Ecosystem Health Assessment (EHA) started to be applied in environmental management in the late 80s. Since then, a considerable number of ecological indicators have been applied for EHA. Jørgensenet al. (2005a) have classified these indicators into eight levels corresponding to: indicators based on specific species, e.g. presence or absence of some characteristic species (level 1); indicators that corresponds to the ratio between classes of organisms (level 2); level 3 uses concentrations of chemical compounds, e.g. total phosphorous (Scheffer et al., 2001); level 4 applies concentration of entire trophic levels, e.g. Chlorophyll-a; level 5 is based on rates of processes in ecosystem, e.g. primary production; level 6 covers composite indicators, e.g. respiration/production; level 7 is based on holistic indicators, e.g. buffers capacity; and, finally, level 8 considers thermodynamic variables able to enclose all ecosystem’s characteristics, e.g. exergy. It is clear that there is no an indicator or even few indicators that may answer all related questions on EHA and that the question to answer as well as the data availability are fundamental on the choice of the most appropriate set of indicators.
Another idea that have gained recognition during the last decades in ecological systems is the existence of thresholds: a critical value of a pressure beyond which a state indicators shifts to a different regime (Schefferet al., 2001; Scheffer and Carpenter, 2003). The existence of such thresholds has been suggested to be used as a conceptual framework for the development of strategies for sustainable management of natural resources (Mudarian, 2001; Huggett, 2005). Even though, the concept of thresholds has been for a long time embedded in ecological risk assessment (Suter, 1993), starting from the dose-response curve, the number of experimental cases for which a threshold of regime shift have been unequivocally detected is still low In addition, industrialized fishing is rapidly depleting fish stocks and some commercial fisheries worldwide are being driven to collapse (Pauly et al., 2002; Myers and Worm, 2003). One of the first major collapses was that of the Peruvian anchoveta in 1971-1972 followed by others collapses, between them, the collapse of cod stocks in New England eastern Canada in the 1990s. Historically and in recent time, fisheries stock assessment and management have mainly been based on singlespecies population models using fisheries catch and effort as input (Beverton and Holt, 1957). By trying to equate the concept of sustainability with the optimum fish catch they produced estimates of maximum sustainable yield. However, this approach has not been able to stop the decline in fisheries and for this reasons, an ecosystem approach to manage fisheries has been proposed (Myers and Worm, 2003) where fish populations are considered as part of the marine ecosystem interacting with the other components of the ecosystem. This approach based on population and ecosystem dynamics is aiming to avoid cascading/secondary effects and destruction of the ecosystem. For example, by fishing top predators one would expect that planktivorous fishes will increase. However, it is often observed is that undesirable species e.g. jellyfish appear (Daskalov, 2002). Furthermore, with an ecosystem approach the environmental fluctuations and its effects on fish stocks (Attrill and Power, 2002) and the interactions between ecosystem compartments may be analyzed in a more appropriate way. Even though this approach goes in the right direction there is still the need to develop a fisheries management approach consistent with the ecosystem view. In this sense the use of EHA indicators as well as the introduction of the threshold
concept could help developing of a strategy of fisheries ecosystem-based management.