Most highly fecund marine fish show a steep decline in natural death rate from egg to first maturity, after which the natural mortality rate remains constant, or may even increase with age for old animals. Relatively few investigations have quantified early life-history mortality vectors for short-lived invertebrate stocks, but this overall picture is also true here for species with planktonic life stages such as penaeid shrimps, and for squids. If M decreases rapidly with age, one logical approach to demographic analysis is by subdividing the lifespan into intervals which increase in duration in proportion to the age up to the start of each interval. This time subdivision strategy is referred to as `gnomonic’. Earlier work (Caddy, 1990) showed that if a reciprocal mortality function applies with age, the product of the instantaneous annual rate of natural mortality and interval duration should be roughly constant for gnomonic intervals. This working hypothesis is shown to produce similar results to the reciprocal function for Mt, but allows a simpler approach to generating realistic life history Mt vectors in the absence of direct estimates of M for stock assessment. Values of a constant probability of death, G = Mt Δt, were used to generate vectors of M-at-age for a gnomonic series of intervals from hatching up to the mean parental age. The value of G is found by iteration that results in 2 survivors from the mean population fecundity by 1 year of age, under the assumption of steady-state population replacement for an unexploited stock. The natural mortality rate in the final, longest interval was assumed to correspond to the ‘constant adult M’ value used in stock assessment. Two extremes of reproductive strategy were chosen by comparison with data from for annual species of cephalopods or penaeid shrimps: cephalopods such as Sepia sp. and Rossia sp., with few, large yolky eggs (and/or parental care), occupy one extreme, and are contrasted with high fecundity penaeid shrimps and at least some Illex squids.

The first category has a low fecundity (130–150 eggs, and a K-selected reproductive strategy). Values of ‘adult M’ of the order of 1.0–1.3 are predicted for the last 60–80% of the annual life span. The high fecundity category (200 000 eggs or more) are opportunistic spawners such as many penaeids and some oceanic squids, and follow an r-selected reproductive strategy. An instantaneous value for pre-spawning M of the order of 2.8–3.4 is predicted for the same period mentioned above. Neither range of values falls outside those in the literature, for which a brief summary is presented.

An important research question relates to the order of magnitude of post-hatching mortality under population stability: it is suggested that irrespective of the specific model used for changes in M with age, this falls rapidly from an initial rate of some 50–75% per day for short-lived, high fecundity species in the 2 days following hatching, unless adult M values are much higher than above, and of the order of 25–40% for the low fecundity organisms over the same initial interval.

When discards data are included in fish stock assessments, it is generally assumed that all discarded animals die. However, there are fisheries in which a significant fraction of discards are able to survive. This case can be treated with a simple extension of the usual analytical model of exploited populations, enabling Virtual Population Analysis (VPA) to be used for parameter estimation. The paper examines the effects on estimates obtained by VPA when discards are allowed to survive, and validates a procedure for deriving partial fishing mortalities in multiple-fleet fisheries where discarding and survival rates are fleet specific.

Dans cet article, nous étudions les associations entre variables qualitatives. Nous présentons une approche basée sur l'utilisation du modèle loglinéaire. Appliqués aux associations entre captures par unité d'effort (cpue) d'aloses (Alosa alosa), coefficient de marée (cm) et débit fluvial (df), différents modèles mettent en évidence que les captures par unité d'effort augmentent proportionnellement à la valeur du coefficient de marée et qu'elles ont des évolutions conjointes moins structurées avec le débit fluvial.

Un système expert simple, reposant exclusivement sur des statistiques de pêche, a été élaboré afin d'estimer des migrations de poissons entre zones. Les flux de migration par trimestre sont calculés à partir de la comparaison de deux estimations distinctes des effectifs par zone et par âge, la première tenant compte du phénomène des migrations, la seconde non. La première estimation est obtenue en répartissant par zone les effectifs par âge issus d'une analyse des cohortes, la seconde est calculée période par période, à partir des effectifs présents dans chaque zone à l'âge précédent. Le modèle est appliqué au listao de l'Atlantique tropical Est. Les flux de migration estimés sont alors analysés, par groupe d'âge puis pour l'ensemble de la population. Bien qu'un certain nombre de limitations inhérentes au modèle existent, les résultats obtenus apparaissent valides par rapport aux schémas connus des migrations du listao.

Quelques aspects de la reproduction du listao (Katsuwonus pelamis) dans deux zones distinctes de l'ouest de l'océan Indien ont été étudiés de février 1989 à février 1994. A partir d'échantillonnages faits, soit à bord des thoniers senneurs sur du poisson frais (1 656 individus), soit à la conserverie de Maurice sur des thons décongelés (4 152 individus), les différents stades du cycle sexuel ont pu être précisés. La taille de première maturité sexuelle est atteinte à 41–42 cm (longueur à la fourche) chez les femelles et à 42-43 cm chez les mâles, correspondant à des individus âgés d'environ 1,5 an. L'évolution mensuelle de la proportion des différents stades macroscopiques de maturation des gonades montre que, quel que soit le mois considéré, il y a toujours plus de 70 % des femelles échantillonnées qui présentent des ovaires en fin de maturation (stade IV). Les variations du rapport gonado-somatique (RGS) montrent que la reproduction du listao a lieu toute l'année mais qu'il existe cependant plusieurs périodes d'activité sexuelle plus intense. L'observation histologique de 737 ovaires indique la présence de follicules post-ovulatoires (indicateur de ponte récente) pendant les 2 périodes de mousson, mousson de nord-ouest (novembre à mars) et mousson de sud-est (juin à août) tandis que la proportion d'ovaires présentant des follicules atrétiques est maximale aux 2 périodes d'inter-mousson (avril-mai et septembre-octobre). L'étude du sex-ratio montre que. pour l'ensemble de la période d'étude et pour les deux zones considérées, la proportion de mâles est significativement supérieure à celle des femelles. Ses variations mensuelles semblent indiquer une accentuation de la prédominance des mâles à chaque période de ponte plus intense.

