Mesophotic reeffish assemblages (30-90 m depth) of the small and remote St. Peter and St. Paul's Archipelago (SPSPA), Mid-Atlantic Ridge, Brazil, were characterized using remotely operated vehicles. Ordination analyses identified distinct fish assemblages in the upper (30-50 m) and lower (50-90 m) mesophotic zones, the former characterized by high abundances of species that are also abundant at euphotic reefs (Caranx lugubris, Melichthys niger, Stegastes sanctipauli and Chromis multilineata) and the latter dominated by two mesophotic specialists (Prognathodes obliquus and Chromis enchrysura). Planktivores dominated fish assemblages, particularly in the upper mesophotic zone, possibly due to a greater availability of zooplankton coming from the colder Equatorial Undercurrent in mesophotic depths of the SPSPA. Turf algae, fleshy macroalgae and scleractinian corals dominated benthic assemblages between 30 and 40 m depth, while bryozoans, black corals and sponges dominated between 40 and 90 m depth. Canonical correspondence analysis explained 74 % of the relationship between environmental characteristics (depth, benthic cover and complexity) and structure of fish assemblages, with depth as the most important independent variable. Juveniles of Bodianus insularis and adults of P. obliquus and C. enchrysura were clearly associated with branching black corals (Tanacetipathes spp.), suggesting that black corals play key ecological roles in lower mesophotic reefs

No abstract

Reef fishes captured by recreational spearfishing on reefs of Bahia State, northeast Brazil. Biota Neotrop. 12(1): http:// www.biotaneotropica.org.br/v12n1/en/abstract?article+bn02012012012Abstract: Although recreational spearfishing is a growing activity, its impacts are poorly understood. This paper aims to present data on reef fishes captured by recreational spearfishing in the Bahia State, Northeastern Brazil. We analyzed 168 photos of spearfishing conducted in the reefs of this region between 2006 and 2008. A total of 1.121 fish belonging to 48 species were captured in three sub-regions of the Bahia coastline: Litoral Norte (LN), Salvador (SSA) and Baixo Sul (BS). The main species caught were: Scomberomorus brasiliensis (n = 191), Sphyraena barracuda (n = 153), Lutjanus jocu (n = 150) and Caranx bartholomaei (n = 141); these four species represented 56.5% of total captures. Over the sampling period, the highest values of catch rates per day were observed at BS, followed by SSA and LN. There were significant between-sites differences in the number of fish in the number of species caught. However, there were no significant between-years differences in the number of fish and numbers of species caught. The trophic classification of fish showed that carnivores (n = 623) were more frequently captured, followed by piscivores (n = 371), herbivores (n = 101) and invertivores (n = 26). The results showed that high sized, top predatory fishes, were the main targets, although other trophic levels were also captured. e Caranx bartholomaei (n = 141); estas quatro espécies representaram 56,5% do total de peixes capturados. Ao longo dos anos amostrados, foram registradas taxas mais elevadas de captura por dia em BS, seguido pelo SSA e LN. Foram encontradas diferenças significativas entre os locais estudados para a quantidade de peixes e o número de espécies capturadas. Entretanto, não foram encontradas diferenças significativas entre os anos para a quantidade de peixes e número de espécies capturadas. A classificação trófica de peixes mostrou que os carnívoros (n = 623) foram mais frequentemente capturados, seguidos de piscívoros (n = 371), herbívoros (n = 101) e invertívoros (n = 26). Os dados apresentados demonstram que grandes peixes predadores de topo são o alvo principal, no entanto outros níveis tróficos também são capturados. Palavras-chave: sobrepesca, recifes de coral, Salvador, Litoral Norte, Baixo Sul. Data AnalysisEach photograph corresponded to one fishery, and was considered as an independent sample. Spatial and temporal variations in total capture and richness (total number of species) were evaluated with nonparametric analysis of variance (Kruskall-Wallis test). Whenever differences were detected in the parameters, the Dunn's test was applied to the data in order to quantify and establish those differences.

No abstract