Search

Stage-dependent spatial distributions of anchovy Engraulis encrasicolus, sardine Sardinops sagax and round herring Etrumeus whiteheadi eggs over the western Agulhas Bank South Africa were examined from samples collected at a fine-scale (1.8 km) resolution using a continuous underway fish egg sampler (CUFES). Spatial structure in egg distributions was assessed using variograms for three developmental stage groups (early-, middle- and late-stage eggs), and indices of aggregation were also computed for those stage groups. No early- or middle-stage anchovy eggs were collected, but indices of aggregation showed that early-stage sardine eggs were more aggregated than those of round herring, suggesting that adult sardine have a more contiguous distribution during spawning than have round herring. Middle-stage eggs were equally aggregated, whereas late-stage eggs were aggregated differently in all three species. Fitted empirical variograms showed that early-stage sardine eggs have low small-scale variability (nugget effect) (⩽10% of total variance) compared with round herring eggs (⩾50% of total variance). The autocorrelation range of both those species extends to 3.7 km. Middlestage sardine eggs have a low nugget effect (⩽10%) compared with round herring eggs (⩽50%), with autocorrelation ranges of 7.4 km and 5.6 km respectively. Late-stage anchovy eggs have a lower nugget effect (⩽20%) than that of sardine (⩽50%) and round herring (⩾100%) eggs, with autocorrelation ranges of 3.7 km for all three species. The spatial structures of the eggs of the three species were heterogeneous at the sampling scale used, indicating differences in spatial organisation of the adults during spawning. The spatial variability of the eggs of the three species is not well captured using the sampling strategy during current ichthyoplankton surveys in the southern Benguela, in which stations are spaced 18.5 km apart. [ABSTRACT FROM AUTHOR]

Identifying and quantifying trophic interactions between organisms is crucial to understand the structure and functioning of food webs. However, many sources of variation are poorly known, such as seasonal and ontogenetic variations. These variations have been studied at several levels of organization, assemblage, species and individual in the Eastern English Channel (EEC) and the Southern North Sea (SNS) ecosystems. These variations were studied by analyzing the stable isotopes of carbon (,13C) and nitrogen (,15N), which provides information on resource and habitat use in a twodimensional space. These analyses were performed on several tissues because isotopic values of tissues reflect the diet over the period during which the tissue was synthesized providing dietary information over different temporal periods. These analyses were coupled with stomach contents analyses providing a short-term picture of feeding patterns. At the assemblage scale, coupling between benthic and pelagic habitats has emerged as an important feature of the EEC ecosystem, due to its shallow depth, as well as to the combination of two ecological processes. First, trophic interaction revealed trophic plasticity and resource partitioning. Second, changes in the composition of fish assemblage did not impact benthic-pelagic couplings as most dominant species were generalist within a period, allowing a complete use of all available resources. Further analyses were carried out on whiting, the dominant species in the EEC-SNS ecosystems in winter, and revealed seasonal and ontogenetic trophic changes for this species. Finally, at the individual scale, species tend to be generalist but composed of specialist individuals along the ,13C and ,15N axes. The species increase their niche mainly through an increase of the niche variation between individuals. This behavior may be a strategy to avoid competition and favored by the important benthic-pelagic coupling offering a wide diversity of resources for species. This study revealed the importance of considering ontogenetic and seasonal changes of trophic interactions. Informing these changes in ecosystem models would increase their ability to capture the complexity of marine ecosystems and inform fisheries management. Modeling these effects at several levels of organization is necessary to predict the effects of global change on ecosystem structure and functioning. [ABSTRACT FROM AUTHOR]