropical Xyleborini (Normark et al., 1999; Farrell et al., 2001; Jordal et al., 2002a). While the origin of sib-mating and haplodiploidy occurred in an ancestor shared by three closely related sib-mating genera of Dryocoetini (160 spp.), the obligate feeding on ophiostomatoid fungi (cultivated in wood tunnels) occurs only in the much more species rich Xyleborini (c. 1300 spp.). Altogether, these sib-mating taxa constitute a monophyletic clade containing nearly onequarter of the 6000 described species of scolytine weevils world-wide (Wood & Bright, 1992; Bright & Skidmore, 1997). Their ecological significance is most prominent in lowland tropical forests, where these insects constitute more than half the species or insect body mass in guilds of early stage wood-decomposers (Schedl, 1956; Browne, 1961). Many of the haplodiploid, sib-mating species are also extremely widespread in the tropics and an increasing number of species have recently been introduced and established outside their native ranges, perhaps promoted by their genetic system in conjunction with sib-mating (Jordal et al., 2001). The wood boring and fungus cultivating Xyleborini pose a significant problem to the timber trade, and much attention has been directed towards suppression of the spread and growth of these beetles. On a brighter side, these beetles have proven extremely informative in studies on a wide range of general biological and ecological systems, which in turn informs management of their impact on diverse forest products. The great potential for culturing lineages of xyleborine species in the laboratory, using artificial diets (Norris & Baker, 1967; Norris & Chu, 1970), creates a rich model system to study the developmental and evolutionary genetics of the group. However, comparative studies on variation within the inbreeding mating system or haplodiploid genetic system, including microbial function in haploid male development from unfertilized eggs (cf. Peleg & Norris, 1972), calls for a reliable hypothesis on their genealogical history. Recent phylogenetic studies have demonstrated difficulties with resolving the monophyly of the fungus feeding Xyleborini, with respect to the sib-mating dryocoetine genera (Normark et al., 1999; Jordal et al., 2000; Farrell et al., 2001; Jordal et al., 2002a). Based on these previous 453


454

B. H. Jordal

results, two hypotheses remain: either (a) obligate fungus feeding is a purely derived feature, implying monophyly of the Xyleborini; or (b) Xyleborini gave rise to some or all of the sib-mating dryocoetines, implying one or several reversals to the ancestral phloem feeding habit. It has been argued (Jordal et al., 2000) that the latter hypothesis is the least probable of the two due to the assumed irreversibility of fungus feeding. Under the latter scenario, fungus feeders are dependent on ergosterol and other steroid components from the fungi to initiate egg hatching and complete development (Kok et al., 1970), a dependence not observed in true bark beetles (e.g. Kukor & Martin, 1989; Six & Paine, 1998). Partial sequences from the nuclear protein encoding gene Elongation Factor 1α (EF-1α) were used to test the proposed hypotheses of xyleborine phylogeny. This is a low-copy gene that promotes the GTP-dependent binding of aminoacyl-tRNA to ribosomes, and shows considerable conservatism in amino acid substitution rate (e.g. Uetsuki et al., 1989). EF-1α has also proven very informative in a wide range of phylogenetic analyses of arthropods (e.g. Shultz & Regier, 2000) and insects (e.g. Cho et al., 1995; Mitchell et al., 1997; Cryan et al., 2000; Cruickshank et al., 2001; Rokas et al., 2001; Danforth, 2002). Used in combination with other nuclear and mitochondrial genes in analyses of other scolytine beetle groups, EF-1α provided crucial data for obtaining well resolved topologies (Sequeira et al., 2000; Cognato & Vogler, 2001; Jordal et al., 2002b; Sequeira & Farrell, 2001). The major radiation in haplodiploid beetles probably occurred rapidly as far back as in the early Miocene – a hypothesis based on a general high level of sequence divergence in combination with the absence of such beetles in Oligocene amber (Jordal et al., 2000; Farrell et al., 2001). The relatively low substitution rate in EF-1α may help to resolve the shallow internodes characterizing the early stages of this group’s phylogenesis. Multiple copies of EF-1α have been described from Drosophila flies (Hovemann et al., 1988) and Apis bees (Danforth & Ji, 1998), and this is probably a universal feature throughout the two insect orders. Normark (1994) found no evidence for multiple copies of EF-1α in his study on Aramigus weevils, but a second putative copy has more recently been suggested for scolytine weevils (Normark et al., 1999). This second copy has not yet been characterized and compared to the commonly used copy. Because ‘cryptic’ paralogous sequences would interfere in phylogeny reconstruction, efforts should be made to ensure that only orthologous sequences are included in a phylogenetic analysis. Hence, DNA sequences from each of the two putative EF-1α copies were obtained from four scolytine species, to test the monophyly of each copy (and shared intron structure) in a phylogenetic framework. Eighty orthologous sequences with identical intron structure (921

