Cryptodacus Main | Tephritidae Main | Diptera Home | SEL Home

Phylogeny of Cryptodacus, Haywardina and Rhagoletotrypeta (Diptera: Tephritidae)

The following text is copied from Norrbom (1994).

Phylogenetic analyses were conducted to test if Cryptodacus, Haywardina, Rhagoletotrypeta, Lezca and Cryptoplagia are monophyletic groups. The latter two names are no longer considered valid, based on the results of the analyses (see below).

The analyses were hindered by the fact that relationships among the genera of Carpomyina are poorly resolved. Because analysis of all carpomyine genera was not within the scope of this study, a working hypothesis was used, that the species with a white medial scutal stripe (i.e., Cryptodacus, Haywardina, Rhagoletotrypeta and Zonosemata) are a monophyletic group.

Relationships were analyzed using the mhennig* and bb* options of Hennig86 (Farris 1988). Table 1 lists the 50 characters used, and Table 2 shows the distributions of their states. Intraspecific variation is shown in Table 2, but only one state at a time was assigned in matrices used in Hennig86. Coding for the variable species is discussed in Table 1. Relationships within Zonosemata were not analyzed, but for the characters analyzed, the full range of states occurring within Zonosemata is represented by the three species included in the matrix. As a convention, in the text and trees, a character number (indicated by "#") refers to state 1 unless followed by a decimal point and an alternative state.

Because relationships among carpomyine genera are uncertain, several different taxa were used as the outgroup in separate analyses. These included: the genus Paraterellia, which is possibly related to the Carpomyina and may indicate the carpomyine groundplan for the characters analyzed; and two Solanaceae-breeding species of Rhagoletis, ferruginea Hendel and striatella Wulp, whose hosts might suggest that they are closely related to Haywardina and Zonosemata.

Depending upon the outgroup used, the resulting cladograms vary greatly in how the genera are grouped. The two trees shown (Fig. 11-12) are included mainly to indicate relationships within each genus, rather than among them. With Paraterellia as outgroup, Cryptodacus is hypothesized as the sister group of Haywardina + Rhagoletotrypeta, with Zonosemata as the sister group of these three genera (Fig. 11). With the species of Rhagoletis as outgroup, almost all possible sister group relationships among the four genera are equally parsimonious. In some trees the xanthogastra and annulata groups of Rhagoletotrypeta are not sister taxa (i.e., Rhagoletotrypeta not monophyletic), and/or Haywardina is not hypothesized as monophyletic. Furthermore, when either or both of species of Rhagoletis is included in the matrix with Paraterellia as outgroup, they are grouped with only some taxa (striatella with Cryptodacus and the annulata group; ferruginea variously), thus the four genera with the medial white scutal stripe might not be a monophyletic group. Considering that the basal nodes of all the trees are defined by only a few characters, and that those characters are highly variable in the Tephritidae (e.g., thorax and abdomen color (#15, #31), dorsocentral seta location (#11), and antennal flagellomere shape (#8)), I have little confidence in any of the trees below the genus level.

Although the relationships among Cryptodacus, Haywardina, Rhagoletotrypeta, and Zonosemata are not resolved by this analysis, Cryptodacus, the R. xanthogastra group, the R. annulata group, and Zonosemata are hypothesized as monophyletic groups in all resulting trees. Each of these groups is based on apomorphies that are unique or nearly unique to it, so no matter what outgroup is used they probably are monophyletic.

Synapomorphies of Cryptodacus include: wing vein Cu₁ bordered by brown band between bm-cu and the discal band (#25), although it is lost in quirozi and variable in tau; and R-M displaced apically (#24), which occurs as homoplasy in H. cuculi. The subapical band faint anteriorly (#29), which occurs as homoplasy in Z. scutellata and H. cuculiformis, and loss of the white apical band on at least tergite 4 (#32), which occurs as homoplasy in R. morgantei and is difficult to code in the species with yellow abdomens, are also interpreted as synapomorphies on all trees, with reversal in #29 in silvai + tigreroi, and in #32 in ornatus. Another possible synapomorphy of all Cryptodacus species is the slightly sagittate and finely serrate aculeus tip (#45), but unfortunately, no females of obliquus are known, so whether this is a synapomorphy for the entire genus or only the seven other species is uncertain. A similar but convergent shape occurs in four species of the Rhagoletis nova group (Foote 1981), in which the tip is sagittate, but with much larger serrations (Frias 1986).

