We calculated the percentage change relative to the baseline run for each model type and case study ( Figures 3 , 4 ). The cross-model comparison of output indicators reveals that the effectiveness of current MPAs in reducing bycatch risk varies depending on the specific case study. While some MPAs may adequately address this issue, others do not; MPAs designed with a purpose (FFP) may mitigate the risk associated with bycatch but also necessitate effort reductions linked to spatial restrictions.

In the short term, fishing effort tended to decrease when displacement of fishing effort or redirection toward other fishing activities is not an option (as in FISHCODE, used for the German fleet in the southern North Sea, see further below). However, if displacement is possible, there was a tendency for increased effort in the short term due to lower catch rates (DISPLACE Ionian Sea) or decreased effort when focusing on more productive areas and/or protecting stocks (DISPLACE North Sea). Implementing an effort reduction scheme could have significantly aided rebuilding spawning stock biomass by drastically cutting catches in the near term (DISPLACE, ECOSPACE, ISIS-Fish). Nonetheless, we lack sufficient data and evidence to ascertain whether these stocks will recover long term. Long-term forecasts are bound by large uncertainty given other factors such as food web dynamics (ECOSPACE), but also especially considering the potential reduction in future fishing opportunities caused by climate change (OSMOSE). ECOSPACE and OSMOSE are tools to investigate the long-term effects on stock recovery.

In the Eastern Ionian Sea, findings with two modelling approaches (DISPLACE and ECOSPACE) regarding fishery management highlight several key points. The DISPLACE model indicates that combining the closure of hake nursery grounds with a reduction in fishing effort has proven to be the most effective strategy for conservation ( Table 1 ). However, there are potential downsides to restricting otter trawl gear within MPAs. This restriction may lead to adverse effects, particularly when exposure and risk scores rise in non-protected areas ( Table 1 ). Hence, reducing the effort is still needed; however, how such a reduction would impact the socio-economy of small-scale fishers remains uncertain, warranting further investigation to balance conservation efforts with the livelihoods of local communities.

The DISPLACE model indicates that annual bottom trawl restrictions in certain MPAs proposed in the 9^th Our Ocean conference have minimal effects on total catches and fisheries economics and demonstrate limited to counterproductive outcomes of the spatial restrictions for the modelled stocks ( Table 1 ; Supplementary Materials ). On the other hand, the broader ban tested of bottom trawling across all MPAs leads to an effort displacement into the second most productive fishing grounds and areas in the vicinity of the MPAs, decreasing SSB and increasing bycatch, and reducing economic performance by -25% to -30% depending on fishing gear and indicator ( Table 1 ). Restricting bottom trawling in sensitive benthic areas (communities with median longevity > 10 years, Smith et al., 2023; Bastardie et al., 2024a), showed minimal impact on catches, bycatch, stock status, or fisheries economics, with potential effects on deep-water species and benthos requiring further investigation. Among the management options tested, combining hake nursery closures while simultaneously reducing overall fishing effort was the most beneficial for conservation of targeted and bycatch species. This approach reduces fishing mortality (~-3% for hake, -2% for red mullet, and -25% for deep-water rose shrimp), increases hake SSB (+13%), and decreases bycatch (-25% in total) (See Supplementary Materials ). This option causes various economic impacts (from low to high, depending on gear type and indicator). The spatiotemporal closure of hake nursery areas increased hake SSB and had various economic consequences, but fewer benefits for other species.

Climate change effects on species distribution, abundance and stock status may affect the effectiveness of management measures, as indicated in our simulations ( Table 1 ).

The ECOSPACE model applied to the Eastern Ionian Sea confirmed that excluding bottom trawling from MPAs led to increased fishing effort outside the MPA, especially in locations with the highest fishing effort. However, climate warming significantly affected the outcomes ( Figure 3 ). Hence, the ECOSPACE and DISPLACE approaches agree that closures cause fisheries redistribution (for most scenarios tested), while climate change may alter the effects of the closures ( Table 1 ). Therefore, the simulation findings suggested that the closure of bottom trawling should be accompanied by effort reduction, particularly a reduction of effort for small-scale fishing to allow the hake stock to recover in the area. A recovered hake stock (+12% in biomass) and mullet (+6% in biomass) would, however, affect other fisheries in the area as, for example, it is anticipated that the increased predation of hake on deep water rose shrimp will reduce fishing opportunities (- 17% catch) on this stock of commercial interest ( Table 1 ), emphasizing the importance of including such trophodynamics effects in simulation models.