aligned characters) were used in the final analyses to test the monophyly of Xyleborini. Results and discussion Multiple copies of EF-1α A majority (fifty-two out of eighty) of the PCR amplified products obtained from Xyleborini and sib-mating Dryocoetini revealed two bands that were clearly separable on a low melting agarose gel. Four of these double-banded products were selected for further comparison between the two putative copies. After sequencing and removal of introns, none of the four sequences from each of the two different-length fragments contained indels or stop codons, and all sequences aligned perfectly with other insect sequences obtained from GenBank. This may suggest that both copies are functional although 498 bp of the coding region (1392 bp in bees and flies) were not sequenced. The inferred amino acid sequences for the two copies showed considerable similarity with each other as well as to other published insect EF-1α sequences. Amino acid distances between the two copies in scolytine beetles ranged between 5.9 and 7.7%, well below the differences found between the two Apis copies and between the two Drosophila copies (9.0%). It is also well below the smallest difference measured between homologous EF-1α copies from different arthropod orders, including apterygote insects (Regier & Shultz, 2001). The corresponding nucleotide sequence divergence between the two copies in beetles ranged between 22.0 and 27.1%. The two copies in beetles differ by the number of introns; the shortest fragment (C1) had one intron and the longest fragment (C2) had three introns in the region sequenced (149–1043, cf. Fig. 1C). All four introns were short, ranging between 50 and 90 base pairs. Both beetle copies shared the placement of the intron in position 753/754, which is equal to the placement of intron 2 in the bee F2 copy (Danforth & Ji, 1998) and in aphids (Normark, 1999). C2 had a second intron in position 1029/1030, equal to the bee and fly F2 copies, and in aphids and booklice [but not lice (Cruickshank et al., 2001)], and a third intron not observed in any other insect, in position 430/431. Because so many of the scolytine beetles did amplify two copies, it is not likely that any of these copies contain another intron adjacent to the primer site (149), as observed in position 143/144 in the bee F2 copy and in Protura (Carapelli et al., 2000). Given the fixed differences in intron structure between the two beetle copies, and the much lower divergence levels within than between the copies at both the nucleotide (10.6 vs. 22%) and amino acid (2.9 vs. 5.9%) levels, it seems likely that EF-1α copies with shared intron structure in beetles are orthologous. However, a phylogenetic analysis of the different copies in insects can more precisely determine the homology of shared intron structure and

© 2002 The Royal Entomological Society, Insect Molecular Biology, 11, 453 – 465


Elongation Factor 1-α resolves the monophyly of Xyleborini

455

Figure 1. Phylogeny of EF-1α sequences from various insects and placement of introns in different copies. (A) Strict consensus of the two most parsimonious trees (CI = 0.38, RI = 0.61) resulting from the analysis of 864 bp of EF-1α with first and second positions weighted ten times third positions. Bootstrap support values higher than 50% are shown above internodes; values in bold mark the monophyly of the haplodiploid beetles, for each copy. The tree was rooted by the remiped crustacean Speleonectes. (B) Same as A, but with the long branched paralogs Apis F1, Drosophila F2 and the beetle C2 sequences removed, resulting in one most parsimonious tree (CI = 0.43, RI = 0.61). (C) An overview of known intron positions (triangles) for some well studied insect groups. The stippled vertical lines indicate positional homology among introns and the length of horizontal lines shows the maximum length sequenced within each insect order.



456

B. H. Jordal

the timing of duplication events. Thus, one crustacean and twenty-three additional insect EF-1α sequences were added to a phylogeny matrix and subjected to a variety of phylogenetic analyses (see methods). All parsimony and likelihood analyses of nucleotides (or amino acid parsimony) resulted in monophyly of each beetle copy, and the two beetle copies were not grouping together in any of the analyses (Fig. 1A). Given these results, duplication of EF-1α in beetles most likely predates the origin of beetles. The likelihood analysis using the GTR+Γ+I model was largely consistent with some of the parsimony analyses (unweighted, 5:10:2, 5:10:1), placing the C2 copy closest to the Apis F1 copy and nested within the Diptera clade, which also contained the Drosophila F2 copy and the Metajapyx sequence. All other weighted parsimony analyses (10:10:5, 10:10:2, 10:10:1) placed the beetle C2 + Apis F1 as sister clade to the Hymenoptera F2 clade (Fig. 1A). When the paralogs with the longest branches (beetle C2, Apis F1 and Drosophila F2 sequences) were removed, some weighting schemes (10:10:1, 5:10:1, 5:10:2) successfully recovered a topology (Fig. 1B) consistent with current phylogeny of insect orders (Wheeler et al., 2001). Replacing the C1 sequences with the C2 paralogs did not recover this topology in any of the analyses, disfavouring C2 as a phylogenetic marker. Although the basal relationships among the different insect EF-1α copies were ambiguously resolved, all phylogenetic analyses suggest frequent intron gains and losses in insects (Fig. 1A,C), consistent with Danforth & Ji’s (1998) previous analysis of insect EF-1α sequences. Intron loss in EF-1α has also been invoked for a variety of apterygote insects (Carapelli et al., 2000), as well as in lice vs. thrips, aphids and booklice (e.g. Cruickshank et al., 2001). In addition, deuterostome taxa show a similar trend of frequent loss of positional identical introns in the same gene (Wada et al., 2002). Also, frequent intron loss has been observed in other nuclear genes, for instance the dipteran white gene (Krzywinski & Besansky, 2002), and recent origins of introns have been detected in the xanthine dehydrogenase gene in Drosophila, favouring the ‘intron-late’ hypothesis (Tarrìo et al., 1998). These findings all argue against the traditional conservative view of intron evolution (e.g. Rokas et al., 1999; Venkatesh et al., 1999).

