Transport and retention of suspended particulate matter and bacteria in the Humber-Ouse Estuary, United Kingdom, and their relationship to hypoxia and anoxia

Data are presented on dissolved oxygen (DO) concentrations and their relati onship to salinity, suspended particulate matter (SPM), concentrations, and the turbidity maximum in the Humber-Ouse Estuary, United Kingdom, during s ummer 1995. Measurements in the upper Humber during March 1995 showed DO in the range 82% to 87% of saturation. Suspended particulate matter concentra tions were <5000 mg l(-1) and salinity was in the range 0.5 to 12. In contr ast, a pronounced DO sag occurred in the upper reaches of the Ouse during m edium and spring tide, summer conditions. The DO minimum was essentially an anoxic level and was associated with the location of the turbidity maximum , at salinities between about 0.4 and 1.5. SPM concentrations at 1 m beneat h the surface reached 25,000 mg l(-1) in the turbidity maximum, between abo ut 20 km and 40 km from the tidal limit. Suspended particulate matter conce ntrations were much lower at neap tides, although dense suspensions of SPM (>60,000 mg l(-1)) occurred within 1 m of the bed in the turbidity maximum region. A spring-neap record showed a dramatic and tidally controlled decre ase in DO at very low salinities as the tides progressed from neaps to spri ngs. An anchor station located down-channel of the turbidity maximum showed that about 95% of the variance in DO, which varied from 28% at low-water s lack to 67% at high-water slack, could be explained in terms of salinity va riation. At the up-channel margins of the turbidity maximum, DO increased f rom zero (anoxia) near high water to 60% near low water slack, in contrast to the behavior down-channel of the turbidity maximum. About 82% of the var iance in DO could be explained in terms of salinity variations atone. Only 43% of the DO variance could be explained in terms of SPM alone. Up-channel of the turbidity maximum, SPM concentrations were relatively low (<3000 mg l(-1)) and DO levels varied from 48% of saturation near high water to 83% near low water slack. About 76% of the variance in DO could be explained in terms of salinity variations alone. Within the turbidity maximum region, D O varied from <2% saturation on the early flood and late ebb and maximized around 7% at high water slack. About 63% of the variance in DO could be exp lained in terms of salinity variation alone. This increased to 70% when sus pended particulate matter was taken into account. Only 29% of the DO varian ce could be explained in terms of suspended particulate matter alone. Becau se bacteria were likely to have been the cause of the observed reduction in DO, the numbers of bacteria, both free-living and attached to particles, w ere measured in the turbidity maximum region. Numbers of free-living bacter ia were low and most of the bacteria were attached to sediment particles. T here was a linear correlation between total bacterial number and suspended particulate matter concentration, suggesting that the strong DO demand was exerted locally as a result of bacterial activity associated with increased suspended particulate matter concentrations. An order of magnitude analysi s of DO consumption within the Ouse's turbidity maximum, based on the premi se that DO depletion was directly related to suspended particulate matter c oncentrations and that DO addition was due to reaeration, indicates that co mplete deoxygenation could have occurred with an oxygen depletion rate of s imilar to 0.01 mg DO h(-1)/g suspended particulate matter during the reside nce time of waters within the turbidity maximum (similar to 7 d). This rate was sufficiently fast that anoxic to aerobic conditions were able to develop a spring-neap periodicity within the turbidity maximum, but too slow to generate substantial intratidal fluctuations in DO. This is in acc ordance with the observations, which show that relatively little of the int ratidal variance in DO could be explained in terms of suspended particulate matter fluctuations, whereas most of the variance could be explained in te rms of salinity, which behaved as a surrogate measure for the proximity of the turbidity maximum.