In the Adriatic and western Ionian Sea, two modelling approaches (ECOSPACE and BEMTOOL) were used to simulate the effects of fisheries management. The ECOSPACE simulations predicted that implementing novel spatial management measures ( Table 2 ), including essential fish habitats, would increase juvenile biomass, especially for red shrimps. When these closures were combined with reductions in fishing efforts to reach F_MSY (fishing mortality at Maximum Sustainable Yield), there was a notable increase in overall species biomass and a decrease in juvenile bycatch of 34% for European hake in GSA 17-18. However, the contribution of the closure was minor compared to the effort reduction that achieving F_MSY would induce. Closing fishing areas without effort reduction increased the bycatch of juveniles. The effort reduction (EffRed) scenario (here corresponding to a Pretty Good Yield, PGY, see e.g. Rindorf et al., 2017) also reduced juvenile catches of hake and red shrimps, further supporting the benefits of combined closures and management strategies. The simulations suggest that closures without effort reduction have negligible effects on ecosystem indicators across guilds and elasmobranchs. Conversely, when closure areas are combined with effort reduction (F_MSY), apex fish predators and, to a lesser extent subapex demersal predators, benthic feeding invertebrates and demersal sharks show an increase in biomass (See Supplementary Materials ).

The BEMTOOL model reveals that fleet fuel consumption varies by location, particularly for the fleets exploiting fishing grounds where the nursery areas of European hake, deep-water rose shrimp and giant red shrimp are located. Some increased costs are shown for specific fleets having to steam longer to reach out to the fishing grounds and savings on expenses for others when fishing is conducted closer to ports. However, implementing F_MSY significantly reduces fleet-wide fuel usage, which is driven by an overall decrease in effort. This indicates that fishing closures effectively shift fleet distributions and influence fuel consumption. The simulations confirmed that reduced fishing pressure, especially on nursery areas, facilitates stock recovery while improving the exploitation pattern for giant red shrimp juveniles in the western Ionian Sea. Indeed, the mean body length of caught giant red shrimp in FAO GSA 19 and, to a lesser extent, of European hake in GSAs 17-18, remarkably improved in the closure scenario because the historically high fishing effort inside and around the nursery areas of the target species in the different regions was displaced. Overall, the results of the two models confirm that the closure areas alone would not be sufficient to revert the stock and ecosystem indicators towards management objectives of sustainability, and a reduction of fishing pressure is needed. The closure areas alone and in combination with effort reduction can contribute to improving the exploitation pattern of the target stocks, when the closure is tailored to the protection of nursery areas.

The DISPLACE model suggests that excluding bottom fishing from MPAs can significantly improve benthic health in protected areas. It also indicates that protecting 30% of areas with currently high RBS (Relative Benthic Status) can benefit long-lived benthic groups while incurring low costs for fisheries ( Table 3 ).

The ECOSPACE model tested the effect of designating no-take zones and offshore wind farms (OWFs) in the North Sea marine ecosystem. Overall, the simulated impact on fish stocks and other marine species of no-take zones (in FFP MPAs that were designed based on relative benthic state and to reduce bycatch) was stronger than for existing MPAs and OWF concessions, as they closed larger coherent areas ( Table 3 ). The simulations showed biomass gains inside closed areas, both on fish biomass (9%) and the deemed vulnerable species under examination (8%). Fish catches increased around closed areas in all cases, suggesting strong edge effects. This was particularly visible when closing regions to protect sensitive benthic areas in combination with OWFs. In this scenario, a great proportion of the southern part of the North Sea is closed for fisheries and the remaining fishable areas are fished increasingly, with fishers “fishing the line” (as in Kellner et al., 2007) (see Supplementary Materials ).

