This thesis examines the forcing and behaviour of oceanographic physical features, relevant to recruitment in fish populations, in the Gulf of Guinea Large Marine Ecosystem, on seasonal and interannual time scales. Remotely sensed sea-surface temperature (SST) data covering the period 1981–1991 was used to identify and describe a number of oceanographic features, including the Senegalese Upwelling influence, the Ghana and Côte d’Ivoire coastal upwelling, river run-off, fronts and the previously unrecorded observation of shelf-break cooling along the coast of Liberia and Sierra Leone during the boreal winter. Interannual variability in SST was observed on an approximate three year scale and an extended warm phase was noted between 1987 and 1991. Principal components analysis (PCA) was used to further investigate the variance structure of these SST data and this technique was shown to be able to accurately define boundaries of the Gulf of Guinea system and its constituent subsystems. River discharge data from throughout the Gulf of Guinea was also investigated using PCA, confirming the hydroclimatic regions identified by Mahé and Olivry (1999). The boundaries between these regions correspond closely to those identified between subsystems in the SST data, suggesting a degree of coupling between oceanographic and meteorological variability in the Gulf of Guinea. To further investigate this coupling, local climate data and global/basin scale indices were compared qualitatively and statistically with remotely sensed and in situ SST data and indices of interannual variability in oceanographic features. A new basin scale index was proposed as a measure of zonal atmospheric variability in the subtropical North Atlantic (SNAZI) and this was shown to be the dominant mode of climate variability forcing SST in the Gulf of Guinea. The implications of these results for fisheries recruitment dynamics are discussed.

The environmental dynamics of the northern Benguela upwelling system was investigated primarily using a time series of mean, weekly satellite images of sea surface temperature (SST), and observed trends were related to patterns in the reproductive activity and recruitment success of the region's pilchard and anchovy stocks. Three main methods were used to look at environmental trends: overall and seasonal trends in the system's behaviour were derived from maps of mean SST and SST variability, charts of mean weekly SST per 0.5° latitude were constructed to show spatio-temporal variability in coastal upwelling activity, and standardised principal components analysis was employed to identify the major modes of spatial SST variability and to quantify the evolution of the system's spatial structure through time. Important new findings include the observation that interannual variability itself is largely due to interannual differences in conditions during February, March and April, and that longshore intrusions of warmer water from both ends of the system and onshore intrusions of oceanic water account for a large proportion of the system's temporal variability in spatial structure. Comparisons of clupeoid reproductive activity with environmental trends further confirms that spawning activity and anchovy larvae abundance tend to be greatest when upwelling activity is reduced and conditions are suitable for the retention of eggs and larvae. As regards recruitment variability, pilchard recruitment was found to have a generally positive relationship with SST conditions, whilst anchovy recruitment appeared to be favoured by cooler conditions than pilchard. Different 'domains' in the relationship between the system's spatial structure, as revealed by the principal components analysis, and inshore SSTs were found to correspond closely with differing levels of anchovy and pilchard recruitment between 1982 and 1987. Should these relationships hold when tested over more years, they might, in the future, provide a basis for predicting clupeoid recruitment success in the region.

Previous research has hinted at changes in the migratory patterns of seabirds nesting in Scotland, including a decreasing number of Northern Gannets Morus bassanus wintering in the North Sea, and an increase in numbers of Northern Gannets and Great Skuas Stercorarius skua spending the winter off north-western Africa (NWA). Both species show increasing numbers of colonies in northern areas, including Norway and Russia. These seabird species move through the North Sea during autumn migration, and from there search for favourable wintering grounds mainly around Iberia, including the Atlantic coast of Portugal, the Bay of Biscay and the Gulf of Cadiz. By means of historic ring recovery data, provided by the British Trust for Ornithology, it was possible to establish that the number of records of adults of both species are increasing in recent years from NWA coasts, despite having to attend their nests in the colonies and, as a result, having limited time to migrate south. Differences were observed in ring recovery locations between years and months. The number of ring recoveries by month coincides with records from observation points along the coast of Western Europe. However, ring recovery data are limited and potentially biased. Using data loggers, it was possible to establish that both species are diurnal in habits during the entire winter period, showing noticeable differences in the times spent flying during the migration months (September-October) and during the wintering and breeding months (January and March respectively), and to confirm the increasing tendency to winter off NWA in recent years. Analyses of fishing landings, discard rates, and sea surface temperature data, show that food available to Northern Gannets and Great Skuas is increasing in NWA coasts where oceanographic conditions are stable; in contrast in the North Sea fisheries are decreasing and the sea surface is warming. Both species are apparently changing their migratory behaviour in order to face the constant changes in the abundance of food. Given the long life-span of Northern Gannets and Great Skuas, genetic changes can be ruled out of an explanation for the changes in migration behaviour, and the fact that the changes in winter distribution appear to be occurring within one generation of the birds. The winter distribution of Northern Gannets and Great Skuas may be due to an ideal free distribution over a wide range, in response to changes in the distribution of fish and the availability of discards.