The evolution of fynbos-endemic Cephalelini leafhoppers specialising on Restionaceae

[摘要] ENGLISH ABSTRACT: Knowledge of the diversity and evolutionary histories of insects in South Africa's fynbos biome lags far behind what is known of the plant groups that make up this global biodiversity hotspot. To address this imbalance, I undertook a molecular phylogenetic study of fynbos-endemic Cephalelini leafhoppers that specialise on restios in the family Restionaceae. My phylogenetic results did not recover the South African Cephalelini as monophyletic, nor did I find monophyly of described species, but several monophyletic clades of species were found within Cephalelini. Furthermore, phylogenetic dating suggested that the divergence between South African and Australian Cephalelini post-dates Gondwanan vicariance, implying intercontinental extreme long distance dispersal of these insects ca. 5-11 MYA. Diversification within the Cephalelini is also much more recent than that of the Restionaceae hosts on which they specialise, negating the possibility of coevolution between plants and insects. Rather, analysis of phylogenetic conservatism of host use reveals that Cephalelini evolution has tracked the evolution of their Restionaceae hosts and that closely related insects feed on the same plant host tribes. A finer scale of tracking of host evolution (such as at the clade or genus level) is expected when taking into account how highly specific I find Cephalelini host use to be, but its absence might be explained by the recent divergence of Cephalelini relative to the age of Restionaceae. Analysis of conservatism of host use was also carried out using a phylogeny of the Restionaceae, and revealed that, overall, Cephalelini host use and avoidance have no phylogenetic bias, indicating many empty potential niches for Cephalelini, or alternately that host use is governed by factors which are phylogenetically unconstrained. Lastly, I also analysed the evolution of specialisation of Cephalelini and find no trend towards increased specialisation within the group, which is contrary to what is expected of the evolution of herbivorous insects. Overall, this study presents the first evidence of intercontinental dispersal of insect fauna between South Africa and Australia and as such highlights an unconsidered factor in the accumulation of faunal diversity in the fynbos biome. I find Cephalelini to be highly specialised in their host preference, but this pattern only becomes apparent at the tribal host level in the evolution of Cephalelini. Although Cephalelini are highly specialised, I find no evidence of evolution towards increasing specialisation within the group.

[发布机构] Stellenbosch University