Combined molecular and morphological evidence on the phylogeny of the earliest lepidopteran lineages

© The Norwegian Academy of Science and Letters • Zoologica Scripta, 31 , 1, February 2002, pp67–81 67 Wiegmann, B. M., Regier, J. C. & Mitter, C. (2002). Combined molecular and morphological evidence on the phylogeny of the earliest lepidopteran lineages. — Zoologica Scripta , 31, 67–81. Agreement among recent morphological and molecular phylogenetic analyses has streng- thened estimates of the relationships among the earliest lineages of the holometabolan order Lepidoptera. For a few major groups, evidence for monophyly and basal relationships remains relatively weak or contradictory — chiefly within the clades of basal Glossata and Heteroneura. Here we assess the support for these controversial areas of lepidopteran classification through molecular systematic investigation of 18S rDNA sequence variation. Parsimony and maxi- mum likelihood analyses are presented for 1379 alignable sites of 18S. These data are then combined with 61 morphological features scored for major lineages of basal Glossata and Heteroneura. Our 18S rDNA data support recent hypotheses for the placement of Micro- pterigidae and Agathiphagidae as the basal-most lineages of Lepidoptera, and support the monophyly of the groups Neolepidoptera and Exoporia. 18S data alone are shown to be insufficient for resolving the monophyly and relationships of the Glossata, and for specifying relationships above the Neolepidoptera. Combination of the 18S data with published morphological ground-plan scorings improves overall support for the morphology-based hypothesis for basal glossatans, but phylogenetic resolution among published alternatives for the basal Heteroneura remains a major question for lepidopteran systematics. Brian M. Wiegmann, Department of Entomology, North Carolina State University, Raleigh, NC 27695, USA. E-mail: [email protected] Jerome C. Regier, Center for Agricultural Biotechnology, University of Maryland Biotechnology Institute, College Park, MD 20742, USA Charles Mitter, Department of Entomology, University of Maryland, College Park, MD 20742, USA Blackwell Science Ltd Combined molecular and morphological evidence on the phylogeny of the earliest lepidopteran lineages B RIAN M. W IEGMANN , J EROME C. R EGIER & C HARLES M ITTER Accepted: 26 September 2001 Introduction Over the last three decades, extensive morphological study of the nonditrysian Lepidoptera has provided a well-established phylogenetic framework for the basal groups (Davis 1975, 1978, 1986; Kristensen 1984; Nielsen 1989; Nielsen & Kristensen 1996; Kristensen & Skalski 1999). Recent separate and combined analyses of the 18S rRNA and phosphoenol- pyruvate carboxykinase (PEPCK) genes sampled from most of the major undisputed lineages, largely confirmed the morphology-based phylogeny (Friedlander et al . 1996; Wiegmann et al . 2000; Fig. 1). The striking congruence found for 18S rDNA with both PEPCK and morphology suggests that the 18S gene should be useful in resolving additional basal lepidopteran divergences originating in the Mesozoic, approximately 100–220 million years ago. Other studies of 18S gene variation in higher Lepidoptera concentrating on questions within the Ditrysia, show that the 18S gene is too conserved to resolve robustly divergences under 70 million years in age (Martin & Pashley 1992; Weller et al . 1992). Phylogenetic relationships whose origins fall between the early Cretaceous and the early Tertiary, however, have proved to be especially difficult to resolve by both molecular and morphological data (Friedlander et al . 2000; Wiegmann et al . 2000). This difficulty could be attributed to rapid clado- genesis in Lepidoptera and other holometabolan insect orders and/or to a paucity of suitably timed character changes. However, it is interesting that there is also a substantial number of ribosomal characters that appear to evolve too rapidly to be useful and, indeed, are difficult to align across taxa (Wiegmann et al . 2000). Here, 18S ribosomal gene sequence data from 39 taxa are analysed to address areas of early lepidopteran phylogeny for which morphological and molecular evidence have been either insufficient or conflicting. The 18S data are also combined with morphological features scored for ground-plans of the major groups by Nielsen & Kristensen (1996) and Krenn & Kristensen (2000). The specific issues on which these data are brought to bear include: (1) the relationships of the basal ZSC_091.fm Page 67 Friday, January 18, 2002 2:05 PM Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 68 Zoologica Scripta, 31 , 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters glossatan families; (2) the phylogenetic position of the Neop- seustidae; (3) monophyly and the relationships of basal Neole- pidoptera; (4) the relationships of basal heteroneuran clades. Relationships of the Lepidoptera: previous hypotheses Despite recent attention, there are several key areas of early lepidopteran phylogeny that morphological data have been unable to resolve adequately. Kristensen (1984) divided the Glossata into four infraorders — Dacnonypha, Neopseustina, Exoporia and Heteroneura. Kristensen’s (1997) recent view is that these higher-level categories should be discontinued in favour of reference to the individual lineages shown in the cladogram of Fig. 1A (Kristensen 1997; Kristensen & Skalski 1999). The composition, monophyly, and relationships of the Dacnonypha and Heteroneura have been particularly con- troversial. ‘Dacnonypha’ are now considered a paraphyletic assemblage (Kristensen 1997; Kristensen & Skalski 1999), but previously included the families Eriocraniidae, Acanthop- teroctetidae and Lophocoronidae (Common 1973; Davis 1978; Nielsen 1982; Nielsen & Kristensen 1996). Nielsen & Kristensen (1996) presented morphological evidence that Acanthopteroctetidae and Lophocoronidae are more recently derived than eriocraniids. These authors also hold that the Neopseustidae (= Neopseustina) and Lophocoronidae are the Trichoptera Agathiphagidae Eriocraniidae Lophocoronidae* Acanthopteroctetidae* Exoporia Heteroneura Neopseustidae* Glossata Neolepidoptera Lepidoptera Amphiesmenoptera 26 10 8 Micropterigidae 21 14 A 10 10 Myoglossata 7 6 11 Heterobathmiidae Incurvarioidea Ditrysia Tischerioidea Nepticuloidea Palaephatoidea C Incurvarioidea Ditrysia Tischerioidea Nepticuloidea Palaephatoidea B Incurvarioidea Tischerioidea Nepticuloidea D Ditrysia Palaephatoidea Fig. 1 Phylogenetic hypotheses for the basal Lepidoptera. —A. Lepidopteran relationships based on morphological evidence (Kristensen 1984; Nielsen & Kristensen 1996; Krenn & Kristensen 2000), 18S rDNA (Wiegmann et al. 2000) and phophoenolpyruvate carboxykinase (Friedlander et al. 1996). Asterisks indicate groups not included in published molecular systematic analyses. The number of assigned morphological synapomorphies is indicated on major group nodes with values below the Myoglossata from Kristensen (1984), Myoglossata and above from Nielsen & Kristensen (1996). —B–D. Published relationships within the Heteroneura based on comparative morphology. —B. Nielsen (1985, 1989). —C. Davis (1986). —D. Kristensen & Skalski (1999) and Krenn & Kristensen (2000). ZSC_091.fm Page 68 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31 , 1, February 2002, pp67–81 69 basal lineages of a monophyletic group termed Myoglossata (Fig. 1A), on the grounds that these families share true intrinsic musculature of the proboscis and normal-type wing scales with the remaining Lepidoptera, or Neolepidop- tera (= Exoporia + Heteroneura) (Kristensen 1984). In general, there are relatively few nonconflicting morpho- logical features that firmly establish the limits and relation- ships of these basal glossatan clades (Nielsen & Kristensen 1996). The relationships of the Neolepidoptera and the phylo- genetic origin of the ditrysian radiation are equally unsettled. The best supported recent view (Friedlander et al . 