M-AMBI revisited: looking inside a widely-used benthic index

M-AMBI is a multimetric index for
assessing the ecological quality status of marine and
transitional waters. It is based on benthic macroinvertebrates and integrates AMBI, a biotic index based on
species sensitivity/tolerance, with diversity and richness, making it compliant with the European Water
Framework Directive. The success of AMBI paved the
way for the introduction of M-AMBI, which was
subsequently incorporated into the regulations of
several European countries. The M-AMBI algorithm
integrates the metrics by means of factor analysis
(FA). In this paper, we first reproduced the algorithm
using the open source R software. This enabled us to
point out that FA is not functional to M-AMBI, and its
omission does not appreciably change the results. We
then enhanced the applicability of the index, making it
independent of the number of samples. In this way,
M-AMBI is closely approximated by the simple mean
of the normalised metrics with no need for multivariate techniques. Finally, we further simplified the
approach, presenting a bivariate version that is still
highly correlated with M-AMBI, in which the constitutive metrics are reduced to a diversity measure and a
species sensitivity index. The properties of this
bivariate version include simplicity, transparency,
robustness, and openness.