Intercalibration of benthic flux chambers I. Accuracy of flux measurements and influence of chamber hydrodynamics

Abstract

The hydrodynamic properties and the capability to measure sediment–water solute fluxes, at assumed steady state conditions, were compared for three radically different benthic chamber designs: the “Microcosm”, the “Mississippi” and the “Göteborg” chambers. The hydrodynamic properties were characterized by mounting a PVC bottom in each chamber and measuring mixing time, diffusive boundary layer thickness (DBL thickness) shear velocity (u ∗) , and total pressure created by the water mixing. The Microcosm had the most even distribution of DBL thickness and u ∗ , but the highest differential pressure at high water mixing rates. The Mississippi chamber had low differential pressures at high u ∗ . The Göteborg chamber was in between the two others regarding these properties. DBL thickness and u ∗ were found to correlate according to the following empirical formula: DBL=76.18(u ∗) −0.933 . Multiple flux incubations with replicates of each of the chamber types were carried out on homogenized, macrofauna-free sediments in four tanks. The degree of homogeneity was determined by calculating solute fluxes (of oxygen, silicate, phosphate and ammonium) from porewater profiles and by sampling for porosity, organic carbon and meiofauna. All these results, except meiofauna, indicated that there were no significant horizontal variations within the sediment in any of the parallel incubation experiments. The statistical evaluations also suggested that the occasional variations in meiofauna abundance did not have any influence on the measured solute fluxes. Forty-three microelectrode profiles of oxygen in the DBL and porewater were evaluated with four different procedures to calculate diffusive fluxes. The procedure presented by Berg, Risgaard-Petersen and Rysgaard, 1989 [Limnol. Oceanogr. 43, 1500] was found to be superior because of its ability to fit measured profiles accurately, and because it takes into consideration vertical zonation with different oxygen consumption rates in the sediment. During the flux incubations, the mixing in the chambers was replicated ranging from slow mixing to just noticeable sediment resuspension. In the “hydrodynamic characterizations” these mixing rates corresponded to average DBL thickness from 120 to 550 μm, to u ∗ from 0.12 to 0.68 cm/s, and to differential pressures from 0–3 Pa. Although not directly transferable, since the incubations were done on a “real” sediment with a rougher surface while in the characterizations a PVC plate simulated the sediments surface, these data give ideas about the prevailing hydrodynamic condition in the chambers during the incubations. The variations in water mixing did not generate statistically significant differences between the chamber types for any of the measured fluxes of oxygen or nutrients. Consequently it can be concluded that, for these non-permeable sediments and so long as appropriate water mixing (within the ranges given above) is maintained, the type of stirring mechanism and chamber design used were not critical for the magnitude of the measured fluxes. The average measured oxygen flux was 11.2 ± 2.7 (from 40 incubations), while the diffusive flux calculated (from 43 profiles using the Berg et al., 1989 [Limnol. Oceanogr. 43, 1500] procedure) was 11.1 ± 3.0 mmol m −2 day −1. This strongly suggests that accurate oxygen flux measurements were obtained with the three types of benthic chambers used and that the oxygen uptake is diffusive.