The FISHCODE model, an agent-based model emphasizing fishers’ decision-making, explored the effect of no-take zones and OWFs on fishing effort and profits of three German fleets in the southern North Sea. Closing potential (but currently not enforced) MPAs had the strongest negative effect on profits, catches, effort, and fuel costs, while other tested scenarios (suggested fit-for-purpose MPAs to reduce bycatch and protect benthic habitats, and OWFs) had weak or moderate effects on profits ( Table 3 ). All scenarios reduced fishing effort for the fleets catching common shrimp and demersal species using bottom trawls, whereas the effort of the fleet targeting plaice and sole increased. Profits decreased for all three fleets, and this was most pronounced for the brown shrimp fleet in scenarios closing potential MPAs, because they heavily overlap with coastal shrimp fishing grounds, and affected agents (fishers) had to travel further, which increased their fuel costs and lowered their fishing time. Potential MPA scenarios also partially closed fishing grounds of trawlers targeting plaice; however, they maintained their level of fishing effort through switching to an alternative métier targeting more Nephrops than plaice ( Supplementary Materials ). Due to all tested fishing closures, fishing effort in the remaining open areas increased. In the scenario with closed areas induced by future OWF and protection of the benthos, which represents the largest closed fishing area, fishing increased by at least 50% in 25% of all fishable grid cells. The impact of fisheries on the ecosystem is not explicitly simulated in FISHCODE; however, more fishing effort is expected to add extra pressure on the still-open-to-fishing seafloor since all simulated fisheries used bottom-contacting gears.

The North Sea OSMOSE model was used to investigate the effects of climate change under a future Representative Concentration Pathways (RCP) 8.5W/m² until 2060 scenario and climate effects in combination with spatial management. Restrictions of bottom trawling in MPA and OWF areas, and pelagic trawling restrictions in OWFs were tested with a proportional redistribution of fishing effort to still-open-to-fishing areas where the respective fishing technique was active ( Table 3 ). An overall reduction in higher trophic level biomass by ~7.5% was simulated for both climate and climate and spatial effort redistribution scenarios ( Supplementary Materials ). This implies that despite spatial effort changes in fishing activity, bottom-up control, governed by climate effects, is likely to drive observed changes at current effort levels. Other biological indicators, including typical length (TyL) and proportion of mature biomass (PropM), show less severe changes, relative to baseline conditions, when spatial closures are implemented, across most species’ groups. In parallel, the vulnerable species (Endangered, Threatened and Protected, ETP, classified species (by the IUCN) were adversely affected when spatial effort is redistributed outside of closed areas. However, a large (>10%) biomass increase of the low trophic level group (benthic planktivorous) was observed inside the closed area ( Supplementary Materials , baseline 2018). This can be explained by the high density of sandeel over the Dogger Bank area, which overlaps with a large MPA designation. By contrast, an overall reduction in catch of higher trophic level fish species by ~7.5% was also simulated for the climate scenario, but more than double the reduction, >15% was simulated for the climate and spatial effort redistribution scenario ( Supplementary Materials ).

The impact of fisheries measures in the Bay of Biscay varied among the combinations of fit-for-purpose closures and scenarios on the overarching management in place. The overarching management was determinant in the contribution of the spatial restrictions to achieve conservation goals ( Figures 3 , 4 , FFP vs FFP+EffRed). Hence, whatever the spatial restriction, the scenario assuming full compliance with the MSY-based management and landing obligation led to a strong reduction in fishing effort. This decrease in fishing pressure benefited the exploited populations with an increase in SSB ( Figure 3 ), along with the decline in catch and profit. In the long term, however, the benefits of stocks rebuilding would partly compensate for the loss of profit induced by effort limitations (see Supplementary Materials ).

The ISIS-Fish simulations showed that the effects of offshore restrictions aiming to reduce bycatch of common dolphin (Delphinus delphis) in the French area of the Bay of Biscay would affect the profit of larger (>15m) vessels, narrowing their fishing grounds to the continental shelf, while coastal fleets using a mixture of gears may benefit economically. Most positive or negative economic impacts of closures on fleets are within -/+ 15% of the gross value added in the short/medium term (10-year horizon, see Supplementary Materials ), mainly reflecting changes in the accessibility of target species to fisheries. If the landing obligation (LO) did not apply, closing the area would have almost no effect on the total fishing effort of individual fleets, meaning they successfully compensate for the loss of effort by increasing activity in the portion of their fishing zone that remains open, whereas, if the LO applied, fisheries are closed early in the year. On the contrary, the spatial restriction under the LO led to an overall increase in the fishing days deployed at sea by making the vessels less efficient (meaning lower catch rates) in the short run (between 0 and +8%, see Supplementary Materials ). All the offshore closures tested had indirect consequences for commercial fish stocks. These closures enhanced the biomass of megrim (Lepidorhombus whiffiagonis) and anglerfish (Lophius piscatorius), typically distributed further offshore, as the displacement of fishing effort away leads to lower mortality in the closed area.

