Contributions to Zoology, 86 (1) – 2017Isabel T Hyman; Irantzu de la Iglesia Lamborena; Frank Köhler: Molecular phylogenetics and systematic revision of the south-eastern Australian Helicarionidae (Gastropoda, Stylommatophora)

To refer to this article use this url: http://contributionstozoology.nl/vol86/nr01/a04

Discussion

Phylogenetic relationships and genus-level taxonomy

next section

The southeastern Australian helicarionid clade has previously been identified as a monophyletic radiation (Hyman et al., 2007; Hyman and Ponder, 2010) containing at least five genera (Helicarion, Brevisentis, Mysticarion, Parmavitrina and Peloparion). Here we expand this group to include Cucullarion and Ubiquitarion gen. nov. Synapomorphies of this clade include a spermatophore with spirally arranged branched spines, an epiphallic flagellum with internal cryptae, the absence of an epiphallic caecum and the presence of at most a very short vagina.

The reproductive characters that define this group occur singly or in combination in a number of other helicarionid taxa. For example, some species from northeastern Queensland have a spiraling, spinose spermatophore (e.g. Pravonitor kreffti, Hyman and Ponder, 2010) but do not share the other morphological characters. There are also several species with no epiphallic caecum (e.g. Levidens, Tarocystis, Sheaia, Hyman and Ponder, 2010). However, whether these similarities are symplesiomorphies of a larger helicarionid radiation or have indeed been acquired in convergence can only be resolved in the framework of a more complete phylogenetic analysis of the Australian Helicarionidae.

The members of the southeastern Australian clade represent a significant proportion of the ecological and morphological diversity of all Australian Helicarionidae. Members of this group are either leaf litter dwellers with fully developed shells (Brevisentis), small, arboreal or semi-arboreal semislugs (Peloparion, Ubi­qutarion, Mysticarion), small to large litter-dwelling semislugs (Helicarion, Parmavitrina), or medium-sized semi-arboreal semislugs with a strongly reduced shell (Cucullarion).

Due to these significant differences in their general morphology and ecology, the south-eastern Australian helicarionids provide a suitable model to study evolutionary phenomena which have also occurred in the broader Australian Helicarionid radiation, such as limacisation.

Within this group, we found that all genera as now delimited formed well-supported (in terms of nodal support) and well-individualized (in terms of branch lengths) clades in the mitochondrial phylogenies. However, the phylogenetic relationships between these clades were poorly supported by bootstrapping, leading to substantially inconsistent topologies produced for different sequence datasets (COI, 16S, concatenated; Fig. 6). This lack of confidence in principal divisions in the phylogenetic tree poses a challenge to the inference of evolutionary trends among the southeastern helicarionid radiation.

However, it is still possible to draw some general inferences relating to character evolution in this group. Brevisentis, the only snail genus included in this study, was never the sister taxon to the rest of the clade. Consequently, under the premise that a fully developed shell represents the ancestral character state in this clade, shell reduction must have either been reversed in Brevisentis or has independently evolved multiple times within this clade. Some signs of slight shell reduction have been observed in the digestive and reproductive systems of Brevisentis jacksoniensis (Hyman, submitted), which may indicate a reversal of limacisation in Brevisentis. In any case, given the variability of body form in this group, shell reduction has evolved to a higher degree in some lineages (e.g., the rainforest semislugs belonging to Cucullarion and the Parmavitrina megastoma group) than in others (e.g., Mysticarion).

FIG2

Fig. 6. Comparison of principal branching patterns of the best maximum likelihood trees produced for the single gene data sets (tips collapsed at genus-level). A. Backbone tree for 16S. B. Backbone tree for COI. Numbers on branches indicate nodal support based on 200 thorough bootstrap replicates for a total of 10 independent runs. Scale bar indicating 3% of modelled sequence divergence.

