Small mesh codends applied in the brown shrimp (Crangon crangon L.) fishery in the North Sea cause large bycatch of fish, invertebrates and undersized shrimp. We investigated the most obvious and enforceable mitigation measure: the increase of mesh size beyond 20 mm, which was the medium mesh size of the fleet before 2016. The predicted selectivity from different codend designs was applied in a yield-per-recruit model able to reproduce the seasonal pattern of landings and biomass by sex, size, and age. The model enables assessment and quantifying population responses to possible management actions, also considering other influential variables such as natural mortality, which vastly decreased during the 1990s to the contemporary low level due to the decrease of cod in the Southern North Sea. Model results show that the selectivity of 20 mm codends leads to the poorest fishery and population indexes. Any increase in mesh size within the range considered and in the present low-predator situation would lead to a larger shrimp stock and higher catches. Increasing mesh size reduces catches of undersized and small commercial-sized shrimps entering the fishery mainly in summer time, thus leading to surplus catches in autumn and spring. Model results reveal that an increase of mesh size diminishes growth overfishing and allows the formation of a larger stock that may better cope a fluctuating natural mortality and the steadily upwards creeping fishing intensity.

English Abstract: The latest ICES WGBFAS report shows that food shortages in the eastern Baltic Sea are causing poor growth and thus a loss of cod productivity. This situation, initially unnoticed, has been worsening for about 30 years. It is essential to avoid starving stocks, because only the portion of assimilated food that exceeds basic needs is used for growth. A reduction of the stock size is necessary! However, the International Council for the Exploration of the Sea (ICES) has recommended the closure of the cod fishery. If we look at fishing through the eyes of aquaculture, we see that this recommendation must be questioned critically. If the "fattened" cod stock is not used sufficiently, potential yields will not be realised either. In addition to this loss, the feed used for fattening is also lost. Because, as an alternative to cod fattening, yields from fishing for herring and sprat would be possible. In contrast, overfishing of the cod stock may result in a loss of yield, but at the same time it opens up greater fishing opportunities for forage fish stocks. A revision of the stock models used and a shift from the one-species approach via the ecosystem approach to approaches that combine economy and ecology are necessary. If one compares the feed costs with the later revenues, it becomes clear that when fattening cod over a body mass of about 1 kg more money is lost in the alternative feed fish fishery than is gained when fishing for cod. A reflex-like demand for stock protection has caused low growth rates for cod in recent decades due to underfishing in the eastern Baltic Sea and thus a decline in productivity for other species as well. You cannot expect high-yield fishing for cod, herring and sprat if you tolerate a starving cod population. In order to catch larger cod, however, the number of recruits must be sufficiently low. We know this from our allotment garden. If we don't thin out the rows of radishes, we can only harvest small radishes. The same applies to cod in the Baltic Sea. If we do not reduce the number of recruits early, we will always catch small cod at unattractive prices.German Abstract: Aus dem aktuellen WGBFAS-Bericht des ICES geht hervor, dass Nahrungsmangel in der östlichen Ostsee Kümmerwachstum und damit einen Verlust an Produktivität bei Dorsch verursacht. Dieser Zustand, zunächst unbemerkt, verschlimmert sich seit etwa 30 Jahren. Hungernde Bestände müssen unbedingt vermieden werden, denn, nur der über den Grundbedarf hinausgehende Anteil der assimilierten Nahrung, wird für das Wachstum eingesetzt. Eine Reduktion der Bestandsgröße ist erforderlich! Der Internationale Rat für Meeresforschung (ICES) hat jedoch die Schließung der Fischerei auf Dorsch empfohlen. Betrachten wir die Fischerei durch die Brille der Aquakultur, stellen wir fest, dass diese Empfehlung kritisch hinterfragt werden muss. Nutzt man den „gemästeten“ Dorschbestand nicht ausreichend werden mögliche Erträge ebenfalls nicht realisiert. Zusätzlich zu diesem Verlust verliert man das zur Mast eingesetzte Futter. Denn, als Alternative zur Dorschmast wären Erträge aus der Fischerei auf Hering und Sprotte möglich. Im Gegensatz dazu verursacht die Überfischung des Dorschbestandes zwar Ertragseinbußen, eröffnet aber gleichzeitig erweiterte Fangmöglichkeiten auf die Futterfischbestände. Die Überarbeitung der verwendeten Bestandsmodelle und die Hinwendung vom Einartenansatz über den Ökosystemansatz, hin zu Ansätzen, die Ökonomie und Ökologie vereinen, sind notwendig. Vergleicht man die Futterkosten mit den späteren Erlösen wird klar, dass bei der Mast von Dorsch über einer Körpermasse von etwa 1 kg mehr Geld in der alternativen Futterfisch-Fischerei verloren wird, als bei der Fischerei auf Dorsch gewonnen wird. Ein reflexartig geforderter Bestandsschutz verursachte in den vergangenen Jahrzehnten geringe Wachstumsraten bei Dorsch durch Unterfischung in der östlichen Ostsee und damit einen Rückgang der Produktivität auch bei anderen Arten. Man kann keine ertragreiche Fischerei auf Dorsch, Hering und Sprotte erwarten, wenn man einen hungernden Dorschbestand duldet. Um nun aber größere Dorsche fangen zu können, muss die Zahl der Rekruten ausreichend gering sein. Wir wissen es aus unserem Kleingarten. Wenn wir die Radieschenreihen nicht ausdünnen, können wir nur kleine Radieschen ernten. Das Gleiche gilt für den Dorsch in der Ostsee. Wenn wir die Zahl der Rekruten nicht frühzeitig reduzieren, werden wir immer kleinen Dorsche zu unattraktiven Preisen fangen.

