Phylogeny, character evolution and tribal status
Our study indicates that mitogenome sequences are useful for resolving phylogenetic relationships of subfamily Typhlocybinae from the species to the tribal level. Analyses of both amino acid and nucleotide sequences yielded similar topologies with the main differences among analyses occurring among a few branches that received low bootstrap support in one or more analysis. Our results agree with the result of a recent analysis by Lu et al. (2021) that incorporated partial sequence data from three nuclear and two mitochondrial genes in recovering Typhlocybinae as a well-supported monophyletic group sister to Mileewinae (Mileewini). A comprehensive analysis of Membracoidea based on anchored hybrid data (Dietrich et al., 2017) failed to recover this relationship but the relationship of Typhlocybinae to other leafhopper subfamilies was poorly resolved by that study, so our results require further confirmation. Within Typhlocybinae, our results consistently supported the monophyly of the tribes Dikraneurini, Empoascini, Erythroneurini and Typhlocybini (Alebrini was represented by a single species in our dataset).
Lu et al. (2021) recovered the three species of Zyginellini included in their dataset as a monophyletic sister group of Typhlocybini (sensu stricto) but our larger sample of Zyginellini genera indicates that this tribe is polyphyletic. This supports the treatment of some recent authors (e.g., Hamilton, 1983; Dietrich, 2013) of Zyginellini as a junior synonym of Typhlocybini. Based on our results, the hind wing character traditionally used to separate Typhlocybini from Zyginellini (absence of the distal segment of vein CuA; Figure 6, f6) is homoplasious and has been derived independently in at least three unrelated lineages. Another hind wing character (veins RP and MA separate versus confluent; Figure 4, character g) was previously used to distinguish Eupterygini from Typhlocybini but our results also indicate considerable homoplasy for this trait.
Details of relationships within individual tribes are difficult to compare between our results and those of Lu et al. (2021) given that the taxon samples of the two studies overlap only partially. Nevertheless, we note that several branches pertaining to relationships among genera of Typhlocybini, as well as a few branches within the other tribes, were extremely short and received low bootstrap support in both studies. This suggests that neither complete mitogenome data nor data from a few nuclear and mitochondrial genes will suffice to completely resolve typhlocybine phylogeny with high confidence, although increased taxon sampling may also help improve phylogenetic resolution in this group.
Our results are generally consistent with morphology-based classifications proposed recently. Young (1965) moved Mileewini (as Mileewanini) to Typhlocybinae but this was rejected by Mahmood (1967). Young (1968) later suggested treating Mileewinae as a separate subfamily. Nevertheless, Mileewinae are similar to Typhlocybinae in having relatively small, slender bodies, reduced forewing venation and slender hind tarsi (Dietrich, 2011). Previous phylogenetic studies have not reached a clear consensus on the sister-group relationship of Typhlocybinae (Balme, 2007; Dietrich, 2013b; Wang et al., 2017; Dietrich et al., 2017). Our results are also equivocal in this regard. Two analyses (Figure 2, c, e) recovered a sister-group relationship between Typhlocybinae and Cicadellinae (sensu stricto), but all other analyses (Figure 2, a, b, d) and the tree topology tests (Table 2) consistently support Typhlocybinae as sister to Mileewinae. Considering the morphological similarities shared by Typhlocybinae and Mileewinae, the latter hypothesis seems to be more plausible.
The tribal relationships in Typhlocybinae have long been controversial (Wagner, 1951; Mahmood & Ahmed, 1968; Zhang, 1990; Balme, 2007; Dietrich et al., 2017). Our analyses consistently support a sister-group relationship between Alebrini and Empoascini, as well as between Erythroneurini and Dikraneurini, in agreement with other recent analyses (Dietrich et al., 2017; Lu et al., 2021). The former relationship is also consistent with Wagner’s intuitive morphology–based hypothesis (1951), but not with those of Mahmood & Ahmed (1968) and Zhang (1990).
The main area of uncertainty appears to be the relationship of Typhlocybini to the other tribes and the relationships of deep internal nodes within Typhlocybini. Different analyses recovered Typhlocybini either as sister to Alebrini + Empoascini (Figure 2, a, b) or as sister to Erythroneurini + Dikraneurini (Figure 2, c, d, e), thus this relationship remains equivocal. Detailed morphology-based cladistics analyses of Typhlocybinae have not been attempted but there appears to be some morphological support for the latter hypothesis (Typhlocybini+(Dikraneurini+Erythroneurini)). For example, all Erythroneurini and most Typhlocybini (except a few Neotropical taxa) have the hind wing submarginal vein absent at the wing apex. The male subgenital plates of Typhlocybini, Dikraneurini and Erythroneurini also have relatively few macrosetae compared to those of Alebrini and Empoascini. These traits are potential synapomorphies supporting the sister relationship of Typhlocybini to Dikraneurini+Erythroneurini recovered in most of our analysis as well as the analyses of Lu et al. (2021) and Dietrich et al. (2017).
Ancestral state reconstructions of key morphological characters previously used to define and diagnose tribes within Typhlocybinae (Evans, 1963, 1971; Hamilton, 1983; Dworakowska, 1979, 1993; Zhang, 1990) (Table 3) indicate that some wing characters traditionally used to diagnose tribes are highly stable. Our analyses generally support the monophyly of tribes Empoascini, Dikraneurini, Erythroneurini and Typhlocybini as defined morphologically by most authors. Thus, our analyses suggest that the wing vein characters traditionally used to define these groups are reliable. Nevertheless, our results also indicate that some morphological characters have undergone homoplastic changes during the evolution of the group. Ocelli may be present or absent within Empoascini and Typhlocybini. In Empoascini, all genera have ocelli, except Beamerana and Paulomanus . The latter genera, which have hind wing venation identical to that of many Empoascini, nevertheless grouped with Typhlocybini in a recent morphology-based phylogeny and were tentatively placed in the latter tribe (Xu et al., 2021). Few genera in Typhlocybini have ocelli, e.g., Hiratettixand Caknesia , but we were able to include only the former in our analyses and it occupies a relatively derived position within the tribe, suggesting that ocelli were lost and regained at least once in this tribe (Figure 4, a). Loss of the hind wing submarginal vein is also apparently homoplastic with independent losses apparently occurring in Erythroneurini and Typhlocybini (Figure 4, e). Hind wing veins RP and MA have also apparently become confluent in the common ancestor of Erythroneurini and Dikraneurini and independently in Empoascini and Typhlocybini (Table 3). Loss of the branched hind wing anal vein occurs only in the Dikraneurini + Erythroneurini lineage but further analyses with a larger sample of taxa will be needed to determine the extent of homoplasy in this character.
The molecular divergence time estimates indicate that Typhlocybinae and extant tribes originated during the Middle and Late Cretaceous, respectively. Thus, diversification of major lineages in this group seems to have roughly coincided with the diversification of angiosperms during the Cretaceous (Hamilton, 1992, 1994; Ledyard, 1974; Foster et al., 2016). Our divergence time estimates for tribes of Typhlocybinae are considerably younger than those reported for the same branches by Dietrich et al. (2017), possibly due to the denser taxon sample of our study. Thus, in our study, the splits between Erythroneurini and Dikraneurini, and between Alebrini and Empoascini are both estimated at 76 MYA, compared to the 95 MYA and 112 MYA, respectively, reported by Dietrich et al. (2017). 95% confidence intervals of both studies are wide and broadly overlap, indicating that the available methods and data are only able to provide very approximate estimates of the times of origin of these major leafhopper lineages. Future analyses incorporating additional fossil evidence (unavailable at present), diversified clock models, dating methods, and rate priors may yield improved estimates.