Comparative analysis and phylogenetic implications of tardigrade musculature architectures.

Most knowledge on tardigrade musculature architecture dates back to the end of XIX century, and the beginning of XX century. It has been only in the last five years that a great deal of new information on tardigrade musculature system has become available, mainly thanks to the use of rhodamine-phalloidin staining of F-actin in combination with three-dimensional microscopical techniques such as confocal laser scanning microscopy (CLSM). In spite of all these information, only few and fragmentary evolutionary considerations on tardigrade musculature system have been done. This is probably due to the relatively low number of analyzed taxa, and to the difficulty in the comparisons of data that often have been obtained with different degree of accuracy, and are presented using different terminologies. In this study we increased the number of analyzed species, by studying CLSM the musculature architectures of 7 species representative of most tardigrade higher taxa: the heterotardigrades Batillipes bullacaudatus (Arthrotardigrada), and Echiniscus testudo (Echiniscoidea), and the eutardigrades Paramacrobiotus richtersi, Dactylobiotus parthenogeneticus (Parachela, Macrobiotoidea), Bertolanius volubilis (Parachela, Eohypsibioidea), Acutuncus antarcticus (Parachela, Hypsibioidea) and Milnesium tardigradum (Apochela, Milnesiidae). We were able to define all the muscular fibers associated with the body movement. The number of fibers and their organization changed among taxa, with heterotardigrades being the least complex. Muscular fibers have been schematically organized into three systems: dorsal, lateral, and ventral. The ventral system was the most conservative, showing a clear metameric pattern and only few differences among taxa, while the lateral system was the most derived and so precise homologies were not always well defined. Using these new morphological information and literature data on Halobiotus crispae (Parachela, Isohypsibioidea), it was possible to analyse the phylogenetic signal of the musculature system. Two matrixes were constructed: a morphological matrix of 94 characters based on musculature data, and a matrix for a total evidence analysis combining the previous data with molecular data (18S, 28S). Both matrixes have been analyzed in Bayesian and parsimony frameworks. The phylogenetic trees, obtained by both analyses using both matrixes, differ only for the position of Eohypsibioidea. Heterotardigrade taxa were the sister group of all Eutardigrada, within this last class, Apochela and Parachela were sister taxa; among parachelan superfamilies the Isohypsibioidea was the most basal, the Macrobiotoidea the most derived, while the Eohypsibioidea changed position according to the analyses. Our data demonstrated that musculature architecture can be used for phylogenetic purposes, but it is only applicable at higher taxonomic levels. Indeed, the musculature of Paramacrobiotus and Dactylobiotus, although belonging to different families of Macrobiotoidea, showed the same muscular organization.