Surprisingly little is known about the influence of salinity on embryo development in crustaceans despite its obvious potential importance to species which breed in inshore estuarine waters. Evidence from the literature is that embryo salinity tolerance may differ from that of later stages in the life history (e.g. Kelley & Burbanck, 1976; Rosenberg & Costlow, 1979; Hartnoll & Paul, 1982; Jones & Simons, 1982; Wear et al., 1986) and such differences may impose restrictions on distribution. Wittmann (1984) referring to mysids, emphasised that an understanding of incubation period and factors influencing it "appears to be a key factor for the understanding of variations in the length and timing of the breeding season, age of maturity, frequency of broods, numbers of young per brood, egg size, and adult body size".

Samples of phytoplankton from nets towed in the surface waters off Plymouth over a period of 24 years were subjected to a correspondence analysis. There was a high species diversity throughout the year, and a strong seasonal cycle. A total of 104 species was divided into eight groups showing different seasonal and year-to-year occurrence. Over the 24 years the seasonal pattern of species changed in a fairly systematic way and there were two periods (1968–70 and 1983–85) when the changes were most marked. These patterns could be related to changes in the weather patterns, and were similar to those found by other workers in other groups of organisms and in other areas.

Macoma carlottensis Whiteaves, 1880 is abundant (up to approximately 350 nr2) in subtidal infaunal communities on the continental shelf of western United States (Lie & Kelly, 1970; Word, 1979) and in British Columbia fjords (Ellis, 1969). Of thirteen different species of Macoma identified from the coastal waters of southern British Columbia, M. carlottensis is the most tolerant to variations in water depth, sediment particle size, current exposure, and salinity (Dunnill & Ellis, 1969). Observations of the species maintained in sediment trays in our laboratory suggest thatM. carlottensis is an infaunal, primarily sub-surface deposit feeder which generally lies at a depth of 1-4 cm with its left valve downwards. Filter feeding was observed in some instances as indicated by extension of the siphons from 0-5-3 cm above the substrate.

The long-standing hypothesis (Thorson, 1946, 1950; Mileikovsky, 1971; Jablonski & Lutz, 1983) that deep-sea invertebrates should reproduce by direct development or by non-feeding (lecithotrophic) larvae is beginning to fall in the light of recent data. Traditional reasoning maintained that planktonic food should be limiting at great depths, and that microscopic, ciliated larvae should be incapable of migration to the euphotic zone. However, in recent years, planktotrophic larvae of two deep-sea gastropods have been collected in surface waters, and planktonic larval development has been inferred from shell morphology and chemistry in several other species (Bouchet & Warren, 1979; Killingley & Rex, 1985). Planktonic larvae have also been collected in the water column of the deep sea (Berg et al., 1985; Smith, 1985; Berg & Van Dover, 1987), particularly near hydrothermal vents. Relatively diverse benthopelagic plankton populations relying primarily on suspended detritus for food are now known from the benthic boundary layer of the deep sea (Wishner, 1980a, b; Gowing & Wishner, 1986). Thus, it is increasingly apparent that planktotrophy may be a common option for deep-sea larval development.

Egg-producing Lernaeocera branchialis (L.) are usually found in the gill cavity of cod (Gadus morhua (L.)) and whiting (Merlangius merlangus (L.)) (Kabata, 1979). The early larval development of the copepod includes two free-swimming nauplius stages followed by the first infective stage, the copepodite (Sproston, 1942). On an intermediate host, usually flounder (Platichthys flesus (L.)), the copepodite moults into a chalimus larva which is attached by its frontal filament to the gill tips of its host. After four successive chalimus stages, the parasite reaches maturity and copulation occurs on the flatfish host.

The periodicity of microgrowth patterns observed in thin sections and acetate peel replicas of radial sections of bivalve shells are well documented for both intertidal and subtidal species (Pannella & MacClintock, 1968; Rhoads & Pannella, 1970; Evans, 1972, 1975; Kennish, 1980; Richardson et al., 1979 et seq.; Richardson, 1987a, b, 1988, b). In many species daily or tidal lines are well preserved in the shell, for example, the hard-shell clam Mercenaria mercenaria (Pannella & MacClintock, 1968), the cockle Cerastoderma edule (Richardson et al., 1979) and the clam Tapes philippinarum (Richardson, 1987a, 1988b), while in others they are poorly preserved or lacking. Such an example is the common mussel Mytilus edulis (L.) where fine growth lines seen in acetate peel replicas have previously been reported to be poorly preserved, with long continuous sequences of such lines seldom occurring (Lutz, 1976). The angular relationship of growth lines to the growing margin and to the crystalline elements of the shell structure are presumed to be the factors influencing the resolution of growth increments (MacClintock, 1967; Pannella & MacClintock, 1968). Pannella & MacClintock (1968) suggest that the larger the angle between depositional surfaces and the structural elements of a shell layer the more evident will be the growth increments. Both oysters and mussels have an unfavourable angular relationship between growth increments and the margin of the shell (Pannella & MacClintock, 1968) and this may explain why the growth lines are difficult to observe. The present paper describes improved methods used to examine the internal micro478 growth patterns in the shell of the common mussel Mytilus edulis and reports on experiments to investigate the periodicity of the patterns.

Perkinsus marinus, a protozoan parasite of oysters, is an important cause of oyster mortality in the Gulf of Mexico and along the south-eastern coast of the United States (Hofstetter, 1977; Quick & Mackin, 1971). Infections are patchily distributed on many reefs, often with uninfected oysters adjacent to infected oysters. The primary mechanism of transmission from one oyster to another is through the water (Ray, 1954; Mackin, 1962; Andrews, 1965). Because dilution rapidly reduces the number of infective elements below the dosage required to initiate new infections (Andrews, 1979), transmission is most efficient over very short distances and declines rapidly within a few metres of the source.

Many field observations have been made on predation by a variety of fish species on the demersally spawned eggs of the Atlantic herring (Clupea harengus L.) (Dunn, 1898; Bowman, 1922; Hempel & Hempel, 1971; Toresen, 1985). Field observations on the predation of herring larvae by fish are comparatively rare although jellyfish medusae have been identified as larval predators (Van der Veer et al, 1983; Moller, 1984). Recently several experiments have been carried out on the predation of medusae and other planktonic organisms (Bailey & Batty, 1983) and of fish on herring larvae (Christensen, 1983; Fuiman & Gamble, 1987). Despite the lack of field observations Hunter (1984) considered that predation, most probably by fish, was the major cause of mortality in fish larvae.

Occasional small specimens of herring have, since at least 1875, been taken by Plymouth anglers in winter under harbour lights. In 1937 larger herrings appeared, and a specialised rod fishery developed which continued, apart from the war years, until 1953, when it collapsed suddenly.