Marine populations are distributed heterogeneously in space and time because of the diversity of habitats and the requirements of species life cycles. Human exploitation of these resources also varies as a function of space, time and the type of fishing activities performed. These three factors determine fishing strategy at different levels of integration. The purpose of this study was to describe and analyse, with respect to different time scales, the relations between the modalities of resource exploitation and the biological or demographic characteristics of the species involved. These investigations relate to the more general task of acquiring the basic knowledge needed for spatialised management of fishing effort. A fleet of trawlers from La Rochelle, operating in the Bay of Biscay, was studied over a 15-year period (1979–1993), which led to the development of a reference resource exploitation scheme for these vessels. The degree of stability of this scheme over time was determined from landing profiles of the 18 most important species fished (94% of landings). An annual cycle for the species composition of landings was determined by multiple factor analysis. Two factors account for more than 55% of the inertia of the data. The first, of biological origin, is closely related to the breeding activity of species and associated migrations between the coast and the open sea. The second is spatial in nature, corresponding to the distribution and availability of resources according to a bathymetric gradient. This organisational scheme persisted over the first 13 years, but showed signs of change toward the end of the study period. Analysis of multi-year trends indicated four periods marked by sustained levels of landings per unit of effort for some species (decreasing for sole and wedge sole, or increasing for Norway lobster, striped red mullet, rays and the smallspotted catshark). These changes are attributable to restrictions on resource access imposed on the fleet (regulations and/or competition among the fishing gears for occupation of space), variations in the abundance of traditionally fished populations (hake, anglerfish) and changes in the professional behaviour of fishermen.

An analysis of 273 samples taken at the filter screens of an estuarine power plant cooling water intake showed that herring and sprat were boldly associated with each other and formed mixed species schools. Herring-dominated schools were typically larger, more frequent and contained larger-sized individuals than schools dominated by sprat. The mean body length of both herring and sprat was biased as a function of their relative importance in the school suggesting that the fish schools were assorted by size. However, switching to a fish school dominated by the other species increased the individual body length variation relative to the group average. Diets of both species overlapped boldly but the feeding rate of sprat was significantly lower than that of herring. Our results suggest that during estuarine residency young sprat suffer the consequences of the school trap, possibly to profit from reduced predation risks in herring schools.

This study was carried out to investigate the effect of feeding diets containing herring or anchvoy oil, on flesh quality parameters of Atlantic salmon (Salmo salar). Two extruded experimental diets with the same basal composition but one coated with herring oil and the other with anchovy oil, were each fed during 24 weeks to salmon with an average initial weight of 1.8 kg. Salmon grew to a final weight of 3.9 kg. Growth, condition factor and biometric parameters were not affected by the dietary treatment. No significant differences were found for intramuscular fat. Monounsaturated fatty acid (MUFA) concentrations were highest in the group fed the diet containing herring oil, in both neutral and polar lipids, while the group fed the diet containing the anchovy oil showed a higher concentration of n-3 fatty acids in both fractions of intramuscular lipids. The n-3/n-6 ratio was higher in the neutral lipid fraction of fish fed the southern hemisphere oil, while no significantly differences were found for the polar lipid fraction. No differences were found on muscle α-tocopherol levels. Muscle homogenates from fish fed the anchovy oil showed the highest thiobarbituric acid reactive substances (TBARS) after 9 days of storage. However no differences were found between groups on the induced oxidation tests. It is concluded that the origin of the fish oil has no effect on growth perfomance, but there is a marked effect on fatty acid composition and susceptibility to lipid oxidation.

The combined effects of feeding time (morning and evening) and dietary fat concentration on feed intake, growth, body composition and lipid tissue distribution were examined in rainbow trout fed on demand. To that purpose, diets with low (LE, 6% lipid) or high energy concentrations (HE, 24% lipid) were used in four treatments that combined provision of the same (HE–HE or LE–LE) or different (HE–LE, LE–HE) diets at morning and evening meals. Digestible energy intakes of the LE–HE (229 kJ kg–1 d–1) and HE–HE (269 kJ kg–1 d–1) groups were significantly different. There was no significant difference in the amount of energy intake between the two meals of the day in any of the treatments. The main effect of dietary treatment was on lipid intake; fish in the HE–HE, HE–LE and LE–HE treatments ingested approximately 3, 2.2 and 1.8 times more lipid than those on the LE–LE treatment. Growth, weight gain and protein gain were not affected by dietary treatment, but lipid gain reflected lipid intake; fish in the HE–HE, HE–LE and LE–HE treatments gained approximately 2.6, 2.1 and 1.7 more lipid than did fish on the LE–LE treatment. Body composition was significantly affected by dietary treatment; whole body lipid content reflected lipid gain, and visceral lipid concentration was affected in a similar way to whole-body lipid. On the other hand, muscle lipid concentrations were similar in fish submitted to the HE–LE, LE–HE, and LE–LE treatments (< 4%), and were lower than in fish on the HE–HE treatment (ca. 5.5%). Consequently muscle lipid concentration was not directly related to lipid intake, because lipid intakes of fish on HE–LE and LE–HE treatments was higher than in those on the LE–LE treatment, while muscle lipid concentration was lower. Thus feeding fish with different diets in the morning and evening might have potential as a tool for manipulating lipid distributions and concentrations without major adverse effects on growth.

