Inhomogeneity detection in phytoplankton time series using multivariate analyses

Abstract

Phytoplankton communities have long been used as water quality indicators within environmental policies. This has fostered the development of national and international phytoplankton monitoring programs, but these networks are subject to sources of uncertainty due to laboratory issues. Nevertheless, studies regarding the interference associated with these aspects are not well-documented. Hence, a long time series (2003-2015) from the Basque continental shelf (southeastern Bay of Biscay) was analyzed to evaluate the uncertainty given by laboratory strategies when studying phytoplankton variability. Variability in phytoplankton communities was explained not only by environmental conditions but also by changes in fixatives (glutaraldehyde and acidic Lugol's solution) and laboratory staff. Based on Bray-Curtis distances, phytoplankton assemblages were found to be significantly dissimilar according to the effect of changes in the specialist handling the sample and the employed fixative. The pair-wise permutational multivariate analysis of variance (PERMANOVA) showed significant differences between the two fixatives utilized and also between the three taxonomists involved. Thus, laboratory-related effects should be considered in the study of phytoplankton time series.