Implementing the Ecosystem Approach to Fisheries is a demanding exercise even though a more or less sophisticated approach might be adopted depending on the fishery or problem concerned and on the capacity available to deal with it. In France, DEMOSTEM, a programme of Ifremer was established to try to practically apply the EAF principles across the high diversity of situations existing in the country. The article summarizes very briefly the lessons learned through this programme implementation and some of the key findings and outcomes. In conclusion, it offers some perspectives on the whole process.

The use of genetic, biochemical and electronic markers in population biology and ecology has been growing tremendously during the last two decades. The first part of this paper aims at reviewing the main principles and advances of these individual markers through a few key applications on exploited marine fish populations. The second part is more prospective and investigates some possibilities that could arise in the near future through: (i) the development of new markers, (ii) the combination of different markers and (iii) the combination of quantitative approaches -whether classical or new- with individual markers. It is finally stressed how crucial individual markers will be to unravel the biocomplexity of wild fish populations and the key role they should play in the implementation of the ecosystem approach to fisheries.

The ecosystem-based approach to fisheries management and, more specifically, the European Marine Strategy Framework Directive require the assessment of the state and dynamics of an ecosystem in order to determine suitable management strategies. This paper takes an analytical approach to assess the state of the Bay of Biscay ecosystem in the early 1990s, chosen as a period of reference because key monitoring data series have been collected since then. To assess the state of the ecosystem, the pressures exerted by six broad categories of human activities were examined. A literature review of the ecosystem components was made and a component tree was tailored according to data availability. Data rich components were subdivided into subcomponents for their assessment while data poor components were assessed at an aggregated level. The component tree of the ecosystem comprised six main branches, four of which further divided into sub-components. In total, assessments were carried out at the level of 19 components. For four of these (fished species, sensitive fish species, marine mammals and turtles) the overall assessments were made combining the status of individual species. Impact from human activities were categorised as (i) “widespread” over the whole Bay of Biscay or “local” and (ii) “possible” when they could be logically expected or “documented” when they were reported in the literature. Fishing appeared to be the only activity exerting widespread documented impacts on several ecosystem components. Terrestrial activities had some possible and documented widespread impacts. With the exception of marine transport impacting seabirds at the regional scale through oil pollution, other activities had only local impacts, mostly nearshore. The reference state in the early 1990s, suggests that continuation of monitoring of vertebrates as well as estuarine and coastal habitats must be central to the monitoring programme and management strategies to be set in the context of the Marine Strategy Framework Directive. Additional monitoring is also required for benthos, substrate and micro-organisms.

This paper provides an overview of the progress Ifremer has made recently in fisheries acoustics and the study of small pelagic fish by: i) pushing observation frontiers using a range of platforms including an autonomous underwater vehicle, AUV, ii) developing measuring instruments and methods and iii) studying fish distributions. Presently, information from several frequencies of single-beam echosounders is routinely collected together with data from the ME70 multibeam echosounder. For onboard data acquisition control the HERMES software was developed. The new MOVIES 3D software includes modules for simultaneous realistic 3D visualisation and post-processing such as bottom detection, school extraction and calculation of descriptors and integration of all acoustic data. Several data analysis methods are being developed and some initial results are presented. Finally, results on the spatial distribution of small pelagic fish schools in the Bay of Biscay illustrate the role that acoustics can play, and are already playing, in the implementation of an ecosystem approach to fisheries.

