Lionfish (Pterois volitans and P. miles) have become a major concern in the western Atlantic and Caribbean since their introduction in the 1980s. Invasive lionfish can reach very high population densities on coral reefs in their invaded range, yet there are few data from their native range in the Indo-Pacific for comparison. We compiled data on the geographical distribution and density of Indo-Pacific lionfishes in their native ranges from published and unpublished underwater visual censuses and field collections. We found that lionfish in their native Indo-Pacific range are unevenly distributed, with higher densities in the Indian Ocean than in the Pacific. Lionfish densities increase significantly with increasing latitude, and are significantly higher in continental areas than around islands. In the Indo-Pacific, lionfishes are found not only on reefs but also on soft bottoms and in nearshore habitats such as seagrass beds and mangroves, and near estuaries. Native lionfish can be found at depths greater than 75 m. Because lionfish can be cryptic and secretive, we estimate that only ~1/8 of Indo-Pacific lionfishes are detected during general underwater visual censuses. In the Pacific Ocean, the relative abundance of lionfish in the catch of reef-fish larvae is of the same order of magnitude as the relative abundance of adult lionfish within reef fish assemblages. Overall the observed densities of lionfishes in the Indo-Pacific are much lower (max. 26.3 fish ha -1 ) than the densities reported in their invaded Atlantic range (max. 400 fish ha -1 ). We found no effects of fishing or pollution on the densities of lionfishes.

No abstract

A total of 212 carnivorous coastal fish species from New Caledonia, represented by 7335 individuals, were analysed for their diet. Fifty two prey items were identified and later grouped into broader taxonomic categories refered as "prey types". For each fish species 6 biological traits were defined: maximum adult size, major biotope, schooling behaviour, home range, nycthemeral behaviour, degree of crypticity. A general linear model was fit to the diet data taking into account these traits, and depth of capture and fish family. This model was applied to the average number of prey types/stomach and to the volume of 5 prey types: nekton, crustaceans, molluscs, echinoderms, worms. A second analysis was performed on the effect of observed size on diet composition, taking into account these traits and taxonomy. This analysis was restricted to species with at least five individuals, representing 113 species from 33 families. More detailed information is given for the three major families, Serranidae, Lutjanidae and Lethrinidae and intra-family variations are illustrated for Lethrinidae. All the factors tested had significant effects on diet, fish size and taxonomy being the two major factors. The five prey types analysed in detail displayed marked differences according to the factors studied. In particular nekton increased in importance with fish size, whereas crustaceans decreased and molluscs presented a dome shaped relationship. Nekton and crustaceans made the bulk of the diet of most species, with molluscs being at times important. Echinoderms and worms were never a major food item and were eaten only by a restricted range of species.Résumé -Régime alimentaire de poissons carnivores des récifs coralliens de Nouvelle Calédonie. Un ensemble de 212 espèces de poissons carnivores côtiers de Nouvelle Calédonie, représentées par 7335 spécimens, ont été analysés pour étudier leur régime alimentaire. Cinquante-deux types de proies ont été identifiées et regroupées par la suite dans des catégories plus vastes ou « type de proies ». Pour chaque espèce, six caractéristiques biologiques ont été définies : taille adulte maximale, principal biotope, grégarité, domaine vital, comportement nycthéméral, niveau de crypticité. Un modèle linéaire général a été ajusté à ces données, en prenant en compte ces caractéristiques ainsi que la profondeur de capture et la famille. Ce modèle est appliqué au nombre moyen de proies par estomac ainsi qu'au volume de cinq types de proies : necton, crustacés, mollusques, échinodermes et vers. Une seconde analyse porte sur les effets de la taille observée sur la composition de l'alimentation, prenant en compte ces caractéristiques vitales et la taxonomie ; cette analyse est restreinte aux espèces comportant au moins 5 spécimens, ce qui représente 113 espèces provenant de 33 familles. Une information plus détaillée est fournie pour les trois familles principales : Serranidae, Lutjanidae et Lethrinidae et les variations au sein d'une même famille sont illustrées pour les Lethrinidae. Tous les facteurs ...