Phylogeny of the semiaquatic bugs (Hemiptera-Heteroptera, Gerromorpha) JAKOB DAMGAARD Damgaard, J.: Phylogeny of the semiaquatic bugs (Hemiptera-Heteroptera, Gerromorpha). Insect Syst. Evol. 39: 431-460. Copenhagen, November 2008. ISSN1399-560X. The phylogeny of semi-aquatic bugs (Hemiptera-Heteroptera: Gerromorpha) was tested in parsimony analyses of 64 morphological characters and approximately 2.5 kb of DNA sequence data from the mitochondrial genes encoding COI+II and 16SrRNA and the nuclear gene encoding 28SrRNA. The taxon sample included representatives of all families and most subfamilies of Gerromorpha and a selection of outgroup taxa representing the two basal infra-orders of Heteroptera, Enicocephalomorpha and Dipsocoromorpha, and two families of Nepomorpha. A simultaneous analysis (SA) of all data, and with gaps scored as fifth state characters, gave a single most parsimonious tree with all families resolved as monophyletic, except the Veliidae, where Microveliinae + Haloveliinae, Veliinae, Rhagoveliinae, Perittopinae, and Ocelloveliinae were resolved as successive sister groups to the Gerridae, thus confirming earlier statements about paraphyly of this family. The Gerridae + Veliidae clade was strongly supported, but otherwise only the Gerridae + Veliidae less Ocelloveliinae and the Gerridae itself had support. These three clades could all be diagnosed on apomorphic morphological characters, although no characters diagnosing the Gerridae were without con-vergences or present in all included taxa. While the Ocelloveliinae, Veliinae and Haloveliinae could not be diagnosed on convincing apomorphies, the Microveliinae + Haloveliinae, and their sister group relationship with the Gerridae, could be diagnosed on rather strong morpho-logical synapomorphies, suggesting that Gerridae could be expanded to include these two veli-id subfamilies, while Ocelloveliinae, and perhaps the remaining veliid subfamilies, could be elevated to new families. In Gerridae, the Ptilomerinae + Halobatinae was sister group to all other subfamilies, while the Rhagadotarsinae + Trepobatinae was sister group to a clade com-prising the Gerrinae, Eotrechinae, Cylindrostethinae and Charmatometrinae. Most relation-ships in this clade were poorly supported and diagnosed, and Cylindrostethinae was surpris-ingly found to be paraphyletic. The sister group to the Gerridae + Veliidae clade was a strong-ly supported clade comprising the Paraphrynoveliidae and Macroveliidae, and this, and the lack of convincing synapomorphies for Paraphrynoveliidae, suggest that these two small fam-ilies could be synonymized. For the basal relationships of Gerromorpha, the Mesoveliidae was strongly supported sister group to all other families, while the Hebridae, Hermatobatidae and Hydrometridae formed a poorly supported and poorly diagnosed sister group to the Gerridae + Veliidae + Paraphrynoveliidae + Macroveliidae clade. The unexpected sister group relation-ship between Hermatobatidae and Hydrometridae was moderately supported, and could be diagnosed on two synapomorphies, thus giving a new hypothesis about the relationships of these very divergent families. Phylogenetic analyses of individual character partitions gave less resolved and less supported relationships, and the mitochondrial genes COI+II and 16SrRNA contributed negative hidden partitioned Bremer support (HPBS) to the simultane-ous analysis tree, probably due to homoplasy caused by saturation effects.