Systematic significance of reproductive characteristics

The southeastern helicarionid radiation is characterized by several morphological characters, including the absence of an epiphallic caecum, a flagellum producing a spinose spermatophore, and a short or absent vagina (Hyman and Ponder, 2010). The function of the epiphallus and epiphallic flagellum is to produce the spermatophore, the tail-pipe forming in the flagellum and the capsule in the epiphallus. The epiphallic caecum is believed to function in turning the spermatophore so that the tail-pipe is moved into the penis first (Dasen, 1933; Van Mol, 1970). In this group, the absence of the epipahllic caecum may be compensated for by the presence of a spermatophore with a relatively short tail-pipe. In Australian helicarionids that do have an epiphallic caecum, a considerably longer flagellum (and therefore a longer spermatophore tail-pipe) is present (e.g., Nitor, Fastosarion, Westracystis; Hyman and Ponder, 2010). There are also a number of species that lack both the epiphallic flagellum and the epiphallic caecum (e.g., Levidens, Sheaia, Tarocystis, Echonitor, Periclocystis). This may indicate a correlation between the presence of the epiphallic caecum and the length of the spermatophore tail-pipe in Helicarionidae.

The presence of the epiphallic flagellum and caecum characterizes the superfamily Helicarionoidea (Helicarionidae, Ariophantidae, Urocyclidae), but both organs have probably been repeatedly lost in all three groups (Hausdorf, 1998; Hyman and Ponder, 2010).

Similarly, genital characters, such as flagellum, epi­phallus, and penial sheath, have repeatedly been lost (and possibly been regained occasionally) by members of Camaenidae (Köhler and Criscione, 2015). These findings highlight the evolutionary plasticity of land snail genitalia in general and underpin the conclusion that while usually useful at lower taxonomic levels, they should be used with care and in conjunction with other characters when defining broader taxonomic groups.

Significance of morphological and mitochondrial characters in species delimitation

The most consistent and reliable character for species delimitation in every group was the internal anatomy of the penis, along with other genital characters relating to the penis, epiphallus and spermatophore. The sculpture of the penis interior was informative at both genus and species level, with a general pattern observed within a genus (for example, longitudinal pilasters and the presence of a penial verge in Mysticarion) and a unique, highly consistent anatomy in each species (for example, number and shape of longitudinal pilasters, size and shape of penial verge). Similarly, the spermatophore provided useful characters at both genus and species level; however, the small number of spermatophores collected limited the value of this character somewhat. Variation in the spermatophore is reflected in the shape of the flagellum, and although less detail is available, in some cases differences in the flagellum shape were sufficient to infer differences in the spermatophore.

Penis morphology shows great variation among land snails and is generally species-specific, supporting the hypothesis that it is an important factor in species recognition during copulation (Gómez, 2001). It is likely that the spermatophore plays a similar role (Gómez, 2001). In the related slug family Milacidae, spermatophores are also heavily sculptured with branching spines and hooks, and are highly species-specific, in some cases providing the most useful means of species discrimination (Wiktor, 1987).

Body colour and shell dimensions, which are of high value in rapid, non-expert identification, were also informative characters in species delimitation in many groups. Body colour was consistent within species, although preserved specimens were faded, making colour markings sometimes difficult to accurately assess. In particular, Mysticarion obscurior, which co-occurs with congeners M. porrectus and M. insuetus and was often misidentified in museum collections, could reliably be distinguished in life by its darker body colour and in faded alcohol-preserved specimens by the disjunct black markings on the right shell lappet. Even in alcohol-preserved specimens body colour could be used to distinguish species in Cucullarion and Brevisentis, but was less informative in Parmavitrina, where colour differences are more subtle.