To conclude, intron structure then may not be a good indicator of deep homology in insect EF-1α phylogenies. However, the strongly supported monophyly of each of the two putative beetle copies (Fig. 1A,B) suggest that intron structure is a sufficient orthology criterion in beetles. In this study, eighty C1 sequences of shared single-intron structure were obtained for further analyses of xyleborine relationships. Given the congruent placement of C1 within the Holometabola (Fig. 1B), and the lower substitution rate in C1 than in C2 (average 8.0 vs. 10.6% in four scolytine spp.), it seems prudent to also select the C1 copy in further work on EF-1α beetle phylogenies. C1 sequence evolution in Xyleborini The base composition of the C1 sequences (Table 1) had a distinct T-bias in third positions and in the intron, a bias also observed in the C2 sequences. Base frequencies were also very similar to those measured in bees (Danforth et al., 1999; Danforth & Ji, 2001), but less so to butterflies (Reed & Sperling, 1999), a seemingly phylogenetic bias in accord with recent holometabolan phylogeny (Wheeler et al., 2001; cf. Fig. 1B). The maximum nucleotide divergence in the coding region for the haplodiploid ingroup was 11.4% (12.6% HKY corrected) between Cnestus suturalis and Cyclorhipidion pruinosum. However, most pair-wise comparisons were below 10% (cf. Fig. 2A), well below the minimum nucleotide divergence between the two different copies (22.0%). The rather low substitution rate in C1 seems advantageous judged by the lack of saturation in exon transversions and transitions for the ingroup (Fig. 2A). More surprisingly perhaps was the similar properties revealed by the single C1 intron. Although the divergence scatter was more dispersed, the C1 intron clearly mimics the C1 exon by showing a steady increase in its uncorrected distances plotted against the HKY corrected exon distances (Fig. 2B). Despite the much higher substitution rate and considerable AT bias in the intron (Table 1), the ratio of transitions to transversions is still higher than unity, suggesting a far from arbitrary substitution pattern. This may seem contrary to the commonly held opinion that intron sequences evolve so rapidly that they should be excluded in phylogenetic studies due to ambiguities with alignments. However, introns have been

Table 1. Properties of the different subpartitions of the EF-1α C1 fragment in scolytine beetles (n = 80). Ti/tv ratios are averaged over the most likely tree topology (see Fig. 5). None of the χ2 values exceeded the critical value in the homogeneity test of base frequencies across taxa Partition

Characters

Variable

Informative

ti/tv ratio

Adenine

Cytosine

Guanine

Thymine

χ2

Pos 1 Pos 2 Pos 3 Intron* All

282 282 281 76 921

40 19 261 72 392

26 12 240 68 346

3.55 0.77 4.12 1.27 –

0.29 0.31 0.21 0.26 0.27

0.17 0.24 0.25 0.15 0.22

0.39 0.16 0.17 0.12 0.23

0.15 0.29 0.37 0.47 0.28

9.8 ns 5.8 ns 157.3 ns 113.0 ns 64.9 ns

*Includes 6 gap-length coded characters.




Figure 2. (A) Pair-wise uncorrected exon distances (all positions) between the eighty C1 sequences (ti = transitions and tv = transversions), plotted against HKY corrected C1 exon distances. (B) Pair-wise uncorrected distances of the C1 intron plotted against HKY corrected C1 exon distances.

used successfully in aphids (Normark, 1999) and proved crucial to resolve certain relationships in halictid bees (Danforth et al., 1999; Danforth & Ji, 2001; Danforth, 2002). Based on the useful properties of the intron in this EF-1α study, these data are included in the phylogenetic study of the C1 sequences. Phylogeny estimation of xyleborine C1 sequences Incorporation of the intron greatly improved the maximum parsimony (MP) and minimum evolution (ME) trees in Table 2. Number of bootstrap supported (BP) nodes with and without the intron included, under the MP and ME optimality criteria. Increase per node was calculated over nodes with more than 50% bootstrap support in the exon analyses Maximum parsimony

Minimum evolution

Partition / BP

50%

90%

Increase per node

50%

90%

Increase per node

Exon Exon + intron

40 46

21 31

6.4%

43 48

26 30

3.3%

457

terms of an increased number of nodes supported and by an average increase in support per node (Table 2). All of the majority bootstrap supported nodes from the exon analysis were also supported in the analyses of all data combined; hence, I only report details from the latter analyses. It also led me to analyse only the full data set in the much more computationally intensive maximum likelihood (ML) analysis. Given the large sequence divergences and many sequences included, it is not surprising to find some discrepancies between the different phylogenetic methods. However, the large proportion of identical clades is striking and indicates secondary support for many of the weakly supported and shallow nodes. Moreover, the longest branches, which could cause long branch attraction of unrelated taxa, did not seem to introduce additional ambiguity. Judged by the similar groupings found in the long branch correcting ML analysis and in the MP analysis, for instance by Eccoptopterus, Webbia and allied taxa (Figs 3–5), long branch problems seem minor. On the other hand, the ME and ML analyses did manage to remove the obviously misplaced Xyleborus and Cyclorhipidion taxa from a paraphyletic Xylosandrus in the MP analysis (Fig. 3). All analyses supported the sistergroup relationship between Dryocoetes and the sib-mating clade, and between Ozopemon and the remaining sib-mating genera. This is in accordance with former results (Jordal et al., 2000; Jordal et al., 2002a), but the more extensive sampling of xyleborine species in this study demonstrated increased resolution in accordance with morphological classification (Wood, 1986). First of all, the Coccotrypes plus Dryocoetiops were monophyletic in the MP and ML analyses and were parapatrically distributed basally in the sib-mating clade in the ME analyses. Species of these two genera combined had identical internal topologies in all analyses, which also were identical to a recent analysis of Coccotrypes based on three genes and morphology (Jordal et al., 2002b). This indicates great reliability of EF-1α at this level of phylogenetic analysis. Second, the Xyleborini appeared monophyletic in the ML analysis, nearly so in the ME analysis (ex. Coptodryas), and monophyly was not contradicted by the MP analyses. Hence, this is the first molecular study to support the simultaneous monophyly of each of the Xyleborini and Coccotrypes (including Dryocoetiops). However, neither of these two clades were supported by synapomorphic amino acid or base substitutions – as would be expected given the shallow divergence between the two groups. Substitutional reversals are also an expected outcome in analyses of this many sequences of Miocene age (Jordal et al., 2000; Farrell et al., 2001), potentially obscuring the synapomorphic signature from ancient cladogenesis. All analyses also supported the monophyly of the following genera: Xyleborinus (related to Coptodryas), Webbia (related to Dryoxylon) and Theoborus (related to



458

B. H. Jordal

Figure 3. Strict consensus tree of twenty-one most parsimonious trees produced with all characters equally weighted (length 2456, CI = 0.29, RI = 0.47). Bootstrap majority support values are written above the internodes. The tree was rooted by the dryocoetine outbreeding genera Thamnurgus, Lymantor and Dryocoetes. Capital letters indicate country of origin for the two species with multiple samples: CR, Costa Rica; J, Japan; U, Uganda; PNG, Papua New Guinea.

