Vertical migration and feeding behavior of Calanus paci$cus females during a phytoplankton bloom in Dabob Bay, U.S

We report the first seasonal time-series observa-tions of porewater composition obtained for deep-sea sedi-ments. We observed considerable temporal variability of O,,NO,-, and Mn2+ profiles at the French JGOFS site DYFA-MED (western Mediterranean) in response to a pulsed inputof organic matter. A delivery of reactive organic matter rep-resenting only 1% of the average organic C content of surfacesediments (0.6% wt/wt) was required in order to fuel the ob-served changes in porewater composition. The perturbation resulted in dramatic changes in integrated reaction rates and sed-iment-water fluxes. O2 uptake rates increased from 1.26 to1.82 mmol m-* d-l, while nitrification rates rose from 0.13 to0.18 mmol m-* d-l. These changes were paralleled by a de-crease in NO,effluxes across the sediment-water interfacefrom 0.07 to 0.01 mmol m-’ d-l and a concomitant rise ofdenitrification rates from 0.06 to 0.17 mmol m2 d-l. Porewater profiles responded rapidly to the sediment pulse depo-sition, returning to their steady-state values within a fewmonths. The main driving force appeared not to be the sea-sonal variability of pelagic production but rather the down-slope transport of resuspended upper-shelf sediments. Numerous studies during the last decade have demonstrat-ed that the deep-sea superficial sediment is an environmentof considerable temporal variability. Seasonality of pelagicprocesses, coupled to rapid delivery of largely unaltered par-ticulate organic matter, and episodic events like benthicstorms and turbidity currents generated by downslope trans-port were identified as potential driving forces of deep-seatemporal variability (Deuser and Ross 1980; Smith andBaldwin 1984; Hollister and McCave 1984; Smith et al.1992; Pfannkuche 1993; Sayles et al. 1994). Resulting dra-matic depositional events have been documented in the past(Lampitt 1985; Gross et al. 1988; Smith et al. 1994), butbecause of the ephemeral nature of the perturbations andlogistical constraints, little is known of the geochemical re-sponse of surface sediments.Pulsed delivery of organic matter to the deep-sea floormay provide an important energy source to deep-sea benthiccommunities (Lochte and Turley 1988; Graf 1989; Gooday1988). The benthic response to variations in particulate or-ganic carbon (POC) flux has been assessed through time-series analyses of sediment community oxygen consumption.These studies have yielded divergent results. While a NorthPacific site indicated varying rates of benthic organic carbonmineralization (Smith et al. 1992), only minor temporal vari-ations in metabolic rates in response to variations in POCflux of similar amplitude were documented for an Atlanticsite near Bermuda (Sayles et al. 1994). Deep-sea benthiccommunities have been documented to react within days tofluctuations in food supply (Graf 1989), generating changesin biomass and burying freshly deposited particles down to9-cm depth. Although such a fast and deep-reaching re-sponse should be reflected by changes in porewater com-position, the temporal variability of porewater chemistry hasnot been established. Modelling results suggest that the ben-thic response will be confined to the sediment-water inter-face, any temporal variability in porewater 0,, and nutrientprofiles being damped out by the large pool of sedimentarvorganic carbon (Martin and Bender 1988). This contrastswith estimates of the reactivity of POC in surface sediments(Harnmond et al. 1996) and sediment mixing rates (Smith et al. 1993), suggesting that fresh organic particles could berapidly incorporated in the sediment and influence porewatercomposition.We investigated surface sediment porewater compositionat the French JGOFS site Dyfamed (43”24.61’N, 7”5 1.68’E,2,300-m depth) between July 1994 and June 1995. Surfacesediments consisted of fine mud with an average porosity of78% (O-O.5 cm), decreasing to 55% in a sandy lamina be-tween 4 and 5 cm. Pebbles also occurred erratically. Sedi-ments collected during November 1994 were overlain by a0.5-cm-thick unconsolidated layer. Numerous animal trackstestified to high biological activity. This distinct layer couldbe identified during the following months as it was consol-idating.Studying temporal variability requires the assessment ofspatial heterogeneity of the sampling area. The latter has tobe low to allow identification of a temporal trend. UsingGPS navigation, the dimension of the sampling site was re-stricted to a circle 500 m in diameter. Except for March