As a first step in assessing the viability of periphyton-based fish production in South Asian pond aquaculture systems, the effects of artificial substrates on development of periphyton and on water quality were evaluated. Earthen ponds (10 × 7.5 m) were provided with an artificial substrate constructed from poles of either bamboo, kanchi or hizol tree branches (1.0 m2 artificial substrate per m2 pond surface). Higher periphyton biomass, in terms of dry matter (DM) (4.9 mg cm–2) and chlorophyll a (11.5 μg cm–2) developed on hizol and bamboo, respectively. Periphyton ash content was higher on hizol (41%) than on the other two substrate types (29%). Protein content of the periphyton growing on bamboo (38% of ash-free dry matter (AFDM)) was 50% higher than that on the other two substrate types. Maximum periphyton productivities of 1.01, 1.38 and 1.03 g C m–2 d–1 were obtained for bamboo, hizol and kanchi substrates, respectively. Taxonomic composition of periphyton showed a rapid development of a relatively stable community with few differences between the substrate types. In total, 56 genera of algal periphyton and 35 genera of phytoplankton were identified. Based on a periphyton productivity estimate of 2.2–2.8 g AFDM m–2 d–1, periphyton alone can sustain an estimated fish production of 5000 kg ha–1 year–1 through the addition of a substrate area equivalent to 100% of the pond surface area.

This paper reports a comparative study of two cupped oysters, Crassostrea gigas and C. angulata, and their hybrids reared in the Marennes-Oléron Sound (France) under usual farming conditions. The relative proficiency of these species was characterized in terms of growth, survival and reproduction, three major concerns for oyster farming. Four populations of each taxon were sampled (C. gigas: Japan and France; C. angulata: Taiwan and Spain) and used as progenitors to produce six different progenies under common hatchery and nursery conditions (transfer of 5 to 6 g of seed oysters at sea and rearing in bags for 2 years). Mortality profiles differed greatly among the progenies. The C. angulata progeny of Taiwanese origin gave the poorest results, with 100% mortality during the second year. For the entire rearing period, the highest yield (+22%) was obtained with the C. gigas progeny of French origin. The sexual maturation index, lipid cycles and dry weight loss after spawning showed that spawning of C. angulata was delayed by about two weeks compared to that of C. gigas. A clear maternal effect was observed for the growth and reproductive characteristics of the hybrids. On the whole, C. gigas of French origin gave the best aquacultural results. However, severe episodes of C. gigas mortality have been observed in France in recent years, indicating that this species might not be entirely suitable for sustained oyster farming in this country.

French mollusc production is based mainly on the Pacific cupped oyster, Crassostrea gigas. Since 1991, annual mass mortality of juveniles has been reported during summer months. These recurring episodes concern professionals who fear that like Portugese oyster, C. angulata, C. gigas could in turn disappear following one of these epizooties. Previously, bacteriological analysis of moribund oyster juveniles yielded an isolate of a Vibrio splendidus biovar II strain, named TNEMF6. This isolate was demonstrated to be pathogenic to Crassostrea gigas spat by experimental challenge. To study the association between summer oyster mortality and presence of TNEMF6 cluster strains, Vibrionaceae fauna were isolated from infected spat along the French Atlantic coast between 1997-1998. Strains related to V. splendidus biovar II were selected. Comparison with TNEMF6 was performed by classical biochemical tests and polymerase chain reaction – restriction fragment length polymorphism (PCR-RFLP) of SSU rDNA, rpoD, and gyrB genes. Genomic similarities were confirmed by DNA/DNA hybridization. Only one strain out of 14, TNNIII7, was found to be closely related to the pathogenic bacteria. Neither the phenotypic nor the genotypic markers used in this study were able to distinguish pathogenic from non-pathogenic strains of the widespread V. splendidus. However, future genetic comparisons of TNEMF6 and TNNIII7 is likely to reveal genes involved in pathogenicity.

The ability to accurately quantify ovarian condition in cultured species of fish is of paramount importance in the understanding of their reproductive biology and the development of appropriate culture conditions. Here, we discuss stereological analysis, a mathematically proven technique that determines 3-dimensional topology and geometry from 2-dimensional thin sections, as a safer and more precise replacement for existing methodology.