Pour décrire la cinétique de l'ovogenèse et la stratégie de ponte chez la limande, Limanda limanda (L., 1758) les ovaires de 74 femelles capturées en baie de Douarnenez et en baie de Seine en 1990 et 1991 ont fait l'objet de prélèvements. Une échelle de développement ovocytaire en six stades a été utilisée pour classer les ovocytes. L'apparition d'alvéoles corticaux à la périphérie du cytoplasme est le premier signe de la vitellogenèse. Ils sont visibles dans les ovocytes dont le diamètre est de 120 μm, taille limite à utiliser lors d'études de la fécondité. Chez les femelles à ponte précoce un hiatus existe avant les pontes entre les ovocytes prévitellogéniques et vitellogéniques. Ceci permet de classer la limande parmi les espèces à fécondité déterminée. Chez ces femelles l'intensité de l'atrésie est faible et leur fécondité annuelle est proche de la fécondité potentielle. Par contre chez les femelles à ponte tardive tous les ovocytes vitellogéniques ne parviendront pas à maturité et ceux non encore émis à la fin du printemps dégénèreront.

The effectiveness of different gonadotropin releasing hormone analogues (GnRHa) and/or pimozide for inducing ovulation in Arctic charr was investigated at two different temperatures: 5 °C, a temperature suitable for spontaneous ovulation in Arctic charr and 10 °C, a temperature which inhibited ovulation in Arctic charr. At 5 °C all the different GnRH analogues tested were able to induce and synchronize ovulation. At 10 °C a sustained release preparation of D-tryptophan6 luteinizing hormone releasing hormone (D-Trp6LH-RH) at 20 μg.kg−1 and an acute release preparation of D-arginine6salmon GnRH (D-Arg6 sGnRH) at 100 μg.kg−1 were able to induce ovulation in 80% of the females within 15 days (versus 2% in controls). Pimozide alone or in combination with a low dose of D-Arg6sGnRH was able to induce some ovulation at 10 °C, suggesting that a dopamine inhibition of gonadotropin secretion could occur at 10 °C. Egg survival in groups receiving GnRHa at 5 °C was comparable to controls (73%) except in the group receiving GnRHa in combination with pimozide (45%). At 10 °C, egg survival was significantly lower than controls at 5 °C, except for the group receiving pimozide alone. At both temperatures, egg survivals of each spawn were negatively correlated with the plasma gonadotropin (GtH2) level of the reproducing females except in groups receiving a sustained release preparation of D-Trp6LH-RH.

Pacifastacus leniusculus has established hundreds of breeding populations in Britain since its introduction in 1976. One such population was established in the River Great Ouse, a major lowland river in mid-to eastern England, from 140 crayfish introduced in 1984. Estimates of growth, density and biomass of the population in 1992 are presented. Under laboratory conditions, the per moult increment (MI) and percentage per moult increment (PCMI) of crayfish were binomially correlated to their carapace length (CL), that is, the MI increased with CL and then decreased as crayfish grew old, and the PCMI decreased with increase in CL. Based on length frequency distribution data collected from the river in October and November and estimated by Bhattacharya's method, seven male cohorts while only four young female cohorts were identified because of insufficient data for old cohorts. The mean CL of each crayfish cohort was 16.6, 31.2, 41.6, 52.0, 57.8, 64.2, and 69.2 mm for 0+ to 6+ year old males respectively, and 15.4, 28.5, 39.7, and 49.8 mm for 0+ to 3+ females. There were no significant differences between growth rates of either CL or WW of males and females at age for 0+ to 3+ (no data of females > 3+ were available for comparison).

In a riffle, the densities and biomass of crayfish were significantly different between four domains with different bottom substrats and highest in the habitat with the highest density of cobbles and thus, greatest density of potential shelters. The mean densities and biomass of crayfish of all sizes in this riffle were between 10.m2 and 61 g.m−2 respectively in July and 15.m−2 and 53 g.m−2 in November. In a pool, mean density and biomass of crayfish > 30 mm CL, estimated by Fish-Ford multiple capture-mark-recapture methods, were 0.8 m−2 and 37 g.m−2 between May and August.

An image discrimination technique was developed to improve specific identification of some toxic Dinophysis cells (marine dinoflagellates involved in diarrhoeic shellfish poisoning), especially the "acuminata" and "sacculus" groups, which can be present at different ratios in natural sea-water samples collected during toxic episodes. This work was performed with image analysis software SAMBA (TITN Alcatel) using preserved cells directly observed through an inverted microscope and recorded with a video camera before further processing. All morphometric parameters used for discrimination of 7 different species or morphotypes of Dinophysis were tested with discriminant analysis. This study indicates that Dinophysis sp. and D. pavillardi seem well classified at the species level, whereas D. cf. acuminata and D. sacculus appear to be morphotypes of D. acuminata.