Unravelling host and symbiont phylogenies of halichondrid sponges (Demospongiae, Porifera) using a mitochondrial marker

We present the first comparative phylogenetic analysis of a selected set of marine sponges and their bacterial associates. The Halichondrida form an important order in demosponge systematics and are of a particular interest due to the production of secondary metabolites. We sequenced a fragment of the cytochrome oxidase subunit 1 (CO1) gene of the sponges and their bacterial associates, compared the reconstructed phylogenies and found evidence for radiation in coevolution. The tree of six host-species associations showed four supported cospeciation events between the sponges and the bacteria. In addition, we present the first gene tree of sponges based on a mitochondrial marker. The tree shows major congruences with previous morphological studies and suggests the applicability of a mitochondrial marker in sponge molecular systematics.     

Reproductive strategies of Mycalecontarenii (Porifera: Demospongiae)

The sexual and asexual phases of the reproductive cycle of the sponge Mycalecontarenii (Martens, 1824) were studied from samples collected over a period of 2 years (June 1994 to May 1996) in a Mediterranean coastal basin (Porto Cesareo, southwestern Apulia). The species is viviparous; year round it shows a discontinuous oocyte production. No males were found, possibly owing to the very short period of spermatogenesis. M.␣contarenii produces asexual buds during autumn and winter, and this process seems to be related to changes in water temperature. Even though, at present, we cannot evaluate the incidence of larval recruitment, we postulate that asexual reproduction in M.contarenii plays a crucial role in maintaining the sponge population. Received: 8 July 1997 / Accepted: 5 November 1997     

Location of toxicity within the Mediterranean sponge Crambe crambe (Demospongiae: Poecilosclerida)

Within-specimen location of toxicity in Crambe crambe (Schmidt) has been addressed by complementary procedures on specimens collected in north-east Spain (Western Mediterranean) in winter of 1993. The toxicity of the distal (ectosome) and basal (choanosome) sponge parts have been analysed and the main cellular types present in these two layers have been studied by light and electron microscopy. The toxicity of the three main cell types, separated by the gradient-density method, has also been analysed. Three main fractions, each of them enriched in a different cellular type, were obtained: Fraction 1 (interface between 2 and 5% Ficoll) contained 90±0.9% (mean±SE) of spherulous cells and 10% of different cell types consisting of choanocytes (5±0.54%), and unidentified cells or cell debris (5±0.84%); Fraction 2 (interface between 5 and 8% Ficoll) was enriched in choanocytes (70±0.95%), and also contained spherulous cells (11.8±0.73%), archeocytes (6.2±0.74%) and unidentified sponge cells (12±0.74%); Fraction 3 (interface between 8 and 11% Ficoll) mainly consisted of archeocytes and archeocyte-like cells (75±0.66%), together with spherulous cells (7±0.74%) and other unidentified sponge cells and cell aggregates mainly formed by choanocytes (18±0.41%). Toxicity [measured in toxicity units, TU, using the Microtox^® procedure] was significantly higher in the sponge ectosome (12.45±1.4 TU) than in the choanosome (2.58±0.92 UT). Only the abundance of spherulous cells in the sponge tissues correlated well with the pattern of toxicity observed, and this was corroborated by the toxic behaviour of the three cellular fractions obtained: the one enriched in spherulous cells was highly toxic (9.08 UT), whereas those enriched in choanocytes and in archeocytes were almost inactive (0.48 UT) or totally innocuous, respectively. All these results point to the spherulous cells being responsible for the storage (and possibly production) of the toxic compounds in C. crambe. Toxicity is concentrated in the sponge periphery. Spherulous cells are also concentrated in this area and can also be observed outside the sponge exopinacoderm. These results correlate well with the assumption of a defensive role of toxicity, since encounters with potential epibionts, predators and competitive neighbours take place through this peripheral zone. However, we found two types of spherulous cells (orange and colourless, respectively) coexisting in the same sponge zones as well as in Cell Fraction 1. Thus, we cannot at present determine whether one or both types are responsible for the toxicity encountered, although it is likely that the two correspond to different states of the same cell type.     