Coptoborus). The ME analysis resulted in a monophyletic Xylosandrus (provided that mutilatus should be placed in Cnestus) and nearly so in the ML analysis (Figs 4, 5). More characters may help to consistently group this morpholo-

gically characteristic genus. Ambrosiodmus and Euwallacea appeared polyphyletic with respect to Xyleborus, within which these were formerly included (Wood, 1986). Although species of the first two genera are sometimes difficult to




459

Figure 4. Single tree with lowest score (3.083) resulting from the Minimum Evolution analysis of maximum likelihood distances. A GTR+Γ+I model was selected by the procedure described and implemented in ‘Modeltest’ (Posada & Crandall, 1998). Parameters estimated from the Neighbour Joining tree in ‘Modeltest’ were refined on an initial ME tree, with marginal changes. The final parameters were: estimated base frequencies, I = 0.554, Γ = 1.346 with four rate categories, and six substitution types with the following estimated frequencies: A↔C 1.07, A↔G 8.20, A↔T 1.56, C↔G 1.30, C↔T 9.78, G↔T 1.00.



460

B. H. Jordal

Figure 5. The most likely tree topology resulting from the Maximum Likelihood analysis using five random additions of heuristic searches (score 12246.130). Model and parameter settings were similar to the ME analysis (see Figure 4). Arrows point to evolutionary changes as follows: sib-mating, all species have strongly female biased broods or male is unknown; haplodiploidy confirmed, all examined species in Coccotrypes, Xylosandrus and Xyleborus are haplodiploid; fungus feeding, all species in the tribe Xyleborini feed upon the ambrosia fungi they cultivate in wood tunnels. Three independent origins of horned males are indicated by ‘H’.



Elongation Factor 1-α resolves the monophyly of Xyleborini separate morphologically, the putatively unrelated Ambrosiodmus colossus is unique by having males with a large horn-like projection on the anterior edge of pronotum. This character seems phylogenetically conservative with only three origins traced on the various topologies (Figs 3–5). That A. colossus groups exclusively with Xyleborus species also having this feature (see Fig. 5), may suggest that this species is more correctly placed by the EF-1α data than in current classification (Wood & Bright, 1992). Furthermore, the EF-1α data clearly rejected any relationship of Ambrosiodmus and Euwallacea to the putative close genera Amasa and Cyclorhipidion (according to Wood, 1986). On the other hand, most species of the latter two genera defined a well supported clade also containing Arixyleborus and one species of Xyleborus. This clade was the largest strongly supported clade in the Xyleborini and occurred in all analyses. All these species have rather narrow protibiae with more than ten socketed lateral teeth and strongly reticulate and dull elytral declivity, characters that may provide some clues towards finding good synapomorphies for this group. The clear evidence for a polyphyletic Xyleborus, Cyclorhipidion, Ambrosiodmus and Euwallacea may be expected based on previous extensive reports on specific as well as generic synonyms, and recent work has emphasized the preliminary nature of the current classification (Wood, 1986; Wood & Bright, 1992; Bright & Skidmore, 1997). Thus, rather than claiming the inadequacy of the EF1α data, it seems more productive to argue for an extensive revision of all genera in Xyleborini. In this regard, the EF1α data point to several presumably misclassified species in addition to A. colossus. One of these, Xylosandrus mutilatus, groups morphologically with Cnestus by the short anterior segment 1 of the antennal club, contiguous procoxae and similarly shaped pronotum and protibiae, and groups with strong support with Cnestus suturalis in all analyses. Another example is Dryoxylon onoharaensum, tentatively placed in Dryocoetini by Bright & Rabaglia (1999), but molecular as well as biological data suggest otherwise (see also Jordal et al., 2000). The single species of this genus is wood boring and possibly cultivate ambrosia fungi for food (Bright & Rabaglia, 1999). More importantly, the suggested dryocoetine diagnostic characters (Wood, 1986) do not exclude membership of the Xyleborini insofar as there are more than twenty species of Xyleborus that also match the tribal description of dryocoetine females (R. A. Beaver, pers. comm.). For instance, most species related to X. dolosus, X. subdentatus and X. fallax (cf. Figs 3–5) do not have depressed pregula and have narrow meso- and metatibiae with few socketed teeth. It is particularly interesting to observe in the ML analysis the basal position of these taxa in Xyleborini (Fig. 5). Thus, it seems conceivable that the shallow internodes distinguishing the sib-mating Dryocoetini and Xyleborini simply reflect the small and gradual morphological differences observed between the two biologically distinct groups.

