Résumé: The Peruvian hake (Merluccius gayi peruanus) stock has been in a delicate state in the last decades due to overexploitation combined with adverse climatic events. The stock is showing certain signs of recovery since 2012. This work analyses the environmental impacts of current fleet operations and its likely trend. The fleet was divided into coherent segments, per holding capacity and engine power. The validity of both segmentations, as well as the presence of an effect of economies of scale driving fuel use intensity (FUI), was tested. Life cycle assessment was used to calculate environmental impacts, per individual sampled vessel and per segment, complemented with indicators of energy efficiency and biotic resource depletion. The fleet is highly fuel-efficient (120 kg fuel per tonne fish) when compared with other reported values, despite a large overcapacity that increases the impact of the construction and maintenance phases. Significant inter-annual FUI variations were observed (80.0 kg t(-1) in 2008 to 210.3 kg t(-1) in 2006), but no clear trend. Neither significant differences in FUI among fleet segments nor a clear effect of economies of scale were found (but FUI analysis was based on a small sample of 32 values for nine vessels, two of which had data for a single year). Only the largest vessels, featuring 242 m(3) holding capacity and 850 hp engine power, were found to have lower FUI than any of the other vessels, but no statistical test could be applied to validate this difference. Differences in environmental impacts of individual vessels are mostly dominated by their relative FUI. Fuel use and, to a lower extent, maintenance are the main sources of environmental impacts. The most contributing impacts to ReCiPe single score are climate change, human toxicity and fossil depletion. The fishery's impacts on the biotic natural resource were orders of magnitude higher than many other global hake stocks, due to overexploitation. The environmental impacts of the national hake fleet are relatively low during the study period, despite an overcapacity of the fleet. With the perspective of expanding its operations and obtaining better yields on the eventuality that the stock fully recovers, these impacts should decrease. More research based on additional FUI data is necessary to effectively compare the performance of these vessels with larger ones (featuring > 180 m(3) and > 500 hp, of which nine existed in 2016) before possibly recommending their preferential use.

Résumé: Ecological indicators are widely used to characterise ecosystem health. In the marine environment, indicators have been developed to assess the ecosystem effects of fishing to support an ecosystem approach to fisheries. However, very little work on the performance and robustness of ecological indicators has been carried out. An important aspect of robustness is that indicators should respond specifically to changes in the pressures they are designed to detect (e.g. fishing) rather than changes in other drivers (e.g. environment). We adopted a multi-model approach to compare and test the specificity of commonly used ecological indicators to capture fishing effects in the presence of environmental change and under different fishing strategies. We tested specificity in the presence of two types of environmental change: “random”, representing interannual climate variability and “directional”, representing climate change. We used phytoplankton biomass as a proxy of the environmental conditions, as this driver was comparable across all ecosystem models, then applied a signal-to-noise ratio analysis to test the specificity of indicators with random environmental change. For directional change, we used mean gradients to apportion the quantity of change in the indicators due to fishing and the environment. We found that depending on the fishing strategy and environmental change, ecological indicators could range from high to low specificity to fishing. As expected, the specificity of indicators to fishing almost always decreased as environmental variability increased. In 55–76% of the scenarios run with directional change in phytoplankton biomass across fishing strategies and ecosystem models, indicators were significantly more responsive to changes in fishing than to changes in phytoplankton biomass. This important result makes the tested ecological indicators good candidates to support fisheries management in a changing environment. Among the indicators, the catch over biomass ratio was most often the most specific indicator to fishing, whereas mean length was most often the most sensitive to change in phytoplankton biomass. However, the responses of indicators were highly variable depending on the ecosystem and fishing strategy under consideration. We therefore recommend that indicators should be tested in the particular ecosystem before they are used for monitoring and management purposes.