Changes in mesh size are often rejected by fisheries management. Fishermen also do not like changes in mesh size. Apparently, a lower, but secure current yield and revenue counts more in the decisions than uncertain benefits in the future. A calculation method is presented that can be used to calculate the immediate changes in yield. It is shown that even large changes in mesh size do not necessarily lead to large changes in yield. Reductions or increases in catches can usually be easily compensated by immediate changes in fishing effort. In addition, it is shown how changes in mesh size affect the biomass of caught undersized fish (discards). As the results depend on the length distribution within the studied area, no general conclusions can be made. The length distribution in the area must therefore be determined in advance each time changes are planned.

A common goal of fisheries management and conservationists is to conserve natural resources productively or to restore the resources to this state. Fisheries management must furthermore ensure that fishing enterprises can operate profitably. In this respect, a lack of recruitment and discarding of small fish is a problem, but also an obvious contradiction and it is assumed that significantly higher yields would be possible. In order to maintain or achieve productivity, feed production of the ecosystem must be or be brought in line with feed consumption of the target stock. This was not the case in the eastern Baltic Sea in particular and there could be observed a density-dependent reduction in the growth rates. An improvement of the food supply in the sea can be achieved by reducing the number of recruits, an established method which is applied by analogy also in agriculture, gardening and forestry. Recruitment management allows the remaining part of the recruits to grow at the maximum rate, ensuring that comparatively little waste is generated. The reduction of recruits by discarding small fish is currently carried out in conjunction with the catch of the target species. In order to achieve the maximum sustainable yield (MSY) the fishing mortality was reduced and the mesh size was increased. This, however, increased also the food requirements of the fish remaining in the sea. Curiously, the food requirement is not taken into account in the model usually used to optimise the fishing pattern. As a consequence the model predicts "ghost yields" which cannot be realized in practice due to the lack of food. Here, two "MSY fishing patterns" are compared with two fishing patterns which, according to conventional definition cause significant growth overfishing. The comparison shows that about three times the amount of feed is required per kg of fish produced if the stock is managed according MSY-criteria. If one compares the fishing pattern currently used with a proposed fishing pattern, minor disadvantages in terms of yield, but advantages in terms of revenue can be expected. The large mesh sizes used in this pattern make the so-called high-grading obsolete, but a reduction in recruits to around 50% is imperative. Therefore, it is proposed in addition to separate the reduction of recruitment from the actual fishing process in order to avoid that each individual fisherman has to process the unwanted small fish on his cutter.

