A taxonomic study of tropical marine sponges (Porifera: Demospongiae: Haplosclerida and Petrosida) using morphological, chemical and reproductive character sets

This study was undertaken to determine whether the separation of sponge species (Phylum: Porifera, Class: Demospongiae, Subclass: Ceractinomorpha) into two orders, the Haplosclerida and Petrosida, was valid. To this end selected species from each order were collected from the Great Barrier Reef, Queensland, Australia. Their taxonomic position was compared using a number of character sets: morphological, chemical and reproductive. The most useful morphological characters for taxonomic purposes were organisation of the internal sponge skeleton, its components and their quantities. These characters allowed separation of the sponge species examined into five families, and also distinguished sponges at the species level. Development of the ectosomal skeleton was found to differentiate genera, and microscleres were useful for distinguishing some species. As a result of the morphological study four new species of Haplosclerida and three new species of Petrosida were described. Analyses of sterol chemistry did not support clear separation of sponge species into two orders but were found to be useful at the species level. Sponge species were found to have both a qualitative and quantitative sterol fingerprint. Sterol complements of species remained remarkably constant with geographic locality and season. This character set provided evidence for a species complex within Xestospongia muta. Three specimens of this species contained significantly different sterol complements, and two of the specimens were more similar to other species in the data set (Xestospongia n.sp.1 and X. testudinaria) than they were to each other. Results of this chemotaxonomic study suggest that most species currently classified in the same genus are generic complexes, as few species of the same genus had similar sterol complements. No cyclopropene or cyclopropane ring structures, previously thought to characterise species of the Petrosida, were found in the side chains of the fifty four sterols found from all species examined. The reproductive biology of six species, three classified in the Haplosclerida and three in the Petrosida, was monitored for two to four years. The reproductive mode of these species clearly separated them into two orders. Species of the Haplosclerida were found to brood larvae which is the normal reproductive mode of the subclass Ceractinomorpha. Species of the Petrosida did not conform to the normal ceractinomorph pattern. The three species examined (Xestospongia exigua, X. testudinaria and Xestospongia n.sp.1) had separate sexes and broadcast eggs and sperm. Reproductive characters also distinguished sponges at the species level. One species complex, in the Petrosida, was differentiated into two species (X. testudinaria and Xestospongia n.sp.1) when temporal isolation of spawning events was discovered. The spawning events of the two species were consistently separated by at least 15 days and diel timing of these events in both species was from 0700 hours. The mechanism, and possible environmental cues, whereby these species consistently separate their spawning events was not determined although X. testudinaria may have a semi-lunar periodicity. No lunar pattern was observed for Xestospongia n.sp.1. Larva of Xestospongia n.sp.1 were seen in aquaria three days after the spawning event in 1989. This is the first reported observation of larva from a species of the Petrosida. Two haplosclerid species found to brood female reproductive products did so in brood chambers situated basally in sponge individuals. This location may be a protective mechanism to avoid damage from turbulence or dehydration. Females of a third haplosclerid species, Haliclona symbiotica, had reproductive products aligned along the central axis of branches of adults. This location of sexual products is most 4-Iy a spatial confinement related to the symbiotic association this sponge maintains with the macroalga, Ceratodictyon spongiosum. All species of both orders examined that were reproductively active released sexual products in spring and summer. As a result of this study, one character set (reproductive mode) clearly supported the division of species into two orders. Sterol chemistry and morphological characters did not, but were useful at lower taxonomic levels. The distrepancies between higher taxonomic groupings, such as genus and family, suggests the taxonomy of these sponges is not currently resolved