Résumé: The monitoring of macroalgae is required by the Water Framework Directive (WFD) to achieve good ecological status for coastal waters and specific questions arise for tropical ecosystems belonging to the outermost European regions. To assess the suitability of macroalgae as a biological quality indicator for La Reunion reef flats (France), we performed multivariate analyses linking the abundance and composition of macroalgae to water physicochemistry. Three hydrological groups of stations were identified according to dissolved inorganic nitrogen (DIN) concentrations and DIN/PO4 ratios. Some indicator species were found at the N-enriched stations (Bryopsis pennata, Caulerpa lamourowcii, Chaetomoropha vieillardii, Derbesia sp1, Blennothrix lyngbyacea, Sphacelaria tribubides), and others at the non-impacted stations (Anabaena sp1, Blennothrix glutinosa, Codium arabicum, Neomeris vanbosseae). Another key result was the significant increase in red algal cover at the most N-enriched station. Our findings are discussed in the context of the application of the WFD in the outermost French regions.

Résumé: Fisheries have had major negative impacts on marine ecosystems, and effective fisheries management and governance are needed to achieve sustainable fisheries, biodiversity conservation goals and thus good ecosystem status. To date, the IndiSeas programme (Indicators for the Seas) has focussed on assessing the ecological impacts of fishing at the ecosystem scale using ecological indicators. Here, we explore fisheries ‘Management Effectiveness’ and ‘Governance Quality’ and relate this to ecosystem health and status. We developed a dedicated expert survey, focused at the ecosystem level, with a series of questions addressing aspects of management and governance, from an ecosystem-based perspective, using objective and evidence-based criteria. The survey was completed by ecosystem experts (managers and scientists) and results analysed using ranking and multivariate methods. Results were further examined for selected ecosystems, using expert knowledge, to explore the overall findings in greater depth. Higher scores for ‘Management Effectiveness’ and ‘Governance Quality’ were significantly and positively related to ecosystems with better ecological status. Key factors that point to success in delivering fisheries and conservation objectives were as follows: the use of reference points for management, frequent review of stock assessments, whether Illegal, Unreported and Unregulated (IUU) catches were being accounted for and addressed, and the inclusion of stakeholders. Additionally, we found that the implementation of a long-term management plan, including economic and social dimensions of fisheries in exploited ecosystems, was a key factor in successful, sustainable fisheries management. Our results support the thesis that good ecosystem-based management and governance, sustainable fisheries and healthy ecosystems go together.

Résumé: Trophic models are key tools to go beyond the single-species approaches used in stock assessments to adopt a more holistic view and implement the Ecosystem Approach to Fisheries Management (EAFM). This study aims to: (i) analyse the trophic functioning of the Celtic Sea and the Bay of Biscay, (ii) investigate ecosystem changes over the 1980–2013 period and, (iii) explore the response to management measures at the food web scale. Ecopath models were built for each ecosystem for years 1980 and 2013, and Ecosim models were fitted to time series data of biomass and catches. EcoTroph diagnosis showed that in both ecosystems, fishing pressure focuses on high trophic levels (TLs) and, to a lesser extent, on intermediate TLs. However, the interplay between local environmental conditions, species composition and ecosystem functioning could explain the different responses to fisheries management observed between these two contiguous ecosystems. Indeed, over the study period, the ecosystem's exploitation status has improved in the Bay of Biscay but not in the Celtic Sea. This improvement does not seem to be sufficient to achieve the objectives of an EAFM, as high trophic levels were still overexploited in 2013 and simulations conducted with Ecosim in the Bay of Biscay indicate that at current fishing effort the biomass will not be rebuilt by 2030. The ecosystem's response to a reduction in fishing mortality depends on which trophic levels receive protection. Reducing fishing mortality on pelagic fish, instead of on demersal fish, appears more efficient at maximising catch and total biomass and at conserving both top-predator and intermediate TLs. Such advice-oriented trophic models should be used on a regular basis to monitor the health status of marine food webs and analyse the trade-offs between multiple objectives in an ecosystem-based fisheries management context.