Sexual and asexual reproduction in two species of Tethya (Porifera: Demospongiae) from a Mediterranean coastal lagoon

The sexual and asexual phases of reproductive cycles of two sponges, Tethya citrina and T. aurantium, living sympatrically in a Mediterranean coastal lagoon (Stagnone di Marsala, NW Sicily) were studied from samples collected over an 18-mo period. Both species are oviparous and gonochoric. They have a summer, partially overlapping, period of oocyte production, although T. citrina appear to mature earlier. No males were found, possibly due to the very short period of spermatogenesis. Both species produce asexual buds during the autumn/winter months. However, they seem to follow different reproductive strategies, with T. citrina showing a significantly lower production of buds than T. aurantium; by contrast, egg production is significantly lower in the latter species. The difference in the reproductive resource allocation is consistent with data reported in the literature on the anatomy features, genetic population structure and ecological distribution.     

Silica content and spicule size variations in Pellina semitubulosa (Porifera: Demospongiae)

 The variations in both silica content and spicular size were studied in two populations of the demosponge Pellina semitubulosa (Lieberkühn). Samples were collected over a period of 1 year (June 1994 to May 1995) in two Mediterranean coastal basins: Porto Cesareo (southwestern Apulia) and Marsala (northwestern Sicily). The values of spicule size (length and width) and sponge silica content were significantly higher in the population of Porto Cesareo, where the highest water silica concentration was recorded. In both Porto Cesareo and Marsala the sponge silica content showed a seasonal trend, positively correlated with water temperature values. In both populations, the smallest spicules were found in specimens collected from summer to late autumn, after sexual reproduction. Secretion of new spicules may be connected with the process of remodelling occurring in sponges after gamete and larval release. Received: 10 October 1999 / Accepted: 13 April 2000     

Modelling growth forms of the spongeHaliclona oculata (Porifera,Demospongiae) using fractal techniques

The radiate-accretive growth process of the spongeHaliclona oculata, under different environmental conditions, is simulated in a two-dimensional model with fractal modelling techniques. In this model material is added in layers to the object, and growth velocities attain highest values at its protrusions. With this model some aspects of the growth process can be explained. It is possible to simulate thin-branching growth forms, which are normally found under sheltered conditions, and plate-like forms, which are typical for sites more exposed to water movement. These simulated forms are compared with actual growth forms in order to test the validity of the model.     

Predictable annual mass release of gametes by the coral reef sponge Neofibularia nolitangere (Porifera: Demospongiae)

Mass release of gametes of the sponge Neofibularia nolitangere (Duch. & Mich., 1864) occurs simultaneosly along the leeward coast of Curaçao over a period of three subsequent days, without any apparent spatial pattern in the sperad of activity. A population of 99 individual sponges was monitored from August through November 1984 for development and subsequent release of gametes. Release started every day of 3 3-d period (12 to 14 October) at about 1400 hrs and lasted until just after sunset (1830 hrs). Ninety percent of the population showed reproductive activity. Exactly one lunar month later (11 to 12 November), a second release of gametes occurred. In the following year the same sequence of events was observed for the original population (2 to 4 October and 1 to 2 November, 1985). In all instances the first gamete release began on the third day after the full moon. These and earlier observations on this phenomenon show a strong correlation between moon phase and the time of gamete release. Histological and field observations show separate development and release of gametes (sex-ratio males: females 1.5:1) with external fertilization. No sex-reversal occurred in the population during release activity the following year, demonstrating a clear gonochoristic and oviparous type of reproduction for N. nolitangere. First development of oocytes and spermatocytes has been observed at 36 and 7 d, respectively, before the date of first release. In both males and females, a major part of the mesohyl of the total sponge was involved in gamete development. Spermatozoa are released through the osculum as a dense white smoke; eggs consisting of oocyte-nurse cell globules, reinforced with spicules, are separately released en masse with the outgoing water stream of the female sponge. Shortly after release the eggs become sticky and show a negative buoyancy. The reproductive strategy of N. nolitangere appears to be directed at maximizing the number of surviving recruits by maintaining a high reproductive output at a short specific time interval.     