1996; Nielsen & Kristensen 1996; Kristensen & Skalski 1999; Wiegmann et al . 2000), places the Exoporia (Mnesarchaeidae + Hepialidae) as a sister group to the Heteroneura, which consists of four ‘monotrysian’ superfamilies and the Ditrysia, or higher moths and butterflies. There are three principal morphology-based hypotheses of heteroneuran relationships (Fig. 1B–D). One links the Ditrysia to a monophyletic group containing the 10 ‘monot- rysian’ families of the Heteroneura (Davis 1986; Fig. 1B). Davis’s (1986) proposed monophyly for the group Monotrysia (Fig. 1B) was based on two characters: the structure of the metafurcasternum and the presence of pectinifers on male valvulae. Davis (1986) also considered a possible relationship between the monotrysian family Palaephatidae and Ditrysia, although synapomorphic character support for this grouping is weak — external larval feeding, reduction of the frenular setae in the female, and loss of pectinifers in the male. A second hypothesis makes ‘Monotrysia’ paraphyletic and favours Palaephatidae + Tischeriidae as the closest relative to Ditrysia (Nielsen 1989; Nielsen & Common 1991; Fig. 1C). Nielsen (1982) had earlier recognized a possible sister-group relationship for Tischerioidea and Ditrysia based on the shared presence of socketed frenular setae. More recently, analyses of morphology and feeding biology (Kristensen 1997; Kristensen & Skalski 1999) support a basal position for Nepticuloidea (Opostegidae + Nepticulidae) relative to the Incurvarioidea and the remaining ‘higher Lepidoptera’ (Fig. 1D). Analyses of dopa decarboxylase sequences (Friedlander et al . 2000) also generally support a basal position for Nepticuloidea relative to other ‘monotrysian’ superfamilies — Incurvarioidea, Nepticuloidea, Tischerioidea, and Palaephatoidea — but their relationship to Ditrysia remains unclear. The current family-level classification of the nonditrysian Heteroneura is listed in Table 1. Materials and methods Taxon sampling We obtained 18S rRNA gene sequences for 37 lepidopteran species and two trichopteran outgroup species. The taxa sequenced, their origin, and their GenBank accession numbers are listed according to their position in the current lepidopteran classification in Table 1. Of the major basal lepidopteran families (Fig. 1A), only the endemic Australian family Lophocoronidae was not available for sequencing. Moths were either fresh frozen in liquid nitrogen and stored at –80 ° C, or collected in 100% ethanol and stored at –20 ° C. Species determinations, in most cases, were made from voucher specimens stored at –80 ° C and deposited in the Department of Entomology, University of Maryland. Laboratory methods The 18S gene was amplified in two fragments from total nucleic acid extracted from whole adults (DNA/RNA Isolation Kit, #US73750; Amersham, Arlington Heights, IL, USA). Polymerase chain reaction amplification of a 5 ¢ frag- ment, approximately 1200 bases in length, was carried out using primers rc18A–18H, and the remaining 3 ¢ fragment, approximately 700 bp, was amplified using primers rc18H and 18L [primer sequences are available in Wiegmann et al . (2000)]. Internal primers were used to sequence overlapping regions, and opposite strands were sequenced for confirma- tion. All amplifications followed standard three-step condi- tions (50 ° C annealing). Approximately 90% of the 18S molecule was sequenced for all taxa. Both manual sequencing and automated sequencing methods were used. Details of the manual sequencing are given in Wiegmann et al . (2000). Automated sequencing was performed using AmpliTaq DNA polymerase FS and automated DNA sequencers (Applied Biosystems, Foster City, CA, USA). Alignment and phylogenetic analysis Sequences were aligned in the multiple-sequence alignment editor in GENETIC DATA ENVIRONMENT 2.2 (Smith et al . 1994). All alignments were carried out by eye. A few areas of ambig- uous alignment due to multiple length variable indels were omitted from the phylogenetic analysis. Omitted positions, comprising approximately 200–250 bases concentrated in divergent domains, were those for which an ad hoc choice of gap placement could affect the phylogenetic outcome. The edges of the deleted regions were set as the last nucleotide position whose alignment preserves local positional homo- logy. Regions within 100 bases of terminal priming sites of amplified fragments were also excluded. The alignments are available electronically at http://www2.ncsu.edu/unity/ users/b/bwiegman/wieglab.html and are archived in the EMBL nucleotide alignment database. Unambiguous insertions and deletions (gaps) that show phylogenetically informative variation were coded as individual states of an unordered transformation series (Lloyd & Calder 1991; Maddison 1994). Sequencer chromatograms were edited, and contigs were assembled using the TREV and GAP 4 programs within the STADEN software package (Staden 1996). ZSC_091.fm Page 69 Friday, January 18, 2002 2:05 PM Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 70 Zoologica Scripta, 31 , 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters Taxon Source 18S rDNA GenBank Accession No. Trichoptera (outgroups) Philopotamidae Wormaldia moesta Maryland AF136861 Brachycentridae Brachycentrus nigrisoma Maryland AF136860 Lepidoptera Micropterigidae Epimartyria auricrinella Canada AF136862 Palaeomicra chalcophanes New Zealand AF421221, AF421245, AF421269 Sabatinca zonodoxa New Zealand AF421222, AF421246, AF421270 Micropterix calthella England AF136863 Agathiphagidae Agathiphaga queenslandensis Australia AF136864 Heterobathmiidae Heterobathmia pseuderiocrania Argentina AF136867 Glossata: Eriocraniidae Heringocrania unimaculella Denmark AF421225, AF421250 Eriocrania sangii Denmark AF421227, AF421252, AF421275 Eriocrania semipurpurella West Virginia AF136866 Dyseriocrania subpurpurella Denmark AF421243, AF421249, AF421273 Dyseriocrania auricyanea California AF421224, AF421248, AF421272 Dyseriocrania griseocapitella Maryland AF136865 Glossata: Acanthopteroctetidae Acanthopteroctetes unifasciata California AF421223, AF421247, AF421271 Glossata: Neopseustidae Apoplania valdiviana Chile AF421226, AF421251, AF421274 Glossata: Exoporia: Mnesarchaeidae Mnesarchaea acuta New Zealand AF421237, AF421263, AF421286 Mnesarchaea paracosma New Zealand AF421238, AF421264, AF421287 Glossata: Exoporia: Hepialidae Korscheltellus gracilis New Hampshire AF136870 Sthenopis quadriguttatus New York AF136871 Glossata: Heteroneura: Nepticulidae Ectoedemia populella Connecticut AF421236, AF421262, AF421285 Opostega saliciella Denmark AF421244, AF421253, AF421276 Glossata: Heteroneura: Heliozelidae Coptodisca kalmiella Connecticut AF421233, AF421259, AF421282 Antispila cornifoliella Connecticut AF421232, AF421258, AF421281 Glossata: Heteroneura: Cecidosidae Cecidoses eremita Argentina AF421234, AF421260, AF421283 Glossata: Heteroneura: Adelidae Chalceopla simpliciella California AF421231, AF421257, AF421280 Adela trigrapha California AF421229, AF421255, AF421278 Adela septentrionalis California AF421230, AF421256, AF421279 Glossata: Heteroneura: Palaephatidae Palaephatus falsus AF421235, AF421261, AF421284 Glossata: Heteroneura: Prodoxidae Prodoxus quinquepunctella