ABSTRACT
Species complexes - groups of closely related species in which intraspecific and interspecific variability overlap - have generated considerable interest and study. Frequently, members of a species complex do not have complete reproductive isolation; therefore, the complex may go through extensive gene flow. In the Caribbean Sea, some encrusting and excavating sponges of the genus Cliona (Porifera, Hadromerida, Clionaidae) are grouped within the great "Cliona viridis" complex because of their morphological similarities. This study examined the evolutionary relationships of the Caribbean members of this complex (C. caribbaea, C. tenuis, C. aprica and C. varians) and related taxa based on nuclear (ITS1 and ITS2) and mitochondrial (3' end of ND6) DNA sequences. The intragenomic ITS variation and its secondary structures were evaluated using a mixed approach of Denaturing Gradient Gel Electrophoresis (DGGE), DNA sequencing and secondary structure prediction. Considerable intragenomic variation was found in all the species, with apparently functional ITS1 and ITS2 secondary structures. Despite the subtle but clear morphological differentiation in these excavating sponges, the intragenomic copies of C. caribbaea, C. tenuis and C. aprica had a polyphyletic placement in the ITS1 and ITS2 genealogies and very low divergence. Therefore, it is clear that these species constitute a species complex (herein called Ct-complex). Genetic distances within the Ct-complex revealed that an important part of the interspecific variation overlapped with intraspecific variation, suggesting either incomplete lineage sorting or extensive gene flow. In contrast, C. varians and an unidentified "Pione" species emerged as monophyletic clades, being the closest sister groups to the Ct-complex. Additionally, our results support that C. laticavicola and C. delitrix conform a monophyletic group, but absence of reciprocal monophyly in these species suggests they may be life stages or ecophenotypes of a single species or they have diverged recently. Our work showed that the 3' end of the ND6 mitochondrial gene was highly conserved and not suitable for phylogenetic analysis at the interspecific level.