461

Conclusion Neither molecular nor morphological markers provide clear cut results to distinguish xyleborine beetles from their sib-mating dryocoetine sister group, despite their fundamental differences in feeding behaviour. Although the phloem based diet of true bark beetles sometimes also contain ophiostomatoid (ambrosia) fungi, the exclusive and obligate feeding upon such domesticated asexual fungi (Beaver, 1989) has probably made the Xyleborini developmentally dependent upon this food resource (Kok et al., 1970). Reversal of fungus feeding under such circumstances seems quite unlikely, if not impossible. Thus, the molecular support for a monophyletic Xyleborini, albeit weak, must be viewed in this light. Taken together with these arguments, the EF-1α data thus provide a reasonably good estimate of haplodiploid beetle phylogeny and xyleborine monophyly. That several mitochondrial and nuclear genes have proven insufficient in previous analyses of the haplodiploid clade (Normark et al., 1999; Farrell et al., 2001; Jordal et al., 2002b), further points to the inherent difficulties in resolving the phylogeny of this group. Nonetheless, the improvements achieved by increased taxon sampling in this study not only shows the importance of the latter, but also demonstrates great properties of EF-1α as a phylogenetic marker for Miocene radiations (see Jordal et al., 2000). Finally, the inclusion of intron sequences proved very informative, and arguments for excluding such partitions in future may not be well founded. Experimental procedures Sampling All sequenced taxa are listed in Table 3, including their GenBank accession numbers. Among these, twenty-six sequences were determined in this study. Twenty-one of the twenty-six described genera nested in the haplodiploid clade sensu (Normark et al., 1999) have been sampled excluding Premnobius, which does not belong to this clade, and Mesoscolytus, which is a synonym of Xyleborus (see Beaver, 1998). For two morphologically variable species, Eccoptopterus spinosus and Coccotrypes advena, two sequences per species were included for intraspecific comparisons. Voucher specimens taken from the same brood as the sequenced specimens were pinned or kept in ice cold ethanol at the Museum of Comparative Zoology, Harvard University, or in the Department of Zoology, University of Bergen. PCR amplification and DNA sequencing Genomic DNA was extracted from single or half (thorax) specimens by the Qiagen DNeasy kit™. Partial EF-1α gene fragments were amplified by the polymerase chain reaction (PCR) technique using the following primers designed by Normark et al. (1999): efs149, 5′-ATCGAGAAGTTCGAGAAGGAGGCYCARGAAATGGG-3′ (forward); efa1043, 5′-GTATATCCATTGGAAAT T TGACCNGGRTGRTT-3′ (reverse); efa923, 5′ACGTTCTTCACGTTGAARCCAA-3′ (reverse). Amplification



462

B. H. Jordal

Table 3. Taxa sampled, including biological features and GenBank accession numbers. Possibly new or unidentified species are denoted by ‘cf.’ which indicates affiliation with the closely related species given by the specific epithet

Tribe Copy 1 Dryocoetini

Dryocoetini? Xyleborini

Species

Mating system

Breeding site

Feeding tissue

GeneBank accession no.

Lymantor coryli (Perris) Thamnurgus senecionis Schedl Thamnurgus lobeliae Eggers Dryocoetes affaber (Mannerhein) Dryocoetes autographus (Ratzeburg) Ozopemon uniseriatus Eggers Ozopemon brownei Schedl Coccotrypes advena Blandford – Japan Coccotrypes advena Blandford – Costa Rica Coccotrypes cardamomi Schaufuss Coccotrypes cf. cardamomi Schaufuss Coccotrypes carpophagus (Hornung) Coccotrypes cyperi (Beeson) Coccotrypes dactyliperda (Fabricius) Coccotrypes cf. distinctus (Motschusky) Coccotrypes gedeanus (Eggers) Coccotrypes graniceps (Eichhoff) Coccotrypes impressus Eggers Coccotrypes litoralis (Beeson) Coccotrypes longior (Eggers) Coccotrypes marginatus (Browne) Coccotrypes medius (Eggers) Coccotrypes petioli Blandford Coccotrypes cf. rhizophorae (Hopkins) Coccotrypes variabilis (Beeson) Dryocoetiops coffeae (Eggers) Dryocoetiops cf. eugeniae (Schedl) Dryoxylon onoharaensum (Murayama) Amasa versicolor (Sampson) Ambrosiodmus aegir (Eggers) Ambrosiodmus compressus (Lea) Ambrosiodmus colossus (Blandford) Arixyleborus cf. granifer (Eichhoff) Arixyleborus medius (Eggers) Arixyleborus cf. sus (Schedl) Cnestus suturalis (Eggers) Coptoborus pseudotenuis (Schedl) Coptodryas eucalyptica (Schedl) Cyclorhipidion agnatum (Eggers) Cyclorhipidion dentatulus (Browne) Cyclorhipidion pruinosum (Blandford) Cyclorhipidion cf. sexspinatum (Schedl) Dryocoetoides cristatus (Fabricius) Eccoptopterus spinosus (Olivier) – Uganda Eccoptopterus spinosus (Olivier) – PNG Euwallacea validus (Eichhoff) Euwallacea wallacei (Blandford) Euwallacea xanthopus (Eichhoff) Leptoxyleborus concisus (Blandford) Sampsonius dampfi Schedl Theoborus ricini (Eggers) Theoborus theobromae Hopkins Webbia bicornis (Schedl) Webbia cf. platypoides Eggers Webbia quattuordecimspinatus Sampson Xyleborinus intersetosus (Blandford) Xyleborinus saxeni (Ratzeburg) Xyleborus affinis Eichhoff Xyleborus annexus Schedl Xyleborus biuncus Browne Xyleborus dolosus Blandford Xyleborus fallax Eichhoff Xyleborus insulindicus Eggers Xyleborus justus Schedl Xyleborus meritus Wood

Outbreeding Outbreeding Outbreeding Outbreeding Outbreeding Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating

bark leaf inflorescence bark bark bark bark bark, seed, petiole bark, seed, petiole bark, seed, petiole bark, seed, petiole seed bark, seed, petiole seed seed bark, seed, petiole seed seed mangrove radicle bark (+ petiole bark) petiole bark, seed, petiole petiole petiole bark, seed, petiole twig twig wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood

phloem leaf inflorescence phloem phloem phloem phloem many many many many endosperm many endosperm endosperm many endosperm endosperm ‘endosperm’ phloem petiole many petiole petiole many pith pith fungus? fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus

AF439743 AF186662 AF186665 AF186661 AF259873 AF439740 AF259870 AF186668 AF444076 AF259869 AF444072 AF259872 AF444074 AF444078 AF444075 AF259867 AF259866 AF259874 AF259864 AF259871 AF186669 AF259875 AF444077 AF444071 AF259865 AF186670 AF439741 AF186660 AF186696 AF259877 AF508869 AF508868 AF508874 AF186695 AF508875 AF186694 AF508880 AF508878 AF259892 AF508877 AF259883 AF508867 AF186687 AF186686 AF508881 AF259878 AF508885 AF259893 AF259886 AF259885 AF186691 AF259881 AF259884 AF508882 AF259882 AF186684 AF259876 AF186688 AF508870 AF508871 AF259887 AF508873 AF508884 AF508876 AF508883




463

Table 3. continued

Tribe

Species

Mating system

Breeding site

Feeding tissue

GeneBank accession no.

Xyleborus metacuneolus Eggers Xyleborus multispinatus Eggers Xyleborus perforans (Wollaston) Xyleborus pfeili (Ratzeburg) Xyleborus pseudopilifer Schedl Xyleborus semipunctatus Eggers Xyleborus spathipennis Eichhoff Xyleborus sphenos Sampson Xyleborus cf. subdentatus Browne Xylosandrus crassiusculus (Motschulsky) Xylosandrus mancus (Blandford) Xylosandrus morigerus (Blandford) Xylosandrus mutilatus (Blandford) Xylosandrus cf. zimmermanni (Hopkins) Xylosandrus sp. n., related to ursa (Eggers)

Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating Sib-mating

wood wood wood wood wood wood wood wood wood wood wood wood wood wood wood

fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus fungus

AF259891 AF186690 AF259880 AF259879 AF259888 AF508887 AF508879 AF186692 AF259889 AF259890 AF186693 AF508866 AF508872 AF186685 AF508886

Copy 2 Coccotrypes advena (A) Coccotrypes impressus (B)

AF508928 AF508923 AF508924 AF508927 AF508925 AF508926

Theoborus ricini Xyleborus sphenos

Additional sequences used to produce Fig. 1 are found under the following GenBank accession numbers: U20125, U90054, U90056, X06869, X06870, X52884, AF015267, AF044815, AF044826, AF063405, AF063416, AF068466, AF068477, AF137388, AF137389, AF137390, AF140324, AF140336, AF147820, AF151624, AF163887, AF264861, AY048510, AY048536.

cycles for the new sequences consisted of 30 s initial denaturation at 94 °C, then forty cycles consisting of 48 °C annealing for 45 s, extension at 72 °C for 60 s and 94 °C denaturing for 30 s. PCR amplifications were performed in a 25 µl volume containing 0.2 M of each primer, 0.2 mM of each dNTP, 2.5 µl 10× Applied Biosystems buffer with additional MgCl2 to a final concentration of 2.0 mM, and 0.75 unit Applied Biosystems AmpliTaq DNA polymerase. PCR products were gel purified using Qiagen gel purification kit. Sequencing reactions followed Perkin Elmer’s recommended thermal profile (but with 10 s annealing at 50 °C) and analysed on a ninety-six well capillary sequencer.

Sequence alignment Sequences were assembled, edited and preliminary alignments performed using Sequencher 3.1 (Gene Codes Corporation). Intron positions were identified by GT (5′) and AG (3′) intron terminals and compared to intron positions in published insect sequences. The putative coding regions were validated by inferred amino acid translations. The intron used in the phylogenetic analysis was aligned in ClustalX (Thompson et al., 1997) under ten different parameter settings: gap cost of 2, 4, 8, 16 and 32, with gap extension cost half or equal to gap cost, and transitions equal to transversions. Consistently misaligned taxa were removed one after another to detect taxon specific regions disturbing the global alignment. Three such sequences were found and putative autapomorphic insertions were deleted to accommodate a less ambiguous alignment: Xylosandrus crassiusculus (4 bases deleted), Coccotrypes litoralis (14 bases [repeats] deleted) and Cyclorhipidion cf. sexpinatum (28 bases deleted). Based on careful comparison of closely related taxa determined from the phylogenetic analysis of the coding region, the final alignment parameters were selected to minimize topological conflict with

the exon analysis. The final alignment was based on gap and extension costs of 16, and the alignment of some sequences were manually corrected to fit separate alignments of close relatives. Gaps were treated as missing data, but gap lengths were coded and used in the parsimony analysis as described by Danforth et al. (1999).

Phylogenetic analyses PAUP* 4.0b4a (Swofford, 1999) was used to calculate distances and perform phylogenetic analyses of the two data sets. In the analyses of different insect EF-1α copies, introns were removed and the remaining 864 coding nucleotides subjected to weighted and unweighted MP analyses and ML analyses. The following weighting scheme was used in the parsimony analyses (pos1:pos2:pos3): 10:10:5, 10:10:2, 10:10:1, 5:10:2, 5:10:1, and amino acid coding. All analyses consisted of heuristic searches with 100 random additions. Parameters for the ML nucleotide analysis were estimated directly during heuristic searches with ten random additions, using a GTR+Γ+I model. In the final analysis of eighty beetle sequences of 845 coding nucleotides and one intron of 70 aligned base pairs, I also used ME analysis of maximum likelihood distances in addition to unweighted MP and ML. In the MP and ME analyses, heuristic searches with 500 random additions was used, all characters weighted equally. Node support was estimated via the bootstrap method (Felsenstein, 1985) with 200 replicates and ten random additions each. Maximum likelihood analysis of eighty taxa is computationally intensive and only five random additions were performed. Parameter settings for these ML and ME searches were estimated using the ‘Modeltest’ software (Posada & Crandall, 1998) in conjunction with PAUP*. The initial parameter settings estimated by ‘Modeltest’ (by Neighbour Joining) were refined over



464

B. H. Jordal

the ME topology, with marginal differences in estimates (slightly increased C-T substitution rate).