Maryland AF136868 Greya politella California AF421228, AF421254, AF421277 Tegeticula yuccasella Maryland AF136869 Glossata: Heteroneura: Tischeriidae Tischeria citrinipennella Connecticut AF421240, AF421266, AF421289 Tischeria badiiella Connecticut AF421239, AF421265, AF421288 Glossata: Heteroneura: Ditrysia: Tineidae Tineola bisselliella Maryland (laboratory colony) AF136873 Glossata: Heteroneura: Ditrysia: Momphidae Mompha cephalonthiella Connecticut AF421242, AF421268, AF421291 Glossata: Heteroneura: Ditrysia: Gracillariidae Phyllonorycter sp. 1 Connecticut AF421241, AF421267, AF421290 Glossata: Heteroneura: Ditrysia: Psychidae Theridopteryx ephemeraeformis Maryland AF136874 Glossata: Heteroneura: Ditrysia: Lymantriidae Lymantria dispar Maryland AF136872 Table 1 Species of Amphiesmenoptera sequenced to reconstruct lepidopteran phylogeny. Classification follows Kristensen (1997) and Kristensen & Skalski (1999) . ZSC_091.fm Page 70 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31 , 1, February 2002, pp67–81 71 Phylogenetic trees were constructed using both parsimony and maximum likelihood methods in the program PAUP * 4.0b2a (Swofford 1999). Parsimony searches were performed by heuristic search with tree bisection–reconnection branch swapping (TBR) and random taxon addition (20 replications). The program MODELTEST 2.0 (Posada & Crandall 1998) was used to identify an appropriate and sufficient substitution model for maximum likelihood analysis of the 18S rDNA data set. MODELTEST applies hierarchical likelihood ratio tests (LRT) and the Akaike information criterion (AIC) to compare likelihood scores for nested and non-nested sets of substitution models for a user tree (we used the default settings in MODELTEST to calculate a neighbour-joining tree as the test topology) (Posada & Crandall 1998; Cryan et al . 2000). The reliability of the parsimony tree estimates was assessed using the Bremer support calculation for each node (Bremer 1988, 1994) using the program TREEROT v.2 (Sorenson 1999), and by bootstrap resampling (200 replicates) (Felsenstein 1985). Due to computational constraints on the maximum likelihood bootstrap calculation, we estimated bootstrap support for the maximum likelihood topology by performing 1000 neighbour-joining bootstrap searches under identical model conditions used for primary maximum likelihood searches. Partitioned Bremer support, or the proportion contributed by each data partition to the overall Bremer value for a node (Baker & DeSalle 1997), was calculated to assess relative support for combined data topologies. Morphological data and data set combination Morphological features scored as hypothesized ground-plan conditions for higher taxonomic groups were taken from Nielsen & Kristensen (1996) and Krenn & Kristensen (2000). These two publications examined the relationships among glossatan moths and did not score variation for more basal lineages outside the Glossata. Also, presumed nonhomoplasious autapomorphies of the individual homoneurous lineages (and of the hypothesized heteroneuran ground-plan) enumerated in Nielsen & Kristensen (1996) were not included, because they were not scored in the published data matrix. All scorings below Glossata were set to the ancestral state. Appendix I lists the morphological features as defined in Nielsen & Kristensen (1996) and Krenn & Kristensen (2000). Following Nielsen & Kristensen (1996) three features were coded as additive multistate transformations (characters 18, 19, 26), all other characters were treated as unordered. The phylogenetic data sets are available electronically from http://www2.ncsu.edu/ unity/users/b/bwiegman/wieglab.html. To combine morpho- logical and molecular data sets, the hypothesized ground- plan condition was assigned to each sequenced species representing that group. Consequently, the morphological data used here can only affect the relative placement of higher groups, but have no bearing on the relationships within them. Results Phylogenetic analysis of 18S rDNA sequences A total of 1379 aligned positions was included in the 18S rDNA phylogenetic data set. Thirteen gaps were numerically coded as part of an unordered transformation series. Of these, 10 were single position gaps and three were multisite. Of the 1392 total characters, 413 were variable (30%) and 246 (18%) were parsimony informative. Parsimony analysis of these data by heuristic search with TBR in PAUP * 4.0 and 20 replicates of random taxon addition found 63 equally parsimonious trees of 732 steps [length = 732, consistency index (CI) = 0.68, retention index (RI) = 0.78]. The strict consensus of these with bootstrap values and Bremer supports are shown in Fig. 2. The monophyly of the Lepidoptera is strongly supported. Within the Lepidoptera, the Micropterigidae branch off first, confirming their position as the most primitive extant lineage of moths. The four sampled micropterigid genera cluster in all trees and their monophyly is supported by a high Bremer support value (36; >4.9% of the total tree length). High boot- strap and character support, indicated by Bremer values greater than 1% of the total tree length (Smith 1989; Cracraft & Helm-Bychowski 1991), also confirm the monophyly of the Exoporia (Mnesarchaeidae + Hepialidae), the Lepidoptera above the Micropterigidae, and the three sampled congeners of Dyseriocrania . The minimum length trees also confirm placement of the Agathiphagidae as the sister group to all other Lepidoptera above the Micropterigidae, and the monophyly of the Neolepidoptera. Most of the variation among the 63 trees is attributable to the low character sup- port for relationships at the base of the Glossata and within the Heteroneura, resulting in many equally parsimonious arrangements in these regions of the phylogeny. The placement of Acanthopteroctetes as basal to all other Glossata + Heterobathmiina, and the paraphyly of the Eriocraniidae relative to Neopseustina, were unexpected from the morphological evidence (Fig. 1A). However, with the exception of the placement of the Neopseustina within the Eriocraniidae (see below), these groupings are supported by very few characters and low bootstrap and Bremer support values. Trees constrained to place the Heterobathmiina as a sister group to a monophyletic Glossata (Kristensen 1984) are only three steps longer (35 trees; length = 735) (Table 2). In these glossatan monophyly constrained trees, the Acanthopteroctetidae are placed as the basal glossatan lineage and the Eriocraniidae remain paraphyletic with respect to Neopseustidae. A similarly small ( Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 72 Zoologica Scripta, 31 , 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters Although the Heteroneura are considered monophyletic in nearly all recent morphological treatments, there is little 18S character support for the relationships within this group. The low support for relationships in this area of lepidopteran phylogeny precludes any strong statement here about heter- oneuran evolution. The results of constraining tree topologies according to three published alternatives are summarized in Table 2. Only two extra steps are required to make the Monotrysia monophyletic. Acceptance of the arrangement of heteroneuran taxa according to Davis’s (1986) monotrysian monophyly hypothesis requires only five additional steps and Nielsen’s (1989) cladogram for the ‘monotrysian’ superfamilies requires four extra steps (Table 2). The more recently proposed hypothesis of a basal Nepticuloidea (Kristensen & Skalski 1999; Krenn & Kristensen 2000; Friedlander et al . 