Résumé: Variable and low egg quality is a major limiting factor for the development of efficient aquaculture production. This stems from limited knowledge on the mechanisms underlying egg quality in cultured fish. Molecular analyses, such as transcriptomic studies, are valuable tools to identify the most important processes modulating egg quality. However, very few studies have been devoted to this aspect so far. Within this study, the microarray-based transcriptomic analysis of eggs (of different quality) of sea bass (Dicentrarchus labrax) was performed. An Agilent oligo microarray experiment was performed on labelled mRNA extracted from 16 batches of eggs (each batch obtained from a different female) of sea bass, in which over 24,000 published probe arrays were used. We identified 39 differentially expressed genes exhibiting a differential expression between the groups of low (fertilization rate < 60 %) and high (fertilization rate > 60 %) quality. The mRNA levels of eight genes were further analyzed by quantitative PCR. Seven genes were confirmed by qPCR to be differentially expressed in eggs of low and high quality. This study confirmed the importance of some of the genes already reported to be potential molecular quality indicators (mainly rnf213 and irf7), but we also found new genes (mainly usp5, mem-prot, plec, cenpf), which had not yet been reported to be quality-dependent in fish. These results suggest the importance of genes involved in several important processes, such as protein ubiquitination, translation, DNA repair, and cell structure and architecture; these probably being the mechanisms that contribute to egg developmental competence in sea bass.

Résumé: Stock-based and ecosystem-based indicators are used to provide a new diagnosis of the fishing impact and environmental status of European seas. In the seven European marine ecosystems covering the Baltic and the North-east Atlantic, (i) trends in landings since 1950 were examined; (ii) syntheses of the status and trends in fish stocks were consolidated at the ecosystem level; and (iii) trends in ecosystem indicators based on landings and surveys were analysed. We show that yields began to decrease everywhere (except in the Baltic) from the mid-1970s, as a result of the over-exploitation of some major stocks. Fishermen adapted by increasing fishing effort and exploiting a wider part of the ecosystems. This was insufficient to compensate for the decrease in abundance of many stocks, and total landings have halved over the last 30 years. The highest fishing impact took place in the late 1990s, with a clear decrease in stock-based and ecosystem indicators. In particular, trophic-based indicators exhibited a continuous decreasing trend in almost all ecosystems. Over the past decade, a decrease in fishing pressure has been observed, the mean fishing mortality rate of assessed stocks being almost halved in all the considered ecosystems, but no clear recovery in the biomass and ecosystem indicators is yet apparent. In addition, the mean recruitment index was shown to decrease by around 50% in all ecosystems (except the Baltic). We conclude that building this kind of diagnosis is a key step on the path to implementing an ecosystem approach to fisheries management

Résumé: Aim Identifying the main factors affecting the spatial distribution of marine predators is essential in order to evaluate their distribution patterns, predict the potential impact of human activities on their populations and design accurate management actions. This information is also valuable from a more general management perspective, as marine predators are often considered indicators of habitat quality. In this context, we aimed to determine the degree to which environmental features, prey availability and human activities interact and influence spatial distribution of two marine mesopredator elasmobranchs, the small-spotted catshark (Scyliorhinus canicula) and the Mediterranean starry ray (Raja asterias), living in a highly human-exploited environment. Location Mediterranean Sea. Methods With information obtained from an extended experimental survey, we investigated the relative importance of environmental variables, prey availability and human activities on the spatial distribution of the abundance, biomass and occurrence rate of these marine mesopredators using deviance partitioning analyses. Results Our results revealed that environmental variables were the most important factors explaining the spatial distribution of Mediterranean starry ray, whereas small-spotted catshark distribution was also influenced by prey availability and human factors. From a management point of view, these findings suggest that Mediterranean starry ray could be a good candidate as an indicator species of demersal environmental quality. On the other hand, the distribution of the small-spotted catshark, which responds in an interactive and complex way to environment, prey availability and particular human activities, may be misleading as an environmental indicator. Main conclusions The spatial distribution of elasmobranchs in highly human-impacted marine areas can reflect the interactive and combined effects of multiple factors. To avoid misunderstandings, attention should be paid to statistical procedures allowing the separation of pure and joint contribution of the factors driving the observed spatial patterns.

