Microhydrula limopsicola is Haliclystus antarcticus

Collins and co-workers [10] did not formally establish a synonym for Microhydrula limopsicola. However, a very close relationship of M. limopsicola with H. octoradiatus and also with “the species of Haliclystus from southern Chile reported on by Zagal (2004)” was noted, although the latter was not formally included in their analysis [10]. Our analyses are obviously constrained by the non-availability of other cultures or samples of M. limopsicola, which as far as we know has only been observed by Jarms and Tiemann [11], but our results are based on the most complete data available at this time. With increased taxon sampling and data from fast-evolving markers, we conclude that M. limopsicola is actually H. antarcticus. Several points support this conclusion.

First, we document 100% identity of three fast-evolving markers of Microhydrula limopsicola with those of H. antarcticus from Antarctica, all of which differ slightly from conspecific samples of H. antarcticus from Chile. While intraspecific variation of these genetic markers is not very well known for species of stauromedusae, available data indicate that some intraspecific genetic variation exists, and that it is smaller than observed interspecific variation (compare Table 3 with Table 2). This would suggest that identity in these genetic markers can only happen if the samples are taken from the same species. Finally, one might question whether there are other Antarctic species of Haliclystus that could confound our identification of Microhydrula limopsicola as H. antarcticus. There is one additional species known from the southern hemisphere, Haliclystus kerguelensis Vanhöffen, 1908 from Kerguelen Island, southern Indian Ocean, but this species is readily differentiated from H. antarcticus by its morphology [12] and thus should have a different genetic signature; no specimens of H. kerguelensis suitable for genetic study were available to us.

With the remainder of the discussion, we synthesize the relevant historical literature to address the implications of our identification of M. limopsicola as H. antarcticus on taxonomy, morphology, and life history of Stauromedusae.

Taxonomy

The implications of this synonymy for the family Microhydrulidae were only briefly touched upon by Collins and co-workers [10]. The family Microhydrulidae (Hydrozoa, Limnomedusae) encompassed three species in two genera: Microhydrula pontica Valkanov, 1965, M. limopsicola and Rhaptapagis cantacuzenei Bouillon and Deroux, 1967. One is now clearly established as a stauromedusa. It remains to be explicitly tested whether M. pontica and R. cantacuzenei are also early stages of the life cycle of local species of Stauromedusae, but we think this is likely to be the case, since similarities between these species and the preserved larvae of stauromedusae have been previously recorded [29]. We note that M. pontica and R. cantacuzenei have been found in abundance living in marine surface biofilms at the Station Biologique in Roscoff, France, on the English Channel, in the same vicinity where several species of Staurozoa including Haliclystus auricula, H. octoradiatus, Depastrum cyathiforme (M. Sars, 1846), Lucernariopsis campanulata (Lamouroux, 1815) and Craterolophus convolvulus (Johnston, 1835) [6], [29], [30], [31] have also been found. Of course, additional data, particularly genetic data, from other species of Microhydrulidae are necessary to test our hypothesis that these species represent stages in the development of species of Stauromedusae.

Morphology

Medusozoan plesiomorphies (e.g., primitive widespread cnidarian characters such as the presence of microbasic euryteles nematocysts), homoplasies (e.g., presence of convergent morphological characters such as periderm and life history characters such as asexual frustules), and the very simple morphology (small solitary hydroids without tentacles and sexual stage [11], [29]) of M. limopsicola evidently worked as obstacles in correctly identifying M. limopsicola when it was discovered. As a result, M. limopsicola bears a closely resemblance to the very simple hydropolyps such as those of Craspedacusta and Monobrachium [11]. No doubt the scarce literature on early stages of staurozoans made it difficult to establish reliable comparisons among taxa. Even though dissimilarities are evident [10], there are several morphological similarities between M. limopsicola and Staurozoa.