2000) requires only two additional steps (Table 2). Consequently, the current views on basal heteroneuran relationships remain equivocal with respect to the small amount of informative variation available in the 18S ribosomal gene. The MODELTEST LRT and AIC calculations indicated that the Kimura 2-parameter model (Kimura 1980) with a correc- tion for significant invariant sites (K80 + I) is the best fitting Brachycentrus Wormaldia Epimartyria Micropterix Palaeomicra Sabatinca Agathiphaga Acanthopteroctetes Heterobathmia Dyseriocrania auricyanea D. griseocapitella D. subpurpurella Apoplania E. semipurpurella Heringocrania Eriocrania sangii Opostega Palaephatus Ectoedemia Lymantria Tineola Thyridopteryx Antispila Coptodisca Tischeria badiiella T. citrinipennella Phyllonorycter Mompha Chalceopla Adela trigrapha A. septontrionalis Cecidoses Tegeticula Prodoxus Greya M. paracosma Mnesarchaea acuta Korscheltellus Sthenopis LEPIDOPTERA 100 100 66 83 60 88 62 100 96 80 59 100 53 80 100 71 57 69 63 36 80 10 6 3 2 3 3 1 1 5 20 4 2 5 1 1 1 1 8 2 1 1 1 NEOLEPIDOPTERA TRICHOPTERA MICROPTERIGIDAE AGATHIPHAGIDAE ACANTHOPTEROCTETIDAE HETEROBATHMIIDAE ERIOCRANIIDAE NEOPSEUSTIDAE ERIOCRANIIDAE NEPTICULIDAE PALAEPHATIDAE DITRYSIA TISCHERIIDAE HELIOZELIDAE ADELIDAE CECIDOSIDAE PRODOXIDAE MNESARCHAEIDAE HEPIALIDAE OPOSTEGIDAE Fig. 2 Phylogenetic estimate of the basal Lepidoptera based on parsimony analysis of 18S rDNA nucleotide data. The strict consensus of 63 equally parsimonious trees [732 steps, consistency index (CI) = 0.68, retention index (RI) = 0.78] is shown with bootstrap support (>50%) indicated above branches and Bremer support indicated below branches. ZSC_091.fm Page 72 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31 , 1, February 2002, pp67–81 73 model for our 18S data. Maximum likelihood analysis of the 18S rDNA data under the K80 + I model yields a similar result to that found by parsimony (Fig. 3: –ln likelihood = 6002.58). Branch lengths and bootstrap values demonstrate the generally low information content of 18S rDNA for resolving relationships within the Neolepidoptera. The branch lengths depicted in Fig. 4 are similar to assigned branch lengths for the parsimony analysis (not shown), and these reveal considerable substitution rate variation across the sampled taxa. Rate heterogeneity is conspicuous for some basal glossatan taxa. In particular, the neopseustid Apoplania valdiviana and the eriocraniid subclade containing Eriocrania and Heringocrania species are each highly autapomorphic in 18S gene sequence — a factor that could influence their grouping (Felsenstein 1978; Huelsenbeck 1997; Wiens & Hollingsworth 2000). Phylogenetic analysis of the morphological data Parsimony analysis of the morphological data in PAUP * 4.0b2a yields two equally parsimonious trees (61 characters, length = 86, CI = 0.83, RI = 0.98). One of the two trees is depicted in Fig. 4. The two trees differ only in their resolution of the eriocraniid genera as a monophyletic group — characters supporting eriocraniid monophyly were not scored in Nielsen & Kristensen (1996) or Krenn & Kristensen (2000). Figure 4 reinforces the point that the morphological data included here apply only to the relationships of the higher- level groups and, hence, should not be viewed as a thorough reanalysis of morphological evidence among the sampled members of specific clades. Phylogenetic analysis of combined 18S rDNA and morphology Parsimony analysis (heuristic search, TBR, 20 random additions) of the 1453 characters of the combined 18S and morphological data yields 684 equally parsimonious trees (Fig. 5; length = 834, CI = 0.68, RI = 0.84). In these trees, support is generated for the monophyly of the Glossata, Eriocraniidae, Myoglossata and placement of Acanthopte- roctetidae and Neopseustidae. As expected from the low levels of 18S support, the morphological evidence is sufficient to overcome the relatively limited contribution of 18S at the base of the Glossata. This is seen in the differential parti- tioned Bremer support for key clades in Fig. 5. Reflecting the support of characters included from Krenn & Kristensen’s (2000) analysis of adult mouthparts, heteroneuran monophyly and the ‘nepticuloids basal hypothesis’ (Fig. 1D) are among the equally parsimonious solutions for the combined data set (Table 2). As only a few characters scored for adult mouthparts (Krenn & Kristensen 2000) were scored for the heteroneuran superfamilies, there remains little support for relationships above the Heteroneura in the results of the combined analysis. Discussion In accordance with an earlier study using fewer taxa, the phylogenetic trees obtained from analysis of the alignable portions of the 18S ribosomal gene are largely concordant with morphology-based relationships for the earliest Lepidoptera. Analyses of the data by parsimony and maximum likelihood methods generally agree on the branching order and monophyly of several of the previously well-established higher-level taxonomic groups. In particular, the 18S gene recovered the monophyly and phylogenetic position of the groups Micropterigidae, Agathiphagidae, Neolepidoptera, and Exoporia. There are some exceptions to the general concordance. Analysis of the 18S data places the Heterobathmiina within the Glossata, making the latter group, which is well supported by morphological data, paraphyletic. Relationships inferred Taxon 18S rDNA 18SrDNA + morpho Lepidoptera 100–0 100–0 Micropterigidae 100–0 100–0 Agathiphagidae + all remaining Lepidoptera 83–0 97–0 Heterobathmiidae + remaining Lepidoptera 57–3 94–0 Glossata 26–2 72–0 Eriocraniidae 8–5 72–0 Myoglossata >5–11 100–0 Neolepidoptera 80–0 100–0 Exoporia 100–0 100–0 Heteroneura 19–1 28–0 Heteroneuran relationships (Davis 1986; Fig. 1B) NA–5 NA–7 Heteroneuran relationships (Nielsen 1985, 1989; Fig. 1C) NA–4 NA–4 Heteroneuran relationships (Kristensen & Skalski 1999; Krenn & Kristensen 2000; Fig. 1D) NA–2 NA–0 NA, not applicable; denotes multiple groupings not describable by a single bootstrap value. Table 2 Relative support for Lepidopteran clades. Taxa follow the classification of Kristensen & Skalski (1999). Scores to the left are bootstrap values for each group; scores to the right are the number of additional steps (above the most-parsimonious tree length for the data set) necessary to recover each grouping with no other relationships constrained. ‘0’ values indicate that the group is supported in the most-parsimonious tree. Values for ‘Heteroneuran relationships’ are the number of extra steps required for most-parsimonious trees compatible with the hypotheses of Fig. 1B–D. ZSC_091.fm Page 73 Friday, January 18, 2002 2:05 PM Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 74 Zoologica Scripta, 31 , 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters Brachycentrus Wormaldia Epimartyria Micropterix Palaeomicra Sabatinca Agathiphaga Acanthopteroctetes Heterobathmia D. auricyanea Dyseriocrania griseocapitella D. subpurpurella Apoplania Eriocrania Heringocrania E. sangii Tischeria badiiella Tischeria citrinipenella Cecidoses Tegiticula Prodoxa Greya Antispila Coptodisca Ectoedemia Chalceopla A.trigrapha Adela septontrionalis Palaephatus Opostega Mnesarchaea paracosma M. acuta Korscheltellus Sthenopis Thyridopteryx Phyllonorycter Mompha Lymantria Tineola 0.005 substitutions/site LEPIDOPTERA NEOLEPIDOPTERA 100 100 100 74 96 100 93 62 83 99 98 69 Fig. 3 Phylogenetic estimate of the basal Lepidoptera based on maximum likelihood analysis of 18S rDNA nucleotide data. The topology and branch lengths were reconstructed under the Kimura 2-parameter model