Résumé: Invertebrate catches are increasing globally following the depletion of many finfish stocks, yet stock assessments and management plans for invertebrates are limited, as is an understanding of the ecosystem effects of these fisheries. Using an ecosystem modelling approach, we explored the trade-offs between invertebrate catches and their impacts on the associated ecosystem on the south coast of Wellington, New Zealand. We simulated exploitation of lobster (Jasus edwardsii), abalone (Haliotis australis, H. iris), and sea urchin (Evechinus chloroticus) over a range of depletion levels—from no depletion to local extinction—to estimate changes in target catches and associated effects on other species groups, trophic levels, and benthic and pelagic components. Exploitation of lobster showed the strongest ecosystem effects, followed by abalone and urchin. In all three fisheries, the current exploitation rate exceeds that which produces maximum sustainable yield, with considerable ecosystem effects. Interestingly, a reduced exploitation rate is predicted to increase target catches (and catch-per-unit-effort), thereby strongly reducing ecosystem effects, a win–win situation. Our results suggest that invertebrate exploitation clearly influences ecosystem structure and function, yet the direction and magnitude of responses depend on the target group and exploitation rate. An ecosystem-based fisheries management approach that includes the role of invertebrates would improve the conservation and management of invertebrate resources and marine ecosystems on broader scales.

Résumé: Fisheries provide critical provisioning services, especially given increasing human population. Understanding where marine communities are declining provides an indication of ecosystems of concern and highlights potential conflicts between seafood provisioning from wild fisheries and other ecosystem services. Here we use the nonparametric statistic, Kendall׳s tau, to assess trends in biomass of exploited marine species across a range of ecosystems. The proportion of ‘Non-Declining Exploited Species’ (NDES) is compared among ecosystems and to three community-level indicators that provide a gauge of the ability of a marine ecosystem to function both in provisioning and as a regulating service: survey-based mean trophic level, proportion of predatory fish, and mean life span. In some ecosystems, NDES corresponds to states and temporal trajectories of the community indicators, indicating deteriorating conditions in both the exploited community and in the overall community. However differences illustrate the necessity of using multiple ecological indicators to reflect the state of the ecosystem. For each ecosystem, we discuss patterns in NDES with respect to the community-level indicators and present results in the context of ecosystem-specific drivers. We conclude that using NDES requires context-specific supporting information in order to provide guidance within a management framework.

Résumé: The contamination of the Gironde Estuary, southwest of France, by polychlorinated biphenyls (PCBs) was assessed using six fish of high ecological and economic importance as bioindicator species. The concentrations of 21 PCB congeners and total fat contents were determined in the muscle and liver of eels (Anguilla anguilla), seabass (Dicentrarchus labrax), flounders (Platichthys flesus), meagres (Argyrosomus regius), mullets (Liza ramada), and soles (Solea vulgaris). In addition, information regarding the trophic ecology of the studied fish was obtained through the analysis of carbon and nitrogen stable isotopes (i.e., delta C-13 and delta N-15) in muscle. Results revealed high PCB concentrations in fish compared to monitored European estuaries. The muscle of eels was by far the most contaminated fish flesh (Sigma(7)PCBs = 1000 +/- 440 ng g(-1), on a dry weight basis), while the higher PCB concentrations in liver were measured in flounder (Sigma(7)PCBs = 2040 +/- 1160 ng g(-1), d.w.). A quantile regression approach allowed to investigate the fate of PCBs in the Gironde estuarine fish assemblage, and revealed a general process of trophic magnification. Finally, most of the analysed fish presented PCB concentrations in muscle meat above the current European maximum limits for sea products, while the derived “Toxic Equivalent Quantity” (TEQ) revealed human health concerns only for high-fat fish consumption.

