A phylogenetic test of Ehrlich and raven's theory of escape and radiation in insects that feed on toxic plants, based on nearctic Depressaria moths (Gelechioidea: Elachistidae: Depressariinae), with discussion of the evolution of genitalia.
Key words: Gelechioidea, Depressariinae, morphology, phylogeny, host-plant evolution.
Moths of the genus Depressaria Haworth 1811 (Gelechioidea: Elachistidae: Depressariinae) are, for the most part, small, brown microlepidoptera that may be best known for their intimate association with their larval host plants. The majority of Nearctic Depressaria species feed on host plants in the families Apiaceae or Asteraceae (Hannemann, 1995; Hodges, 1974; McKenna and Berenbaum, 2003), with the preponderance of Nearctic species feeding on plants in the family Apiaceae (Hodges, 1974). These plants are known to contain a class of noxious secondary plant chemicals called furanocoumarins, some of which exhibit an increase in toxicity when exposed to ultraviolet light. Some Depressaria pull together portions of the host umbel to create a shelter protecting the larva from UV sunlight, hence protecting itself in part from the toxic effects of the chemicals. Furanocoumarins are also known from plants in Rutaceae (Berenbaum and Zangerl, 1991), Moraceae (Abegaz et al., 2004) and Fabiaceae (Innocenti et al., 1997). The ability of species in Depressariinae to metabolize toxins led Berenbaum (1983) to speculate that populations of Depressaria pastinacella Stainton 1849 could have coevolved with populations of its toxic hostplant Pastinaca sativa L. based on Ehrlich and Raven's (1964) classic mechanism of coevolution through escape and radiation. Based on this research, Berenbaum and Passoa (1999) later explored the possibility that genera within Depressariinae could have coevolved with their host plants. They found no overall pattern suggesting a potential case of coevolution, and suggested that a detailed analysis of species within Depressaria, and associated sister groups, is the next logical step testing Berenbaum's widely-cited hypothetical example of Ehrlich and Raven's model of coevolution in Depressaria.
Most evolutionary scenarios require a certain order of appearance of traits, hence they are implicitly statements about the phylogeny of a group. Phylogenetic tests are able to either confirm or refute such proposals, including when they are overlapping or non-exclusive (Bucheli et al., 2002; Bucheli, 2009). Here, we investigate larval host-plant preferences in a phylogenic framework. If escape and radiation were the driving forces behind patterns of modern day host-plant utilization, we would expect to see basal lineages feeding on less toxic plants while more derived lineages were feeding on more toxic plants. We would also expect to see a preference for certain plant taxa rather than a preference for certain plant tissues (flowers versus leaves) and growth forms (herbaceous versus woody), if the plant groups in question contain toxic furanocoumarins (Ehrlich and Raven, 1964; Berenbaum, 1983; Bucheli et al., 2002). Alternatively, tissue specialization may be the driving force, producing a pattern of host plant use that reflects more closely fidelity for certain plant tissues (flowers versus leaves) and growth forms (herbaceous versus woody) rather than taxa.
We address species-level systematics of Depressaria Haworth 1811 (Gelechioidea: Elachistidae: Depressariinae), commonly studied because of interesting physiological and behavioral reactions exhibited in response to furanocoumarins produced by their host plants. We include an in-depth study of morphological evolution, host-plant selection, and a discussion of geographical distribution. A generic-level phylogeny for the subfamily Depressariinae (Gelechioidea: Elachistidae) is presented. This study is the first modern phylogeny of Depressaria and Depressariinae, and includes New World as well as Old World species.
Taxonomy of Depressaria
The taxonomy and classification of species of Depressaria have varied, sometimes drastically, between authors (Hannemann, 1953; Hodges, 1974; Fetz, 1994; Berenbaum and Passoa, 1999). Clarke (1941b) delineated five species groups of Nearctic Depressaria based mainly on genital characters. These groups were: atrostrigella, artemisiae, and palousella in the first group; juliella, eleanorae, heracliana (now pastinacella in part), and einereocostella in the second group; artemisiella and alienella in the third group; togata, angustati, and multifidae in the fourth group, and lastly, maculatella, betullela, and grotella in the fifth group. Working with European taxa, Hannemann formally named species groups in 1953, which he later revised in 1995. His scheme was similar to Clarke's, but included two additional groups: the diseipunctella-group and the erinacella-hirtipalpis-group. Hodges (1974) later applied Clarke's and Hannemann's classification to all known Nearctic species. He created the betina group, which includes only two North American species that are morphologically distinct from European species. In addition, he removed five species from Depressaria (three of which, maculatella, betullela and grotella, were in Clarke's fifth group) to form the genus Nites. Currently, the six groups of Depressaria we recognize are: the artemisiae group, pastinacella group, thomaniella group, betina group, douglasella group, and discipunctella group (see Hodges, 1974 and Hannemann, 1953 for a diagnosis of these taxa).
Although the species groups of Depressaria are defined entirely on adult features, some publications contain detailed morphological information on the immature stages. Mosher (1916) separated the pupa of D. pastinacella from two other species of North American "Oecophoridae." Clarke (1941) characterized the larva and pupa of Depressaria at the genus level. Fetz (1994) in the Old World, and Passoa (1995) using mostly New World taxa, each studied larvae and pupae of Depressaria as part of a larger survey of the Gelechioidea. These last two works contained chaetotaxy maps of selected species and some pupal drawings. Fetz (1994) proposed the genus Hasenfussia based on larval features and partially described 24 species of Depressaria from Europe. Passoa (1995) provided a key to North American genera of larval and pupal Depressariinae and described in detail three species of Depressaria feeding on Apiaceae. The pupal study by Patocka and Turcani (2005) treated 23 species of Depressaria from Central Europe, including illustrations of taxonomic features used in their key. In contrast, there are no published larval keys to species of Depressaria in either Europe or North America, and many taxa remain poorly known.