of nucleotide substitution with a correction for invariant sites (K80 + I); –log likelihood = 6002.84. Bootstrap support from 1000 replicate neighbour-joining searches under the K80 + I model (>50%) is indicated above branches. ZSC_091.fm Page 74 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31 , 1, February 2002, pp67–81 75 from the 18S sequence for the basal glossatan families Eriocraniidae, Neopseustidae, and Acanthopteroctetidae also differ from proposed morphological arrangements. Inter- pretation of these results may be clouded by observed 18S substitution rate heterogeneity among these taxa. Finally, the alignable 18S data showed limited ability to resolve relation- ships within the Heteroneura. In parsimony and maximum likelihood analyses, the molecule varied too little to place the ‘monotrysian’ families with confidence, and observed place- ments often differed from morphological treatments. The following sections summarize this observed variability in level of agreement between 18S data and previous hypotheses, focusing on the relative strength of the two sources of evi- dence, possible causes for disagreement, and areas requiring further molecular study. The primary lineages of the Lepidoptera The 18S data confirm the position of the Micropterigidae as the basal lepidopteran clade. All analyses united the four sampled micropterigid species, each representing a different genus, as the primary lineage within the Lepidoptera. The alternative hypothesis that Agathiphagidae, rather than Micropterigidae, are the most primitive lepidopterans, was revived by Shields (1993), but his argument rested only on 12 8 5 24 4 14 9 1 3 1 1 1 1 2 Brachycentrus Wormaldia Epimartyria Micropterix Palaeomicra Sabatinca Agathiphaga Acanthopteroctetes ACANTHOPTEROCTETIDAE Heterobathmia D. auricyanea D. griseocapitella D. subpurpurella Apoplania NEOPSEUSTIDAE Eriocrania Heringocrania E. sangii Opostega OPOSTEGIDAE Palaephatus PALAEPHATIDAE Ectoedemia NEPTICULIDAE Lymantria Tineola Thyridopteryx Phyllonorycter Mompha DITRYSIA Antispila Coptodisca HELIOZELIDAE Tischeria badiiella T. citrinipennella TISCHERIIDAE Chalceopla Adela trigrapha A. septontrionalis ADELIDAE Cecidoses Tegeticula Prodoxus Greya CECIDOSIDAE PRODOXIDAE M. parcosoma M. acuta Korscheltellus Sthenopis MNESARCHAEIDAE HEPIALIDAE TRICHOPTERA MICROPTERIGIDAE AGATHIPHAGIDAE HETEROBATHMIIDAE ERIOCRANIIDAE Fig. 4 Phylogenetic estimate of the basal Lepidoptera based on 61 morphological characters (Appendix I). One of two equally parsimonious trees [86 steps, consistency index (CI) = 0.83, retention index (RI) = 0.98] is shown with the number of assigned character state changes under ACCTRAN optimization in PAUP* 4.0b2a indicated above each branch. Character state assignments and synapomorphy lists are published in Nielsen & Kristensen (1996), Kristensen & Skalski (1999) and Krenn & Kristensen (2000). ZSC_091.fm Page 75 Friday, January 18, 2002 2:05 PM Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 76 Zoologica Scripta, 31 , 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters several plesiomorphic similarities that agathiphagids share with a mecopteroid fossil. In contrast, 16 unambiguous sub- stitutions in the ribosomal data set support Kristensen’s (1984) list of 10 morphological characters placing the Agathiphagidae as a sister group to the Heterobathmiina + Glossata. This branching order is consistent with the hypothesis that diversification of the earliest lepidopteran groups tracked that of higher plants (Powell et al . 1999). Larval micropterigids feed on liverworts (Bryophyta), detritus, or fungal hyphae (Heath 1976; Scoble 1992), and larval Aglossata live in the seed cones of Araucaria pines (Pinophyta: Araucariaceae), while the Glossata, most derived, have undergone extensive radiation primarily as angiosperm feeders. Thus, gymnosperm feeding seems to have preceded angiosperm association in lepidop- teran evolution, as expected if Lepidoptera and higher plants evolved more or less contemporaneously (Powell et al . 1999). Brachycentrus Wormaldia Epimartyria Micropterix Palaeomicra Sabatinca Agathiphaga Acanthopteroctetes Heterobathmia D. auricyanea D. griseocapitella D. subpurpurella Apoplania Eriocrania Heringocrania E. sangii Opostega Palaephatus Tischeria badiiella T. citrinipennella TISCHERIIDAE PALAEPHATIDAE Ectoedemia Lymantria Tineola Thyridopteryx Phyllonorycter Mompha DITRYSIA Antispila Coptodisca Chalceopla Adela trigrapha A. septontrionalis ADELIDAE Cecidoses Tegeticula Prodoxus Greya PRODOXIDAE M. parcosoma M. acuta Korscheltellus Sthenopis MNESARCHAEIDAE HEPIALIDAE TRICHOPTERA MICROPTERIGIDAE AGATHIPHAGIDAE ACANTHOPTEROCTETIDAE HETEROBATHMIIDAE ERIOCRANIIDAE NEOPSEUSTIDAE NEPTICULIDAE HELIOZELIDAE CECIDOSIDAE 100 80.0 / NA 97 9.0 / NA 94 6.0 / NA 91 –2.0 / 3.0 –2.6 / 5.6 100 –1.2 / 9.2 100 7.1 / 8.9 85 2.4 / –0.4 100 1.5 / 1.5 100 25.5 / 0.5 92 3.9 / –0.972 2.3 / –0.3 100 36.0 / NA 57 1.0/ NA 100 13.4 / 5.6 82 1.1 / –0.1 85 1.1 / 0.9 100 –0.2 / 1.2 100 –0.5 / 1.5 100 4.9 / 3.1 52 0.8/ 0.2 96 3.0 / 0 OPOSTEGIDAE Fig. 5 Phylogenetic estimate of the basal Lepidoptera based on parsimony analysis of combined 18S rDNA and morphology. The strict consensus of 684 equally parsimonious trees [834 steps, consistency index (CI) = 0.68, retention index (RI) = 0.78] is shown with bootstrap support (>50%) indicated above branches and partitioned Bremer support indicated below branches (18S/morphology). ZSC_091.fm Page 76 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31 , 1, February 2002, pp67–81 77 Glossatan monophyly and the position of the Heterobathmiina The 18S data consistently group Heterobathmiina with Glossata, and therefore support the five morphological char- acters taken from adults, larvae and pupae that unite these groups, as opposed to likely convergent features that formerly placed Heterobathmiidae with the Micropterigidae (Kristensen 1984). Nevertheless, the 18S data reverse the morphology-based positions of the suborder Heterobath- miina and the basal glossatan Acanthopteroctetes, rendering the Glossata paraphyletic. This switch, however, is not strongly supported. Removing the Heterobathmiina from Glossata to its usual position as a sister taxon to that group added only two steps to the minimum length tree estimate. Trees constructed omitting the acanthopterctetid did not resolve the monophyly of the Glossata relative to Hetero- bathmiidae, but the expected sister-group relationship among them was included in the equally parsimonious trees for that analysis (72 trees, length = 1042, CI = 0.64, RI = 0.76, RC = 0.48). Too few characters in the data differ- entiate the heterobathmiid from early glossatans. A similar result was found in a previous study of fewer taxa, which cor- rectly placed Heterobathmiina according to morphology, but the support for that placement was low and easily lost under taxon or character subsampling (Wiegmann et al . 