Résumé: * Seabird proxies have the potential to act as useful and cost-effective indicators of the state of the marine environment. Seabird time-activity budgets, in particular, reflect short-term changes in prey conditions. * We tested an automated technique for long-term continuous recording of Cape gannet, Morus capensis, time-activity budgets using coded very high frequency (VHF) transmitters allowing for simultaneous monitoring of a large sample of study birds. * Radiotransmitters attached to leg-rings had no impact on adult foraging trip and nest attendance durations, breeding success or chick growth. Furthermore, frequencies of nest attendance and foraging trip durations estimated by the VHF logging system were no different to those estimated from hourly direct observations. * Using time-depth recorders, the relationship between the time that birds rested on the sea surface in relation to foraging trip duration was assessed. Trip duration during chick rearing was clearly an accurate proxy for foraging effort. * The VHF monitoring system provides a simple method of accurately assessing the time-activity budgets of colonial seabirds, which can be expanded to a range of other colonially breeding taxa. In the case of seabirds, this approach can potentially provide sensitive, real-time indicators of prey abundance for fisheries management.

Résumé: Ensuring the sustainability of fish resources necessitates understanding their interaction with coastal habitats, which is becoming ever more challenging in the context of ever increasing anthropogenic pressures. The ability of coastal lagoons, exposed to major sources of disturbance, to provide resources and suitable habitats for growth and survival of juvenile fish is especially important. We analysed three lagoons with different ecological statuses and habitat quality on the basis of their eutrophication and ecotoxicity (Trix test) levels. Fish abundances were sampled using fishing and horizontal beaming acoustic surveys with the same protocols in the same year. The relative abundance of Anguilla anguilla, Dicentrarchus labrax or the Mugilidae group was not an indicator of habitat quality, whereas Atherina boyeri and Sparus aurata appeared to be more sensitive to habitat quality. Fish abundance was higher in the two lagoons with high eutrophication and ecotoxicity levels than in the less impacted lagoon, while fish sizes were significantly higher in the two most severely impacted lagoons. This leads us to suggest low habitat quality may increase fish growth rate (by the mean of a cascading effect), but may reduce lagoon juvenile abundance by increasing larval mortality. Such a hypothesis needs to be further validated using greater investigations which take into account more influences on fish growth and recruitment in such variable environments under complex multi-stressor conditions. (C) 2013 Elsevier Ltd. All rights reserved.

Résumé: In order to assess the impact of bacterial input loads in Bizerte lagoon (Northern Tunisia), a seasonal survey was done for total coliform bacteria covering the whole lagoon area, to reproduce and prevent dispersion of input loads of terrestrial discharges in the lagoon. Effect of solar radiation on Escherichia coli strains was investigated using seawater microcosms comparatively exposed to sunlight and dark conditions. The results revealed (i) high charges of coliforms in the sediment rather than in the water column, (ii) total coliforms increased periodically with season in both compartments, (iii) anthropogenic discharges influence coastal areas of the lagoon, and (iv) solar radiation has an effect on decay rate of E. coli maintained in microcosm experiments.

Résumé: Two stocking densities, “low” (L, between similar to 19 and similar to 25 kg m(-3)) and “high” (H, between similar to 75 and similar to 100 kg m(-3)) were compared for effects on specific growth rate (SGR), feed conversion, energetics and welfare of rainbow trout reared at 14 degrees C either in static water (S) or swimming in a gentle current of similar to 0.9 bodylengths s(-1) (C). Trout (initial mass similar to 110 g) were reared for 9 weeks in circular tanks (volume 0.6 m(3)), in triplicate of four conditions (LS, LC, HS, HC). Fish were fed ad-libitum daily: waste pellets were swirl-collected at the outflow to calculate feed intake. SGR was measured each three weeks for the last six weeks of the trial. The tanks functioned as intermittent-stopped flow respirometers, to permit metabolic rate to be measured as instantaneous oxygen uptake once per hour. Mean (+/-SD) SGR was significantly lower at H than L (1.51 +/- 0.03 vs 1.44 +/- 0.04% day(-1), respectively, n = 6) and lowest in HC. When compared over a similar interval of mass gain, H groups had approximately 25% higher metabolic rates than L, with the highest rates in the HC condition. As a result, fish in the H groups dissipated a greater amount of feed energy as metabolism and, across all groups, there was a direct negative relationship between the quantity of energy dissipated and their SCR. There was no evidence of a neuroendocrine stress response, plasma cortisol was around 1 ng ml(-1) in all conditions. An acute crowding stress increased plasma cortisol to above 120 ng ml(-1) in all groups, but C groups recovered to control levels within 8 h whereas S groups required 20 h. Respirometry on individuals revealed that H fish had approximately 14% higher metabolic rates than L fish, indicating that increased metabolic rate in rearing tanks was in part physiological. The H groups had approximately 15% lower critical swimming speeds than the L groups which, together with their raised metabolic rate, indicated a physiological impairment Thus, high density reduced SGR by raising energy dissipation, at least partially as a physiological response by the fish, although there was no evidence of an endocrine stress response. The only beneficial effect of C was in recovery from acute stress. (C) 2012 Elsevier B.V. All rights reserved.