Genetic, morphological, and chemical divergence in the sponge genus Latrunculia (Porifera: Demospongiae) from New Zealand

We undertook a comprehensive study of Latrunculia in New Zealand to determine the relationship between taxonomic, environmental, and chemical variation within the genus. Sponges were collected from five locations around New Zealand: Three Kings Islands, Tutukaka, Wellington, Kaikoura, and Doubtful Sound. Allozyme electrophoresis at nine polymorphic loci indicated that sponges from each geographic location were genetically distinct, and that they displayed genetic differences of the magnitude usually associated with reproductively isolated species (Nei's D between locations =0.375-2.476). Additionally, the comparisons revealed that the green and brown colour morphs of Latrunculia that are sympatric at Three Kings Islands and Kaikoura are distinct from each other, and that there are two genetic groups within the green sponges in Doubtful Sound. On the basis of genetic data we conclude that there are at least eight species of Latrunculia in New Zealand waters, not one to four as had been previously thought. Morphological comparisons of the eight genetic species based on skeletal characters (i.e. skeletal organisation of the choanosome, spicule composition, size, and geometry) indicated that the eight Latrunculia species fell into only two morphological groups that could be easily diagnosed on the basis of discorhabd type. Within these two primary morphological groups, skeletal characteristics among the eight species largely overlap and are not diagnostic. These findings emphasise the limitations of traditional taxonomic methods based solely on skeletal characters for distinguishing species of Latrunculia. However, multivariate analysis (MANOVA and CDA) based on six measured skeletal variables did reveal significant morphological variation among the species (Pillai's Trace=3.28, F=6.90, P<0.0001), supporting the division of the genus into eight species. Comparisons of chemical extracts from Latrunculia also showed that the amounts of five different bioactive compounds (discorhabdins A, B, C, D, and J) varied predictably among the eight species. This finding suggests that discorhabdin variation within Latrunculia, previously thought to be associated with intra-specific environmental variability, is more likely to reflect differences among species rather than phenotypic plasticity. Our results also highlight the importance of thorough taxonomic studies associated with marine natural products research to understand fully the variation in bioactive properties among individuals. The potential processes underlying the unusually high speciation rates in New Zealand Latrunculia that are indicated in our study are discussed.     

Sessile and non-sessile morphs of Geodia cydonium (Jameson) (Porifera, Demospongiae) in two semi-enclosed Mediterranean bays

Morphological plasticity and ecological aspects of the demosponge Geodia cydonium (Jameson) were studied from seasonal samples collected over 1 year in two semi-enclosed Mediterranean bays of the Southern Italian coast (Marsala lagoon and Porto Cesareo basin). Sponge specimens present two morphs: sessile and non-sessile, both of which showed constant size distribution and density over the studied year. Sessile specimens were larger in size than non-sessile ones. This feature is particularly evident at Porto Cesareo, where these sponges have a more compact skeletal network than at Marsala (evident both in the cortical spicule size and sponge silica content). Sessile specimens adhere to hard rocky substrates (Porto Cesareo) or phanerogam rhizomes (Marsala); non-sessile ones occur on soft bottom areas. Several morphological and structural features of the non-sessile forms differ in the two environments, but the difference in body shape seems to play the most relevant role in enhancing the colonization of incoherent substrates. Indeed, at Marsala, where the large amount of silt and clay determines the occurrence of a markedly reduced anoxic layer just below the surface of the sediment, non-sessile specimens of G. cydonium are fairly spherical and thus able to roll, dragged by slow circular currents. In addition, the usual association with the red alga Rytyphlöea tinctoria, which almost constantly forms a thick and continuous layer around the sponge, allows them to avoid contact with the substrate. The non-sessile specimens from Porto Cesareo inhabit sandy soft bottoms and are flattened. In such an environment, affected by moderate wave turbulence, the flattened shape widens the contact surface between the body and the substrate, thereby reducing the risk of stranding. The evident signs of abrasion, provided by scanning electron microscopy investigations, on both cortical spicules and outermost sponge surface suggest that sponges rub on the bottom. Sediment, epibiontic organisms, and the phanerogam leaves protect this sciaphilous sponge from high solar radiation, allowing the specimens to live in these shallow environments.     