The feed conversion ratio (FCR) or its reciprocal, the feed efficiency ratio (FER), is currently used to estimate the feed requirements in aquaculture. This has a serious disadvantage. The FCR is a function of feeding rate, initial and final mass at fattening, temperature and quality of feed. Therefore, the FCR is specific for any combination of these parameters and has to be adjusted whenever a parameter has changed. The feed costs and the fattening time influence the total costs significantly — a precise analysis of the relationships is necessary. For this purpose, the FER is divided into a feed efficiency factor, describing the assimilated percentage of consumed feed and a fish efficiency function, describing the percentage of assimilated food used for growth. The equation for the fish efficiency is derived from a novel growth equation, which involves the relationship between assimilation rate and mass growth. In a further step, an equation for the economic feed conversion ratio (eFCR) was derived. While doing this, it was noticeable that many relationships are already known. Aquaculture facilities can be regarded as ecosystems. It is obvious that corresponding equations should be used, and so the Beverton and Holt yield-per-recruit model and a very similar assimilated-food-per-recruit model were used for the calculations. It has been found that FCR is a special case of eFCR for M = 0 per year, therefore the term FCR is sufficient. The numerical relationships are illustrated by the example of Bluefin tuna. The production of biomass by fishing is compared with the biomass production in aquaculture facilities. The competition for food is compared, as the feed for wild and farmed fish comes from the same ecosystem. Similar to the FCR, the faecal loss ratio (FLR) and the non-faecal loss ratio (NLR) are derived. These functions can be used to compute the quantities of waste products while fattening. Even if aquaculture has a bad reputation in the press, the biomass production by aquaculture is superior to the biomass production by fishing except for the workload. Finally, an important hint. The considerations presented here provide a closed theoretical picture. However, since the measurements of individual parameters are extremely difficult and error-prone, practical evidence and experiments are still lacking. The usefulness of the model will be proven hopefully by the use of the results in practice. The risk of errors can be minimized by crosschecking with other models.

The brown shrimp (Crangon crangon) fishery is of great socio-economic importance to coastal communities on the North Sea. The fishery is exploited by beam trawlers often using codends with very small mesh sizes, leading to concerns about catch rates of undersized shrimp. However, little information is available on codend size selection, making it difficult to provide scientifically based advice on alternative codend designs. Therefore, this study establishes a predictive framework for codend size selection of brown shrimp, based on a large selectivity dataset from 33 different codend designs tested during four experimental fishing cruises, during which more than 350,000 brown shrimp were length measured. Predictions by the framework confirm concerns about the exploitation pattern in the fishery, because the retention probability of undersized shrimp reaches 95% with the currently applied designs. The framework predictions allow the exploration of obtainable exploitation patterns depending on codend design. For example, increasing codend mesh size to 25-29 mm would reduce the retention rate of undersized shrimp to a maximum of 50%, depending on codend mesh type.

Marine ecosystems evolve under many interconnected and area-specific pressures. To fulfil society's intensifying and diversifying needs while ensuring ecologically sustainable development, more effective marine spatial planning and broader-scope management of marine resources is necessary. Integrated ecological-economic fisheries models (IEEFMs) of marine systems are needed to evaluate impacts and sustainability of potential management actions and understand, and anticipate ecological, economic and social dynamics at a range of scales from local to national and regional. To make these models most effective, it is important to determine how model characteristics and methods of communicating results influence the model implementation, the nature of the advice that can be provided and the impact on decisions taken by managers. This article presents a global review and comparative evaluation of 35 IEEFMs applied to marine fisheries and marine ecosystem resources to identify the characteristics that determine their usefulness, effectiveness and implementation. The focus is on fully integrated models that allow for feedbacks between ecological and human processes although not all the models reviewed achieve that. Modellers must invest more time to make models user friendly and to participate in management fora where models and model results can be explained and discussed. Such involvement is beneficial to all parties, leading to improvement of mo-dels and more effective implementation of advice, but demands substantial resources which must be built into the governance process. It takes time to develop effective processes for using IEEFMs requiring a long-term commitment to integrating multidisciplinary modelling advice into management decision-making.

