Phylogenies are a cornerstone to the study of evolution: without them we are unable to reconstruct and understand pattern and process of evolutionary change. To obtain a robust phylogenetic hypothesis requires the gathering of multiple independent and complementary data sets, because usually no single data set is sufficiently powerful to simultaneously resolve older and more recent cladogenic events.
The genus Triturus (European newts) has been the subject of extensive phylogenetic analysis with various data sets [osteological (Bolkay, 1928; Rafinski and Pecio, 1989), immunological (Busack et al., 1988), morphological (Giacoma and Balletto, 1988), biochemical (Rafinski and Arntzen, 1987), behavioral (Arntzen and Sparreboom, 1989), cytogenetic (Macgregor et al., 1990), and is considered by some to be the phylogenetically best studied genus in the world (Halliday and Arano, 1991). Despite these concerted efforts, its phylogeny is not fully resolved and several competing hypotheses are available, none of them with unambiguous overall support. We take the study by Arntzen and Sparreboom (1989) as the basis of our work because the phylogeny they present is based on two independent and complementary data sets. Moreover, their hypothesis resolves the earlier as well as the later events in the Triturus radiation and is robust under the jack-knife test.
Nine or twelve Triturus species are currently recognized, depending on the criteria used for species recognition. Here we refer to twelve species, with four of them (Triturus cristatus, T. carnifex, T. karelinii and T. dobrogicus) grouped together in the T. cristatus superspecies (Wallis and Arntzen, 1989). The ‘big’- and ‘medium-sized’ newts (T. cristatus, T. marmoratus, T. vittatus and T. alpestris) are organized in the subgenus Triturus and the ‘small-bodied’ newts (T. boscai, T. helveticus, T. italicus, T. montandoni and T. vulgaris) are placed in the subgenus Palaeotriton. The detailed configuration is given in Fig. 1, alongside with the nomenclature that we adopt. For the following sections of the phylogenetic tree the support is limited or contradictory: 1) the sister taxon status of T. vulgaris and T. montandoni is supported by allozyme data only; 2) the monophyly of T. boscai and T. italicus hinges on the interpretation given to behavioral characters; 3) the status of T. vittatus as the sister taxon of the T. marmoratus species group may be called into question on the basis of its overall similarity to some small bodied species, T. vulgaris in particular; 4) the support for the position of T. alpestris in the subgenus Triturus is relatively weak due to the non-independence of some behavioral synapomorphies; 5) the monophyly of Triturus, traditionally taken for granted, has been put in doubt by molecular data (Titus and Larson, 1995).
Fig. 1. Phylogeny for the genus Triturus based on available biochemical and behavioral characters. Numbers refer to specific questions as addressed in the text. Outgroup taxa are taken from the genera Cynops, Neurergus, and Paramesotriton.
We tested the phylogenetic hypothesis of Triturus with a newly generated set of independent data. Two fragments of the mitochondrial (mt)DNA molecule experiencing different evolutionary rates were studied. The slowly evolving 12S rRNA gene (Kocher et al., 1989; Hickson et al., 1996) was partially sequenced to test some of the supposedly earlier events of the Triturus radiation, while the fast evolving ATPase gene (Kumar, 1996) was partially sequenced to test the branching order among some supposedly closely related species in the T. vulgaris species group. Sequence data for cytochrome b produced ambiguous phylogenetic results, perhaps due to the comparison of non-homologous sequences (Caccone et al., 1997; cf. Zhang & Hewit, 1996) and were discarded.