Biochemical genetic divergence and systematics in sponges of the genera Corticium and Oscarella (Demospongiae: Homoscleromorpha) in the Mediterranean Sea

The sponge sub-class Homoscleromorpha is generally considered to include just two families, the Oscarellidae (without spicules) and the Plakinidae (with simple spicules). In May 1990, an unusual sponge was found deep inside a submarine cave in the western Mediterranean Sea. On the basis of externally visible characters this sponge appeared indistinguishable from the common plakinid species Corticium candelabrum Schmidt, 1862. However, on closer examination in the laboratory the new sponge proved to be devoid of spicules. Therefore, despite great morphological similarities to C. candelabrum, the new sponge should, by taxonomic convention, have been placed in the Oscarellidae. On the basis of other criteria, the similarities to C. candelabrum were great and the new sponge was at first considered to be conspecific. Thus, the taxonomic position of the new sponge and its relationship to C. candelabrum are highly confusing. It could be an aspiculate morph of C. candelabrum, or a new and undescribed related species or, lacking spicules, it could justifiably be placed in a different family (Oscarellidae). The relationship of the new sponge to C. candelabrum and also to two species of Oscarella (Oscarellidae) was assessed by the use of enzyme electrophoresis to estimate genetic divergence between species. It was found that the new sponge was reproductively isolated from sympatric C. candelabrum, with 6 of 16 loci proving diagnostic. Thus it is clear that the new sponge belongs to a different biological species. Surprisingly it was also found that, although this new species was fairly closely related to C. candelabrum (level of genetic identity, I0.47), the two Oscarella species were similarly closely related to C. candelabrum (I0.31 to 0.41) and rather less closely to the new species (I0.17 to 0.28). Indeed from genetic identity estimates, O. tuberculata is more closely related to C. candelabrum than it is to O. lobularis. It is concluded that all homoscleromorph sponges should be placed in the single family Plakinidae.     

Ultrastructural aspects of the symbiosis between two species of the genus Aphanocapsa (Cyanophyceae) and Ircinia variabilis (Demospongiae)

The association between two species of the genus Aphanocapsa (Cyanophyceae) and the sponge Ircinia variabilis has been studied by electron microscopy. A. feldmanni is localized in the mesohyl or inside the cells of the sponge, while the larger A. raspaigellae is located only in an extracellular position inside cavities of the mesohyl. Both algae differ from other symbiotic Cyanophyceae in having a normal cell wall. They are able to reproduce in symbiotic condition, but also undergo, in their various extracellular and intracellular positions, a massive process of disintegration. A large amount of algal material is dispersed in the sponge tissues, which is a confirmation, at the ultrastructural level, of trophic relationships in the symbiosis Aphanocapsa-Ircinia.     

Relationships between sponges and crabs: patterns of epibiosis on Inachus aguiarii (Decapoda: Majidae)

The association between sponges and the crab Inachus aguiarii Brito Capello, 1876 was studied by analysing the relationships between sponge distributional patterns on the crab carapaces and several morphological and biological crab parameters. Juveniles, mature females and mature males were differentiated on the basis of sex dimorphism and terminal pubertary moult. All three groups were fouled to different degrees by sponges. Percent sponge cover was related to carapace size only in mature males whereas mature females, all but one of which were ovigerous, were extensively covered regard-less of their size. It is proposed that some behavioural patterns unique to females, such as long resting periods in sponge-rich microhabitats, are responsible for these high sponge covers in females. Sponges showed two trends in the colonization of the carapace, leading to either a monopolizing or a sharing of the available carapace surface. The sponge species found on the carapaces studied are not obligatory epibionts of crabs, but are believed to reflect the sponge population characterizing the crab home range.     

Unique and species-specific microbial communities in Oscarella lobularis and other Mediterranean Oscarella species (Porifera: Homoscleromorpha)

Homoscleromorph sponges such as Oscarella spp. are characterized by unique morphological features, and Homoscleromorpha were therefore recently proposed as the fourth class of sponges. The microbiology of these sponges was mainly studied by electron microscopy while molecular studies are scarce. The aim of this study was to characterize the bacteria in Oscarella sponges using molecular tools. Denaturing gradient gel electrophoresis revealed distinct bacterial profiles in five Oscarella species and several color morphs of Oscarella lobularis. These profiles are characteristic of low microbial abundance (LMA) sponges. This was further confirmed by analysis of a 16S rRNA clone library from O. lobularis that yielded a low phylum-level diversity with dominance of Alphaproteobacteria. Bacterial communities in O. lobularis were very similar among different individuals (collected at the same site and time), five different color morphs, and specimens from different depths and locations, indicating a species-specific association. These results allow novel insights into the microbiology of the first known LMA sponge genus within the new class Homoscleromorpha.     

Apoptosis in marine sponges: a biomarker for environmental stress (cadmium and bacteria)

The marine demosponge Suberites domuncula is abundantly present on muddy sand bottoms, both in the open sea and in harbors. In the present study it is shown that exposure of S. domuncula to cadmium (CdCl₂) in concentrations ranging from 0.01 to 5.0 g ml⁻¹ for up to 5 d results in apoptotic fragmentation of DNA. Kinetics experiments revealed that after 24 h a significant increase of DNA fragmentation already occurred. Besides cadmium a second stimulus was identified to also cause apoptosis in this species, namely exposure to heat-treated Escherichia coli. In order to support the finding that both cadmium and E. coli induce apoptosis in the sponge, expression of the apoptotic gene MA-3 was studied. The cDNA, SDMA3, was isolated and found to be 2247 nucleotides long. The deduced amino acid sequence (M_r 50 765) shares high similarity with the corresponding mouse molecule. Like the mouse gene, the sponge MA-3 gene undergoes increased expression in response to apoptotic stimuli. While the specimens remained alive after treatment with cadmium, the sponges treated with E. coli died after approximately 12 d. The E. coli-treated animals started to form gemmules 10 to 12 d after addition of the bacteria. Hence, the process of apoptosis in sponges is triggered by two different pathways, one which is initiated by exogenous factors, e.g. heavy metals, and a second one, caused by endogenous factors, which leads to gemmule formation and a shift of the presumably immortal cells to mortal cells. The latter assumption is supported by the finding that during the process of bacteria-induced apoptosis, which results in the death of the specimens, the activity of the telomerase drops. It is concluded that the cells which appear to be immortal and telomerase-positive undergo apoptosis during the process of gemmule formation. In consequence cells not involved in the production of gemmules become mortal. Based on these data, it is proposed that apoptosis is a suitable biomarker in the bioindicator organism S. domuncula to monitor unfavorable environmental conditions, at least in this animal phylum. Received: 10 November 1997 / Accepted: 6 March 1998     

Morphological evidence for vertical transmission of symbiotic bacteria in the viviparous sponge<Emphasis Type="Italic"> Halisarca dujardini</Emphasis> Johnston (Porifera,Demospongiae, Halisarcida)

All stages of vertical transmission of symbiotic bacteria, from the penetration into oocytes to the formation of rhagon, were investigated in the White Sea (Arctic) representatives of Halisarca dujardini Johnston (Demospongiae). Small populations of free-living specific symbiotic bacteria inhabit the mesohyl of H. dujardini. They are represented by a single morphotype of small spiral gram-positive bacteria. Vertical transmission of symbiotic bacteria between generations in sponges may occur in different ways. In the case of H. dujardini the bacteria penetrate into growing oocytes by endocytosis. A part of the bacteria plays a trophic role for oocytes and the other part remains undigested in membrane-bound vacuoles within the cytoplasm. In cleaving embryos bacteria are situated between the blastomeres or in the vacuoles. In the blastula all bacteria are disposed in the blastocoel. The symbionts are situated in intercellular spaces in free-swimming larvae and during metamorphosis. Symbiotic bacteria do not play any trophic role in the period of embryonic and postembryonic development of H. dujardini. No signs of destruction and digestion of bacteria were revealed at any stage of development.Communicated by O. Kinne, Oldendorf/Luhe     

Genetic differentiation between morphotypes of the marine sponge Suberites ficus (Demospongiae: Hadromerida)

Sponges of three morphotypes of Suberites ficus (Johnston, 1842) were collected during February and March 1985 off the south-west of the Isle of Man, and were compared by using spicule size distributions and genetic allele frequencies of isozyme loci. The populations did not show any significant differences of spicule size or type, but could be easily differentiated into three separate species based on isozyme patterns. Samples of pale orange S. ficus growing on gastropod shells inhabited by hermit crabs (Pagurus spp.) were reproductively isolated from the redorange and the pale yellow colour morphs encrusting the bivalve Chlamys opercularis. These latter two colour morphs were genetically similar, but significant differences were observed at two of the 19 gene loci assayed. All the sponges studied were sympatric, and therefore the genetic differences, indicating reproductive isolation, are strong evidence for separate gene pools and, hence, that they are different species. The genetic identity between the two colour morphs of S. ficus on C. opercularis shells was 0.977, whilst between each of these and S. ficus on hermit crabs it was about 0.65. In all three species genetic variability was high, with mean expected and observed heterozygosity values per locus ranging from 0.17 to 0.36.     

A 3-year investigation of sexual reproduction in <Emphasis Type="Italic">Geodia cydonium</Emphasis> (Jameson 1811) (Porifera,Demospongiae) from a semi-enclosed Mediterranean bay

The reproductive cycle of Geodia cydonium in a semi-enclosed Mediterranean bay (Porto Cesareo, SW Apulia) was studied with monthly frequency over a 3-year period. The investigation was carried out by utilizing a technique consisting of tagging ten individuals with a PVC stick and cutting off, by means of a metal cork borer, small samples (cylinders about 5 cm³ in volume) from each of them for histological analysis. Sexual reproductive elements were detected in all individuals, but in the third year some specimens showed reduced reproductive activity or complete infertility. Spermatogenesis occurred in a short period (from June to August) whereas oogenesis lasted longer (from spring to late summer). Spermatic cysts occurred when the frequency of specimens with oocytes reached its maximum values. A relationship between water temperature and the onset of gamete differentiation was observed. G. cydonium is here confirmed oviparous and gonochoric with a sex ratio in favour of the females. However, in contrast with current literature on Porifera—which suggests that only a limited number of sponges, all belonging to the same species are sexually active—the sexual reproduction of this species involves all the examined individuals. This finding can be explained by the methodological approach used in this research, which differs from the traditional way of assessing the sponge reproductive cycle by analysing randomly collected specimens within a population. The study of a series of individuals over time represents a better “model technique” for investigating sponge sexual reproduction and the effect of environmental parameters on gamete differentiation.     

Pluri-annual study of the reproduction of two Mediterranean Oscarella species (Porifera, Homoscleromorpha): cycle, sex-ratio, reproductive effort and phenology

This study presents the phenology of two common Mediterranean sponges belonging to the genus Oscarella (Porifera, Homoscleromorpha). Oscarella tuberculata and Oscarella lobularis are two sibling species, dwellers of shallow benthic communities which tend to have distinct ecological behavior, respectively, euryecious and rather stenoecious. The comparative study of their reproductive cycle showed that both Oscarella species have a seasonal reproductive cycle with a successive phase duration differing from one species to another. In both species, there is a continuous oogenesis, with new oocytes appearing in spring, whereas the spermatogenesis generally starts later with the early warming of the sea. The embryonic development and the larval release are restricted to the warmest months of the year. We also observed a shift in the period of gametogenesis and larval emission depending on species and differences in their sensitivity to changes in thermal regime. It appears that an increase in seawater temperature can affect sex determination, with mainly a shift toward males in both species. Their reproductive efforts are variable in time, and can be in some cases influenced by the temperature regime. This is especially the case of O. lobularis which seems to be the most thermosensitive, its phenology responding significantly to changes in thermal regime, whereas O. tuberculata seems to be less sensitive and/or reactive. By detecting phenological changes among sponges, this study demonstrated the relevance of such monitoring to assess the possible biological response to climate change.     

Degrees of vacuolation of the absorptive intestinal cells of five Sagitta (Chaetognatha) species: possible ecophysiological implications

The cell organization and ultrastructure of the intestine have been studied in five species of Sagitta, one epiplanktonic (S. bipunctata Quoy and Gaimard, 1827), four mesoplanktonic (S. megalophthalma Dallot and Ducret, 1969, S. decipiens Fowler, 1905, S. minima Grassi, 1881, S. zetesios Fowler, 1905). The intestinal epithelium is composed of two ciliated cell types. The first (S-cell) principally occurs in the anterior intestine and represents typical secretory cells. The second (A-cell) occurs in the median and posterior intestine and displays ultrastructural features involved in absorption and intracellular digestion, that is, coated pits, endocytotic vesicles, cytoplasmic tubules, and two distinct types of digestive vacuoles. Whereas S-cells exhibit few ultrastructural differences among the five species, the vacuolar volume of the A-cells located in the mid portion of the intestine is higher in the mesoplanktonic species. In S. zetesios and S. megalophthalma, each side of the median intestine comprises several hypervacuolated A-cells visible in transverse section; their degree of vacuolation is inversely proportional to their number. The increased volume of the intestinal vacuoles is most marked in S. decipiens and S. minima; in both these species, each lateral side of the median intestine displays a single ultravacuolated A-cell. The possible ecophysiological implications of intestinal vacuoles are discussed in relation to patterns of the vertical distribution of several Sagitta species. The vacuoles are presumed to regulate buoyancy, enabling the mesoplanktonic species to make vertical diel migrations. Received: 22 January 2000 / Accepted: 28 July 2000     

Sexual reproduction,larval development and release in <Emphasis Type="Italic">Spongia officinalis</Emphasis> L. (Porifera,Demospongiae) from the Apulian coast

The reproductive cycle of the bath sponge Spongia officinalis L. has been studied over 1 year on 11 tagged specimens of different sizes (from 82 to 886 ml, in volume) from Ionian coasts of Apulia (SE Italy). According to literature data, the sponge is viviparous. All the monitored specimens showed sexual reproduction, even if the process usually involved small portions of the sponge tissue. Ten specimens were gonochoric (sex ratio 1:1), one specimen showed successive hermaphroditism, with alternate production of oocytes and spermatic cysts in the same reproductive season. Young oocytes occur almost all year round, whereas large mature eggs show a peak in October–November, concomitantly with the appearance of spermatic cysts. No relationships were observed between the sponge size and the presence of sexual elements within the range of the sponge size considered in this research. Embryo development occurs in patches of choanosomal tissue, which contain four or more elements. Cleavage is total and equal; it starts in November and in May leads to a solid stereoblastula, which develops into a parenchymella larva, from May to July. The stereoblastula lacks flagella and its surface is delimited by elongated cells well segregated from the internal ones. Parenchymella larvae are released from June to July, asynchronously, either at the individual or population level, with a few days of de-phasing. Up to 523 larval elements/48 h for a sponge specimen were counted. The free-swimming larvae are ovoid and uniformly flagellated. Flagella are longer at the posterior region, than on the rest of the larval body. Flagellated cells form a pseudo-stratified epithelial layer, delimiting the outermost larval surface, and are filled with electron-lucent vesicles showing a homogenous content. No choanocyte chambers, pinacocytes or skeletal elements are present in the newly released larvae.