Characters relating to the shell, commonly used in land snail identification, are now thought to be significantly influenced by environmental factors and as a result can show high levels of both convergence and conservatism (Stankowski, 2011; Criscione and Köhler, 2013a; Köhler and Criscione, 2015). Within the south-eastern Australian helicarionid clade there is great diversity in shell form, ranging from snails with a complete shell to semislugs with a highly reduced shell. Gross shell shape was therefore an important character in the separation of genera, although convergence was evident in (for example) the semislugs Peloparion helenae and Ubiquitarion iridis. At the species level, differences in shell size and shape (primarily height-to-diameter ratio) were informative, particularly in Brevisentis, Mysticarion and Parmavitrina, although differences were slight and in many cases could not reliably be used as a sole means of identification. Protoconch microsculpture has been used to delimit genera in many Australian land snails including the Helicarionidae (Stanisic et al., 2010); however, we did not find significant differences in shell microsculpture, with the exception of a unique sculpture of notched spiral grooves on the protoconch of Mysticarion.

Species delimited by their morphology were consistently found to form monophyletic clusters in the mitochondrial trees, which usually were well separated from each other by long basal branches. The amounts of interspecific genetic differentiation observed here were at the lower end of the rather wide spectrum of average interspecific differences among stylommatophoran land snails (Davison et al., 2009; Criscione et al., 2012; Köhler and Johnson, 2012; Criscione and Köhler, 2013b; Burghardt and Köhler, 2014). In addition, there has been considerable overlap between the ranges of intraspecific and interspecific distances in the southeastern Australian helicarionids, particularly in 16S (Fig. 5). This overlap is attributable primarily to the high intraspecific variation in the widespread Ubiquitarion iridis, and the comparatively low interspecific variation among P. megastoma, P. flavocarinata and P. maculosa. However, we consider the differences in the penial complex as sufficient to justify their separation as distinct species, particularly given that two of them occur in sympatry with each other. The particularly low interspecific distances among these three species are probably indicative of their rather recent origin.

The highest intraspecific differences among the studied species were observed in Ubiquitarion iridis (COI 2.8%, 16S 2.2%) and Brevisentis atratus (COI 2.8%, 16S 1.6%). In neither case was this reflected by any morphological differentiation nor were these distances remarkable when compared with other land snail groups (e.g., Davison et al., 2009). In contrast to the abovementioned species, we found that the two sympatric species of Cucullarion exhibit a high degree of genetic distinctiveness while being distinguished by rather subtle anatomical differences.

Our findings of such inconsistent amounts of morphological and mitochondrial differentiation among the examined species confirm that species are best delimited by combining the appraisals of their morphological and mitochondrial differentiation.

Species distributions

Southeastern Australian helicarionids exhibit varied distributional patterns. We have found very broad ranges of up to 500 km in Mysticarion and 750 km in Ubiquitarion (although Ubiquitarion is thought to be introduced into the southern part of its range, and its natural range is probably closer to 400 km; Stanisic, pers. comm.). Species belonging to both genera are arboreal and when not actively crawling, sit in a resting posture on the underside of leaves. They are likely to be subjected to greater passive dispersal by wind than litter-dwelling species, such as Brevisentis. While M. hyalinus is allopatric with respect to its congeners, the other three species of Mysticarion are sympatric throughout much of their range. There is some evidence that M. porrectus inhabits higher altitude, moister areas and may not occur in micro-sympatry with the others; however, M. obscurior and M. insuetus have indeed been collected from the same locality. These two species are not sister taxa in the phylogenetic tree and their sympatry is probably secondary.

In contrast, the semi-arboreal taxa Peloparion and Cucullarion have relatively narrow ranges of around 30-50 km in diameter. Peloparion helenae lives under bark on trees or under grass (Stanisic et al., 2010). Cucullarion generally rests in the base of palm fronds or in leaf litter, only crawling on palm fronds and trunks during rain. Both are less prone to passive dispersal by wind than fully arboreal species. In addition, both groups have a high degree of shell reduction and may be more susceptible to desiccation, rendering them rather poor dispersers.

The litter-dwelling taxa Brevisentis and Parmavitrina also exhibit restricted distributions, with ranges of about 30 km (B. kaputarensis) up to about 300 km (B. atratus, P. planilabris, P. rubrica).