Acknowledgements I want to thank R. A. Beaver for identifying many of the specimens, B. B. Normark and D. J. Rees for helpful discussions and comments on the manuscript, and B. D. Farrell and L. R. Kirkendall for access to lab and computer facilities. This paper was supported by a Norwegian Research Council grant 123588/410, and a Marie Curie fellowship HPMF-CT2001-01323. References Beaver, R.A. (1989) Insect–fungus relationships in the bark and ambrosia beetles. In: Insect–Fungus Interactions (Wilding, N., Collins, N.M., Hammond, P.M., Webber, J.F., eds), pp. 121– 143. Academic Press, London. Beaver, R.A. (1998) New synonomy, new combinations and taxonomic notes on Scolytidae and Platypodidae (Insecta: Coleoptera). Ann Naturhist Mus Wien 100B: 179 –192. Bright, D.E. and Rabaglia, R.J. (1999) Dryoxylon, a new genus for Xyleborus onoharaensis Murayama. Coleopt Bull 53: 333–337. Bright, D.E. and Skidmore, R.E. (1997) A Catalog of Scolytidae and Platypodidae (Coleoptera), supplement 1 (1990 – 94). NRC Research Press, Ottawa. Browne, F.G. (1961) The biology of Malayan Scolytidae and Platypodidae. Malay for Rec 22: 1–255. Carapelli, A., Frati, F., Nardi, F., Dallas, R. and Simon, C. (2000) Molecular phylogeny of the apterygotan insects based on nuclear and mitochondrial genes. Paedobiologia 44: 361– 373. Cho, S., Mitchell, A., Regier, J.C., Mitter, C., Poole, R.W., Friedlander, T.P. and Zhao, S. (1995) A highly conserved nuclear gene for low-level phylogenetics: Elongation Factor 1-α recovers morphology-based tree for heliothine moths. Mol Biol Evol 12: 650–656. Cognato, A.I. and Vogler, A.P. (2001) Exploring data interaction and nucleotidealignment in a multiple gene analysis of Ips (Coleoptera: Scolytinae). Syst Biol 50: 758 – 780. Cruickshank, R.H., Johnson, K.P., Smith, V.S., Adams, R.J., Clayton, D.H. and Page, R.D.M. (2001) Phylogenetic analysis of partial sequences of Elongation Factor 1α identifies major groups of lice (Insecta: Phtiraptera). Mol Phyl Evol 19: 202 – 215. Cryan, J.R., Wiegmann, B.M., Deitz, L.L. and Dietrich, C.H. (2000) Phylogeny of the Treehoppers (Insecta: Hemiptera: Membracidae): evidence from two nuclear genes. Mol Phyl Evol 17: 317 –334. Danforth, B.N. (2002) Evolution of sociality in a primitively eusocial lineage of bees. Proc Nat Acad Sci 99: 286 – 290. Danforth, B.N. and Ji, S. (1998) Elongation Factor-1α occurs as two copies in bees: implications for phylogenetic analysis of EF-1α sequences in insects. Mol Biol Evol 15: 225 – 235. Danforth, B.N. and Ji, S. (2001) Australian Lasioglossum + Homalictus form a Monophyletic group: resolving the ‘Australian enigma’. Syst Biol 50: 268–283. Danforth, B.N., Sauquet, H. and Packer, L. (1999) Phylogeny of the bee genus Halictus (Hymenoptera: Halictidae) based on parsimony and likelihood analyses of nuclear EF-1α. Mol Phyl Evol 13: 605–618. Farrell, B.D., Sequeira, A., O’Meara, B., Normark, B.B., Chung, J. and Jordal, B. (2001) The evolution of agriculture in beetles

