A spatial model to estimate gear efficiency and animal density from depletion experiments

Depletion experiments are conducted to estimate efficiency of sampling gear and density of organisms. Traditional models for analyzing these experiments make restrictive assumptions that are often violated. We present a new spatial model, suitable for sessile benthic invertebrates, that does not depend on these restrictive assumptions. The new model (i) allows flexibility during the experiment in choosing the spatial location of successive samples, (ii) does not require organisms or successive samples to be randomized over the entire area of the experiment, and (iii) permits target organisms to be lost or added during the experiment. The model treats total catch per sample as a sum of catches from smaller cells with different, but known, sampling histories. A negative binomial model is used to describe the distribution of catches from tows made during the depletion experiment. Maximum likelihood methods are used to estimate parameters, derive confidence regions for parameters, and evaluate goodness of fit between data and the model. Data from an experiment involving Atlantic surfclams (Spisula solidissima) are used to demonstrate the model.