Contaminant dispersal on the Palos Verdes continental margin II. Estimates of the biodiffusion coefficient, DB, from composition of the benthic infaunal community

Sea bed mixing has generally been modeled as a one-dimensional, vertically diffusive process in which a biodiffusion coefficient is estimated by fitting regression lines to vertical profiles of tracer concentrations, typically radioisotopes. In this paper, we describe an alternative approach to deriving time- and space-dependent biodiffusion coefficients in the sea bed, in which the coefficient is estimated directly from the composition and distribution of the benthic infaunal community. We have used a random walk model to describe diffusion driven by benthic organisms. The model is applied to an infaunal population sampled in 13 cores from the 60-m isobath, each sliced into 10 segments. The biodiffusion coefficient, DB, is evaluated for a species by means of this model as the product of characteristic length and velocity scales, L2T−1. The mixing frequency, T−1, is the product of the estimated velocity of sediment displacement and mean cross-sectional area, multiplied by the number of individuals in the core segment, and divided by the volume of the segment. The areal coefficient, L2, is equal to the mean cross-sectional area of the species. The aggregate DB curve is the sum of the species curves. Much of the bioturbation is due to the behavior of a few large species which are incompletely sampled due to the limited core volume. Mixing values for these species are redistributed through all of the core segments by probabilistic methods. The mixing coefficients computed in this manner are considered relative in view of uncertainty in determining sediment velocity, but correspond closely to values estimated independently from DDT gradients. The computations indicate that bioturbation extends through the effluent deposit in the Palos Verdes shelf, into the zone contaminated by DDT.