2000). Of the 11 morphological characters shared by all glossatans exclusive of Heterobathmiina (Kristensen 1984), four are associated with the modification of the mouthparts into a coilable proboscis, and five are convergent in Agathiphagidae or Micropterigidae or both. In both its morphology and 18S gene sequence, Heterobathmia is a plesiomorphic taxon. That so few characters uniting all Glossata can be found may be attributable to the subsequent rapid radiation of that group and the observed evolutionary rate heterogeneity of its earli- est lineages. Each of these potentially affect observable synapomorphy through subsequent change: multiple hits for molecular data or further modification for morphology. Nevertheless, the morphological evidence for Glossata is com- pelling, especially given that the support comes from multiple character systems: mouthparts, larval spinnerets, and nervous system (Kristensen & Skalski 1999). Clearly, increased sam- pling for nucleotides and taxa in the basal glossatan lineages, and perhaps from additional heterobathmiid species, will be required to improve estimates of common ancestry for Glossata and potentially correct for spurious grouping due to rate heterogeneity. Phylogenetic position of the basal glossatan families Discerning phylogenetic relationships among the basal families of the Glossata based on morphology has been problematic for lepidopterists and is similarly difficult using the 18S gene. The concept of the higher taxon Dacnonypha has been particularly fluid in light of new characters and broader taxon surveys (Davis 1975, 1978; Kristensen 1984; Nielsen 1989; Nielsen & Kristensen 1996). The Dacnonypha have at some point contained the families Eriocraniidae, Acanthopteroctetidae, Neopseustidae, Lophocoronidae, Agathiphagidae, and Mnesarchaeidae. As discussed earlier, the Agathiphagidae have been transferred to their own sub- order, Aglossata, whose position outside Glossata is fairly certain. The Mnesarchaeidae are now securely placed in the Exoporia based on synapomorphies of female reproductive morphology (Dugdale 1974; Kristensen 1984). Relationships among the remaining four families are less certain, however, as few synapomorphies support any of the proposed relation- ships (Davis 1978; Kristensen 1984; Nielsen 1989). The Acanthopteroctetidae have been associated with the Eriocra- niidae by Davis (1978) as Dacnonypha, and as a possible sister group to the Myoglossata (Nielsen, personal communica- tion). The position of Acanthopteroctetidae implied by the 18S-based trees as basal to all other Glossata and Hetero- bathmiina contradicts both morphological hypotheses, and is not strongly supported. The low Bremer support values for each of the nodes specifying early Glossata suggest that alter- native positions for these taxa are not ruled out. In fact, the near parsimonious topologies (+ two steps) in which Glossata are constrained to be monophyletic place the Acanthopte- roctetidae with the Eriocraniidae + Neopseustidae. Davis (1978) listed five synapomorphies uniting Eriocraniidae and Acanthopteroctetidae, and the 18S data are not significantly incompatible with that grouping. More recently, the Acanthopteroctetidae and Neopseu- stidae have been removed from the Dacnonypha and placed as separate basal lineages of the Myoglossata (Nielsen & Common 1991; Nielsen & Kristensen 1996; Krenn & Kristensen 2000) or Lepidoptera with true intrinsic musculature of the proboscis, a feature lacking in the primitive glossatan family Eriocraniidae. These two families also share ‘normal-type’ hollow wing scales with all other Glossata, as opposed to the more primitive condition of solid scales found in the Eriocra- niidae, Agathiphagidae and Micropterigidae (Kristensen 1984). The placement of the Neopseustidae within the Eri- ocraniidae favoured in all analyses of the alignable 18S gene sequence data is thus highly suspect. Davis (1978) listed three morphological synapomorphies defining the Eriocraniidae, and while he recognized a possible connection of these to his Neopseustoidea (at a time when much less was known about the latter and about the Acanthopteroctetidae and Lophoc- oronidae), there is no reason to believe that the Eriocraniidae are paraphyletic. In addition, the grouping of Eriocrania sangii with Heringocrania was also unexpected as E. sangii is highly similar to E. semipurpurella in adult morphology. In light of the evidence for eriocraniid monophyly and relation- ships, the grouping of Apoplania and E. semipurpurella , each ZSC_091.fm Page 77 Friday, January 18, 2002 2:05 PM Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 78 Zoologica Scripta, 31 , 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters highly apomorphic in its 18S sequence, could be a case of spurious grouping by ‘long-branch’ attraction rather than true phylogenetic relationship. It is important to note, however, that this relationship was also found by maximum likelihood analysis, a method considered less sensitive to the effects of long-branch attraction (Huelsenbeck 1997). Place- ment of the neopseustid and acanthopteroctetid as the first lineages of Myoglossata according to Nielsen & Kristensen (1996) required 11 extra steps, only an approximate 1% increase in tree length. In sum, while the 18S data fail to support published hypotheses for the phylogenetic relationships among the basal glossatan families, these are hardly disproved and, indeed, are confirmed when the 18S and morphological evidence are analysed simultaneously. It appears that if the 18S gene is to be useful at clarifying relationships among these groups, greater sampling of taxa from the Eriocraniidae and Neopseustidae may be required to help place highly diver- gent taxa. Exemplars from the potentially important family Lophocoronidae would also probably improve estimates. Neolepidopteran relationships The monophyly of the Neolepidoptera, comprised of the groups Exoporia and Heteroneura, is confirmed in all ana- lyses of the 18S data set. The two families of Exoporia sampled here, Mnesarchaeidae and Hepialidae, consistently clustered as one of the better supported nodes in the most-parsimonious trees (Bremer support = 17) and were found in 100% of the bootstrap replicates. At least four morphological characters have been found uniting these taxa, one of which is the exopo- rian female genitalia, or separate egg-laying and mating genital openings with sperm transferred ‘externally’ along a furrow, for which the group is named (Bourgogne 1949; Scoble 1992). Sequence variation was low among the broadly sampled, but numerically limited, survey of taxa from the monotrysian superfamilies and Ditrysia. The small number of character changes assignable to branches within the Heteroneura probably indicates that the 18S gene is too conserved to resolve relationships for these taxa. The families for which multiple species were sequenced were recovered — Tischerii- dae, Adelidae, and Heliozelidae; however, no individual node had a Bremer support value greater than 3, indicating a lack of persuasive resolution in the data. Recent morphological (Krenn & Kristensen 2000) and nucleotide data (dopa decar- boxylase; Friedlander et al. 2000) tentatively support the basal nepticuloid arrangement of Fig. 3D, but the 18S data do not as yet strengthen these arguments. Future combined and separate data analyses will be greatly improved by includ- ing morphological features scored for exemplars, multiple genes scored for identical sets of taxa, and increased taxon sampling of critical taxa. Perhaps most importantly, data from additional genes with slightly higher levels of sequence variation will be necessary to test adequately the contrasting hypotheses on heteroneuran relationships. Acknowledgements We dedicate this work in memory of our departed friend and close colleague, Ebbe Schmidt Nielsen. Ebbe’s collaboration and impressive body of work in lepidopterology both inspired and significantly improved this study. We also gratefully acknowledge the support of N. P. Kristensen with whose expertise, data, and freely offered collaboration, this study was made more complete. We thank L. L. Deitz, H. Neunzig, S. Winterton and two anonymous reviewers for comments on the manuscript. B. Cassel and S. Zhao provided laboratory