Résumé: Time-area closures have become a frequently used tool to control fishing effort and protect feeding and spawning areas. However, because time-area closure strata are mainly based on biological and ecological considerations, and do not accounts for fishermen's behavior-at-sea, this type of regulation tool may not entirely achieve its objectives. With the aim of comparing the impact of two different time-area regulations: (1) a moratorium on Fish Aggregating Devices (FAD) sets (1997-2005) and (2) a no-take area for surface fleets (2005-2010) on the dynamics of the European (EU) tuna purse seine fleet operating in the eastern tropical Atlantic, several fishery indicators were evaluated through a Before-After, Control-Impact (BACI) approach. The results showed that prior to any regulation, the fleet used to be concentrated within the Gulf of Guinea area. During the first years of the moratorium on FAD (from November to January within a large region in the eastern Atlantic) there was a movement towards outside the protected area, increasing the total sets on FAD (restricted fishing activity). In general, this moratorium fulfilled its objectives; however, it was not respected during the last years of this regulation. The no-take time-area closure restricted all tuna catches for the surface fisheries but only in November and within a small area (i.e., the Picolo zone). As a result, there was an increase in activities on free schools outside the no-take area. Our findings suggest the use of some simple fishery indicators to understand fleet dynamics as a complement of ecological information before implementing new time area closures. Furthermore, since tunas are highly mobile species, anticipating the possible re-allocation of effort of purse seiners to adjacent areas in response to the spatial regulation is required to design different candidate time-area closures and to evaluate their effectiveness to protect juvenile tunas.

Résumé: Marine ecosystems have been exploited for a long time, growing increasingly vulnerable to collapse and irreversible change. How do we know when an ecosystem may be in danger? A measure of the status of individual stocks is only a partial gauge of its status, and does not include changes at the broader ecosystem level, to non-commercial species or to its structure or functioning. Six ecosystem indicators measuring trends over time were collated for 19 ecosystems, corresponding to four ecological attributes: resource potential, ecosystem structure and functioning, conservation of functional biodiversity, and ecosystem stability and resistance to perturbations. We explored the use of a decision-tree approach, a definition of initial ecosystem state (impacted or non-impacted), and the trends in the ecosystem indicators to classify the ecosystems into improving, stationary, and deteriorating. Ecosystem experts classified all ecosystems as impacted at the time of their initial state. Of these, 15 were diagnosed as “ugly”, because they had deteriorated from an already impacted state. Several also exhibited specific combinations of trends indicating “fishing down the foodweb”, reduction in size structure, reduction in diversity and stability, and changed productivity. The classification provides an initial evaluation for scientists, resource managers, stakeholders, and the general public of the concerning status of ecosystems globally.