MATERIALS AND METHODS
We follow the taxonomy presented in Hodges 1998 for Gelechioidea unless otherwise noted. This study examines the ecological hypothesis of escape and radiation in the North American Depressaria (Berenbaum, 1983; Ehrlich and Raven, 1964), so we focus on monophyly of Depressaria and its species groups. Our taxon sampling was broad and included the following representative genera from within the subfamily Depressariinae: Levipalpus Hannemann 1953, Nites Hodges 1974, Apachea Clarke 1941, Depressaria Haworth 1811, Agonopterix Htibner 1825, Bibarrambla Clarke 1941, Exaeretia Stainton 1849, Semioscopis Hfibner 1825, and Himmacia Clarke 1941. We also include genera from the subfamily Amphisbatinae (Machimia Clemens 1860, Amphisbatis Zeller 1870, and Psilocorsis Clemens 1860) because Hodges' (1974) treatment of the Depressariinae included this taxon as a tribe of Depressariinae (Amphisbatini). We added representatives from the Ethmiinae (Pyramidobela Braun 1923 and Ethmia Hfibner 1819) because Hodges (1998) included this subfamily with the Depressariinae in his enlarged concept of Elachistidae. We used Antaeotricha Zeller 1854 (Stenomatinae) as a distant outgroup and the root for this analysis. See Table 1 for a detailed list of taxa examined. This study represents one of the more detailed species-level works to date in this superfamily, and we offer an analysis of character evolution in Depressaria, especially regarding genital evolution. Broader analyses of related moths are underway (e.g., Kaila, 2004; Bucheli and Wenzel, 2005; Bucheli, 2009).
For the ingroup Depressaria, we examined adults of 20 Nearctic species, 3 Holarctic species, and 26 Palearctic species. Specimens were borrowed from the United States National Museum of Natural History, Washington, D.C. and the Mus6um National d'Histoire Naturelle, Paris, France, and supplemented with material from private collections (See Table 1). Undescribed species of Depressaria have been discovered in the western United States (McKenna and Berenbaum, 2003) as well as other new species of depressariines from other areas (Bucheli, unpubl.), but it is difficult to predict how many total species of Depressaria remain to be described.
A list of morphological characters is presented in Table 2. These were used to generate the data matrix of Table 3. Many of the specimens borrowed were previously dissected by other researchers. Given their historical significance, and because the vesica is often too brittle to evert from slide mounted material, we chose not to remount any preparations. When dissections were needed, specimens were prepared according to Clarke (1941a) using a standard 10% potassium hydroxide solution. In some cases, structures were not stained with Mercurochrome as this was not necessary to see the character in question. Preparations were mounted in Euparal on slides following Robinson (1976) and viewed at 100400x magnification. Illustrations were drawn first as pencil sketches with the aid of a drawing tube mounted on a compound microscope, digitized by scanning, and finalized using Adobe[R] Illustrator CS to create scalable vector files (Figs. 1-8.) All illustrations in this manuscript are original artwork done by SRB unless otherwise noted.
We compiled a morphological matrix consisting of 66 terminals and 47 characters with 113 states (all characters were non-additive except for Character number 1: condition of the second segment of labial palpi) using WinClada version 1.00.08 (Nixon, 1999) (Table 3). Adult characters were generated from personal observation of S.R.B. using published accounts by Clarke (1941b), Hannemann (1995), and Hodges (1974) as reference. Data on the pupal femur was taken from Patocka and Turcani (2005) and Berenbaum and Passoa (1999). When possible, characters were coded with existing terminology. If no appropriate term was available, we used a descriptive phrase to avoid introducing inappropriate homology into the literature (see Bucheli, 2009 for discussion). As is standard, non-applicable characters were coded with a dash, "-" and missing characters were coded with a question mark, "?."
We analyzed the matrix with two search programs, NONA ver. 2.0 (Goloboff, 1994) and TNT (Goloboff et al., 2000). We ran the matrix with "rs 0; hold 1,000; mult* 50" and obtained 102 trees, which we saved with "ksv*." Exiting, restarting, and performing "best; ksv*" twice gave sequentially 50 trees and 15 trees. These last 15 trees were stable to additional "best; ksv*" and form the basis of the consensus tree presented here. The consensus collapses 5 nodes and has length 156. TNT run under Asado (WinClada extended functionality) with the following parameters: hold 1,000 trees, 200 ratchet iterations, 4% upweight, 4% downweight, 50 iterations of Drift per replicate, 5 rounds of Tree Fusion, TBR-max. This procedure yields 105 trees of length 148, and if these are entered into Nona for 2 rounds of "best; ksv*" the same 15 trees result as from the original Nona searches. Taking the original 105 trees, consensus collapses five nodes for a length of 156. We show the consensus based on the 15 best trees because these trees have the highest internal consistency.
To investigate scenarios of coevolution and biogeography in a phylogenetic framework, characters of larval host preference, host-plant growth form, and geographic distribution of Depressariinae were optimized in nonadditive fashion onto the consensus phylogeny after the parsimony searches were implemented by allowing WinClada (1999) to objectively assign the fewest steps.
Toxicity of plants was used to test a scenario of coevolution by means escape and radiation (Berenbaum, 1983) by using plant family preference as a surrogate for plant toxicity: (0) larval host-plant unknown, (1) larvae feed on plants in the family Lamiaceae (do not produce furanocoumarins), (2) larvae fees on plants in the family Asteraceae (do not produce furanocoumarins), (3) larvae feed on plants in the family Apiaceae (produce furanocoumarins), and (4) larvae feed on multiple plants families known to produce furanocoumarins but is not a member of Apiaceae (such as Rutaceae and Fabaceae). If escape and radiation were the driving force behind patterns of modern day host-plant utilization, we would expect to see basal lineages feeding on less toxic plants while more derived lineages were feeding on more toxic plants and preference for certain plant taxa (Asteraceae, Apiaceae, or other) rather than a preference for certain plant tissues (flowers versus leaves) and growth forms (herbaceous versus woody) with exploitation of different plant taxa (Ehrlich and Raven, 1964; Berenbaum, 1983; Bucheli et al., 2002). To investigate a scenario of tissue and growth form specialization (Bucheli et al., 2002), we mapped the characters of larval host-plant tissue preference: (0) herbaceous-plant feeding or (1) woody-plant feeding. Generally speaking, woody-plant feeders are also leaf-tiers or leaf-rollers while herbaceous-plant feeders are flower-feeders (see discussion).
To superficially investigate patterns of global distribution, we mapped the characters of current location: (0) palearctic, (1) nearctic or (2) holarctic. For patterns of distribution, only members of the genus Depressaria were considered in this analysis because this is the only group with thorough species sampling, at greater than 50%.
Each analysis resulted in 15 equally most parsimonious topologies for 66 taxa (L = 148, C.I. = 0.44, R.I. = 0.84). For 50 taxa, Felsenstein reports the number of rooted, bifurcating trees as 2.75 x 1076 (2004:24). Thus, our data contain strong phylogenetic signal. A strict consensus collapsed five nodes (L = 156, C.I. = 0.42, R.I. = 0.83). Figure 6 shows the consensus phylogeny and character evolution for species of Depressaria. Bremer Support values (Bremer, 1988, 1994) up to five steps longer than the most parsimonious tree were calculated and are given in Figs. 7 and 8 (to the left of the node). Figures 7 and 8 show trends in morphological evolution for the Depressariinae. Characters of larval host preference, hostplant growth form, and geographic distribution of Depressariinae were optimized on the consensus phylogeny using WinClada (1999) to assign the fewest steps objectively (Figs. 9, 10, 11).
