Quantifying bioirrigation in aquatic sediments: An inverse modeling approach

An inverse model was developed to quantify the depth distributions of bioir rigation intensities in sediments based on measured solute concentration an d reaction rate profiles. The model computes statistically optimal bioirrig ation coefficient profiles; that is, profiles that best represent measured data with the least number of adjustable parameters. A parameter reduction routine weighs the goodness-of-fit of calculated concentration profiles aga inst the number of adjustable parameters by performing statistical F-tests, whereas Monte Carlo simulations reduce the effects of spatial correlation and help avoid local minima encountered by the downhill simplex optimizatio n algorithm. A quality function allows identification of depth intervals wh ere bioirrigation coefficients are not well constrained. The inverse model was applied to four different depositional environments (Sapelo Island, Geo rgia; Buzzards Bay, Massachusetts; Washington Shelf; Svalbard, Norway) usin g total CO2 production, sulfate reduction, and Rn-222/Ra-226 disequilibrium data. Calculated bioirrigation coefficients generally decreased rapidly as a function of depth, but distinct subsurface maxims were observed for site s in Buzzards Bay and along the Washington Shelf. irrigation fluxes of O-2 computed with the model-derived bioirrigation coefficients were in good agr eement with those obtained by difference between total benthic O-2 fluxes m easured with benthic chambers and diffusive fluxes calculated from O-2 micr oprofiles.