expertise. We also thank the many individuals who provided taxa to the project, including J. Brown, T. Friedlander, M. Gentili, P. Gentili, G. Gibbs, N. Kristensen, B. Landry, E. Nielsen, L. Pena, M. Scoble, and D. Wagner. This project was supported by grants from the University of Maryland Agricultural Experiment Station, National Science Foundation (DEB-9212669, DEB-9509174), United States Department of Agriculture, The University of Maryland Center for Agricultural Biotechnology and the North Carolina Agricul- ture Research Service. References Baker, R. H. & DeSalle, R. (1997). Multiple sources of character information and the phylogeny of Hawaiian drosophilids. System- atic Biology, 46, 674–698. Bourgogne, J. (1949). L’appreil génital femelle de quelques Hepialidae (Lépidoptères). Bulletin Société Zoologique de France, 74, 284–291. Bremer, K. (1988). The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution, 42, 795–803. Bremer, K. (1994). Branch support and tree stability. Cladistics, 10, 295–304. Common, I. F. B. (1973). A new family of Dacnonypha (Lepidoptera) based on three new species from southern Australia, with notes on the Agathiphagidae. Journal of the Australian Entomological Society, 12, 11–23. Cracraft, J. & Helm-Bychowski, K. (1991). Parsimony and phylo- genetic inference using DNA sequences: some methodological strategies. In M. M. Miyamoto & J. Cracraft (Eds) Phylogenetic Analysis of DNA Sequences (pp. 184–220). New York: Oxford University Press. Cryan, J. R., Wiegmann, B. M., Deitz, L. L. & Dietrich, C. H. (2000). Phylogeny of the treehoppers (Insecta: Hemiptera: Mem- bracidae): evidence from two nuclear genes. Molecular Phylogenetics and Evolution, 17, 317–334. Davis, D. R. (1975). Systematics and zoogeography of the family Neopseustidae with the proposal of a new superfamily (Lepidop- tera: Neopseustoidea). Smithsonian Contributions in Zoology, 210, 1–45. Davis, D. R. (1978). A revision of the North American moths of the superfamily Eriocranioidea with the proposal of a new family, Acanthopteroctetidae (Lepidoptera). Smithsonian Contributions in Zoology, 251, 1–131. ZSC_091.fm Page 78 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 1, February 2002, pp67–81 79 Davis, D. R. (1986). A new family of monotrysian moths from austral South America (Lepidoptera: Palaephatidae), with a phylogenetic review of the Monotrysia. Smithsonian Contributions in Zoology, 434, 1–202. Dugdale, J. S. (1974). Female genital configuration in the classifica- tion of Lepidoptera. New Zealand Journal of Zoology, 1, 127–146. Felsenstein, J. (1978). Cases in which parsimony or compatibility methods will be positively misleading. Systematic Zoology, 27, 401– 410. Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39, 783–791. Friedlander, T. P., Regier, J. C., Mitter, C. & Wagner, D. L. (1996). A nuclear gene for higher level phylogenetics: phosphoenolpyru- vate carboxykinase tracks Mesozoic-aged divergences within Lepidoptera (Insecta). Molecular Biology and Evolution, 13, 594– 604. Friedlander, T. P., Regier, J. C., Mitter, C., Wagner, D. L. & Fang, Q. Q. (2000). Evolution of heteroneuran Lepidoptera (Insecta) and the utility of dopa decarboxylase for Cretaceous- aged phylogenetics. Zoological Journal of the Linnean Society, 130, 213–234. Heath, J. (1976). Micropterigidae. In J. Heath (Ed.) The Moths and Butterflies of Great Britain and Ireland 1 (pp. 151–155). Oxford: Blackwell Scientific Publications and The Curwen Press. Huelsenbeck, J. P. (1997). Is the Felsenstein zone a fly trap? System- atic Biology, 46, 69–74. Kimura, M. (1980). A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleo- tide sequences. Journal of Molecular Evolution, 16, 111–120. Krenn, H. W. & Kristensen, N. P. (2000). Early evolution of the proboscis of Lepidoptera (Insecta): external morphology of the galea in basal glossatan moth lineages, with remarks on the origin of the pilifers. Zoologischer Anzeiger, 239, 179–196. Kristensen, N. P. (1984). Studies on the morphology and systematics of primitive Lepidoptera (Insecta). Steenstrupia, 10, 141–191. Kristensen, N. P. (1997). Early evolution of the Lepidoptera + Trichoptera lineage: phylogeny and the ecological scenario. Memoirs de Museum Nationale Histoire Natural, 173, 253–271. Kristensen, N. P. & Skalski, A. W. (1999). Phylogeny and paleonto- logy. In N. P. Kristensen (Ed.) Lepidoptera: Moths and Butterflies. 1. Handbook of Zoology, Vol. IV, Part 35 (pp. 7–25). New York: De Gruyter. Lloyd, D. G. & Calder, V. L. (1991). Multi-residue gaps, a class of molecular characters with exceptional reliability for phylogenetic analyses. Journal of Evolutionary Biology, 4, 9–21. Maddison, D. R. (1994). Missing data versus missing characters in phylogenetic analysis. Systematic Biology, 42, 576–581. Martin, J. A. & Pashley, D. P. (1992). Molecular systematic analysis of butterfly family and some subfamily relationships (Lepidoptera: Papilionoidea). Annals of the Entomological Society of America, 85, 127–139. Nielsen, E. S. (1982). Review of the higher classification of the Lepidoptera, with special reference to lower heteroneurans. Tyô to Ga, 33, 98–101. Nielsen, E. S. (1985). Primitive (non-ditrysian) Lepidoptera of the Andes: diversity, distribution, biology and phylogenetic relation- ships. In G. Lamas (Ed.) Second Symposium on Neotropical Lepidoptera. Arquipa, Peru, 1983. Journal of Research on the Lepi- doptera Supp 1., 1: 1–16. Nielsen, E. S. (1989). Phylogeny of the major lepidopteran groups. In B. K. Fernholm, K. Bremer & H. Jörnvall (Eds) The Hierarchy of Life (pp. 281–294). Amsterdam: Elsevier. Nielsen, E. S. & Common, I. F. B. (1991). Lepidoptera (moths and butterflies). In I. D. Naumann (Ed.) Insects of Australia, 2nd edn (pp. 817–915). Carlton, Victoria: Melbourne University Press and University College of London Press. Nielsen, E. S. & Kristensen, N. P. (1996). The Australian moth fam- ily Lophocoronidae and the basal phylogeny of the Lepidoptera Glossata. Invertebrate Taxonomy, 10, 1199–1302. Posada, D. & Crandall, K. A. (1998). MODELTEST: testing the model of DNA substitution. Bioinformatics, 14, 817–818. Powell, J. A., Mitter, C. & Farrell, B. (1999). Evolution of larval food preferences in Lepidoptera. In N. P. Kristensen (Ed.) Lepidoptera: Moths and Butterflies. 1. Handbook of Zoology, Vol. IV, Part 35 (pp. 403–422). New York: De Gruyter. Scoble, M. J. (1992). The Lepidoptera: Form, Function and Diversity. Oxford: Oxford University Press. Shields, O. (1993). Is Micropterigidae or Agathiphagidae the most primitive lepidopteran family? Tyô to Ga, 44, 152–156. Smith, A. B. (1989). RNA sequence data in phylogenetic reconstruc- tion: testing the limits of its resolution. Cladistics, 5, 321–344. Smith, S. W., Overbeek, R., Woese, C. R., Gilbert, W. & Gillevet, P. M. (1994). The genetic data environment and expandable GUI for multiple sequence analysis. Computer Applications in Bioscience, 10, 671–675. Sorenson, M. D. (1999). Treerot, Version 2. Boston: Boston University. Staden, R. (1996). The Staden sequence analysis package. Molecular Biotechnology, 5, 233–241. Swofford, D. L. (1999). PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods), Version 4. Sunderland, MA: Sinauer Press. Weller, S. J., Friedlander, T. P., Martin, J. A. & Pashley, D. P. (1992). Phylogenetic studies of ribosomal RNA variation in higher moths and butterflies (Lepidoptera: Ditrysia). Molecular Phylogenetics and Evolution, 1, 312–337. Wiegmann, B. M., Mitter, C., Regier, J. C., Friedlander, T. P., Wagner, D. M. & Nielsen, E. S. (2000). Nuclear genes resolve Mesozoic-aged divergences in the insect order Lepidoptera. Molecular Phylogenetics and Evolution, 15, 242–259. Wiens, J. J. & Hollingsworth, B. D. (2000). War of the iguanas: conflicting molecular and morphological phylogenies and long-branch attraction in iguanid lizards. Systematic Biology, 49, 143–159. Appendix I Morphological characters and character states Characters 0–46 are from Nielsen & Kristensen (1996), 47– 60 are from Krenn & Kristensen (2000). Characters 18, 19 and 36 were treated as linear ordered transformations. All other characters were treated as unordered. 