The temperature evolution over the past 40 years in the Bay of Biscay (North-East Atlantic) is investigated from an in situ data analysis, completed with a satellite SST (Sea Surface Temperature) analysis over the last 20 years. The in situ dataset is an interannual version of the BoByClim climatology, covering the Bay of Biscay area with a 10-km horizontal resolution and a 5-m vertical step. The satellite dataset is the European COastral Sea Operational Observing and Forcast System Program (ECOOP) daily analysis, covering the IBIROOS (Iberian-Biscay-Irish Sea) area with a 4-km resolution. The study area (43°N-50°N/12°W-1°W) extends over the intersection of the domains covered by these two analyses. In the 0–200 m layer of this domain, a heat budget has been computed from an ocean circulation model over the period 1965-2004. Heat is essentially imported from the west by the North Atlantic Drift, then exported southward and northward, or transferred downward into deeper layers. The annual average of air-sea flux is weak, with a zero isoline crossing the domain from the northwest to the southeast. Ocean and atmosphere, forming a coupled system, have comparable contributions and may both result in a heat sink or source. At the interannual scale, variability is mainly controlled by the air-sea flux, with a secondary contribution from oceanic transport. A strong warming of $\sim $ 0.2 °C/decade for the period 1965-2004 is apparent from the surface down to the 200-m depth. The decadal evolution is consistent with several analyses over the North Atlantic. Meanwhile, warming is approximately twice faster than in the whole basin. This trend is not steady, as a cooling period occurred until the early 1970's, and the temperature increase has been significantly stronger during the last 20 years ( $\sim $ 0.3 °C/decade between 0 and 100 m, from 1986 to 2005). The trend and the interannual variability are maximal over the northern part of the shelf. The typical duration of interannual anomalies is two years and their penetration depth is ~100 m, although they occasionally exceed 200 m. The interseasonal anomalies are also stronger within the mixed layer (root-mean-square, RMS = 0.6 °C at the surface) than below (RMS = 0.3 °C at 100 m). Their geographic distribution is radically different between the surface and the deep levels. A seasonal dependence of the warming rate is evident in the upper layer, as the temperature trend is stronger in summer than in winter (by a factor of 2 at the surface). Some of the potential implications for the ecosystem are discussed, particularly in terms of species distribution and dynamics of species exploited by fisheries.

This review aims at comparing the fisheries management systems existing in New Zealand and in the European Union. The involvement of stakeholders at all stages of the management process is generally more transparent and better established in New Zealand than in the EU. Both systems aim at achieving an adequate balance between sustainability and utilisation and consider the precautionary approach as a founding principle. The socialobjectives are probably more explicit in the EU management system. In New Zealand, B MSY is a legal management target for all stocks in the quota management system (QMS), but management strategies were poorly explicituntil most recently. In the EU, there have not been any legal management targets or strategies until 1999. Since 1999, a number of multi-annual recovery and management plans have been established, including bothmanagement targets and strategies. Both management systems include conservation and access regulation measures. The EU management measures aim at regulating fisheries outputs and inputs, and discarding is tolerated. New Zealand management is almost exclusively output-based, and discarding practices are banned. In the EU, while individual quotas (IQs) are implicit in several countries, there is no consistent pattern across Member States for allocating TACs. In New Zealand, individual transferable quotas (ITQs) are implemented, and some flexibility in catch-quota balancing is provided by a carry-over allowance and the payment of a landing tax, the deemed value, for every fish landed above quota. If rights-based management were introduced in the EU based on, e.g., the New Zealand model, we suggest that concentration rules be set in accordance with the social objectives of the Common Fisheries Policy, and also that the deemed value should be set based on science and economics.

Managing adequately the exploitation of commercial species subject to technical interactions is a key step towards the ecosystem approach to fisheries. We evaluate here, using a bio-economic modelling approach building on the ISIS-Fish simulation platform, the relative impact of total allowable commercial catches (TACC) and of taxes applicable to over-quota landings (deemed value) on the sustainability of a selection of species exploited by the New Zealand hoki (Macruronus novaezelandiae) fishery. We investigate some aspects of the hoki mixed fisheries, consisting of four fleets and nineteen métiers, by considering the technical interactions between hoki and Southern hake (Merluccius australis). The dynamics of effort allocation were modelled using gravity model, using value per unit of effort (VPUE) as attractivity coefficient. Several management scenarios, based on different levels of TACC, effort limits and deemed value have been defined. The impact of these strategies on spawning biomass and catches has been investigated. The results confirm that the deemed value may, in combination with TACC, be anefficient management tool. By increasing the hoki deemed value, we could somehow limit the decrease in the hoki TACC needed to harvest Western hokisustainably. Constraining hoki fishing also restricts Southern hake fishing in the Chatham Rise and in the West Coast South Island, but leads to increased targeting of this species in the Sub-Antarctic area. The relative costs of renting/buying quota or paying the deemed value are most constraining when the hoki TACC is at relatively low level and when the hoki deemed value is more than 1.5 times the current value.

