This paper develops models of the initial impact of marine aggregate extraction on a benthic assemblage. We predict the effect of dredging on species numbers and abundance assuming spatial randomness of individuals. We extend the model to allow for spatial clustering of individuals using a Matern process. Data from a controlled field experiment are used to develop a framework for estimating species reduction. This involves modelling the spatial pattern of individuals before dredging using a Matern process, the impact of dredging at an individual level, and the probability that a species is not seen in a post-dredging survey. The framework was used to estimate that, of the 41 species that were seen in a pre-dredging survey but not in a post-dredging survey, between 0 and 14 were eliminated (with 95% likelihood) rather than escaped detection. The most likely number eliminated was 4.