Résumé: Under the context of an ecosystem approach to fisheries (EAF), there is keen interest in providing insights into the evolution of exploited ecosystems using simple ecosystem indicators. Many nations have long-term scientific research surveys, originally driven by conventional approaches in fisheries assessment and management. The aim of this study is to address the practical concerns linked to current objectives of monitoring simple EAF indicators, using data from surveys that were not historically designed for the purpose. Based on the results of an expert survey designed to collect expert knowledge on research surveys from scientists working on different ecosystems worldwide, a list of challenges faced during indicator estimation is highlighted, along with associated concerns and constraints. The work provides additional information useful in the interpretation of the results obtained on the state and trends of ecosystems using EAF indicators by the IndiSeas WG. Further, the related discussion provides potential pathways that could be useful for future research and development aiming to improve the ecosystem indicator approach in the operational context of EAF. The question of the utility for EAF of using historical dataseries of scientific trawl series is also discussed. Such long-term series are concluded to be useful, that they are even inescapable (since the past cannot be resamplied), and that EAF therefore brings a supplementary reason for continuing such monitoring and to incorporate new insights in how research surveys may be conducted.

Résumé: Shannon, L. J., Coll, M., Yemane, D., Jouffre, D., Neira, S., Bertrand, A., Diaz, E., and Shin, Y-J. 2010. Comparing data-based indicators across upwelling and comparable systems for communicating ecosystem states and trends. – ICES Journal of Marine Science, 67: 807-832.

Résumé: Within the IndiSeas WG, the evaluation of exploited marine ecosystems has several steps, from simple binary categorization of ecosystems to a more-complex attempt to rank them and to evaluate their status using decision-tree analyses. With the intention of communicating scientific knowledge to the public and stakeholders, focus is on evaluating and comparing the status of exploited marine ecosystems using a set of six ecological indicators and a simple and transparent graphic representation of ecosystem state (pie charts). A question that arose was whether it was acceptable to compare different types of marine ecosystems using a generic set of indicators. To this end, an attempt is made to provide reference levels to which ecosystems can be objectively compared. Unacceptable thresholds for each indicator are determined based on ecological expertise derived from a questionnaire distributed to a group of scientific experts. Analysis of the questionnaires revealed no significant difference in the thresholds provided for different ecosystem types, suggesting that it was reasonable to compare states directly across different types of ecosystem using the set of indicators selected.

Résumé: Ecosystem models provide a platform allowing exploration into the possible responses of marine food webs to fishing pressure and various potential management decisions. In this study we investigate the particular effects of overfishing on the structure and function of the southern Benguela food web, using two models with different underlying assumptions: the spatialized, size-based individual-based model, OSMOSE, and the trophic mass-balance model, Ecopath with Ecosim (EwE). Starting from the same reference state of the southern Benguela upwelling ecosystem during the 1990s, we compare the response of the food web to scenarios of overfishing using these two modelling approaches. A scenario of increased fishing mortality is applied to two distinct functional groups: i) two species of Cape hake, representing important target predatory fish, and ii) the forage species anchovy, sardine and redeye. In these simulations, fishing mortality on the selected functional groups is doubled for 10 years, followed by 10 years at the initial fishing mortality. We compare the food web states before the increase of fishing mortality, after 10 years of overfishing and after a further 10 years during which fishing was returned to initial levels. In order to compare the simulated food web structures with the reference state, and between the two modelling approaches, we use a set of trophic indicators: the mean trophic level of the community and in catches, the trophic pyramid (biomass per discrete trophic level), and the predatory/forage fish biomass ratio. OSMOSE and EwE present globally similar results for the trophic functioning of the ecosystem under fishing pressure: the biomass of targeted species decreases whereas that of their potential competitors increases. The reaction of distant species is more diverse, depending on the feeding links between the compartments. The mean trophic level of the community does not vary enough to be used for assessing ecosystem impacts of fishing, and the mean trophic level in the catch displays a surprising increase due to the short period of overfishing. The trophic pyramids behave in an unexpected way compared to trophic control theory. because at least two food chains with different dynamics are intertwined within the food web. We emphasize the importance of biomass information at the species level for interpreting dynamics in aggregated indicators, and we highlight the importance of competitive groups when looking at ecosystem functioning under fishing disturbance. Finally, we discuss the results within the scope of differences between models, in terms of the way they are formulated, spatial dimensions, predation formulations and the representation of fish life cycles.