Eutrophication is one of the most severe and widespread forms of disturbance affecting
coastal marine systems. Whilst there are general models of effects on benthos, such as the PearsonRosenberg
(P-R) model, the models are descriptive rather than predictive. Here we first review the
process of increased organic matter production and the ensuing sedimentation to the seafloor. It is
shown that there is no simple relationship between nutrient inputs and the vertical flux of particulate
organic matter (POM). In particular, episodic hydrographic events are thought to be the key factor
leading to high rates of sedimentation and accompanying hypoxia. We extend an earlier review of
effects of hypoxia to include organisms living in the water column. In general, fishes are more sensitive
to hypoxia than crustaceans and echinoderms, which in turn are more sensitive than annelids,
whilst molluscs are the least sensitive. Growth is affected at oxygen concentrations between 6.0 and
4.5 mg O2 l
–1, other aspects of metabolism are affected at between 4 and 2 mg O2 l
–1 and mortality
occurs where concentrations are below 2.0 to 0.5 mg O2 l
–1. Field studies, however, show that complex
behavioural changes also occur as hypoxia increases, and these are described herein. The areas
where hypoxia occurs are frequently areas that are stagnant or with poor water exchange. Thus
again, hydrographic factors are key processes determining whether or not hypoxia and eutrophication
occur. Tolerance to ammonia and hydrogen sulphide is also reviewed, as these substances are
found at near zero concentrations of oxygen and are highly toxic to most organisms. However, the
effects of interactions between oxygen, ammonia and hydrogen sulphide only occur below oxygen
concentrations of ca. 0.5 mg O2 l
–1, since only below this concentration are hydrogen sulphide and
oxygen released into the water. Models of eutrophication and the generation of hypoxia are discussed,
and in particular the P-R model is analysed. Although agreement with the model is widely
reported the actual predictions of the model have rarely been tested. Our review suggests that the
major effects on benthic fauna result from hypoxia rather than organic enrichment per se and suggests
that the P-R model is descriptive rather than predictive. Finally, a managerial tool is proposed,
based on the stages of effects of hypoxia and organic enrichment suggested by the P-R model and on
an earlier study. The proposed strategy involves rapid assessment tools and indicates where more
detailed surveys are needed. Managers are advised that remedial action will not produce rapid
results and that recovery from eutrophication will probably take decades. Thus it is essential to detect
potential hypoxia and eutrophication effects at early stages of development.