Left aside some exceptions, Ariantinae shells are rather monomorphic (Fig. 3, Appendix). Because of the limited number of conchological characters, many authors studied the genital tract for morphological characters that could discriminate species and especially higher taxa. However, the genital morphology within this subfamily is also very homogeneous, what is uncommon among pulmonates. The form of the accessory glands, which are either undivided or more or less completely split (Fig. 4, Appendix), has been used by some authors as a (partial) basis for the systematics of the Ariantinae (Sturany and Wagner, 1914; Schileyko, 2006), although according to other authors both types of accessory glands can occur within the same genus, or even species (Hesse, 1931; Knipper, 1939; Schileyko, 2013). A classification of the species of Ariantinae in only two genera, as for example Helicigona and Campylaea (sensu Sturany and Wagner, 1914; Table 1), or Chilostoma and Campylaea (sensu Zilch, 1960; Table 1) is an oversimplification according to all modern authors, but what classification should be accepted alternatively remains a matter of dispute. Recently it has been suggested that the structure of the penial papilla might be a useful character to clarify the phylogenetic relationships between the (sub)genera within the subfamily (Schileyko, 2013), but that view still has to be confirmed.
Obviously, given the actual situation, a new approach is necessary, as was realized most recently by Groenenberg et al. (2012) and Cadahia et al. (2013), who tried to escape from the confusion by the use of molecular phylogenetics. Despite its shortcomings in the completeness of the molecular data, this article expands the reliability of the molecular phylogeny reconstructions, enabling a still better founded discussion regarding the subdivisions of the Ariantinae.
Our analyses do not support an evolutionary significance of the transformation series based by Schileyko (2013) on the structure of the penial papilla in several Ariantinae genera. Dinarica and Cattania are not closely related to Helicigona, for example, so that the depicted morphocline Cattania - Helicigona - Dinarica (Schileyko, 2013) cannot be interpreted in an evolutionary context.
Aiming at a general classification of the Ariantinae, the shape of the accessory glands is equally uninformative. The transition from undivided to split gland(s), or the other way round, must have occurred more than once.
When the phylogeny reconstructions obtained with this study are compared to generic classifications based on conchology and geography, nearly all the named (sub)genera are recovered as distinct clades. A few additional (sub)generic groups were discovered and described, viz. Campylaea (Oricampylaea), Chilostoma (Achatica), Cattania (Cattaniella), Pseudotrizona, Kollarix (Table 5, Appendix). The phylogenetic relationships above the genus-level, as indicated by the lower posterior probabilities, remain less certain in many cases. For a limited number of genera, sister-group relationships were disclosed, i.e. Arianta-Cylindrus, Causa-Isognomostoma, Josephinella-Thiessea and Kosicia-Faustina (PP ≥ 0.95; Figs 1, 2, S5, S6). In particular the close relationship between Cylindrus and Arianta is intriguing. Clearly both genera are part of a lineage that was less restricted in the development of conchological novelties. The classification of Cylindrus as a member of the Ariantinae is now confirmed genetically.
We agree with Cadahia et al. (2013) that indications of evolutionary age are uncertain, to say the least. The fossil record is very incomplete indeed, and a molecular clock model is also not easily applicable. The unattractive alternative would have been to omit such speculations altogether.