DOES COSMOPOLITANISM RESULT FROM OVERCONSERVATIVE SYSTEMATICS? A CASE STUDY USING THE MARINE SPONGE CHONDRILLA NUCULA

The sponge species Chondrilla nucula has a simple morphology and a very wide geographical distribution. To verify whether the latter might be an artifact of the former, samples of this species were collected across 10,000 km of its range, in the Mediterranean, the Caribbean, and the southwestern atlantic. The classical (spicule morphology) and molecular (allozymes) systematic approaches were compared, to try to define the geographic limits between populations and detect possible cryptic species. We found five distinct genetic forms within C. nucula that sometimes showed morphological homogeneity and other times plasticity. The difference in size of spicules could not be related to the clear-cut genetic differences, suggesting that the use of spicule sizes for sponge systematics should be reappraised. The population of one of the genetic forms along 3000 km of the Brazilian coast was highly structured (F_ST = 0.21; N_em = 0.96). Our results reject the null hypothesis of cosmopolitanism of C. nucula and indicate that the putative worldwide distribution of some marine sponges, and possibly many other benthic invertebrates, may be the result of overly conservative systematics. Cryptic species appear to be particularly prevalent when genera are well defined but species are characterized by only a few morphological characters.