Effects of coastal restrictions aiming to reduce the bycatch of Balearic shearwaters (Puffinus mauretanicus) were much more limited, and no significant consequences for commercial fish stocks were observed, which is a positive outcome. Despite some coastal fleets and especially longliners having a substantial proportion of their fishing zone covered by the MPAs, all fleets directly affected by the MPAs can reallocate their effort outside, but with interactions between them ( Supplementary Materials ). Notably, the longliners under 10 m, with the highest overlap between fishing zone and the MPAs, are also the most negatively impacted by the MPAs. No substantial differences were observed between short/medium and long-term projections. The overall low impact of the coastal MPAs in comparison with offshore restrictions is consistent with the smaller proportion fleets concerned. However, the fishing grounds for small, coastal vessels were possibly assumed in the model to be larger thanthey are, which may overestimate vessel displacement opportunities. In summary, closures have low impacts compared to effort reduction but provide interesting synergies, notably to mitigate long-term economic losses.

EU fisheries face serious challenges threatening the marine space available to fishing. Besides capture activities themselves threatening future fishing opportunities whenever degrading supportive habitats, human factors like habitat degradation (Patti et al., 2022), marine pollution (Vagi et al., 2021), eutrophication (Solidoro et al., 2009), acidification and warming waters (Bahri et al., 2021), invasive species (Galanidi et al., 2023), and fishing practices jeopardize future exploitation (e.g., a review in Bastardie et al., 2021). This multitude of pressures necessitates an adapted governance of EU fisheries (Drouineau et al., 2023; Bastardie et al., 2024a). Effective area closures would ideally translate into enhanced fish stocks, reduced bycatch and less seafloor damage and thereby may help preserve marine biodiversity and vulnerable species. Such an ideal situation would also mean safeguarding sensitive habitats essential for breeding and growth of many marine organisms (Kraufvelin et al., 2018). Protecting these areas would contribute to more resilient marine ecosystems. Closed or restricted areas would improve spatial selectivity beyond the usual gear selectivity, which would enhance the efficiency of fishing practices in ensuring long-term yield on commercial species while minimizing the impact on non-target species (Dunn et al., 2011). However, conservation measures limiting specific fishing techniques might come with reducing opportunities in some areas while pushing fishing efforts to nearby locations. This redirection would alter catch composition and selectivity, potentially increasing operating expenses or, conversely, requiring less effort to maintain profit.

The last reform of the Common Fisheries Policy (CFP 2013, Art. 11) introduced the possibility of aligning fisheries management in Europe with its environmental protection objectives as defined in the Marine Strategy Framework Directive (MSFD), EU Habitats Directive, EU Birds Directive, and other directives, which are inherently implemented at the national level. To implement cross-border protection, the European Commission (EC) can propose new legislation that follows joint recommendations issued from a colloquium of EU Member States in the CFP regionalization process (CFP Art. 18). While the responsibility of implementing the ecosystem approach ultimately lies with Member States, a growing number of delegated acts are being adopted to include recommendations for more selective gears, less impactful gears, or closed areas to maintain seafloor integrity and monitor trophic guilds to ensure the functioning of marine food webs (see EC COM (2023/520)). This represents a well-intended opportunity to establish case-by-case restrictions on fishing in existing and future MPAs that are purposeful and well-agreed upon by the Member States, after consultation with various stakeholders, including the EU Advisory Councils (Art. 3 of the CFP about principles of good governance).