0 Epistomal suture: well developed, with pronounced cutic- ular thickening (0); a zone of close-set grooves, cuticle of individual grooves without pronounced thickening (1); absent, clypeus not in any way demarcated from frons (2). 1 Dorsal cranial condyles at antennal base: absent (0); present (1). 2 Ocelli: present (0); absent (1). ZSC_091.fm Page 79 Friday, January 18, 2002 2:05 PM Phylogeny of the earliest lepidopteran lineages • B. M. Wiegmann et al. 80 Zoologica Scripta, 31, 1, February 2002, pp67–81 • © The Norwegian Academy of Science and Letters 3 Labrum: generalized (0); without frontal retractors (1). 4 Mandibular musculature: strongly developed (0); reduced, nonfunctional in postpharate stage, or absent (1). 5 Intrinsic galea musculature: absent (0); present (1). 6 Postlabial area: with arched, setose/scaly sclerotization (0); membranous and naked (1). 7 Sclerotized postgenal bridge proximad from postlabial sclerite/membrane: absent (0); present (1). 8 Posterior/upper sitophore portion: with truncate margin, not backwards bent/curved (0); medially with short, backwards bent marginal area (1); medially more or less pronouncedly produced and backwards bent (2). 9 Anterior/lower sitophore portion: almost parallel-sided (0); markedly tapering (1). 10 Ventral salivarium dilator: absent (0); prominent, arising in prelabium (1); arising in ligula, fibres very short (2). 11 Lateral frontal muscles inserting on the sucking pump (roof or sitophore margin): two pairs (0); one pair (1). 12 Proprioceptive setae on apical arm of laterocervicale: present (0); absent (1). 13 Posterior corner of laterocervicale: at least partly cover- ing anepisternal tooth (0); distinctly dorsad to anepisternal tooth/pleurocoxal bridge (1). 14 Anteromedian pronotal sclerite that articulates with anterodorsal propleural corners: absent (0); present (1). 15 Proprecoxal bridge: absent (0); present (1). 16 Muscle from proanepisternal tooth to cranium: absent (0); present (1). 17 Base of free profurcal arm: distinctly separate from pleural hind margin (0); touching or more or less extensively fused with pleural hind margin (1). 18 Mesophragma index: low, 0.45 and 0.55 (2). 19 Anteromedial margin of mesobasisternum: straight or with weakly convex curvature (0); with slightly produced, angularly convex outline (1); markedly produced anteriorly, but apex without strengthening ridge (2); markedly produced anteriorly, strengthening ridge extending to apex (3); markedly produced, apex strengthened and articulating with prospinasternum (4). 20 Transverse basisternal sutures, branching off from pre- coxal suture and delimiting the anterior ‘mesoclidium’ from the posterior basisternum s. str: absent (0); present (1). 21 Anterodorsal marginal area of mesepimeron: simple (0); anteriorly curling, concealing upper portion of pleural suture (1). 22 Mesotibial spur complement: two apical spurs (0); one apical spur (1). 23 Forewing radius: forked (0); unforked (1). 24 Forewing vein Rs4: preapical/apical (0); distinctly postapical (1). 25 Mesothoracic median notal wing process: short, 1/100 wing length (1). 26 Forewing jugum: a more or less prominently produced lobe, posterior margin (i.e. the margin between the lobe apex and the end of the jugal fold) concave (0); a blunt lobe, pos- terior margin almost straight (1); strongly reduced (2). 27 Wing surface scales: ‘primitive’ type (0); ‘normal’ type (1). 28 First thoracic spiracle: ‘primitive’ type (0); ‘normal’ type (1). 29 Acrotergite I: nearly horizontal (0); upturned, hence more or less vertical, at least in lateral part (1). 30 Posterior insertions of metathoracic indirect wing depressors: on anterior margin of acrotergite I (0); on topo- graphically anterior surface of upturned acrotergite I (1); on transverse antecosta = I-phragma (2). 31 Ligamentous connection between primary and secondary metafurcal arms: present (0); absent (1). 32 Sternum V gland: present (0); absent (1). 33 Subterminal segmental unit of female postabdomen (segment VIII or composite VIII + IX): not a synscleritous cone (0); a more or less extensively synscleritous cone (1). 34 Terminal segmental unit of female postabdomen: not dorsoventrally flattened, without finely serrated lateral ‘saw’ (0); dorsoventrally flattened, with finely serrated lateral ‘saw’ (1). 35 Spermathecal duct: simple, without distinct spiral coiling (0); at least on part of its course distinctly spiralled (1). 36 Gut configuration: stomodaeal crop undeveloped or small, midgut beginning in thorax (0); stomodaeal crop large, midgut restricted to abdomen (1). 37 Malpighian tubules: discharging separately into gut (0); arranged in two clusters with common ducts (1). 38 Deutocerebral lobes: ‘normal’, i.e. distinctly separate (0); contiguous in midline, forming a ‘deutocerebral loop’ around the posteromedian pharyngeal sucking pump dilator (1). 39 Oesophageal passage in the cephalic central nervous system: at least as wide as the lateral brain–suboesophageal ganglion connective mass (0); narrower (0.8 m m or less) than connective mass (1). 40 Optic lobe of brain: not appressed against suboesophageal ganglion (0); appressed against suboesophageal ganglion, completely enclosing a portion of the anterior tentorial arm (1). 41 Abdominal ganglion II in adult: shifted forward in front of metafurca and incorporated in posterior-most thoracic mass (0); remaining behind metafurca (1). 42 Dorsal thickening of abdominal nerve cord sheath: mainly cellular (0); mainly connective tissue (1); reduced (2). 43 Ventral diaphragm fibres in segment II: arising fan-wise from anterior margin (0); arranged in rows (1); reduced or absent (2). 44 Metathoracic aorta: touching wing heart (0); dissociated from wing heart (1). 45 Larval prolegs: absent (0); present on III–VI and X, with crotchets and extrinsic muscles (1). ZSC_091.fm Page 80 Friday, January 18, 2002 2:05 PM B. M. Wiegmann et al. • Phylogeny of the earliest lepidopteran lineages © The Norwegian Academy of Science and Letters • Zoologica Scripta, 31, 1, February 2002, pp67–81 81 46 Pupa: exarate, without abdominal spines (0); obtect, with spiny abdominal segments (1). 47 Galea elongate: absent (0); present (1). 48 Nonspinose food groove bearing ventral legulae: absent (0); present (1). 49 Dorsal legulae: absent (0); present (1). 50 One single-pointed ventral legula per food groove plate: absent (0); present (1). 51 Numerous sensilla trichodea proximally: absent (0); present (1). 52 Secondary ventral legulae: absent (0); present (1). 53 Zip scales link the galeae: absent (0); present (1). 54 Ribbed sensilla styloconica: absent (0); present (1). 55 Pilifer bristles: absent (0); present (1). 56 Scales on the basal proboscis: absent (0); present (1). 57 Prominent sensilla trichodea in the tip region: absent (0); present (1). 58 Regularly ribbed distal region of galea: absent (0); present (1). 59 Double-tube proboscis: absent (0); present (1). 60 Spines on the medial galea wall: absent (0); present (1). ZSC_091.fm Page 81 Friday, January 18, 2002 2:05 PM ZSC_091.fm Page 82 Friday, January 18, 2002 2:05 PM