Marine ecosystems evolve under many interconnected and area-specific pressures. To fulfil society's intensifying and diversifying needs while ensuring ecologically sustainable development, more effective marine spatial planning and broader-scope management of marine resources is necessary. Integrated ecological–economic fisheries models (IEEFMs) of marine systems are needed to evaluate impacts and sustainability of potential management actions and understand, and anticipate ecological, economic and social dynamics at a range of scales from local to national and regional. To make these models most effective, it is important to determine how model characteristics and methods of communicating results influence the model implementation, the nature of the advice that can be provided and the impact on decisions taken by managers. This article presents a global review and comparative evaluation of 35 IEEFMs applied to marine fisheries and marine ecosystem resources to identify the characteristics that determine their usefulness, effectiveness and implementation. The focus is on fully integrated models that allow for feedbacks between ecological and human processes although not all the models reviewed achieve that. Modellers must invest more time to make models user friendly and to participate in management fora where models and model results can be explained and discussed. Such involvement is beneficial to all parties, leading to improvement of mo-dels and more effective implementation of advice, but demands substantial resources which must be built into the governance process. It takes time to develop effective processes for using IEEFMs requiring a long-term commitment to integrating multidisciplinary modelling advice into management decision-making.

The efficiency of the production chain of aquaculture plants is influenced among others by the fish efficiency – the mass surplus per assimilated food. A new indicator – the stock efficiency as a function of fishing mortality and mesh size – was introduced for wild fish stocks, which describes more or less the same context. The stock efficiency provides a measure of how much yield per assimilated food quantity can be harvested. Over-utilization leaves food resources unutilized. Under-utilization causes a larger stock size and consequently a lack of food for the remaining fish. The fish cannot feed at the maximum consumption rate because the ecosystem is not able to produce enough food. This is compensated by lower growth rates or by cannibalism. In the case of starving fish, we know from aquaculture that the fish efficiency and thus the productivity of the aquaculture plant decreases. Similar results are obtained for wild fish stocks. Fishing with small meshes and high fishing mortalities results in a high stock efficiency, and therefore, in larger catches per assimilated food quantity. Fisheries targeting predator and prey fish at the same time take place in many marine ecosystems. The interdependencies between predator and prey fish were analyzed based on a theoretical example. It was quantitatively shown that the achievement of high total yields is only possible if the predator-fish stock is overexploited. In this case, it is possible to fish the part of the prey stock, which has been used before exclusively as a feed for the predator fish stock. It is clear that recruitment must be closely watched when applying such an exploitation pattern. Fishing with small meshes and high fishing mortalities were successfully used for a long time within the so-called balanced harvesting approach. It is known that much larger total yields can be expected when employing balanced harvesting compared to yields actually obtained by single stock optimization.

The selectivity properties of trawls are traditionally given as a function of fish length. This saves time during selectivity trials at sea and enables rapid calculation of the selectivity properties of the trawls. The tedious age readings are not necessary. If we assume a constant Fulton’s condition factor of fish then the length-based selectivity properties for the specie under consideration are an exclusive characteristic of the used trawl described by the parameters length-selection factor, range-selection factor and mesh opening. The length-based selectivity function is independent of the growth rates of fish. This, however, is a disadvantage when it comes to exploitation models. The length based results can not immediately be used in management models. It is important also to know when growth rates change, for example, by changing forage availability. Fast-growing fish in good years is accessible earlier to fishery than slow-growing fish in starvation periods. An equation describing the selectivity properties of trawls as a function of time is needed. This can easily be accomplished by inserting the fish length as a time-dependent function. However, age-dependent selectivity functions are difficult to handle within the derivation of stock equations. They inflate the results and prevent partly closed analytical solutions. The goal here is not exciting new biological knowledge, but a new simple mathematical function which describes the retention probability of trawls with sufficient accuracy while allowing sequential calculations in a simple manner. An example illustrates at the end of the manuscript in an exemplary fashion how an overfished (food in abundance – constant growth) and an underutilized fish stock (limited food – low growth rate) responds on changes in recruitment which causes variable nutritional requirements.