First, the hemispherical shape of the settled planulae of Haliclystus octoradiatus [6] is similar to the general shape of Microhydrulidae “polyps” [11] (Figure 4). Moreover, both lack mouth and a permanent gastrovascular cavity necessitating intracellular digestion [6], [11], [29]. A further similarity is the production of frustules [6], [11]. More specifically, the settled, rounded up planula of H. octoradiatus produces lateral protuberances, which become buds for the process of frustulation [6], similar to the “young polyps of M. limopsicola”, which can also produce frustules [11] (Figure 4). Likewise, the frustules of Microhydrulidae are formed from a lateral budding of the body [29]. The planulae of H. octoradiatus can also have small pronounced expansions, more or less regular, not related to the process of frustulation. These protuberances, which are provided with numerous nematocysts, result from a local thickening of the ectoderm and seem to play an important role in prey capture prior to the development of tentacles [6]. Homology between the protuberances, found in H. octoradiatus planulae larvae and the similar morphology in M. limopsicola is not clear. One possible interpretation is that these expansions correspond to the cauliflower structure bearing numerous nematocysts found at one end of “elder” individuals of M. limopsicola [11]. However, the cauliflower structure can also be a simple result of a strong aggregation of larvae (see below) [6] since at a more advanced stage, the larva of H. octoradiatus increases considerably in size, and its contour, previously more or less rounded, develop four lobules [8] in a similar arrangement to the cauliflower structure (Figure 4). Another interpretation is that the development of H. antarcticus differs from that of H. octoradiatus, with the cauliflower structure corresponding to the early development of the primary tentacles. Recently metamorphosed individuals of Haliclystus borealis Uchida, 1933, H. stejnegeri, and H. auricula, as well as their juvenile medusae stages, have eight primary tentacles [32], which might be correlated to these cauliflower structures.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Comparisons between the M. limopsicola polyp and the H. octoradiatus settled planula.

(A–C) different stages of M. limopsicola [11]: A) newly settled “polyp”; B) closely attached “polyps”, with expansions provided with nematocysts; C) later stage, with a cauliflower-shaped head. (D–E) process of frustulation observed in both species: D) planula of H. octoradiatus producing lateral protuberances, which become frustules [6]; E) frustules [11]. (F–G) possible correspondences of stages of both species: F) a group of H. octoradiatus larvae, capturing a nauplius [6]; G) superior view of a settled planula of H. octoradiatus at an advanced stage, showing four lobes [8].The hemispherical shape and the production of frustules (A, D, E) are similar in settled planulae of H. octoradiatus and “polyps” of M. limopsicola. The same gregarious behavior to feeding was observed in both species (B, F). At a more advanced stage, the larva of H. octoradiatus presents four lobes (G), that might be associated with the cauliflower structure seen in later stages of M. limopsicola (C), which possibly is an aggregation of more than one individual. Figures modified from [6], [8], [11].

https://doi.org/10.1371/journal.pone.0010182.g004

Histological similarities between M. limopsicola and H. antarcticus are also of note. The “polyps” of M. limopsicola are attached to the substratum with a slightly widened base, whose ectoderm produces a flat thin fibrous periderm plate [11]. Stalks of adult individuals of the genus Haliclystus also present fibrillar components at the attachment sites [33]. In addition, the endoderm of M. limopsicola and of the settled planula of H. octoradiatus is composed by vacuolar cylindrical cells, which touch at the terminal end, without leaving any space [6], [11].

One feature remarked by Jarms and Tiemann [11] and Collins and co-workers [10] is the cnidome, with microbasic euryteles being present in M. limopsicola. This nematocyst type is a common feature of H. antarcticus [12], [34], but microbasic euryteles are plesiomorphic for Staurozoa [3], [32], [35], [36], [37], [38], [39], [40] and not particularly useful for staurozoan taxonomy, besides its ubiquitous presence in other medusozoan groups (see data matrix in [2]). Adults of H. antarcticus also possess isorhizas [12], [34], which were not recorded for M. limopsicola [11]. Accepting the identity of M. limopsicola with H. antarcticus suggests that the cnidome of this species varies ontogenetically [38], [41], [42]. Indeed, nematocysts of the creeping planula larvae of Haliclystus salpinx Clark, 1863 are different from the adults of the same species [43]. The planula larval stage of H. salpinx also has only microbasic euryteles, whereas the adults have both microbasic euryteles and isorhizas [43].

Life cycle inference

Based on our finding that M. limopsicola is synonymous with H. antarcticus we propose that M. limopsicola is an early life cycle stage of H. antarcticus (Figure 5). Thus far, early stages of the life cycle of H. antarcticus have never been recorded. In fact, few staurozoan pre-adult stages are known. Only the creeping benthic planula stage of H. “sanjuanensis” (misidentified as H. stejnegeri) and H. salpinx [43], [44], the post-metamorphosis stages of some stauromedusae [32] and the complete development of H. octoradiatus, S. riedli and S. variabilis [6], [7] have been documented. Therefore, it is presently impossible to assert that the Microhydrula stage is present in all species of Staurozoa. Similarly, it is unclear how widespread the presence of frustule stages is in other stauromedusae. Nevertheless, we suggest that the study of sediments and potential associations with overlooked substrata (e.g., bivalves) is likely to reveal a hidden diversity of life cycle stages and strategies in Staurozoa, which are generally assumed to live only on rocks and macrophytes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Putative scheme of the life cycle of H. antarcticus, including the “microhydrula” phase.

