Negative binomial distribution

We construct credible intervals to estimate the mean organism (zooplankton and phytoplankton) concentration in ballast water via a decision-theoretic approach. To obtain the required optimal sample size, we use a total cost minimization criterion defined as the sum of the sampling cost and the Bayes risk either under a Poisson or a negative binomial model for organism counts, both with a gamma prior distribution. Such credible intervals may be employed to verify whether the ballast water discharged from a ship is in compliance with international standards. We also conduct a simulation study to evaluate the credible interval lengths associated with the proposed optimal sample sizes.

We consider the computation of sample sizes for estimating the mean concentration of organisms in ballast water. Given the possible heterogeneity of their distribution in the tank, we adopt a negative binomial model to obtain confidence intervals for the mean concentration. We show that the results obtained by Chen and Chen (2012) in a different set-up hold for the proposed model and use them to develop algorithms to compute sample sizes both in cases where the mean concentration is known to lie in some bounded interval or where there is no information about its range. We also construct simple diagrams that may be easily employed to decide for compliance with the D-2 regulation of the International Maritime Organization (IMO).

Ballast water sampling is one of the problems still needing investigation in order to enforce the D-2 Regulation of the International Convention for the Control and Management of Ship Ballast Water and Sediments. Although statistical “representativeness” of the sample is an issue usually discussed in the literature, neither a definition nor a clear description of its implications are presented. In this context, we relate it to the heterogeneity of the distribution of organisms in ballast water and show how to specify compliance tests under different models based on the Poisson and negative binomial distributions. We provide algorithms to obtain minimum sample volumes required to satisfy fixed limits on the probabilities of Type I and II errors. We show that when the sample consists of a large number of aliquots, the Poisson model may be employed even under moderate heterogeneity of the distribution of the organisms in the ballast water tank.