Resolution of Depressaria, and its species groups
There has been a great deal of species rearrangement between Agonopterix, Exaeretia and Depressaria by authors working on these genera (Clarke, 1941b; Hannemann, 1995; Hodges, 1974), suggesting a degree of historical uncertain ty about their boundaries. Clarke (1941) begins his revision of North American Oecophoridae by saying, "In the beginning I had intended only to do a specific revision of the genera Agonopterix and Depressaria. It soon became apparent, however, that it would be necessary to carefully study all the species known from North America together with many from other parts of the world." As Clarke had done, we intended to study only the systematics of North American Depressaria. It became clear to us as well that phylogenetic relationships of some species from a widespread genus cannot be based on examinations of species from one region alone. Therefore, we enlarged our study to include both North American and European depressariine taxa. This approach had both advantages and drawbacks. One important advantage was the ability to compare species groups of Depressaria proposed by North American and European authors in the same matrix. This methodology is especially critical because not all species groups of Depressaria occur in each region. In addition, a complete phylogeny with all the species groups represented is a powerful tool for tests of ecological and behavioral hypotheses. A major drawback was the inability to include larval characters in our expanded study. There were several reasons for this. Of the approximately 100 species of Depressaria known from the Holarctic Region, we were only able to examine 17 of them in the larval stage, with nearly all of them being from the New World. Large amounts of missing data for numerous Old World taxa in our data matrix would weaken our analysis. Another problem is that for the Old World larvae we did examine, generally only one or two individual specimens were available, and many of these were never associated with reared adults. Both Fetz (1994: 83) and Passoa (1995: 145) acknowledged variability in the chaetotaxy of Depressaria. This suggests that adding larval characters based on one or two specimens might be inappropriate, and even though many Depressaria have been reared (Hodges, 1974), there is still a need to preserve a large series of associated larvae in collections.
We were not able to solve the above problems by using the published literature. Fetz (1994) used groundplan coding, not explicit investigation of taxa, for some key characters of Depressaria (e.g., SV group of A1). He did diagnose species using the anal proleg chaetotaxy, but as pointed out by Stehr (1987), this segment has the most confusing homology of any part of the caterpillar body. Without specimens or illustrations to consult, we were not able to add these characters to the matrix. In summary, our survey largely agreed with Hodges (1998) who stated that Fetz (1994) produced a massive amount of novel information on larval Oecophoridae, but the phylogenetic significance of this information is currently difficult to evaluate.
In spite of the practical problems listed above, our brief survey of the literature and preserved immatures of Depressaria suggests a series of characters that would merit attention for phylogenetic studies. Small differences in mandible morphology will separate members of the betina group (Passoa, unpubl.). Fetz (1994: Fig. 55) noted that two species of Depressaria have secondary setae on the prothoracic shield and there was variation in the spacing of prothoracic setae. Several Depressaria species in the pastinacella group have exceptionally large pigmented body pinacula (Passoa, 1995). The most significant character may be the SV setae on A1. At least in the New World, members of the pastinaeella (D. cinereocostella, daucella, juliella, pastinacella) and artemis&e (D. artemisiae) groups have a bisetose SV condition which differs from other species of Depressaria (D. angustati, angelicivora, armata, betina, constancei, leptotaeniae, multifidae,) that have the SV group trisetose (Clarke, 1941; Passoa, unpubl.). This suggests the loss of a SV seta on A1 might be a character uniting the pastinacella and artemisiae groups as shown in our phylogeny (Figs. 6 and 8). In the Old World, Fetz (1994) also noted variation in the SV group of A1, but he correlated this to larval biology and host plants (flower-feeder or leaf-tier on Apiaceae or Asteraceae) instead of taxonomic status. As a final example, the spacing and angles formed by setae on the anal proleg varies among species of Depressaria (Fetz, 1994), and D. cinereocostella has six instead of the normal four setae on the lateral portion of the anal proleg (Passoa, 1995).
The most important pupal characters studied by Clarke (1941) to characterize Depressaria were in the labial palpi, prothoracic femur, cuticular texture and cremaster. Patocka and Turcani (2005) defined Depressaria in their key to Central European genera of "Depressariidae" by their exposed prothoracic femur. To diagnose three species in the pastinacella group, Passoa (1995) used the prothoracic cuticular texture (smooth or spiny), width of the appendages, shape of the dorsum on A1 and the presence or absence of ventral anal lobes. Patocka and Turcani (2005) used similar characters and added the form of the maxillary palpi and the profile of the anal area, especially in lateral view.
Our data matrix emphasizes male genitalia (66%), which other authors may criticize as inclusion of too many potentially homoplasious characters. Depressaria differ from each other mainly in features of the male genitalia; there is little variation in other adult features. Taxonomists such as Hannemann (1995) and Hodges (1974) have traditionally relied on features of male genitalia for species descriptions because they are a rich source of variation as well as being indicative of hierarchical relationships (as shown in this study). Hodges (1974) states, "I have been unable to find suitable characters of pattern or appearance to separate the species in most instances, so the key is based mainly on the male and female genitalia." Additionally, Song and Bucheli (2009) demonstrate that, with respect to the cladistic measures of fit of the Consistency Index (CI) and Retention Index (RI), phylogenetic signal between genital and non-genital data sets are statistically similar and that ad hoc dismissal of so many characters simply because they may be potentially homoplasious at supraspecific levels is unscientific. Based on the findings of Song and Bucheli (2009), we conclude that for this research, there is no a priori reason to exclude characters, and their value at resolving phylogenetic relationships speaks for itself regarding their hierarchical nature.
[TABLE 3 OMITTED]
Whereas Berenbaum's (1983) hypothesis of the evolution of host choice was a kind of scala naturae approach, lining up extant species in a series, Ehrlich and Raven's (1964) original theory of escape and radiation requires certain ancestral conditions and certain cladistic structure to the data. Ecological theories of the evolution of host choice may be confirmed through phylogenetic analysis (Bucheli et al., 2002; Bucheli, 2009) or not, but there is no way to evaluate their status outside of a phylogenetic examination. Ehrlich and Raven's hypothesis (1964) and Berenbaum's (1983) later detailed scenario came before cladistic thinking; however, parsimony is well suited to test these hypotheses. Our phylogeny of Depressaria and its near relatives permit discovery of ancient patterns of host use through character optimization. To investigate scenarios of coevolution of Depressaria and their larval host-plant preferences in a phylogenic framework, patterns of character evolution for fidelity to toxicity of host plant and utilization of host-plant tissue type were investigated.