The main life cycle was based on [6], for H. octoradiatus. Stauropolyp stage and its ability to create frustules (white arrows) are hypothesized based on observations of Stylocoronella [7]. Dotted gray arrows corresponding to the “microhydrula” stage, derived from this study. Figures modified from [6], [7], [11].

https://doi.org/10.1371/journal.pone.0010182.g005

The non-sexually-reproductive “microhydrula” stage most likely occurs before the development of the stauropolyp and its apical metamorphosis into an adult medusa in the life cycle. We hypothesize that this stage occurs right after the settlement of the planula larvae (or sometime after the settlement of the frustule), adding a stage (Figure 5) to the hypothetical life cycle proposed for staurozoans, since the best known life cycle of a staurozoan, the odd psammic Stylocoronella spp., apparently does not contain a “microhydrula” stage [7]. Settlement of planulae occurs in groups of 3–20 individuals in H. octoradiatus [6]. Experimental procedures restricting larval aggregation of H. octoradiatus to groups of 1–3 larvae demonstrated that none of these larvae were successful in undertaking further development, probably due to not being able to capture sufficient food when growing in small aggregations [6]. The same gregarious behavior has been described for “adults” of Microhydrulidae, also hypothesized to enhance efficiency in prey capture [11], [29] (Figure 4). Furthermore, similar to planula settlement in H. octoradiatus, the frustules of Microhydrulidae attach to the substrate by adhesion of a portion of the surface lacking nematocysts [6], [11], [29].

It is important to note, however, that the juvenile stauropolyps of S. riedli and S. variabilis can produce frustules by budding of the long filiform tentacles [7]. Such a polyp has not been described for any other staurozoan. Asexual reproduction via frustulation found in the “microhydrula” stage (and possibly also in the polyp, since cnidarians demonstrate different kinds of budding [45], [46]) would increase the potential of large populations in isolated areas of a larger fragmented seascape, which is consistent with the patchy distribution of H. antarcticus, for example. In fact, 130 specimens of H. antarcticus were found in an area of ca. 150m² in Copacabana, King George Island, Antarctica [12], and ca. 385 individuals/m² in Valdivia, Chile [14]. Further, intense asexual reproduction would lead to low genetic diversity. This is consistent with our finding of just single 16S and ITS haplotypes (Table 2) in each sampled H. antarcticus population (Antarctica and Chile).

Ecological constraints of the “microhydrula” stage may restrict the distribution of H. antarcticus. Microhydrula limopsicola was described living attached to the upper valve of the very small, subtidal lamellibranch bivalve Limopsis hirtella on King George Island [11]. Jarms and Tiemann [11] suggested that the association between the clam Limopsis hirtella and the hydroid Microhydrula limopsicola is to be regarded as “highly specific”. Jarms and Tiemann have kept M. limopsicola alive on glass for nearly 20 years and, to our knowledge, no development other than asexual frustulation has been observed [11]. Since M. limopsicola can be kept alive in the lab, the association between M. limopsicola and L. hirtella may be regarded as a coincidence of co-distribution. However, the absence of further development in the lab leads us to believe that L. hirtella provides vital cues enabling further development of M. limopsicola. Such relationships among epibiont and host are not uncommon in marine ecosystems [47]. Limopsis hirtella is spread across the Magellanic Province, the Falkland Islands and the western part of the Antarctic [48]. Coincidentally, this is the area where H. antarcticus is recorded: Antarctica Peninsula [12], [34], [49], [50] and southern South America (Figure 1) [12], [13], [14], [15], [51].

Conclusions

Documenting medusozoan life cycles is an enormous challenge. In this work we show that molecular data can be a useful tool to identify an unknown life cycle stage by tying together different life history stages with the same haplotype and from it derive a hypothesis about the life history of the species. Similarity between the haplotypes of three markers of Microhydrula limopsicola and Haliclystus antarcticus settles the identity of these taxa, expanding our understanding of the staurozoan life cycle, which was thought to be more straightforward and simple. Frustulation was recorded for Haliclystus for the second time, in a different life stage from the one recorded for Stylocoronella. This knowledge sheds light on morphological, biogeographical, and evolutionary issues, mainly because Haliclystus is the most diverse genus in Staurozoa. However molecular analysis will not replace additional investigations. Continued exploration of the meiofauna and integrated analysis encompassing morphology, ecology, molecules, life cycles and biology will be needed to solve outstanding evolutionary and biogeographical questions like those addressed here.