Contributions to Zoology, 85 (1) – 2016Dick S.J. Groenenberg; Peter Subai; Edmund Gittenberger: Systematics of Ariantinae (Gastropoda, Pulmonata, Helicidae), a new approach to an old problem

To refer to this article use this url: http://ctoz.nl/vol85/nr01/a03

Results

In total 483 sequences were obtained: 161, 149, 91 and 82 for the markers H3, COI, CytB and 16S, respectively. A summary of the character statistics (as calculated with Paup 4.0b10 for Unix; Swofford, 2002) for each dataset is given in Table 3. Table 4 (Appendix) gives a summary of uncorrected p-distances.

FIG2

Table 3. Information content per dataset. Inf. char. = The number of parsimony informative characters. Perc. inf. = The percentage of informative characters, calculated as 100 × (Inform. char. / Total char.).

Both the phylogeny reconstructions for the individual markers (Figs S1-S4; online supplementary information), as well as those for the concatenated datasets (Figs 1, 2, S5, S6) distinguish most of the (sub)genera that were traditionally characterized by subtle conchological differences and geographic origin (e.g. Arianta, Cattania, Corneola, Chilostoma, Cingulifera, Dinarica, Faustina, Helicigona, Josephinella, Kosicia and Liburnica). Although sister-group relationships between some taxa were shown explicitly (e.g. Josephinella - Thiessea, Ariantopsis - Wladislawia), deeper nodes were hardly supported, particularly in the phylogenies based on the individual markers (Figs S1-S4). The H3 dataset differs most from the other datasets by its relatively low percentage of parsimony informative characters (Table 3). The phylogeny based on this marker is not discriminative below the genus level, but can be useful for the assignment of species (or subgenera) to genera (e.g. Ariantopsis pelia, Campylaeopsis moellendorffii, Superba spec., Wladislawia sztolcmani). The Histone gene cluster is multicopy (slight intercopy variation might exist and H3 pseudogenes have been reported; Rooney et al., 2002) and has been used in higher-level phylogenetics (Armbruster et al., 2005; Colgan et al., 2007 and references therein). For some species of mostly Chilostoma, double peaks were observed at a few positions within the H3 sequence. This genus which is shown as monophyletic in nearly all phylogeny reconstructions (Figs 1, 2, S2, S3, S5 and S6), turns out paraphyletic in the phylogeny for H3 (Fig. S1). It might be argued that the more extensive sampling of Chilostoma (Cingulifera) increased the chance of observing this apparent intercopy variation, but it was not observed in other genera for which multiple species were sequenced (e.g. Arianta, Cattania, Josephinella and Liburnica) either. In agreement with Colgan et al. (2000) we therefore conclude that intercopy variation in H3 will not significantly interfere with the phylogeny reconstructions.

FIG2

Fig. 1. MrBayes phylogeny based on the ‘stringent’ H3-COI-CytB dataset. Branch values show posterior probabilities.

FIG2

Fig. 2. BEAST phylogeny based on the ‘relaxed’ H3-COI-CytB-16S dataset. Branch values show posterior probabilities. Node values indicate divergence estimates in MYA.

Initially no amplicons were obtained with the COI mini-barcode primers of Meusnier et al. (2008); to get these working for Ariantinae, the reverse primer was modified (Table 2). Hajibabaei et al. (2006) showed (in silico) that COI mini-barcodes (109 bp; compared to the full length barcode of 654 bp) are 3% less effective in the correct identification of closely related species and pointed out that mini-barcodes might be less useful for the classification of specimens in larger species assemblages. Based on those taxa for which both a complete and a mini-barcode sequence were obtained (C. (Cattania) faueri, Corneola desmoulinsii, Helicigona lapicida andorrica and Vidovicia caerulans), we conclude that the mini-barcodes (despite their short length) are placed correctly in the COI phylogeny (Fig. S2).

The phylogeny reconstructions for the concatenated datasets (Figs 1, 2, S5-S6) show a basal split within the Ariantinae, differentiating the ancestor of the genera Causa, Isognomostoma and Helicigona, referred to here as group A, from that of all other genera (except Campylaea and Corneola), collectively referred to as group B. It remains unclear whether Campylaea and Corneola belong to A (Fig. 2) or B (Figs 1, S5), or should be considered a group on their own (Fig. S6).

All of the obtained sequences were deposited in GenBank; a complete list of taxa, sampling information and accession numbers is given in Table S7.