Laboratory experiments tested the effects of two polychaetes, a surface deposit feeder Eupolymnia heterobranchia, and a head-down deposit feeder Abarenicola pacifica, on silicate dynamics in sediment porewaters and overlying waters. Experimental chambers of sediment containing an individual deposit feeder, and controls with no macrofauna, were studied over a one month period during the summers of 1989 and 1990. Measurements included temporal changes in vertical depth profiles of pore water silicate concentrations, concurrent determinations of silicate accumulation in the water column, and the activities of experimental. organisms. A diffusion-nonlocal exchange-reaction model was devised to determine, from the pore water profiles, both the magnitude of, and variability associated with, organism effects on pore water silicate. Model results within chambers containing worms indicate that the changes in silicate concentrations due to worm activity varied by as much as an order of magnitude at a given point in the sediment column, over time periods of several days. Biologically-driven fluxes calculated from the sediment model indicate that fluxes attributable to macrofauna were positively correlated with the frequency of new burrow or tube construction, and were strongly related to the activity of the organism. Variability in the rate of silicate transport due to worms likely was related to the mechanism of habitat construction, as well as the relative distances involved in tube/burrow relocation. Silicate fluxes calculated from the model were compared with direct measures of silicate flux via the accumulation of silicate in the water column. These comparisons show that benthic diatoms at times exerted a significant effect on silicate removal from sediments and the water column. The relative importance of this effect was dictated by differences in the activity of diatoms and surrounding macrofauna between experiments, and the type of macrofaunal organisms involved. Strong interactions between the surface deposit feeder and benthic diatoms significantly affected the magnitude and direction of silicate flux across the sediment-water interface. Such interactions were lacking in experiments with the head-down deposit feeder. In combination, the sediment model and water column measurements showed that net silicate fluxes to the water column may be masked by diatom activity, even when bioirrigation serves to actively transport silicate directly from depth across the sediment-water interface.