A karyosystematic analysis of some water beetles related to Deronectes Sharp (Coleoptera, Dytiscidae)

An account is given of the karyotypes of five species and one additional subspecies of Deronectes Sharp, 1882, three species of Stictotarsus Zimmermann, 1919, one species of Trichonectes Guignot, 1941, four species of Scarodytes Gozis, 1914 and 17 species of Nebrioporus Régimbart, 1906. Deronectes species are characterised by a neo-Xy system of sex chromosomes and autosome numbers ranging from 60 (Deronectes ferrugineus Fery et Brancucci, 1987 and Deronectes wewalkai Fery et Fresneda, 1988) through 48 (Deronectes latus (Stephens, 1829), Deronectes angusi Fery et Brancucci, 1990) to 28 (Deronectes costipennis Brancucci, 1983, Deronectes costipennis gignouxi Fery et Brancucci, 1989 and Deronectes platynotus (Germar, 1834)). The three species of Stictotarsus, Stictotarsus duodecimpustulatus (Fabricius, 1792), Stictotarsus procerus (Aubé, 1838) and Stictotarsus bertrandi (Legros, 1956), all belonging to the Stictotarsus duodecimpustulatus group of species, have karyotypes comprising 54 autosomes and neo-Xy sex chromosomes. Trichonectes otini Guignot, 1941 has 48 autosomes and an X0 system of sex chromosomes, an arrangement shared with the 17 species of Nebrioporus Régimbart. The four Scarodytes species, Scarodytes halensis (Fabricius, 1787), Scarodytes nigriventris (Zimmermann, 1919), Scarodytes fuscitarsis (Aubé, 1838) and Scarodytes malickyi Wewalka, 1997, all have 54 autosomes and X0 sex chromosomes. The karyotypes of the various species are found to be distinctive and to support separation of these species from one another. In two cases (Nebrioporus martinii (Fairmaire, 1858) and Nebrioporus sardus (Gemminger et Harold, 1868), and Scarodytes halensis and Scarodytes fuscitarsis) the karyotypes require the recognition of the taxa as full species, not subspecies. The implications of these data for the generic classification are considered. The data are found to be compatible with the DNA-based phylogeny proposed by Ribera (2003), where the enlarged Stictotarsus proposed by Nilsson and Angus (1992) is found to be unsatisfactory.