When comparing ecosystem structures and functions between non-fished areas (like MPAs or offshore windmill farms) and fished waters, it is important to determine whether any benefits remain limited to the local area or spread and extend to a broader region, potentially creating new fishing opportunities. In this endeavor, when assessing the effects and potentials of spatial restrictions achieving conservation targets or purpose, several shortcomings and pitfalls can be identified (see Table 5 for a non-exhaustive list of them we identified alongside our findings). Hence, area-based management options require scientific documentation of their performance ahead of the implementation to be effective. Usually those evidence-based insights are not issued from field observations and experiments, given the difficulty in finding and analyzing real-life fishing effort displacement and in collecting relevant data. Hence, in the field, there is limited possibility for applying rigorous approaches and appropriate metrics, as some indicators may not respond to fishing impacts. This raises concerns about how protected areas are assessed, habitat and biodiversity preservation are proven, and subsequent fisheries benefits are supported ( Table 5 #4). Capturing the closure effects suggests using B(efore) A(fter) C(ontrol) I(mpact) designs in studies to evaluate management measure impacts more effectively. Often, the BACI scheme is ignored, and the adopted methods confound pre-existing differences with the impact (e.g., see the meta-analysis in Sciberras et al., 2018 on macrobenthos or evidence for spillover effects in Van Hoey et al., 2024). Also, closed areas should not be compared solely to adjacent open areas, which can have fishing effort displaced into them ( Table 5 #4). Lack of empirical evidence highlights the necessity of using spatial modelling to advise on the performance of spatial fishing restrictions.

Describing fisheries’ effort allocation in response to factors influencing decision-making (social, economic, ecological, institutional) is also challenging. Both environmental and economic effects depend on the mobility of the fishing fleet itself, which requires modelling of how fleets would be affected by changing the fishable areas. The effect may differ among small-scale fishing that are often more polyvalent on their catch portfolio but less mobile, and large-scale fishing vessels that can steam longer distances but are often more specialized in the species they target (Salas & Gaertner, 2004). Moreover, fishing behavior may differ depending on cultural background and company structure (Schadeberg et al., 2021). On the ecological side, the effect may vary if the closure is seasonal or only spatial. For example, if vessels are very mobile, seasonal closures, where presumably all areas are closed in a season, could increase the homogeneity of overall disturbance by redistributing the effort on other seasons of the year. On the contrary, spatial closures could intensify the pressure on the remaining areas, making the pressure on habitats uneven. Both spatial and seasonal closures could lead to the redistribution of fishing activity to environmentally sensitive or previously unfished or relatively preserved areas. Bioeconomic models with more flexibility in simulating human behavior and fishers’ decision-making, both in space and time, are a way to evaluate the potential effects of including fisher behavior in model simulations. Models deployed in this study have captured some of these dynamics with more realism in describing and anticipating change in fisheries, such as changes in catchability per species (Bastardie et al., 2020; Maina et al., 2021), spatially and seasonally (Bastardie et al., 2020; Maina et al., 2021), and feedback from population dynamics (Bastardie et al., 2014; Maina et al., 2021) or foodweb dynamics (Romagnoni et al., 2015; Püts et al., 2023), a possible change in background economic landscape and fish market dynamics (Russo et al., 2017), complex fisher behavior beyond profit maximization and allows for the adaptation of métier choices (Letschert et al., 2025), etc. (See Supplementary Materials ).

The crucial aspect of evaluating MPAs is determining spatial modelling approaches, like the ones we presented in this work, to provide complementary guidance for management decisions that includes such indirect effects of closures. Using several different models increases our confidence in the estimated effects, where these align and highlight areas where further investigation is needed where they differ. However, if the same models were applied in different regions (DISPLACE and ECOSPACE) they yielded various outcomes depending on the area, which makes it likely that the main differences do not arise from the models used, but rather because of inherent differences between fisheries and local circumstances. Comparisons across models within a region have been possible in two of the four regions investigated here, strengthening the results but highlighting the difficulty in developing and deploying appropriate models.

As the percentage of closed regions increases (to reach the 30% commitment of the EC Marine Action plan), the effort displacement effects will likely change. Therefore, area-based management should only be used in combination with other technical measures such as non-spatial gear selectivity improvement and continuation of effort control, including effort reduction and the transition toward alternative fishing techniques and implementing catch limits ( Table 5 #7). Our findings are significant in the current debate about excluding specific fishing techniques from marine protected areas in EU waters to attain the aims of the Habitat Directive, MSFD and the recently adopted nature restoration regulation. In such an endeavor, several factors should be accounted for: