Many symbiotic organisms are narrowly distributed on one or a few host species. These associations are intriguing, as they invite the development of hypotheses regarding the pattern and process of speciation and serve as laboratories for the testing of methods of phylogenetic reconstruction (Kraus, 1978; Futuyma & Slatkin, 1983; Stone & Hawks worth, 1986). The evolution of host-specificity in the sea may be expected to be severely constrained by the difficulty of achieving reproductive isolation in taxa whose gametes are freely released into the water column and/or whose larvae are potentially widely distributed (Scheltema, 1977). Yet this difficulty may well be overestimated, given the recent demonstrations of limited gamete (Pennington, 1985; Yund, in press) and larval dispersal (Knight-Jones & Moyse, 1961; Ryland, 1981; Olsen, 1985; Jackson & Coates, 1986; Grosberg, 1987). Indeed, if gamete and larval dispersal are as limited as has recently been contended (Jackson, 1986), local isolation of populations may be a routine occurence, offering repeated opportunities for speciation.

Between 1980 and 1985 ninety-seven stations were sampled by Smith-Mclntyre grab from the offshore northern section of the North Sea. Four hundred and nine infaunal species were identified from the 76 selected macrofaunal stations. The number of species per station varied from 25 to 80 with a maximum abundance of 9,600 individuals m−2. The biomass ranged from 0.13 to 18.86 g dry weight m−2. At most stations, however, biomass varied between 1 and 4 g dry weight m−2. Diversity and abundance were highest in the 120–140 m zone, characterised by fine sand containing variable amounts of silt. The highest biomasses were recorded in two areas; firstly where stronger currents predominate and the sediments are coarser (east of Shetland and west of the Norwegian Trough), and secondly in the fine sandy deposits of the centrally located area. In the silty sediments (Fladen Ground and smaller depressions) there was a predominantly subsurface deposit-feeding community, whereas in the coarser area east of the Shetlands carnivores predominated. Over the remaining area surface deposit feeders were dominant.

The demographic parameters of a population (the number of age-classes present; growth rates; mortality as a function of age and recruitment levels) are of considerable interest to marine biologists. If individuals can be aged from growth rings in their hard parts, then the estimation of demographic parameters is relatively straightforward. If this is not possible, the next best alternative is to tag or mark individuals and use data on the recapture of these to give the information required. For many marine invertebrates, neither of these options is practical and we must resort to estimating the demographic parameters by making assumptions about recruitment and the size variation between individuals of the same age and then infer the age structure of the population from its size structure. This was first done by Petersen (1891) who interpreted each mode on a size/ frequency histogram as representing a single age-class. More recently, extensive use has been made of methods which assume that the sizes of individuals of the same age will be normally distributed. The size/frequency histogram can then be decomposed into a number of normal distributions, each of which represents a single age-class. This can be done graphically (Harding, 1949; Cassie, 1954; Bhattacharya, 1967) or with computerbased numerical methods (Macdonald & Pitcher, 1979). The graphical methods seem to be the most popular and are frequently taught to undergraduate students. The same methods can be used to dissect a size/frequency distribution into components other than age-classes (Harding, 1949), but the principles are the same.

The eggs of the polyopisthocotylean monogenean Plectanocotyle gurnardi (van Beneden & Hesse, 1863) Llewellyn, 1941, from the gills of gurnards (Triglidae) develop and hatch readily when maintained in the laboratory. However, the impression was gained that some or all of this hatching might be the consequence of inadvertent shadowing or accidental mechanical disturbance of the eggs by the investigators. Since similar stimuli are known to induce hatching in some other monogeneans (see Kearn, 1982; Whittington & Kearn, 1988) and since such stimuli are likely to be generated by the activities of potential hosts, it was decided to embark on a more detailed investigation of hatching in P. gurnardi and to attempt to learn more about the behaviour of gurnards. During these investigations, the availability of freshly hatched larvae of P. gurnardi presented an opportunity to study their behaviour and to expand our knowledge of their anatomy.

Records of fish captures on power station intake screens were used to analyse the structure of the English and Welsh inshore fish community. The study was undertaken as part of a programme to predict fish captures at future power station sites. It was found that 118 of the 122 fish species known to live inshore were recorded from the screens of 12 coastal power stations. The maximum number of species at one site was about 80, achieved at fully marine localities along the English Channel coast. This number declined with increasing latitude and decreasing salinity.

Fish subdivide their use of the inshore habitat both spatially and temporally and the proportion of species complement within the major subdivisions of the habitat was remarkably similar at all sites. On average, 28 species were present simultaneously and these would include 4 pelagic, 8 demersal and 16 benthic species.

The horse-mussel Modiolus modiolus (L.) is a bivalve molluse inhabiting coastal marine environments down to about 200 m, where it may occur in very large communities (Wiborg, 1946; Tebble, 1966; Comely, 1978). It can also be found low in the intertidal zone where it lives in rock pools or in the hold-fasts of Laminaria (Wilson, 1977; Davenport & Kjørsvik, 1982). It is a boreal species. In the Atlantic it ranges from the White Sea to the Bay of Biscay, off Iceland, the Faroes and down the east coast of North America to North Carolina. In the Pacific it occurs from the Bering Sea to Japan and California (Wiborg, 1946). Comparatively little work has been done on Modiolus modiolus, probably because it is of little commercial importance and lives predominantly subtidally in relatively inaccessible environments.

To understand the working of the marine pelagic ecosystem one needs to study the logical frame of the complex food web. This study is much easier in an enclosed water column than in the open sea where biological interactions are often concealed by physical events.

Descriptions of the polychaete Nephtys hombergii Savigny, 1818 have encompassed a wide range of apparently intraspecific variation in taxonomic features such as the segment on which interramal cirri first occur, and the form and size of pre- and postsetal lamellae (Fauchald, 1963; Woolf, 1968; Hartmann-Schröder, 1971). During the examination of museum collections and of extensive ecological collections from the North Sea (Rainer, 1990) and from Danish waters, it was found that the name N. hombergii has been applied to three morphologically similar species, viz. N. hombergii (s.s.), N. assimilis Örsted and N. kersivalensis Mclntosh. Both are locally abundant species that may occur with N. hombergii. This paper designates a neotype of N. assimilis, provides a full description of N. assimilis and N. kersivalensis, and removes these species from synonymy with N. hombergii.

More than 65% of the demersal fish and shellfish caught by British vessels are landed at Scottish ports, most of the catch being taken from waters north of 56°, i.e. in the northern North Sea. The major resource supporting these fisheries is the benthos, but the distributions of both infaunal and epifaunal benthic invertebrates in the North Sea are poorly documented. Moreover, the environmental factors determining the composition and structure of North Sea benthic assemblages are not well understood. There have been several limited surveys of the benthos (Stephen, 1934; Ursin, 1960), and recently Dyer et al. (1982, 1983), Cranmer (1985) and Cranmer et al. (1984) attempted a more quantitative and systematic survey. Various schemes have been proposed to explain observed regional differences in types of benthic assemblage, usually based on the biological characteristics of water masses, and / or a limited and fragmentary knowledge of the sediments of the North Sea {e.g. Glemarec, 1973; Dyer et al, 1983).

Bacteria have been isolated into pure culture from the gills of Lucinoma borealis, Myrtea spinifera, Thyasira flexuosa and T. sarsi collected from several British and Scandinavian coastal sediments. It is proposed that these bacteria are symbiotic with the animals. The bacteria were either autotrophs, obtaining energy from oxidation of inorganic sulphur compounds, or methylotrophs using methanol or methylamine as the source of carbon and energy. The latter could also oxidize thiosulphate to tetrathionate and precipitate elemental sulphur. The pure cultures have been characterised in some detail, showing that (1) their DNA base composition falls in the range 47–58% guanine + cytosine; (2) although all grow well at the natural environment temperatures of 8–15°C, their optima lie between 30 and 35°C; (3) all produce copious amounts of extracellular slime when grown at low temperatures; (4) the autotrophic isolate is facultatively heterotrophic, containing a repressible ribulose bisphosphate carboxylase, and excreting up to 30% of its carbon dioxide fixation into the culture medium; (5) the methylotrophs include some containing high activities of hexulose phosphate synthase for assimilation of formaldehyde by the ribulose monophosphate cycle; others may have novel assimilatory metabolism. The possibility is discussed that environmental factors may determine whether a methylotroph-animal or autotroph-animal symbiosis is established rather than the symbiont type being an invariable species-specific character.

Feeding by marine invertebrates affects dietary lipids as they pass through the gut (Volkman et al, 1980s; Tanoue et al, 1982; Prahl et al, 1984a, b,1985; Neal et al, 1986; Harvey et ah, 1987,1989). Not only do animals appear to alter the dietary lipids but they also contribute their own lipids to the egested material. Faecal pellets are thus likely to have a lipid composition which has contributions from the ingested food material, the animal itself and the microbial populations residing in the animal's alimentary system.

Predation is now considered the main cause of mortality in larval and juvenile fish (Hunter, 1984) and is therefore the most important factor controlling recruitment to the adult population. Marine fish larvae are prey for many different types of predator including medusae, crustaceans and larger fish. When predatory attacks are sensed both adult and larval fish may respond by making a 'C-start', a very fast, simultaneous contraction of the trunk musculature that deforms the body into a C-shape within 20 ms (Eaton & Hackett, 1984). This startle response is mediated by the Mauthner cells, a pair of prominent neurones in the hind brain or by other reticulo-spinal cells located in the same region. As a result of reciprocal inhibition and decussation of the cell axon, stimulation on one side of the body results in contraction of all the muscle on the contralateral side.

Experiments on intact retinae from the eyes of the deep-sea fish Diretmus argenteus show that the region of the retina that receives light from above and possesses very long rods has different photosensitive pigments from the region that receives light from below. In both of these regions the retina has several banks of rods. The optical densities of the photosensitive pigments in the retina at the peaks of absorption are almost the same (about one) in these two regions. However, the region of the retina with very long rods also contains a stable yellow pigment that absorbs heavily in the blue and near ultra-violet (optical density about three at the wavelength 390 nm). If, as seems certain, this stable pigment is largely in the long rods it will filter the light reaching the other layers of rods and act, for this region, as yellow lenses do for whole retinae in other fish. The lens of Diretmus was shown to be transparent in the visible and near ultra-violet.

Comparison of early shell morphological features of six mytilids from the north-western Atlantic revealed differences useful for species identification and classification. Brachidontes exustus, lschadium recurvum, Geukensia demissa, Amygdalum papyrium, Mytilus edulis, and Modiolus modiolus larvae and post-larvae were cultured in the laboratory. Scanning electron micrographs of the shell and hinge during early ontogenetic stages showed that all species had a long provinculum with taxodont dentition. In addition, provinculum length and number of teeth increased during the larval period in the six species. The small, numerous provincular teeth of Mytilus edulis and the bold, comparatively few provincular teeth of Amygdalum papyrium clearly differentiated these two species. Most species had a low umbo, round posterior margin, and more pointed anterior margin, although Amygdalum papyrium was distinguished by a high, prominent umbo. Distinction of Geukensia demissa and lschadium recurvum larval shells was difficult due to similarity in their shapes and hinge dentition. However, discriminant analysis using larval shell length, shell height, provinculum length, and number of teeth aided in classification of these and other sympatric species.

Lincoln (1979) listed four species of Hyale Rathke from British and Irish waters. A fifth species Hyale grimaldii Chevreux has recently been collected from among goose barnacles, Lepas anatifera L., attached to drift timber washed ashore on a rocky beach in Co. Galway, west coast of Ireland. This record was referred to previously in an abstract by McGrath (1984) but no details of the occurrence or material collected was given. An earlier unconfirmed record of H. grimaldii from Scotland is listed among the Reid collection by Sanderson (1973). H. grimaldii was originally described by Chevreux (1891) from two males found among algae attached to drift timber at sea in the Atlantic, north east of the Azores. Subsequently, the species has been recorded on a number of occasions in the Mediterranean (Chevreux & de Guerne, 1893; Cecchini, 1928; Ruffo, 1946; Berner, 1976) and from the Azores, on a coconut cast ashore (Mateus & Afonso, 1974). All records of the species are from objects floating at sea or washed ashore.

The demersal fish populations of the slopes of the Rockall Trough (250 to 3000m) have been sampled by a number of different types of trawl (Gordon & Duncan, 1985; Gordon, 1986). The diets of most species have been described (Mauchline & Gordon, 1980, 1983a,b, 1984a,b,c) and aspects of the trophic ecology discussed (Mauchline & Gordon, 1985, 1986). Rays were relatively uncommon in the early sampling but the more recent use of a Marinovitch semi-balloon trawl with mud rollers and chain on the footrope has increased catches. This paper describes the distribution and diet of 12 species of ray from the slopes of the Rockall Trough.

Vance (1973a, b) argued that among the possible range of developmental strategies available to marine invertebrates, only the extremes of obligate planktotrophy and obligate lecithotrophy are evolutionarily stable. Vance's model, relating reproductive 'efficiency' to egg size (in terms of energetic content), predation rate, and prefeeding (lecithotrophic) vs feeding (planktotrophic) larval periods, has been a source of much discussion and debate since its inception (e.g. Underwood, 1974; Vance, 1974; Christiansen & Fenchel, 1979; Obrebski, 1979; Williams, 1980; Jablonski & Lutz, 1983; Strathmann, 1978, 1985; Todd, 1985). Subsequent publications have continued to dwell mainly on potential selective factors and the extremes of larval developmental type (i.e. obligate planktotrophy or obligate non-pelagic lecithotrophy). For the most part, these investigations have ignored questions concerning how a transition from one larval type to another would be accomplished in morphological and functional terms. Nonetheless, the consensus persists that small eggs and planktotrophy are the primitive (or ancestral) condition, and that lecithotrophy is the more advanced evolutionary derivative (see Strathmann, 1978,1985).

The disposal of wastes to sea is regulated internationally by the Oslo and Paris Conventions, which were ratified by the United Kingdom in 1975 and 1978 respectively. Under these Conventions the direct disposal of certain substances, for example mercury and cadmium compounds, is prohibited, unless they are present in wastes as trace constituents. Of particular concern in the case of mercury is its possible incorporation in the aquatic food chain and transfer to man via marine foodstuffs. There are many examples of accumulation of mercury by fish in contaminated areas e.g. Liverpool Bay (Norton & Murray, 1983; Franklin, 1987); Haifa Bay (Hornung & Krumgalz, 1984); Minimata Bay (Fujiki, 1980) and in river systems such as the Ems (Essink, 1980,1988).

Previous studies with the marine gastropod Littorina littorea have shown that cadmium is concentrated in the digestive gland and that cytosolic metallothionein-like proteins sequester a large proportion of the accumulated burden. In the current investigation, an electron microscope cryo-preparation method and X-ray microanalysis have revealed that some cadmium, in association with sulphur, is localised in the region of the limiting membrane of the mineralised granules which detoxify certain metals as an insoluble phosphate salt. It appears, therefore, that there could be interaction between the sulphur and phosphate systems for binding metals within the basophil cells of the digestive epithelium.