Systematics of the blackfly subgenus Trichodagmia Enderlein (Diptera: Simuliidae: Simulium) in the new world

Systematics of the blackfly subgenus Trichodagmia ENDERLEIN (Diptera: Simuliidae:Simulium) in the New WorldThe systematics of the New World subgenus Trichodagmia has been reassessed by employing anintegrated taxonomic approach based upon revisionary taxonomy, phylogenetic (cladistics) analysis, andDNA barcoding. This subgenus included several species of great medical importance, which are allmorphologically very similar. The history of the taxonomy and classification of the subgenusTrichodagmia has been put into context with other subgenera within New World Simuliidae, whiledescriptions and keys to the identification of species in this subgenus are also given.The subgenus Obuchovia is here considered a new junior synonym of Trichodagmia, and all itsconstituents’ species are now placed in the ALBELLUM species group to represent a Palaearcticelement within this subgenus. Three new junior synonymies are here proposed: Simulium chiriquienseFIELD is a synonym of S. ethelae DALMAT n. syn.; S. biuxinisa COSCARÓN & IBÁÑEZ-BERNAL is asynonym of S. paynei VARGAS n. syn.; and S. keenani FIELD is a synonym of S. earlei VARGAS, MARTÍNEZPALÁCIOS &DÍAZ NÁJERA n. syn. A neotype is designated for S. lahillei (PATERSON & SHANNON) and alectotype for S. pulverulentum KNAB.1 Simulium falculatum ENDERLEIN is transferred from theTARSATUM species group of to the CANADENSE species group based on the morphology of thefemale genitalia. Two species, S. rivasi RAMÍREZ PÉREZ and S. oviedoi RAMÍREZ PÉREZ, are transferredfrom the TARSATUM species group to the subgenus Psilopelmia based on the morphology of the malegonostyle and the ventral plate. Keys to separate all species groups and species based on the adults,pupae and larvae are also provided.The phylogeny and classification of the subgenus Trichodagmia is delineated using a cladisticanalysis of 63 taxa based on males, females, pupae and larvae, including two species belonging to thesubgenus Aspathia and two species of the subgenus Simulium s.str. that served as outgroups. Analysis ofthe original full data set [67 taxa and 67 characters] with multistate characters treated as unorderedunder equal weights led to poorly resolved trees, with many polytomies within TARSATUM [= oldsubgenus Hemicnetha] and CANADENSE [= old subgenus Hearlea]. Nonetheless, the ALBELLUM [=old subgenus Obuchovia] and PICTIPES [= old subgenus Shewellomyia] species groups, and some cladeswithin the CANADENSE species group were well supported. In the most parsimonious cladograms,the position of S. falculatum was problematic as it was placed basal to Trichodagmia. The position of S.jeteri, albeit within the ORBITALE [= old subgenera Trichodagmia + Thyrsopelma of MIRANDA-ESQUIVEL& COSCARÓN, 2001] clade, was also poorly resolved. This was certainly due to the numerous missingdata in these two taxa. Therefore, they were removed from the data set together with other taxa inwhich three life stages (> 70% of characters) were missing (e.g. S. paracarolinae and S. tarsale). A secondanalysis was then performed with 63 taxa and 67 characters. In this analysis, the Strict Consensus Treewas better resolved and certain clades within the expanded concept of Trichodagmia (sensu SHELLEY et al.,2010) were recovered as monophyletic with high support values. The ALBELLUM species group ismonophyletic in a sister-group relationship with the other species groups in Trichodagmia (sensu SHELLEYet al., 2010). The ORBITALE species group clade was recovered as monophyletic by a uniquecombination of seven characters with 89% bootstrap support. In this clade, all species close to S.guianense s.l. were better diagnosed by a combination of four characters, one of which (male ventral platewith a globular median process) was unique to this group. The position of S. hirtipupa is better resolvedin the latter clade by the presence of black spiniform setae in the frontoclypeus and thorax of the pupa.In contrast, the TARSATUM and CANADENSE species groups were diagnosed by only fourand five characters, respectively. Within the CANADENSE species group only species with larvaehaving sclerotized plates in the posterior region of the abdomen were well resolved. Species in theTARSATUM group were homoplastic. The PICTIPES group is only diagnosed by homoplasies, but thecombination of these characters is unique to this clade (polythetic taxon). In general, this study supportssome of the taxonomic changes proposed in SHELLEY et al. (2010), in which the subgeneric-namesHearlea, Hemicnetha, Shewellomyia, Trichodagmia + Thyrsopelma (sensu MIRANDA-ESQUIVEL & COSCARÓN,5022001) are treated as species groups within the subgenus Trichodagmia. Moreover, this study also supportsthe proposal of Obuchovia as a junior synonym within the clade Trichodagmia to represent theALBELLUM species group.The utility of the COI DNA barcoding methodology for identification of species in the subgenusTrichodagmia and related taxa has been tested. In total, 24 morphospecies within the current expandedmorphological concept of Trichodagmia were analyzed. In addition, three species of the subgenusAspathia and 10 species of the subgenus Simulium s.str. were also included in the analysis because of theirputative phylogenetic relationship with Trichodagmia. Within the barcoding neighbour-joining tree, mostof the specimens were grouped together according to morpho-taxon (species groups and species).Mean genetic distance amongst groups (morphospecies) averaged 11.2% (ranged 2.8-19.5%), whereasintraspecific genetic divergence within morphologically distinct species averaged 0.5% (range 0-1.3%).In known species complexes, maximum values of genetic divergence (3.28-3.79%) indicate the probablepresence of cryptic diversity. DNA barcoding achieved nearly 100% success in identifying all specimensof the subgenus Trichodagmia and related taxa.The existence of well defined groups within S. piperi, S. duodenicornium, S. canadense and S. rostratumhighlighted the possible presence of species complexes in these taxa. In addition, the suspectedpresence of a sibling species in S. paynei and S. tarsatum among populations of Belize, Costa Rica, andthe USA is confirmed. The use of shorter barcodes (midi and minibarcodes) from specimens held incollections was problematic with regards to the DNA quality and PCR success. However, in the casesthat a readable sequence was obtained, they were sufficient for reliable species identification. Withregards to the different extraction and preservation techniques tested, larvae preserved in dilutedCarnoy’s (10% acetic acid) provided full DNA barcodes. Furthermore, legs added directly to the PCRmix from freshly collected individuals provided full length barcodes sequences. However, specimens ofmore than 10 years old did not yield good PCR products. In short, I conclude that DNA barcoding incombination with a morphological benchwork platform is an effective approach for identification anddelineation of species in the subgenus Trichodagmia, and the discovery of hidden diversity in this taxon.