Studies On The Molecular Phylogeny Of Tetracneminae (Hymenoptera:Encyrtidae) And Taxonomy Of Chinese Species Of Ericydnus Haliday(Hymenoptera:Encyrtidae)

Tetracneminae is a subfamily of Encyrtidae, belonging to Chalcidoidea, Hymenoptera. To date, there are107genera and more than800species recognized. The species of Tetracneminae parasitize various insect pests injurious to agriculture and forestry, and frequently used in the biological control projects of the insect pests.First, nine Chinese species of Ericydnus are studied. E. elongatus sp. nov., E. huangi sp. nov., E. liaoi sp. nov., E. infuscatus sp. nov. and E. gigas sp. nov. are described as new to science. E. apterogenes Mayr, E. japonicus (Tachikawa) and E. ventralis (Dalman) are recorded from China for the first time. A key to species is given and62photomicrographs are provided to illustrate morphological characters of the species. Each of these taxa is accompanied with synonymy list, main references, all the known hosts and geographical distribution.Then, a molecular phylogenetic study of the Tetracneminae was carried out. Utilizing97terminal taxa in27recognized genera, through extraction of DNA, amplification of18S rDNA,28S rDNA and COI, followed by sequencing. Members of the closely related families Eupelmidae (Tineobius sp., Anastaus sp.), Tanaostigmatidae (Tanaostigmatus sp.) and Pteromalidae(Gastrancistrus sp., Epistenia sp., Coruna sp., Cleonymus sp. and Nasonia vitripennis) are included as outgroups. Our results are inferred from Mitochondrial Genome locus COI, and the Ribosomal DNA regions18S and28S. The single gene and combined data showed that18S rDNA<28S rDNA<28S rDNA+COI<18S rDNA+28S rDNA+COI<18S rDNA+28S rDNA for resolving the relationship of Tetracneminae. We used Clustal-X, MUSCLE, BlastAlign, T-Coffee, PRANK and manual (by eye) methods to align the matrix of18S rDNA and28S rDNA. Comparing RAxML trees inferred from the difference alignment methods, we find PRANK and manuals alignment may be the most appropriate methods for this subfamily. Using TNT, PAUP and MrBayes trees to compare the PRANK and manually aligned matrices, manual alignment generated better resolved relationships of Tetracneminae than PRANK. Parsimony analysis was performed on the complete dataset. Where gaps were treated as fifth states. phylogenetic performance was improved over than where gaps were treated as missing data. Bayesian analysis was also performed, with a partitioned evolutionary model resulting in good resolution of the relationships of the ingroup, with high posterior probabilities.The results are compared with the current classification as based on morphology. Although our conclusions are partially congruent with the accepted hypotheses of the Tetracneminae relationships, with some of the currently adopted assemblages were recovered as monophyletic. The newly proposed classification of Tetracneminae is as follows:Aenasiini, Ericydnini, Mirini, Tetracnemini (including thress subfamilies:Tetracnemina, Arhopoideina and Tricladiina), Pauridiini, Anagyrini and Dinocarsiini. Aeneasiini with a bolder hypothesis suggesting that Aenasiini are monophyletic, and maybe the primitive encyrtids. The Mirini tribe was recoved, which included Charitopus, Savzdargia, Eotopus, Mira, Paraclausenia and Clausenia. The results from the different methods of analysis agreed with the hypothesized split of the Tetracnemini into three subtribes, Arhopoideina, Tetracnemina and Tricladiina. Arhopoideina was notable as having a very large subtending branchlength. The Anagyrini and Dinocarsiini are resolved as sister taxa. To broadened the scope of Dinocarsiini, including Dicarnosis, Dinocarsiella, Dusmetia, Anagyrietta and Dinocarsis. Anagyrini included Anagyrus, Anomalicornia, Apoleptomastix, Anusia, Callipteroma, Leptomastidea, Leptomastix and Rhopus. Ten topological hypotheses were constrained during tree searching, to test the statistical likelihood of each. The hypotheses tested included taxonomic groupings suggested by the molecular results, and previous morphology based groupings. The results showed that the monophyly of Dinocarsiini and Anagryini were dependent on the composition. Monophyletic Mirini and Tetracnemini were also well supported. Based on the genera sampled in the current study, we found the structure of the male antenna was in general accordance with tribal positions in the molecular hypotheses, and could be useful for understanding evolution in the Tetracneminae.