Calcaronea, Bidder, 1898

Alvizu, Adriana, Eilertsen, Mari Heggernes, Xavier, Joana R. & Rapp, Hans Tore, 2018, Increased taxon sampling provides new insights into the phylogeny and evolution of the subclass Calcaronea (Porifera, Calcarea), Organisms Diversity & Evolution (New York, N. Y.) 18 (3), pp. 279-290 : 286-287

Phylogeny of Calcaronea using C-region (28S rRNA)

The C-region dataset provided better phylogenetic resolution within Calcaronea than the 18S rRNA, and in most cases, our phylogeny was congruent with previous studies ( Dohrmann et al. 2006; Voigt et al. 2012; Voigt and Wörheide 2016). For instance, the basal position of the genus Leucosolenia was also supported in our phylogeny but with the inclusion of more species, this genus could not be recovered as monophyletic and was recovered in three different clades. Unfortunately, the relationships between these clades are not well resolved, and more data is needed to assess the status of the genus. L. botryoides , which is the type species of the genus and family, is recovered in the most basal of the three Leucosolenia clades alongside Leucosolenia sp. (Accession NO AY026372). Although this sequence has been used in various previous phylogenetic studies, we could not obtain additional information on the origin and morphology of the material from which it was generated.

Additional differences resulting from the higher taxon sampling in the present study can also be highlighted. For example, species belonging to the Baerida sensu stricto were recovered together with the genus Achramorpha in a well-supported clade, and this result was obtained using both markers. This close relationship may indicate that the pugioles (small harpoon-shaped tetractines) characteristic for the order Baerida may be homologous to the larger chiactines which are present in Achramorphidae . These unusual tetractines found in the atrial skeleton may possibly serve as synapomorphy for the well-supported clade of Baerida and Achramorpha sp. We consider that Baerida should be abandoned as an order, and its member taxa should be reassigned to the order Leucosolenida . This is in agreement with previous results where the actual ordinal classification of Calcaronea has been rejected ( Dohrmann et al. 2006; Voigt et al. 2012).

Another interesting result was that although we recovered two well-supported clades comprising several Paraleucilla spp. , we could not recover the monophyly of the genus. The inclusion of the two other genera, Leucilla and Amphoriscus , would allow to test the hypothesis of non-monophyly of the family Amphoriscidae . Furthermore, we can also mention that the species P. magna , well-known for being the first true invasive species among calcareous sponges in the Mediterranean Sea ( Longo et al. 2007), is also present in St. Helena Island in the South Atlantic and in the southern coast of Portugal. With these results, we expand on the distribution of P. magna , which was originally described from the Atlantic coast of Brazil ( Klautau et al. 2004; Cavalcanti et al. 2013).

According to our phylogeny, there were two instances in which sequences exhibiting low divergence were not recovered as monophyletic but rather formed poorly supported clades (support values <49) – Breitfussia spp. (in clade VIII) and Sycon abyssale (in clade X). It has been previously suggested that speciation events closely spaced in time lead to short internal branches that are difficult to solve ( Saitou and Nei 1986).

The close relationship between Grantiidae with giant diactines and the family Heteropiidae suggested by Voigt et al. (2012) is also supported by our results (clade IV). However, upon addition of more taxa, this relationship seems to include also some species of Grantiidae and Sycettidae with long radially arranged diactines, such as Grantia arctica , G. capillosa , and Sycon ciliatum . Interestingly, the relationship between these three families ( Grantiidae , Sycettidae , and Heteropiidae ) is repeated and well-supported also in the clades VII and X. These results confirm that the pseudosagittal spicules, considered a characteristic trait for the family Heteropiidae , have convergently evolved more than twice or were lost in closely related species, as also suggested by Voigt et al. (2012). This scenario can also explain the evolution of other morphological traits found in these clades, like long protruding diactines and long longitudinal diactines.

Interestingly, clade X also contains the type species of the monotypic genus Sycandra ( Sycandra utriculus ) from the Grantiidae . This genus has traditionally been characterized by an internal atrial network of diactines. In our material, we identified an additional and new species of Sycandra from Greenland, which present the main morphological characteristics of the genus but also displaying long diactines ornamenting the distal cones and an oscular membrane supported by embedded and longitudinally arranged diactines.

One of the main challenges of phylogenetics is to interpret the causes underlying incongruent reconstructions based on independent (e.g., morphological vs. molecular) data sources. In the case of sponges, this is particularly complex because their taxonomy and classification is largely based on a limited set of morphological characters, i.e., spicules and skeleton organization. Moreover, in some taxa, these characters present high levels of homoplasy, making it even more difficult to interpret the evolutionary history ( Manuel et al. 2003; Cárdenas et al. 2011; Morrow et al. 2013). Convergent evolution and secondary loss of spicules and skeletal characters have been shown to explain the incongruences found in various groups, e.g., in Heteroscleromorpha, Astrophorida, and Polymastiida ( Cárdenas et al. 2011; Plotkin et al. 2012; Morrow et al. 2013; Plotkin et al. 2017).

In the case of calcaronean sponges, it also seems like convergent evolution and secondary loss play a key role in the evolutionary history of the group. In fact, after increasing the taxonomic coverage, some morphological characters which were thought to be characteristic for some taxa turn out to be homoplasious. For example, giant longitudinal diactines, presence/absence of a cortex, pseudosagittal triactines, or long radially arranged diactines are characters that have been acquired independently in different taxa. Another example is the presence of a stem without choanocyte chambers in all the Sycon and Sycetta species from clade X, which is also found in Grantia sp. nov. 1 from clade VII, suggesting that it has evolved independently in different lineages. However, the absence of the B stem^ character in the rest of the species from clade X could represent a case of secondary loss, which, according to Jenner (2004), is a process that can occur often, if not more often, than independent evolution of a character.

Our study highlights that even a much-increased taxonomic coverage does not provide a congruent classification for this group, especially when the main morphological characters present such high levels of homoplasy. Therefore, we agree with Voigt et al. (2012) and Manuel (2006) that the evolution of Calcaronea does not follow a clear trajectory, is not oriented from simple to more complex structures, and is characterized by frequent secondary loss and convergent evolution.