A severity index to assess and monitor the incidence of pollution-related pathological conditions in marine organisms

A normalised matrix of the occurrences and co-occurrences of n-pathological conditions in a hypothetical population is termed a severity matrix and was used to define a volume in an n-dimensional vector space representing the severity or extent of the spread of the conditions in the population. The changes in this volume resulting from varying the numbers of occurrences and co-occurrences in the severity matrix are described by the varying determinant of the matrix. The natural logarithm of this determinant was shown to be a consistent measure or index of severity for comparing the spread of the same pathological conditions among different populations or among chronological samples from the same population. While the eigenvalues of the severity matrix are used to determine the relative contribution of each pathological condition to the index of severity, the eigenvectors serve to investigate possible associations among the conditions. Actual data from a population of the soft-shell clam Mya arenaria were used to demonstrate the application of this methodology to characterise the incidence of five pathological conditions assumed to be indicators of pollution related stress on the clam population.