If escape and radiation were the driving forces behind patterns of modern day host-plant utilization, we would expect to see basal lineages feeding on less toxic plants while more derived lineages were feeding on more toxic plants. We would also expect to see a preference for certain plant taxa rather than a preference for certain plant tissues (flowers versus leaves) and growth forms (herbaceous versus woody), if the plant groups in question contain toxic furanocoumarins (Ehrlich and Raven, 1964; Berenbaum, 1983; Bucheli et al., 2002). We do not see this pattern of evolution. Rather, we see a pattern of host plant use that reflects more closely preference for certain plant tissues (flowers versus leaves) and growth forms (herbaceous versus woody) with exploitation of different plant taxa, rather than preference for certain plant taxa with exploitation of different plant tissues, suggesting that tissue specialization is the driving force similar to evolution of host-plant selection found in Coleophora (Bucheli et al., 2002).
Species of Apachea feed on Ptelea (Figs. 9, 10), a genus within the family Rutaceae that produces furanocoumarins. Species of Agonopterix feed on many plant families, including the Apiaceae, Rutaceae, and Fabaceae, all of which also produce furanocoumarins as secondary metabolites. Species of Exaeretia are an exception in that they are not known to feed on plants containing furanocoumarins (Fig. 9). Species of Depressaria primarily feed on plants in the family Apiaceae (Fig. 9). Our data support Passoa (1995) in suggesting an ancestral state of feeding on plants with furanocoumarins, an ancestral state of feeding on Apiaceae, followed by a subsequent switch to less toxic species in Asteraceae (Fig. 9). Accordingly, at least three and no more than four derivations of feeding on plants with furanocoumarins have occurred within Depressariinae. Species of Depressaria primarily feed on plants in the family Apiaceae (Fig. 9). According to our results, there are six or seven independent derivations to asteraceous-plant feeding. In only one species group, the thomaniella Group, does asteraceous-plant feeding form a monophyletic clade in all most parsimonious topologies. Hasenfussia hirtipalp& feeds on plants in the Lamia ceae, another switch to a plant that does not produce furanocoumarins. These data also suggest that feeding on Apiaceae is the ancestral condition and that once feeding on apiaceous plant species evolves, it is easy to switch to asteraceous (or lamiaceous) plant species.
[FIGURE 1 OMITTED]
Among Depressariinae, the basal lineage Himmacia huachucella is a tree-feeding species (Fig. 10); specifically, larvae are recorded as leaftiers on species of oak (Quercus sp). Species of Agonopterix and Exaeretia both feed on a wide variety of woody and herbaceous plants families. Species of Agonopterix and Exaeretia are either leaf-tiers or leaf-folders; occasionally feeding in flowers of host plants; however, larval biology for several species of these genera remains unknown, including larval host preference. Nites maculatella and Apachea barberella, the most closely related genera to Depressaria, are leaf-tiers of trees (Hodges, 1974). All species of Depressaria whose larval hosts are recorded feed on herbaceous plants (Fig. 10). The larvae enter the umbels or meristems of the host plant after the first instar constructs a small silk shelter (Hodges, 1974). This trend would indicate that tree-feeding is the ancestral condition for Depressariinae with at least three switches to herbaceous-plant feeding. Additionally, it appears that well-known leaftying habits form the origin of the behavior that produces a shelter by tying the umbel above the larva protecting it from the UV that would activate furanocoumarins (Berenbaum, 1983; Berenbaum and Passoa, 1999; Berenbaum and Zangerl, 1991; Berenbaum and Zangerl, 1994).
[FIGURE 2 OMITTED]
The results of this investigation in general support the findings of Passoa (1995) and Berenbaum and Passoa's (1999) generic level phylogeny: co-speciation between genera of Depressariinae and their host plants is not the best explanation for patterns of host choice. However, this analysis is different in several regards. Our analysis investigates species of Depressaria rather than genera of Depressariinae. It is therefore a more detailed, and by nature, more specific test of a coevolutionary scenario. In general, within Depressariinae, species that feed on trees tend to use a higher number of host-plant genera than species that feed on herbaceous plants (Hodges, 1974; Hannemann, 1995). Woody-plant feeding species also tend to use host plants from different families, whereas if herbaceous-plant feeding species have multiple host plants, they tend to use genera within the same families.
This analysis included 16 outgroup taxa who are members of Depressariini (both Palearctic and Nearctic species were used when possible). Depressaria is Holarctic with approximately 100 described species. Twenty-one described species are endemic to North America, three species are Holarctic (one of which (daucella) may have been introduced to North America), and the remainder are from the Palearctic region (Hodges, 1974). Within Depressaria, this analysis included two Holarctic species, one introduced species, 19 Nearctic species, and 25 Palearctic species. Figure 11 shows the geographic distribution of Depressaria mapped onto the phylogeny. There are four clades of Depressaria that contain Nearctic taxa: The thomaniella group, the betina group, the clade of the douglasella group that contains the species (angelicivora, leptotaeniae, yakimae, angustati, togata, armata (whitmani, schellbachi, multifidae)(moya, pteryxiphaga, besma)), and the clade of the pastinacella group that contains the species (cinereocostella (daucella, juliella, eleanorae)). In some equally most parsimonious solutions, atrostrigella is sister to the Holarctic species artemisiae. Palearctic species are, generally speaking, the basal most lineages in the majority of the clades of Depressaria. This overall trend suggests Depressaria may be originally Palearctic in distribution with multiple colonization events into the Nearctic region.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Systematics of depressariinae
The subfamily Depressariinae (Figs. 6, 7 (Clade C)) is supported by having an anal lobe in the hindwing (8:1; later lost in the ancestor of Bibarrambla allenella), having the distal process of the male genitalia linear to curved at the tip, bifid, or quadrate (31:1; the distal process is later lost in the ancestor of the artemisiae group, the pastinacella group), and having the valvae broadest at base or nearly parallel sided (36:1 and 36:0; evolving multiple times within Depressariinae), and the parallelism of having a rounded vinculum (43:1).
Hodges (1974) defined Depressariinae as containing: Himmaeia, Semioscopis, Nites, Apachea, Bibarrambla, Agonopterix, Exaeretia (=Martyrhilda), and Depressaria. Passoa (1995) and Berenbaum and Passoa (1999) suggested that Himmacia and Semioscopis were more closely related to genera within Amphisbatinae than to genera within Depressariinae. Kaila's (2004) analysis suggested Semioscopis is more closely allied with other Depressariinae taxa (Exaeretia, Levipalpus, Depressaria, and Agonopterix) than to Amphisbatinae. Our analysis indicates that Semioscopis parckardella and S. avellanella are not closely related to other Depressariinae genera as they ally with genera belonging to Amphisbatinae; and that Himmacia huachucella is sister to the remaining members of Depressariinae. This relationship is supported by three uncontroverted synapomorphies (features of the hindwing (8:1), distal process (31:1) and shape of the valve (36:1)) and a moderate Bremer Support value of 3.
[FIGURE 5 OMITTED]
Passoa (1995), Berenbaum and Passoa (1999), and the current analysis find that Agonopterix and Exaeretia are sister clades, unlike Kaila's (2004) analysis, where Agonopterix, Depressaria, Exaeretia, and Levipalpus are all part of an unresolved polytomy. Passoa (1995), Berenbaum and Passoa (1999), and the current analysis also agree in the placement of Nites and Apachea as closely related to Depressaria. This analysis suggests that Bibarrambla allenella and Levipalpus hepatariella are closely related and that Hasenfussia hirtipalpis may be a species of Depressaria with modified larvae (see below for further discussion) (Figs. 6-11).
Depressaria (Figs. 6, 7 (Clade G), and 8 (Clade G)) is monophyletic and is supported by having the prothoracic femur of the pupae exposed (0:1; shared with Amphisbatis incongruella and Psilocorsis reflexella), having ocelli present (5:1; shared with Bibarrambla allenella, Levipalpus hepatariella, Agonopterix flavicomella, A. gelidella, Exaeretia fulvus and E. lutosella), having the uncus reduced (12:1; shared with Agonopterix flavicomella and A. gelidella but lost secondarily in D. pastinacella), and having the socii present (14:2; shared with Semioscopis packardella, S. avellanella, Agonopterix flavicomella, A. gelidella, Exaeretia fulvus and E. lutosella). All species groups of Depressaria are monophyletic.
Morphological trends within Depressaria
The derived condition for Depressariinae is to have the vesica of the aedeagus with cornuti (Fig. 7, clade B, I), later lost in the ancestor of the betina and douglasella groups (Fig. 8, clade K, XXI). The aedeagus becomes more highly curved in the ancestor of the (betina group, douglasella group) (Fig. 8, clade J, XXIII). The presence or absence of a flange is homoplastic, evolving several times in this analysis. A single flange develops in the ancestor of the (betina group, douglasella group) (Fig. 8, clade J, XXII) (shared with Nites maculatella). The flange is lost in the ancestor of clade M (Fig. 8, XXVI) and a double flange evolves in the ancestor of clade P (Fig. 8, XXIX).
There is a trend for the development of the transtilla from absent to present and sclerotized in the ancestor of clade D (Fig. 7, VII), with two transformations to a membranous structure; once in the ancestor of (cinereocostella (juliella, daucella, eleanorae) clade b (Fig. 8) and once in the ancestor of clade H (Fig. 8, XIX). The transtilla is later reversed to the ancestral condition of being sclerotized in the ancestor of the thomaniella group (Fig. 8, clade d) and the ancestor of clade Q (Fig. 8). The vinculum transforms to a more pointed condition in the ancestor of Depressariinae (except for Himmacia) (Fig. 7, clade D, V) and is later reversed in the ancestor of clade O (Fig. 8, XXVIII).
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
There is a trend towards a reduction of the uncus as seen in the ancestor of Depressaria (Fig. 8, clade G, XIII). In the taxa we examined, socii are either reduced (clade U) or developed. Socii are reduced as in clade U (Fig. 8) or present within the taxa in this analysis. The presence of socii is homoplastic and evolved at least three times in this analysis: once in the ancestor of Semioscopis (Fig. 7, clade S), once in the ancestor of (Exaeretia, Agonopterix) (Fig. 7, clade V), and once in the ancestor of Depressaria (Fig. 7, clade G, XIV). A spined gnathos evolves in the ancestor of clade B (Fig. 7, II).
The ancestral condition for Depressaria is to have a distal process that is more or less straight. The presence of a branched distal process evolves in the ancestor of clade F (Fig. 8, XXV) and the presence of an elbowed distal process evolves in the ancestor of clade N (Fig. 8, XXVII). The basal process is plesiomorphically absent but evolves twice within Depressaria, once in the ancestor of the pastinacella group (Fig. 8, clade Y, XVII), where only the tip is covered with stout scale-like projections, and once in the ancestor of the douglasella group (Fig. 8, L, XXIV), where the entire process is covered with long scale-like projections. The ancestor of clade Q, containing only North American taxa, has a basal process that is highly curved with the inner margin convex (Fig. 8, XXX).
[FIGURE 8 OMITTED]
Identity of Hasenfussia Fetz, 1994
Hannemann (1953) placed Depressaria hirtipalpis Zeller 1854 the erinacella-hirtipalpis-Group based on the presence of a well developed tuft on the labial palpi as well as features of the male genitalia, such as" absence of the basal process and the socii; a pointed vinculum; and presence of a sack-like distal process. Fetz (1994) placed Depressaria hirtipalpis in a newly created genus, Hasenfussia, based on larval traits. Features important for this taxonomic change were: larva feeding internally on Salvia officianalis (Labiatae); cuticle of the body with polygonal densely sclerotized areas; pinacula strongly reduced; second subventral setae (SV2) on abdominal segments 3 through 6 located opposite of first and third subventral setae (SV2 and SV3); and the third lateral seta (L3) absent from the ninth abdominal segment. The host being outside of the Apiaceae or Asteraceae, loss of L3 on A9, and the cuticle texture are especially unusual compared to other known Depressaria.
Hannemann (1995) does not discuss Hasenfussia hirtipalpis in his latest treatment of Depres saria. In "The Lepidoptera of Europe," Karsholt and Razowski (1996) place H. hirtipalpis in Depressaria without comment and there is no formal synonymy of Hasenfussia under Depressaria. Our phylogenetic analysis indicates that, in order to preserve monophyly of Depressaria, Hasenfussia hirtipalpis belongs within Depressaria.
[FIGURE 9 OMITTED]
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
Depressaria hirtipalpis syn. nov. exhibits the following characters: second segment of labial palpi with a tuft or well-developed brush with a deep furrow; anellus without lateral lobes or projections; ocelli present; antennal pectin present; forewing with [Cu.sub.1] and [Cu.sub.2] separate; hindwing with anal lobe present, [M.sub.3] and [Cu.sub.1] stalked, removed from Cu2; abdominal tergites curving over sternites (abdomen flattened); uncus reduced, short and rounded; socii present and separate from uncus; gnathos broadly triangular and broadly joined to tegumen; aedeagus blunt and straight or nearly so, lacking a flange; vesica of aedeagus with numerous scalelike cornuti located in a patch; sacculus lacking basal lobe or process; valve lacking lobe; distal process straight or slightly curved, the apex rounded or pointed; costal margin of valve straight or nearly so; valve broadest base to nearly 5/6 its length or nearly straight; cucullus rounded; costal margin of valve lacking a process; transtilla membranous and lacking lateral lobes and vinculum pointed without well developed anterodorsal process.
This analysis represents the most comprehensive phylogeny of Depressaria to date. Investigation of host-plant preference suggests that woodyplant feeding is the ancestral condition for genera in the subfamily Depressariinae. It also suggests that feeding on plants which contain furanocoumarins is the ancestral condition for Depressaria and there are several independent colonization events onto Asteraceae from an ancestor that fed on Apiaceae. Therefore, the findings of our research do not support a classic "escape and radiation" coveolutionary scenario; rather, our results suggest that even though many species of depressariine, as well as the majority of species of Depressaria, feed on toxic host plants, members of Depressariinae in general and Depressaria in particular are tissue and habitat specialists.
From the systematic perspective, our results suggest that Semioscopis is a not a member of Depressariinae, and that Himmacia is a basal lineage of Depressariinae. To maintain monophyly of Depressaria, Hasenfussia Fetz 1994 is synonymized with Depressaria Haworth 1811. Historical species groups within Depressaria designated by Hodges (1974) are monophyletic.
We are grateful to two anonymous reviewers and our communicating editor for their carful attention to this manuscript while in review. We are grateful to David Adamski at the National Museum of Natural History, Washington, D.C., Joel Minet at the Museum national d'Histoire naturelle, in Paris, France, and John Rawlins at the Carnegie Mellon Museum of Natural History, in Pittsburgh, PA, who lent specimens for comparison. We thank Rick Hoebeke for Amphisbatis incongruella on loan to Steven C. Passoa from Cornell University Insect Collection. We are also grateful to Christopher P. Randle who translated German text. We are appreciative of comments made by colleagues regarding this manuscript in its earliest drafts (including a poster and submitted talk) and to SRB's SHSU graduate students for helpful final edits. This project was funded by NSF# 0416051.
Received 11 June 2010; accepted 1 October 2010.
Abegaz, B. M., B. T. Ngadjui, G. N. Folefoc, S. Fotso, P. Ambassa, M. Bezabih, E. Dongo, F. Rise and D. Petersen. 2004. Prenylated flavonoids, monoterpenoid furanocoumarins and other constituents from the twigs of Dorstenia elliptica (Moraceae). Phytochemistry 65(2): 221-226.
Berenbaum, M. R. 1983. Coumarins and caterpillars: a case for coevolution. Evolution 37: 163-179.
Berenbaum, M. R. and S. Passoa. 1999. Generic phylogeny of North American Depressariinae (Lepidoptera: Elachistidae) and hypotheses about coevolution. Annals of the Entomological Society of America 92: 971-986.
Berenbaum, M. R. and A. R. Zangerl. 1991. Acquisition of a native host-plant by an introduced oligophagous herbivore. Oikos 62:153-159.
Berenbaum, M. R. and A. R. Zangerl. 1994. Costs of inducible defense: protein limitation, and detoxification in parsnip webworms. Ecology 75: 2311-2317.
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.
Bucheli, S. R. 2009. Annotated review and discussion of phylogenetically important characters for families and subfamilies of Gelechioidea (Insecta: Lepidoptera). Zootaxa 26(1): 23 35.
Bucheli, S. R., J-F. Landry and J. Wenzel. 2002. Larval case architecture and implications of host-plant associations for North American Coleophora (Lepidoptera; Coleophoridae). Cladistics 18: 71-93.
Bucheli, S. R. and J. W. Wenzel. 2005. Gelechioidea (Insecta: Lepidoptera) systematics: A reexamination using combined morphology and mitochondrial DNA data. Molecular Phylogenetics and Evolution 35: 380-394.
Clarke, J. F. G. 1941a. The preparation of slides of the genitalia of Lepidoptera. Bulletin of the Brooklyn Entomological Society 36: 149-161.
Clarke, J. F. G. 1941b. Revision of the North American moths of the family Oecophoridae, with descriptions of new genera and species. Proceedings of the United States National Museum 90:33 286.
Ehrlich, P. R. and P. H. Raven. 1964. Butterflies and Plants: a study in coevolution. Evolution 18: 586-608.
Felsenstein, J. 2004. Inferring Phylogenies. Sinauer Associates Inc, Sunderland, Massachusetts, 664 pp.
Fetz, R. 1994. Larvalmorphologische Beitrage zumphylogenetisehen System der ehemaligen Oecophoridae (Lepidoptera, Gelechioidea). Neue Entomologische Nachrichten 33:1-273.
Goloboff, P. A. 1994. NONA reference: documentation for version 1.15. Program and documentation available from http://www.zmuc.dk/public/ phylogeny/Nona-PeeWee/.
Goloboff, P. A., J. S. Farris and K. C. Nixon. 2000. TNT: Tree Analysis Using New Technology. Program and documentation available from http://www.zmue.dk/public/phylogeny/TNT/.
Hannemann, H-J. 1953. Naturliche Gruppierung der europeaischen Arten der Gattung Depressaria s.l. (Lep. Oechop.). Mitteilungen aus dem Zoologishen Museum in Berlin 29(2): 270-372.
Hannemann, H-J. 1995. Kleinschmetterlinge Oder Microlepidoptera IV. Flachleibmotten (Depressariidae). Die Tierwelt Deutschlands, 69. Teil. G. Fischer Verlag, Jena, Stuttgart. 192 pp.
Hodges, R. W. 1974. Gelechioidea: Oecophoridae (in part). In: R. B. Dominick, T. Dominick, D. C. Ferguson, J. G. Franclemont, R. W. Hodges and E. G. Munroe (eds.), The Moths of America North of Mexico. E. W. Classey Ltd. and The Wedge Entomological Research Foundation, London, 142 pp.
Hodges, R. W. 1998. The Gelechioidea. In: N. Kristensen (ed.), The Handbook of Zoology/Handbuch der Zoologic. 1999. Walter de Gruyter, Berlin & New York, 132-158.
Innocenti, G., A. Piovan, R. Filippini, R. Caniato and E. M. Cappelletti. 1997. Quantitative recovery of furanocoumarins from Psoralea bituminosa. Phytochemical Analysis 8: 84-86.
Kaila, L. 2004. Phylogeny of the superfamily Gelechioidea (Lepidoptera: Ditrysia): an exemplar approach. Cladistics 20:303-340.
Karsholt, O. and J. Razowski. 1996. The Lepidoptera of Europe: A distributional checklist. Apollo Books, Stenstrup, Denmark, 380 pp.
McKenna, D. D. and M. R. Berenbaum. 2003. A field investigation of Depressaria (Elachistidae) host plants and ecology in the western United States. Journal of the Lepidopterist's Society 57(1): 36-42.
Nixon, K. C. 1999. Winclada (BETA) ver. 0.9.9 Published by the author, Ithaca, NY. Program and documentation available from http://www. cladistics.com/about_winc.htm.
Passoa, S. C. 1995. Larval and pupal systematics of Nearctic Amphisbatinae and Depressariinae (Lepidoptera; Oecophoridae) [dissertation]. University of Illinois, Urbana, 300 pp.
Patocka, J. and M. Turcani. 2005. Lepidoptera Pupae: Central European Species. Apollo Books, Stenstrup, Denmark, 863 pp.
Robinson, G. S. 1976. The preparation of slides of Lepidoptera genitalia with special reference to the microlepidoptera. Entomologist's Gazette 27: 127-132.
Song, H. and S. R. Bucheli. 2009. Comparison of phylogenetic signal between male genitalia and non-genital characters in insect systematics. Cladistics 26(1): 23-35.
Stehr, F. W. 1987. Superfamily Gelechioidea. In: F. W. Stehr (ed.), Immature Insects. Kendall/Hunt, Dubuque, Iowa, 379-399.
SIBYL R. BUCHELI (1,4), STEVEN PASSOA (2) AND JOHN W. WENZEL (3)
(1) Sam Houston State University, Department of Biological Sciences, Box 2116, Huntsville, TX 77341
(2) United States Department of Agriculture, Animal and Plant Health Inspection Service, Plant Protection and Quarantine, Museum of Biological Diversity, 1315 Kinnear Road, Columbus, OH 43212
(3) The Ohio State University, Department of Entomology, Museum of Biological Diversity, 1315 Kinnear Road, Columbus, OH 43212
(4) Email address for correspondence: bucheli@shsu. edu
Table 1. Species of Depressariinae and outgroups used in cladistic analysis. USNM, National Museum of Natural History, Washington, D.C.; GSMNP, Great Smoky Mountains National Park voucher collection; in part, Columbus, OH; SPIC, Steven Passoa Insect Collection, Columbus, OH; SRBC, Sibyl Rae Bucheli Collection, Huntsville, TX. Specimen examined Collection STENOMATINAE Antaeotricha schlaegeri Zeller, 1854 SRBC ETHMIINAE Ethmia longimaculella (Chambers, 1872) SPIC Pyramidobela angelarum Keifer, 1936 SPIC AMPHISBATINAE Machimia tentoriferella Clemens, 1860 SPIC Psilocorsis refexella Clemens, 1860 GSMNP Amphisbatis incongruella (Stainton, 1849) SPIC DEPRESSARIINAE Agonopterix favicomella (Engel, 1907) GSMNP Apachea barberella (Busck, 1902) USNM Bibarrambla allenella (Walsingham, 1882) USNM Depressaria absynthiella Herrich-Schaffer, 1865 USNM Depressaria albipunctella (Denis & Schiffermuller, USNM 1775) Depressaria alienella Busck, 1904 USNM Depressaria angelicivora Clarke, 1952 USNM Depressaria angustati Clarke, 1941 USNM Depressaria armata Clarke, 1952 UNSM Depressaria artemisiella McDunnough, 1927 USNM Depressaria artemisiae, Nickerl, 1864 USNM Depressaria atrostrigella Clarke, 1941 USNM Depressaria badiella (Hubner, 1796) USNM Depressaria beckmanni Heinemann, 1870 UNSM Depressaria besma Clarke, 1947 USNM Depressaria betina Clarke, 1947 USNM Depressaria bupleurella Heinemann, 1870 USNM Depressaria cervicella Herrich-Schaffer, 1854 UNSM Depressaria chaerophylli Zeller, 1839 USNM Depressaria cinereocostella Clemens, 1864 USNM Depressaria constancei Clarke, 1947 USNM Depressaria daucella (Denis & Schiffermuller, 1775) USNM Depressaria depressana (Fabricius, 1775) USNM Depressaria discipunctella Herrich-Schaffer, 1854 USNM Depressaria douglasella Stainton, 1849 USNM Depressaria eleanorae Clarke, 1941 USNM Depressaria heydenii Zeller, 1854 USNM Depressaria hofmanni Stainton, 1861 USNM Depressaria juliella Busck, 1908 USNM Depressaria libanotidella Schlager, 1849 USNM Depressaria leptotaeniae Clarke, 1933 USNM Depressaria leucocephala Suellen, 1884 USNM Depressaria moya Clarke, 1947 UNSM Depressaria multifidae Clarke, 1933 USNM Depressaria nemolella Svensson, 1982 UNSM Depressaria olerella Zeller, 1854 USNM Depressaria pastinacella (Duponchel, 1838) SRBC Depressaria pimpinellae Zeller, 1839 USNM Depressaria pulcherrimella Stainton, 1949 USNM Depressaria pteryxiphaga Clarke, 1952 USNM Depressaria schellhachi Clarke, 1947 USNM Depressaria silesiaca Heinemann, 1870 USNM Depressaria sordidatella Tengstrom, 1848 USNM Depressaria togata Walshingham, 1889 USNM Depressaria ultimella Stainton, 1849 USNM Depressaria ululana R6ssler, 1866 USNM Depressaria velox Staudinger, 1859 UNSM Depressaria whitmani Clarke, 1941 USNM Depressaria yakimae Clarke, 1941 USNM Exaeretia lutosella (Herrich-Schdffer, 1854) USNM Exaeretia fulvus Walsingham, 1882 UNSM Hasenjussia hirtipalpis Zeller, 1854 USNM Himmacia huachucella (Busck 1908) USNM Levipalpus hepatariella (Lienig & Zeller, 1846) USNM Nites maculatella (Busck, 1908) GSMNP Semioscopis packardella (Clemens, 1863) SPIC Semioscopis avellanella (Hubner, 1793) SPIC Table 2. Morphological characters used to generate the data matrix in Table 3. Character State 0. Prothoracic femur of pupa 0. hidden (Berenbaum and Passoa, 1999) 1. exposed 1. Second segment of labial 0. lacking tuft or brush palpi [additive] 1. with moderately developed tuft or brush 2. with well developed tuft or brush (broadly triangular) 2. Furrow of tuft or brush 0. poorly developed or lacking l. well developed 3. Anellus 0. without lateral lobes or projections 1. with lateral lobes 2. with sclerotized projections 4. Mesothoracic leg with scale 0. absent tufts 1. present at halfway and at apex 5. Ocelli 0. absent l. present 6. Antennal pecten 0. absent 1. present 7. Forewing with [Cu.sub.1] and 0. separate [Cu.sub.2] l. stalked 8. Hindwing 0. without anal lobe 1. with anal lobe 9. Hindwing with [M.sub.3] and 0. separate [Cu.sub.1] 1. stalked or connate 10. Hindwing with [M.sub.3] and 0. connate with [Cu.sub.2] [Cu.sub.1] 1. removed from [Cu.sub.2] 11. Abdominal tergites 0. not curving over sternites (abdomen rounded) 1. curving over sternites (abdomen flattened) 12. Uncus 0. absent (Fig. 1) 1. reduced 2. present 13. Uncus 0. short and rounded 1. well developed and triangular 2. conical 14. Socii 0. absent 1. reduced 2. present (Fig. 1.) 15. Uncus and Socii 0. separate 1. fused 16. Gnathos 0. narrowly ovoid (Fig. 1) l. broadly triangular 2. not well developed or absent 3. bifid 17. Gnathos 0. narrowly joined to tegumen (Fig. 1) 1. broadly joined to tegumen 18. Gnathos 0. naked 1. spined (Fig. l) 19. Aedeagus 0. tapering to a point (Fig. 4, A-0, Q, R) l. blunt (Figs. 2, P) 20. Aedeagus 0. straight or nearly so (as in Leripalpus, Haseufussia, and Depressaria Figs. 2 E, F, H, J, L, N, P) 1. moderately to strongly c-shaped (Figs. 2 A-D, G, 1, K, M, O, Q, R) 21. Aedeagus flange 0. absent (Figs. 2 A, B, D-F, H- L, N-P) 1. single (Fig. 2 G) 2. double (Figs. 2 C, M, Q, R) 22. Vesica of aedeagus 0. bare (Figs. 2 B, C, G, M, O, Q, R) 1. with cornuti (Figs. 2 A, D-F, H-L, N, P) 23. Cornuti 0. scale-like and numerous located in a patch (as in Semioseopis and Bibarrambla) l. stout and finger-like and either single or a few in a row (as in Depressaria Figs. 2 A, D-F, H-L, N, P) 2. saw tooth-like and numerous located at base of vesica (as in Apachea) 24. If cornuti finger-like 0. limited to middle of vesica (Figs. 2 D-F, H, J, N) 1. extending near tip of vesica (Figs. 2 A, 1, K, L, P) 25. Sacculus 0. without lobe or process basally (Figs. 3, 4 G-I, K) 1. with a small indistinct distinct basal process (Figs. 3 A) 2. with a short or long basal process (Figs. 3, 4 B, C, E, F, J, L-R) 26. Basal process with scale- 0. absent (Figs. 3 A) like projections 1. present only at tip (Figs. 3 E) 2. covering entire process (Figs. 3, 4 B, C, F, G, J, L -N, P-R) 27. Scale-like projections of 0. fine (Figs. 3, 4 B, C, E, F, L, basal process N) 1. robust (Figs. 3, 4 G, M, O-R) 28. Basal process of sacculus 0. straight or nearly so or with inner margin slightly concave (Figs. 3, 4 C, G, L -Q) 1. highly curved with inner margin convex (Figs. 3 B, E, F, J, L, M, R) 30. Valva with medial lobe 0. absent (Figs. 3, 4 B, C, E-G, 1, J, M, O, Q, R) 1. clearly distinct from sacculus (Figs. 3, 4 A, K, P) 2. not clearly distinct from sacculus (Figs. 3, 4 D, H, L, N) 31. Distal process 0. absent (Figs. 3, 4 A, D, E, F, H, J, K, L, N) 1. linear to curved and sometimes bifid at tip (Figs. 3, 4 B, C, G, 1, M, O, Q, R) 2. quadrate (as in Nites and Exaeretia) 32. Distal process of valve 0. free (Figs. 3 G) 1. not free (Figs. 3, 4 B, C, 1, M, O, Q, R) See Clarke, 1941b and Hodges, 1974 for complete explanation of 'free" 33. Distal process of valve 0. straight or slightly curved (Figs. 3, 4 G, O) 1. elbowed (Figs. 3, 4 B, C, M, Q, R) 34. Apex of distal process 0. rounded or pointed (Figs. 3, 4 of valve B, C, G, I, M, Q, R) 1. bifid (Fig. 4 O) 35. Costal margin of valve 0. straight or nearly so (Figs. 3, 4 A-F, H-R) 1. noticeably concave (Figs. 3 G) 2. noticeably convex 36. Shape of valve 0. broadest at base and tapering at 1/2 to 1/3 length (Figs. 3, 4 G, O) 1. broadest at base to nearly 5/6 its length or nearly straight (Figs. 3, 4 A-F, H-N, P-R) 2. broadest at middle 37. Cucullus of valve 0. rounded (Figs. 3, 4 A, C-F, G, H, J, L -N, P-R) 1. pointed (Figs. 3, 4 B, 1, O) 38. Costal margin of valve 0. without process (Figs. 2 B, C, F-J, M, O-R) 1. with process located at least 3/4 middle of costa to tip of costa (Figs. 3, 4 A, D, K, L, N) 2. located basally or at least 3/4 middle of costa 39. Valve hairs 0. diffuse and covering most of valve (Figs. 3, 4 H, J) 1. primarily located at cucullus (Figs. 3, 4 A D G K P) 2. primarily located along costa, sacculus, and cucullus (Figs. 3, 4 B, C, E, F, 1, L-N, Q, R) 3. centrally located (Fig. 4 O) 40. Transtilla 0. absent 1. present (Fig. 1) 41. Transtilla 0. membranous (Fig. 1) 1. sclerotized (lightly or heavily) 42. Lateral lobes of transtilla 0. absent 1. weakly developed (diffuse) (Fig. 1) 2. well developed (sclerotized) 43. Vinculum 0. rounded (Fig. 5, C) 1. pointed (Fig. 5, A and B) 44. Vinculum anterodorsal 0. not well developed (membranous process or absent) (Fig. 1) 1. well developed (sclerotized) 45. Ostium bursae 0. present anteriorly 1. near middle 46. Ductus bursae meets corpus 0. at right angle bursae 1. straight (or slightly curved)
|Printer friendly Cite/link Email Feedback|
|Author:||Bucheli, Sibyl R.; Passoa, Steven; Wenzel, John W.|
|Date:||Jul 1, 2010|
|Previous Article:||Ross H. Crozier (1943-2009).|
|Next Article:||Description of the gyne of the ant thaumatomyrmex feroxmann, 1922 (hymenoptera: formicidae).|