Univariate or multivariate approaches for histopathological biomarkers in the context of environmental quality assessments?

Abstract

Although the simplification of multivariate histopathological data into univariate indices can be useful for the assessment of environmental quality, this implies a great loss of information. The objective of the present study was to evaluate the effectiveness, in the context of environmental quality assessment, of an approach that integrates individual histopathological responses in a discriminated manner with the results of contaminants by means of multivariate analyses. This analysis was compared to the diagnosis of environmental quality provided by the use of the univariate Bernet histopathological index. Contaminant loads (sediments and fish) and the liver histopathology of Cathorops spixii were integrated through multivariate analysis. Integrated individual histopathological responses allowed classifying environmental quality from more to less impacted sites, while the univariate index showed some inconsistencies with chemical loads and allowed identifying only the most impacted site.