The eastern English Channel, the narrow channel of water separating northern France and southeast England is an area of intense human use of the array of resources concentrated into its relative small area. The vulnerability of living resources and their habitats brought together French and British maritime experts within a common project (called CHARM): to create an atlas of marine resource habitats in the eastern English Channel so as to provide planners and decision-makers with the necessary information to help managing the use of its living and non-living resources. This multidisciplinary and richly illustrated atlas provides abundant information on the legal framework and physical environment; benthic invertebrates, fish and their habitats; fishing activities; and a first attempt at developing a trophic network model (using ECOPATH software) and a marine conservation planning exercise (using MARXAN software, at a spatial resolution of 25 km $^{2})$ . Although most of the data used were collected elsewhere, some were collected especially for the project. Similarly, most of the analyses performed on the data where entirely original for this geographical area. The CHARM atlas has significantly improved the knowledge about the eastern Channel while contributing to the recognition that such holistic or multidisciplinary approaches to exploited marine systems are necessary to efficiently and durably manage their resources use.

The European eel (Anguilla anguilla) is drastically declining in all its distribution area and listed in the red list of the IUCN (International Union for Conservation of Nature). A rebuilding plan was adopted at European level in 2007, Regulation R(CE) 100/2007, to restore eel abundance to the level observed during the seventies. Its implementation started on the 1st of January 2009. This species is heavily threatened by numerous activities including fishing, and its management can only be effective through a systemic approach minimizing the whole range of human-induced impacts on the resource and its habitat. In the framework of the European interregional programme INTERREG IIIB-Atlantic Area– the INDICANG project aimed at elaborating abundance indicators of the European eel in the central part of its distribution area (http://www.ifremer/fr/indicang/). A methodological guide was elaborated by this project to define the indicators needed for this resource assessment. In this framework, Ifremer (Institut Français pour l'Exploitation de la Mer) and the UPPA University (Université de Pau et des Pays de l'Adour) have developed a method to estimate the daily and seasonal biomass of glass eels with the view to evaluate the fishery's impact on the estuarine eel recruitment to the Adour catchment. The estimation method uses observations from nightly scientific surveys to estimate glass eels' densities in the water column during various flood tides characterized by different hydrodynamic conditions. Information on these conditions allows the estimation of the glass eels biomass migrating during the night. From these estimates and reported catches made by the fishery during the same night, the exploitation rate applied by the fishery on the flow of glass eels progressing upstream during night flood tides is estimated. The relationship between the exploitation rate, fishery catches and hydrodynamic conditions allowed the estimation of the exploitation rate and nocturnal biomass fluctuations during the fishing season, from November 1st to March 31st of the following year. Finally, from the chronological series of biomass migrating at night, the total biomass migrating every day and the total recruitment into the estuary, during the main glass eel migration period, are estimated. Estimations made between 1998 and 2005 indicate that the overall rate of exploitation of the marine and continental fisheries, on average, is of 15.7%, ranging between 8 and 25% according to fishing seasons.

The studies carried out on the Adour estuary on the migration of glass eels lead to the characterisation of their swimming behavior. Individuals migrate passively with the flood tide current behind the dynamic tide front, and their movements into the water column are linked to the surrounding luminosity and water turbidity. A numerical model was built from observations gathered on glass eel densities during scientific surveys. It uses the outputs of a hydrodynamic model accounting for the variations in the river flow and tide coefficient. It allows the simulation of the displacement of a particle in the longitudinal axis of the estuary. The variation of the vertical movements through the water column takes into account the cloud cover, the moon phases, the alternation of days and nights and the water turbidity. The results allow displaying the migration speed of groups of glass eels entering the Adour estuary. These simulations are validated by the comparison with in situ observations and the outputs of the behavioral model make possible, on the one hand, to simulate the migration speed of glass eels according to hydrological data and, on the other hand, to define the environmental conditions that stop or slow down their displacement into the estuary.

