ABSTRACT1. Mass mortality events are becoming more common all over the world, both in tropical and temperate seas. An extensive mortality occurred in the Mediterranean Sea in 1999, affecting many benthic species, mainly sponges and gorgonians.2. The recovery of a population of the sea fan Paramuricea clavata, for a period of 3 yr, from 1999 to 2002, was studied by both line transects and fixed frames. The average size of the colonies decreased, indicating a size-dependent mortality episode, but their density, as a result of successful recruitment, was not altered after 3 yr.3. P. clavata showed three recovery patterns: (i) sexual reproduction, (ii) coenenchyme regeneration and (iii) fragmentation of affected branches. Moreover, the growth rates of small colonies varied in the different years. The sex ratio of the population was also altered, with females being more affected than males; the population studied showed a significant male bias (3.3:1, n ¼ 150), varying greatly from the typical sex ratio (1:1) previously recorded in the same population before the mass mortality event.

The m e r a l o g i c a l features of the substrate were generally cons~dered a rmnor factor m structunng manne benthic communities The alm of t h~s work 1s to venfy whether the presence of quartz nunerals in rock may exphcate dlfferences, usually explamed m terms of substrate roughness or other factors In epibenthic cornrnun~hes Laboratory tests on the hydroid Eudendnum glomeratum showed that ~t s planulae settle preferentially on carbonat~c, rather than quartzltic substrates To test the Influence of quartz on established communities, we analysed the specles composition and quantitat~ve structure of subhttoral sesslle assemblages on Wferent rocks In several locahhes of the L~gurian and Tyrrhenian seas The observed dlfferences appeared to be related to the presence of quartz in the substrate rock The ~nteractions between organisms and minerals (bio-mmeralogy) mlght play a slgnlficant role on benthic communities affechng not only the inihal colonlsahon but also later assemblages This potential role has been largely neglected to date and further studies are needed to prove ~t s Impor tance KEY WORDS: Substrate colonisation . Mineral composition . Marine benthos distnbution . Hard substrates . Bio-mineralogy INTRODUCTIONThe spatial distribution and structure of marine benthic communities are due to numerous abiotic and biotic factors which, in turn, are influenced by the presence of the organisms, in a mutual exchange of inputs. Among the abiotic factors, the mineralogical features of the substrate were generally considered of scarce importance, but recent studies by Cerrano et al. (1998) have shown that the presence of quartz in the sand may affect the initial steps of infauna colonisation. Cerrano et al. (1998) introduced the term bio-mineralogy to explicate the interrelationships between biological systems at different hierarchies (cell, organism, species, community) and minerals.Bio-mineralogy could influence hard-bottom assemblages and explain some 'anomalies' in the structure of communities growing on rocks of different nature. A species assemblage, which may be slightly more attracted to a particular substrate, could affect succession by its subsequent interaction with later assemblages. A similar effect was evidenced in the colonisation of artificial substrates, with respect to both species composition and abundance (Anderson & Underwood 1994, Holm et al. 1997). Less information is available for natural substrates (McGuiness 1989), but it is common knowledge that the softness and asperity of a rock can favour or hamper biotic colonisation through selective larval settling, retention of water (in the littoral) and organic matter, and provision of refuges from predation or grazing (Den Hartog 1972, Levinton 1982, Walters & Wethey 1996.More is known about the influence of substrate mineralogy on bioboring, which is prevented by high percentages of quartzitic or pelitic components in the rock. Sublittoral endolitic communities are charac-

Temperate reefs, built by multilayers of encrusting algae accumulated during hundreds to thousands of years, represent one of the most important habitats of the Mediterranean Sea. These bioconstructions are known as “coralligenous” and their spatial complexity allows the formation of heterogeneous microhabitats offering opportunities for a large number of small cryptic species hardly ever considered.Although sponges are the dominant animal taxon in the coralligenous rims with both insinuating and perforating species, this group is until now poorly known. Aim of this work is to develop a reference baseline about the taxonomic knowledge of sponges and, considering their high level of phenotypic plasticity, evaluate the importance of coralligenous accretions as a pocket for biodiversity conservation.Collecting samples in four sites along the coast of the Ligurian Sea, we recorded 133 sponge taxa (115 of them identified at species level and 18 at genus level). One species, Eurypon gracilis is new for science; three species, Paratimea oxeata, Clathria (Microciona) haplotoxa and Eurypon denisae are new records for the Italian sponge fauna, eleven species are new findings for the Ligurian Sea. Moreover, seventeen species have not been recorded before from the coralligenous community. The obtained data, together with an extensive review of the existing literature, increase to 273 the number of sponge species associated with the coralligenous concretions and confirm that this habitat is an extraordinary reservoir of biodiversity still largely unexplored, not only taxonomically, but also as to peculiar adaptations and life histories.

The new endoperoxyketal polyketides manadoperoxides A-D (2-5) have been isolated from the Indonesian sponge Plakortis cfr. simplex and their stereostructures established by means of spectroscopic data and semisynthetic transformations. Manadoperoxides were assayed in vitro against D10 and W2 strains of Plasmodium falciparum and showed moderate antimalarial activity compared to that of plakortin (1) and peroxyplakoric B(3) ester (9), the latter differing from manadoperoxide B only by minor structural details. This unexpected difference in the antimalarial activity has been rationalized on the basis of our recently published model for the interaction of 1,2-dioxanes with heme and production of C-centered radicals toxic to the parasite. For the manadoperoxides, either the endoperoxide linkage is inaccessible to the heme iron or the O1 radical cannot evolve to produce a C-centered radical.