ELOISE research has greatly increased our understanding of the biogeochemical processes in the near-coastal pelagic and its interactions with the benthos. Important physical drivers of natural dynamics have been incorporated in habitat models of benthic systems (HIMOM, ECOFLAT). The biogeochemistry of nutrients, heavy metals and several persistent organic pollutants has been clarified substantially over the course of ELOISE. Effects of eutrophication have been comparatively well studied, particularly for the pelagic, and for benthic plants, such as seagrasses and seaweeds. Poorly known complex second order interactions, however, still hamper our predictive capacity of crucial phenomena such as the possible existence of thresholds for habitat state transitions and the differential nutrient absorption capacity of foodwebs.

Major changes in driving pressures that interact along the European coasts, such as climate change, increased urbanisation, trade and tourism, fisheries, aquaculture and agriculture, will lead to considerable changes in extent and environmental quality in coastal habitats. Different seas witness differences in strength of these societal pressures and their impacts (Table 4 in 4. Drivers and Pressures). Indicators required to monitor these state changes have been developed in a considerable number of ELOISE projects (Table 1(a) in 1. Introduction). Integration of disciplines is only beginning to materialise in the last generation of ELOISE projects (DANUBS, EUROCAT). In addition, Ledoux et al. (2004) and Herman et al. (2004) have concluded that ELOISE research projects have been distributed unequally over the research themes identified in the ELOISE science plan (Cadée et al., 1994).

Several habitats, notably the pelagic of the open coastal sea, lagoons, as well as tidal mud- and sandflats have been represented quite favourably in the ELOISE research portfolio (Table 1(a) in 1. Introduction). Hence for these we can draw on a sound knowledge base. Other European coastal habitats, such as cliffs, dunes, saltmarshes and other coastal wetlands have been studied less extensively, and thus probably are less well understood. For saltmarshes and dune complexes, indicator and decision support models have been developed within ELOISE, based on existing knowledge. Overall, the Atlantic seaboard has received little attention within ELOISE compared to the other European seas.

Future European coastal research would certainly augment the effort reviewed here in a useful way. We identify four larger areas of research:

• comparative surveys of spatial pattern . Larger-scale, comparative surveys of European seas have been carried to a limited extent only (COMWEB, EUROTROPH). An understanding of the overall relevance of observed patterns will be greatly enhanced by such a comparative approach. Larger-scale spatial variability such as habitat mosaics and their spatio-temporal dynamics have remained poorly investigated. These, however, would enable firmer conclusions on state-transitions within and among habitat types.
• fisheries and aquaculture . The effects of fisheries on coastal foodwebs is comparatively well-known, though data have not been collected within ELOISE. Those of aquaculture may be less well clarified, which is particularly relevant in view of the foreseen expansion of aquaculture into all European seas.
• EC-wide stock-taking . Stock-taking and assessment of habitat areas covering the coastal zone of Europe has been imbalanced so far, both with respect to countries as well as habitats. With the presently available CORINE and EUNIS typologies, the necessary defining stage is passed. Calibration among member countries as well as discussion on the level of aggregation for EC-wide stock-taking reviews requires attention and coordination between EEA, national authorities, and the coastal research community.
• disciplinary integration . Multidisciplinary assessment of interacting coastal issues with partners from natural and social sciences is developing in ELOISE only recently. Here also stakeholder involvement mechanisms, the roles of government, institutional organisations, law and socio-cultural contrasts are worthwhile to be developed.

We suggest that subsequent ELOISE project clusters:
(a) build on achievements and tempts to cover identified gaps in content;
(b) more carefully organise a mechanism that ascertains the EC-wide priority research issues to be resolved;
(c) develop a joint EU knowledge base on the dynamics of coastal habitats insofar as these are relevant for the implementation of EU environmental framework legislation vis-à-vis our notion of a sustainable provision of coastal resources and habitats to future generations of Europeans.
A joint, comparative effort to assess the extent and state of European coastal habitats appears necessary. This can complement and build on existing environmental surveys such as those of the EEA (cf Ærtebjerg & Carstensen, 2003).