Emerging understanding of the variability of natural coastal hypoxia is divided between two main hypotheses: the biogeochemical oxygen demand linked to locally‐driven organic matter decay or to supply of low‐oxygen waters by physical processes. The precise role of either mechanism in triggering hypoxia has remained elusive. A combined methane and oxygen high resolution year‐long hourly data time series in a coastal upwelling system suggests that these systems may be responding to a complex interaction between the two. The data show how anoxia is initially triggered by remote equatorial hypoxic waters after which it can be sustained by a local biogeochemical flux of exported production. Crucially, without a remote trigger the local forcing could not develop anoxic conditions because the physical flux of oxygen would be too high. The robustness of this interpretation is tested using a 10‐year data set (1994–2003) with seasonal and interannual scales of variability.

Abstract. As part of the Bonus-GoodHope (BGH) campaign, 15 N-labelled nitrate, ammonium and urea uptake measurements were made along the BGH transect from Cape Town to ∼60 • S in late austral summer, 2008. Our results are categorised according to distinct hydrographic regions defined by oceanic fronts and open ocean zones. High regenerated nitrate uptake rate in the oligotrophic Subtropical Zone (STZ) resulted in low f -ratios (f = 0.2) with nitrogen uptake being dominated by ρurea, which contributed up to 70 % of total nitrogen uptake. Size fractionated chlorophyll data showed that the greatest contribution (>50 %) of picophytoplankton (<2 µm) were found in the STZ, consistent with a community based on regenerated production. The Subantarctic Zone (SAZ) showed the greatest total integrated nitrogen uptake (10.3 mmol m −2 d −1 ), mainly due to enhanced nutrient supply within an anticyclonic eddy observed in this region. A decrease in the contribution of smaller size classes to the phytoplankton community was observed with increasing latitude, concurrent with a decrease in the contribution of regenerated production. Higher f -ratios observed in the SAZ (f = 0.49), Polar Frontal Zone (f = 0.41) and Antarctic Zone (f = 0.45) relative to the STZ (f = 0.24), indicate a higher contribution of NO − 3 -uptake relative to total nitrogen and potentially higher export production. High ambient regenerated nutrient concentrations are indicative of active regeneration processes throughout the transect and ascribed to late summer season sampling. Higher depth integratedCorrespondence to: W. R. Joubert (wjoubert@csir.co.za) uptake rates also correspond with higher surface iron concentrations. No clear correlation was observed between carbon export estimates derived from new production and 234 Th flux. In addition, export derived from 15 N estimates were 2-20 times greater than those based on 234 Th flux. Variability in the magnitude of export is likely due to intrinsically different methods, compounded by differences in integration time scales for the two proxies of carbon export.

Estuaries are dynamic transition zones acting as filters and transformers of nutrients passing from catchments to the sea. We propose an extension to an existing southern hemisphere model on nutrient dynamics in estuaries to include the relatively constricted, microtidal estuaries located along wave-dominated coasts in the region, specifically focusing on the limiting macronutrients nitrogen (N) and phosphorus (P) and on key processes, including physical (e.g. flushing, mixing and sedimentation), geochemical (e.g. flocculation), biochemical (e.g. remineralisation) and biological (e.g. primary production) processes. A simplified model of the physical states (primarily controlled by hydrological characteristics) is used as the basis for the qualitative model, where these are defined in terms of characteristic salinity-induced stratification of the water column, flushing time and the mouth condition.Four physical states are identified: a freshwater-dominated state, freshwater pulsed/recovery state, marine-dominated state and the closed mouth state. The states and their physical characteristics largely resemble that of the earlier model, except that the extended model reflects the dynamics of restricted inlets and introduces the closed mouth state. This model specifically explores the variation within southern African estuarine systems to better inform research and management programmes on the appropriate trophic, temporal and spatial scales at which uncertainties in ecosystem functioning need resolving. It can also be applied to other regions in the southern hemisphere, and even the northern hemisphere, with similar hydrological and estuarine geomorphological characteristics (e.g. the Mediterranean coast, the west coasts of North and South America, and the south-west and south-eastern coasts of Australia).

As part of the Bonus-Good Hope (BGH) campaign, ¹⁵N-labelled nitrate, ammonium and urea uptake measurements were made along the BGH transect from Cape Town to ~ 60° S in late austral summer, 2008. Our results are categorised according to distinct hydrographic regions defined by oceanic fronts and open ocean zones. Nitrogen uptake (&rho;N) in the oligotrophic Subtropical Zone (STZ) was dominated by &rho; urea, which contributed up to 70 % of &rho;N. High regenerated &rho;N in the STZ resulted in low <i>f</i>-ratios (<i>f</i> = 0.2). Size fractionated chlorophyll data showed that the greatest contribution (>50 %) of picophytoplankton (<2 μm) were found in the STZ, consistent with a community based on regenerated production. The Subantarctic Zone (SAZ) showed the greatest total integrated &rho;N (10.3 mmol m⁻² d⁻¹), mainly due to enhanced light and nutrient supply within an anticyclonic eddy observed in this region. A decrease in the contribution of smaller size classes to the phytoplankton community was observed with increasing latitude, concurrent with a decrease in the contribution of regenerated production. Higher <i>f</i>-ratios observed in the SAZ (<i>f</i> = 0.49), Polar Frontal Zone (<i>f</i> = 0.41) and Antarctic Zone (<i>f</i> = 0.45) relative to the STZ (<i>f</i> = 0.2), indicate a higher contribution of &rho;NO₃ relative to total &rho;N and potentially higher export production. Greater contribution of regenerated uptake to ∫&rho;N in the northern sector of the cruise resulted from increased ambient regenerated nutrient concentrations, shallow mixed layers in the north (~40 m) relative to the regions further south (~100 m). Higher &rho;∫N rates also correspond with higher surface iron concentrations. No clear correlation was observed between carbon export estimates derived from new production and ²³⁴Th flux. In addition, export derived from ¹⁵N estimates were 2–20 times greater than those based on ²³⁴Th flux. Variability in the magnitude of export is likely due to intrinsically different methods, compounded by differences in integration time scales for the two proxies of carbon export