Within Cryptodacus, the hypothesis that obliquus is the sister taxon of the other species is well supported. The positions of the other species are consistent in all trees, except for ornatus and tau. C. ornatus is variously hypothesized as the sister group of silvai + tigreroi, of parkeri + quirozi, or of all four species. C. tau is either the sister group of those five species (Fig. 11), which are always grouped together, or the sister group of lopezi (Fig. 12).

I consider Lezca a subjective synonym of Cryptodacus. Continued recognition of Lezca as a monotypic taxon for tau, would make Cryptodacus paraphyletic. C. silvai and the five new species instead could be placed in Lezca, leaving only obliquus in Cryptodacus, but since the close relationship of all of these species is well supported, I see no reason to divide them into separate genera.

Haywardina is not well defined by synapomorphies as are the other three genera. Support is weak for the inclusion of obscura (Fig. 11), but the female of this species is unknown and it otherwise has mostly plesiomorphic character states, so there are few apomorphies grouping it with any other taxon. Discovery of the female and host plant may clarify its relationships. The shape of the lateral scutal white stripe (#17.2) suggests that obscura is related to bimaculata and cuculi, but if so, there is reversal in this character because the stripe is broader and more extensive anteriorly in cuculiformis. All four species have a similar abdominal color pattern (#31.0), but this is plesiomorphic with Paraterellia as outgroup. It is apomorphic with either Rhagoletis species as outgroup, but occurs as homoplasy in the R. xanthogastra group. With Paraterellia as outgroup, the shape of the antennal flagellomere (#8.0) appears to be a synapomorphy for Haywardina, but given that this state is present in other carpomyine genera except Cryptodacus, Rhagoletotrypeta, and Goniglossum, it probably is not a true synapomorphy at this level. With Rhagoletis ferruginea as outgroup, in some trees obscura is hypothesized as the sister taxon of a clade including Cryptodacus and the xanthogastra and annulata groups, supported by the mostly brown thorax (#15); in other trees it is placed as the sister group of the xanthogastra group, based on #15, #31 and #23.0 (scutellum white area small), or within the xanthogastra group (but this assumes states 44.1, 47.1, and 49.0 for obscura). With all three outgroup taxa included in the matrix, obscura is placed as the sister taxon of the xanthogastra group (Fig. 12), but this is supported only by #13 and #23.0, in which there is reversal within the clade.

The close relationship of bimaculata, cuculiformis, and cuculi (Fig. 11) is suggested by their scutal markings. These species have isolated brown marks bordering the lateral white stripe (#19). Some species of Zonosemata may have brown markings in this area, but if so they are usually connected posteriorly to form a broad U. The brown lateral marks on the scutellum (#23.2) may be interpreted as a synapomorphy for bimaculata + cuculiformis + cuculi, with reversal to state 1 in cuculi (Fig. 11), or as homoplasy in bimaculata and cuculiformis. The unique pattern of brown spots on the thoracic pleuron (#20), the loss of the accessory costal band (#30), which occurs as homoplasy in some species of Cryptodacus and Rhagoletotrypeta, and the slender apical lobe on the glans (#42) are synapomorhies indicating the close relationship of cuculi and cuculiformis.

The aculeus tip is notched in Paraterellia, Zonosemata, and H. bimaculata (#46.0), although the different shapes of the tip in these taxa may indicate that this is convergence. With all three outgroup taxa included in the matrix, in some trees (Fig. 12) H. bimaculata is hypothesized as the sister group of all the other taxa except Paraterellia and Zonosemata, for which #46 is the only synapomorphy. Trees with bimaculata as the sister group of cuculi + cuculiformis are equally parsimonious, however, and the former hypothesis of relationship is not supported with any of the other taxa as outgroup. With either species of Rhagoletis as outgroup, #46.0 could be interpreted as a synapomorphy for bimaculata and Zonosemata, but other character distributions do not support that hypothesis. If obscura has the same type of aculeus tip, this state could be the groundplan condition for a clade including Haywardina and Zonosemata (with reversal in cuculi + cuculiformis). H. cuculiformis also shares an apomorphic state with Zonosemata, the absence of scutal microtrichia (#12.0), which is also interpreted here as homoplasy because of other character state distributions. Either species possibly could be a plesiomorphic member of Zonosemata, but, based on the characters analyzed, no trees result with either of them grouped with Zonosemata.

I consider Cryptoplagia a subjective synonym of Haywardina. By the generic limits proposed by Aczél, bimaculata (and perhaps obscura) would be included in Cryptoplagia, but the latter would be paraphyletic, because cuculiformis, the type species of Cryptoplagia, is more closely related to cuculi, the type species of Haywardina, despite the distinctive autapomorphies of the latter species.

Rhagoletotrypeta includes two well defined monophyletic species groups, whose relationship to each other (Fig. 11) is suggested mainly by the short medial surstylus (#39), a character that also occurs in Zonosemata. The fact that species of both species groups breed in Celtis species, a host taxon not known to be attacked by any other Carpomyina, also suggests their relationship. With either Rhagoletis species as outgroup, in some trees the xanthogastra group forms a clade with Zonosemata and Haywardina, or it is hypothesized as the sister group of the annulata group + Cryptodacus. Also, with all three outgroup taxa included in the matrix, the xanthogastra and annulata groups are not hypothesized as sister taxa (Fig. 12).

The xanthogastra group includes parallela, pastranai, and xanthogastra. Its monophyly is indicated by at least three synapomorphies: aculeus with lateral barbs (#44); aculeus tip deeply trilobed (#47); and 3 spermathecae present (reversal to #49.0). The shape of the apex of the lateral surstylus (#36), which is unknown in parallela, and the abdominal color pattern (#31), which varies in pastranai, may be additional ones. The relationships among these three species are not well resolved. In most trees, either parallela is the sister taxon of pastranai + xanthogastra (Fig. 11), for which #23.0 and perhaps #36 are synapomorphies, or xanthogastra is the sister group of parallela + pastranai (Fig. 12), if parallela is assumed to have state #36.1. With either Rhagoletis species as outgroup, if pastranai is coded state 2 for #31, the same trees result as when it is coded state 0, but there are additional trees in which pastranai is the sister taxon of the other two species.

The annulata group of Rhagoletotrypeta includes annulata, argentinensis, intermedia, morgantei, rohweri, and uniformis. Synapomorphies for this group include: male sternite 6 with a medial lobe (#35); lateral surstylus with mesal lobe not sharply differentiated (#37); and aculeus tip length more than half the length of the aculeus (#43). There is homoplasy in #37 in C. tau. The absence of the lateral scutal stripe (#17.0), also absent in C. obliquus and Rhagoletis, may be another synapomorphy. Within the annulata group, intermedia is the sister group of morgantei + annulata + argentinensis. The shape of the aculeus tip, flattened in the sagittal plane and trilobed in lateral view (#48), may be a synapomorphy of annulata and argentinensis (Fig. 11), but this character is unknown in morgantei, so it could be a synapomorphy of all three species (Fig. 12). If annulata is coded state 0 for #10, the shortest trees are one step longer, and additional trees result with annulata and morgantei reversed in their positions in Fig. 11. In the latter trees, state 1 is assumed for #48 in morgantei, and #10 is a synapomorphy for it and argentinensis. The relationships of uniformis and rohweri are not well resolved. Only in trees in which Cryptodacus and the annulata group are hypothesized as sister taxa (Fig. 12), state #10.0 can be interpreted as a synapomorphy for these two species.

Zonosemata is a well defined monophyletic taxon of seven species that breed in Solanaceae (Bush 1965, Norrbom 1989b, Hernández 1989). Autapomorphies for it include: 4 or more frontal setae (#2), which occurs as homoplasy in some specimens of Rhagoletis striatella; scutum without microtrichia (#12.0), which occurs as homoplasy in H. cuculiformis and R. xanthogastra; lateral surstylus short and without medial mesal lobe (#38.2) (its apex is rugose and may be homologous with the mesal lobe); glans relatively simple, evenly sclerotized, with little distinct internal sclerotization (#40); and spermathecae large, elongate, and weakly sclerotized (#50). Other possible synapomorphies (i.e. states of characters whose polarity depends upon the outgroup) are the posterior location of the dorsocentral seta (#11) and the notched aculeus tip (#46.0). There is homoplasy in some Cryptodacus and some Rhagoletis species in the former character, and because the aculei of Paraterellia and H. bimaculata have similar notches, the latter character state could have evolved at a lower level (see discussion of Haywardina). Bush (1966) also suggested that there are apomorphic traits in chromosome shape for Zonosemata, but karyotypes have not been studied in other carpomyine genera except Rhagoletis.

Berlocher and Bush (1982) indicated that R. striatella and perhaps other solanaceous breeding species of Rhagoletis may be more closely related to Zonosemata than to their congeners. Their hypothesis does not necessarily conflict with hypotheses of Zonosemata's relationships discussed here, because Berlocher and Bush were unable to include species of Cryptodacus, Rhagoletotrypeta, or Haywardina in their electrophoretic study.

See the Fruit Fly Literature Database for full information for cited references.