(Curclionidae: Scolytinae and Platypodinae). Evolution 55: 2011– 2027. Felsenstein, J. (1985) Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39: 783 –791. Hovemann, B., Richter, S., Walldorf, U. and Cziepluch, C. (1988) Two genes encode related cytoplasmatic elongation factors 1α (EF-1α) in Drosophila melanogaster with continuous and stage specific expression. Nucl Acids Res 16: 3175 – 3194. Jordal, B.H., Beaver, R.A. and Kirkendall, L.R. (2001) Breaking taboos in the tropics: Inbreeding promotes colonization by wood-boring beetles. Glob Ecol Biogeogr 10: 345 – 358. Jordal, B.H., Beaver, R.A., Normark, B.B. and Farrell, B.D. (2002a) Extraordinary sex ratios and the evolution of male neoteny in sib-mating Ozopemon beetles. Biol J Linn Soc 75: 353 – 360. Jordal, B.H., Normark, B.B. and Farrell, B.D. (2000) Evolutionary radiation of an inbreeding haplodiploid beetle lineage (Curculionidae, Scolytinae). Biol J Linn Soc 71: 483 – 499. Jordal, B.H., Normark, B.B., Farrell, B.D. and Kirkendall, L.R. (2002b) Extraordinary haplotype diversity in haplodiploid inbreeders: Phylogenetics and evolution of the sib-mating bark beetle genus Coccotrypes. Mol Phyl Evol 23: 171– 188. Kok, L.T., Norris, D.M. and Chu, H.M. (1970) Sterol metabolism as a basis for a mutualistic symbiosis. Nature 225: 661– 662. Krzywinski, J. and Besansky, N.J. (2002) Frequent intron loss in the white gene: a cautionary tale for phylogenetics. Mol Biol Evol 19: 362 – 366. Kukor, J.J. and Martin, M.M. (1989) Nutritional ecology of fungus feeding arthropods. In: Nutritional Ecology of Insects, Mites, Spiders and Related Invertebrates (Slansky, F. and Rodriguez, J.G., eds), pp. 791– 814. Wiley, New York. Mitchell, A., Cho, S., Regier, J.C., Mitter, C., Poole, R.W. and Matthews, M. (1997) Phylogenetic utility of Elongation Factor1α in Noctuoidea (Insecta: Lepidoptera): the limits of synonymous substitution. Mol Biol Evol 14: 381– 390. Normark, B.B. (1994) Phylogeny and Evolution of Parthenogenesis in the Aramigus tesselatus Complex (Coleoptera: Curculionidae). PhD Thesis, Cornell University, Ithaca, NY. Normark, B.B. (1999) Evolution in a putatively ancient asexual aphid lineage: recombination and rapid karyotype change. Evolution 53: 1458 –1469. Normark, B.B., Jordal, B.H. and Farrell, B.D. (1999) Origin of a haplodiploid beetle lineage. Proc R Soc Lond B 266: 2253 –2259. Norris, D.M. and Baker, J.K. (1967) Symbiosis: effects of a mutualistic fungus (Fusarium solani ) upon the growth and reproduction of Xyleborus ferrugineus. Science 156: 1120 –1122. Norris, D.M. and Chu, H. (1970) Nutrition of Xyleborus ferrugineus, II: a holidic diet for the aposymbiotic insect. Ann Ent Soc Am 63: 1142 –1145. Peleg, B. and Norris, D.M. (1972) Bacterial symbiote activation of insect parthenogenetic reproduction. Nature New Biol 236: 111–112. Posada, D. and Crandall, K.A. (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817 – 818. Reed, R.D. and Sperling, F.A.H. (1999) Interaction of process partitions in phylogenetic analysis: an example from the Swallowtail butterfly genus Papilio. Mol Biol Evol 16: 286 – 297. Regier, J.C. and Shultz, J.W. (2001) Elongation Factor-2: a useful gene for Arthropod phylogenetics. Mol Phyl Evol 20: 136 –148. Rokas, A., Kathirithamby, J. and Holland, P. (1999) Intron insertion as a phylogenetic character: the engrailed homeobox of Strepsiptera does not indicate affinity with Diptera. Insect Mol Biol 8: 527 – 530.



Elongation Factor 1-α resolves the monophyly of Xyleborini Rokas, A., Nylander, J.A.A., Ronquist, F. and Stone, G.N. (2001) A maximum-likelihood analysis of eight phylogenetic markers in gallwasps (Hymenoptera: Cynipidae): implications for insect phylogenetic studies. Mol Phyl Evol 22: 206 – 219. Schedl, K.E. (1956) Breeding habits of arboricole insects in Central Africa. Xth Int Congr Ent. Montreal 1: 183 –197. Sequeira, A.S. and Farrell, B.B. (2001) Evolutionary origins of Gondwanan interactions: How old are Araucaria beetle herbivores? Biol J Linn Soc 74: 459 – 474. Sequeira, A.S., Normark, B.B. and Farrell, B.D. (2000) Evolutionary assembly of the conifer fauna: distinguishing ancient from recent associations in bark beetles. Proc R Soc Lond B 267: 2359–2366. Shultz, J.W. and Regier, J.C. (2000) Phylogenetic analysis of arthropods using two nuclear protein-encoding genes supports a crustacean + hexapod clade. Proc R Soc Lond B 267: 1011–1019. Six, D.L. and Paine, T.D. (1998) Effects of mycangial fungi and host tree species on progeny survival and emergence of Dendroctonus ponderosae (Coleoptera: Scolytidae). Ann Ent Soc Am 27: 1393–1401. Swofford, D. (1999) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods), Version 4. Sinauer Associates, Sunderland, Massachusetts.

465

Tarrìo, R., Rodrìguez-Trelles, F. and Ayala, F.J. (1998) New Drosophila introns originate by duplication. Proc Natl Acad Sci 95: 1658 –1662. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. and Higgins, D.G. (1997) The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nuc Acid Res 24: 4876 – 4882. Uetsuki, T., Naito, A., Nagata, S. and Kaziro, Y. (1989) Isolation and characterization of the human chromosomal gene for polypeptide chain elongation factor-1α. J Biol Chem 264: 5791– 5798. Venkatesh, B., Ning, Y. and Brenner, S. (1999) Late changes in spliceosomal introns define clades in vertebrate evolution. Proc Natl Acad Sci 96: 10267–10271. Wada, H., Kobayashi, M., Sato, R., Satoh, N., Miyasaka, H. and Shirayama, Y. (2002) Dynamic insertion – deletion of introns in Deuterostome EF-1α genes. J Mol Evol 54: 118 –128. Wheeler, W.C., Whiting, M., Wheeler, Q.D. and Carpenter, J.M. (2001) The phylogeny of extant hexapod orders. Cladistics 17: 113 –169. Wood, S.L. (1986) A reclassification of the genera of Scolytidae (Coleoptera). Gt Basin Nat Mem 10: 1–126. Wood, S.L. and Bright, D. (1992) A catalog of Scolytidae and Platypodidae (Coleoptera). Part 2: Taxonomic index. Gt Basin Nat Mem 13: 1–1553.