Thirty French fleets fishing primarily in the Bay of Biscay were defined based on techno-economic input criteria: i) 18 fishing gear or gear combinations ii) the location of fishing grounds visited during a year (<12 nautical miles from coast, beyond 12 nm or both) and iii) total vessel length (<20 m for trawlers and other gears <24 m). The fleets were then characterised in terms of horse power, economic dependence on a list of nine dominant commercial species (sole, Nephrops, sea bass, hake, monkfish, anchovy, cuttlefish, sardine and squid) and their contribution to the landings of these species from the Bay of Biscay. Most fleets derived the majority of their income from one or two among the nine species. Species dependencies changed little during the study period, except for those vessels targeting anchovy before the fishery restrictions starting in 2005. Fish community impacts were compared based on fleet landings' profiles. Twelve landings profiles were determined using hierarchical cluster analysis. Fleets using similar gears had the same landings' profiles and therefore exercised similar pressures on the ecosystem, though not always in the same areas. The depletion of anchovy and the closure of the fishery lead to a change in landings profile for offshore pelagic trawlers from anchovy dominance to one with a larger proportion of albacore and sea bass. During the study period, the overall vessel number decreased from 2163 to 1845. Few vessels changed fleets, but if they did, they primarily started to fish closer to the coast, which might be a result of increasing fuel prices.

Individual transferable quotas (ITQ) were implemented in the Tasmanian red rock lobster fishery in 1998 and ten years later we assessed the impacts on the fishery. Particular attention was devoted to investigating the performances of the fishery with regard to three features identified as major impacts in the literature: rationalization of the fishing fleet, change in fishing strategy in order to maximise the fisher's profit and concentration of fishing rights and activity. On average, the fishery reacted as expected and reached its objective in terms of reconstruction of the biomass, but the overall assessment in terms of resulting profitability is not very conclusive. There is no evidence of decrease of the profitability over the period of the study but the fishery is more reactive to external factors on its export market in China than to changes in its own structure.

When reviewing the various fish-based community indicators aiming at assessing the ecological status of marine communities and/or ecosystems, a typology consisting of three major components emerges. The first component highlights the choice of relevant metrics associated with a level of organization (e.g. population or community). The second relies on the method used to combine the metrics (an aggregated indicator or a synoptic table). The third refers to the type of analysis (direct or indirect) that is used to establish the link between the metrics and a given pressure. In this paper we use the Vilaine coastal-estuarine fish communities as a case study to discuss and relate two different approaches to the suggested typology. The first approach (time-trend approach) is based on historical series of data, whereas the second approach (multi-metric index, MMI) is based on a geographic series including various French estuaries. They were developed or adapted from French studies but are representative of the approaches used in Europe. When applied to the case study, they differed in their diagnosis: the time-trend approach indicated that the bay of Vilaine habitat was deteriorating, whereas the MMI, which compares many French estuaries, indicated a good environmental state. Differences and complementarities between the approaches are thoroughly discussed with respect to the three mains components of the typology. This work appears at a particular moment, when several fish-based indicators are being developed in France and the rest of Europe. It intends to serve as an element of the ongoing reflection concerning the limitations of the various approaches that can be used to develop such indicators.

In the Northeast Atlantic fishery for roundnose grenadier Coryphaenoides rupestris, discards account for about 30% of the catch in weight. Data on discards are scarce; however, length distributions and discard rates appear to be relatively stable from year to year. In contrast, landings data available since 1990 show that the average pre-anal fin length has decreased from 20.7 cm in 1990 to 15.2 cm in 2008, resulting in a 58% reduction of the mean individual weight (850 g in 2008) and an increasing occurrence of overlapping class sizes between landings and discards in recent years. For stock assessment, the method of separable virtual population analysis (SVPA) was used. However, because of the lack of discard data for many of the years covered by the study, the catch data used as input to the assessment model had to be reconstructed from the available information (landings, discards, fishing effort and bathymetric distribution of the stock) using two methods. The first method relied on the assumption that the recent length distributions of discards were applicable to the earlier years (1990-1997). It resulted in unrealistic bimodal length distributions, suggesting a change in discarding practices through time, with larger individuals being discarded in the early days. The second method, based on the fishing effort and length distribution by depth strata, produced unimodal distributions for the whole period and confirmed that the average length of discarded fish was higher in the early days of the fishery. In both cases, the estimates of biomass follow parallel trends, suggesting a strong decline in the population. The available information and methods are discussed.