For profitable operation of aquaculture facilities, it is necessary to feed fish preferably with the optimum feeding rate. The appetite and thus the optimum feeding rate depend on the instantaneous body mass and increase with temperature up to a certain limit. However, the food consumption continues to rise even if the temperature exceeds this limit, but the food is no longer completely digested. Fish need to be fed less than they will eat. Overfeeding and excessive temperatures generate unnecessary costs; also decreasing the water quality and finally the profit. An optimization of those relationships regarding profits and value chain is complicated because a precise functional relationship between feeding and growth does not exist yet. Feeding data are provided usually in tabular form and the relationship between food intake and growth is modeled by the feed conversion ratio. However, the feed conversion ratio is a function of body mass and temperature, thus calculations based on those relationships are inaccurate. To gain precise estimates of feeding data a feeding function was developed based on the approach of Pütter (1920) and the approach of Ivlev (1939). Key parameters of this model are the reference assimilation rate, its temperature constant, the limit temperature and the food efficiency. The first three parameters were determined applying the least squares method on data of feeding experiments of Hinshaw (1999) with rainbow trout (Oncorhynchus mykiss). Further experiments are necessary (Bethke et al. 2013) for the estimation of the food efficiency and maintenance rate. In addition to the precise control of food intake, the new model should also allow the calculation of the percentage of food which is not utilized. A more complete picture of the value chain is available.

The selection properties of gears are mainly determined by the selection properties of the codend, frequently described by the logistic function. This function reflects the retention probability of fish in the gear as a function of length caught from the length distribution of fish in that area. Certain conditions must be met for the measurements of the selection properties of trawls. According to the approach of Millar and Walsh (1992) the selection properties for the reference-trawl codend have to be chosen in a way that the frequencies of fish caught by the reference gear have a retention probability pr =1 for the length range within the test gear catches fish. We need small meshes in the reference codend. But small meshes in the reference trawl can reduce dramatically the flow rate in the codend (Thiele et al., 1997). This may change the fish behavior and may lead to measurement errors. A compromise must be found. In order to devastate not too much young fish, in practice the meshes of the reference-trawl codend are usually chosen as large as possible. Here, errors may occur, especially since the requirements for the selection properties of the reference-trawl codend depend on the results of the measurements which are obtained later. Applying the traditional approach but using too large meshes in reference-trawl codend can cause measurement errors. In the chosen example we get the largest measuring error for the selection-range factor SR with 36 % but for all other parameters lower errors. The new approach addresses this issue and avoids the occurrence of these errors.

Animals to be fattened must be fed with nearly the maximum feeding rate to achieve maximum yields – food supply and food consumption must coincide. This we know for centuries from agriculture. Yields decline immediately when the food supply is limited. Are there parallels between agriculture and fisheries? The only way to control the food consumption of fish stocks is by influencing the population size by fishing. In standard yield per recruit analyses fixed growth parameters are assumed. The effect of food supply on growth is not considered and thus, the implications of changes in growth for yield- and spawner-per-recruit curves due to variations in the availability of food are neglected. Under stock recovery density dependent changes in growth may severely bias biological reference points derived from such curves. Here, we present a revised Beverton and Holt yield per recruit model including a revised von Bertalanffy Growth model that is triggered by available food for assimilation and demonstrate the implications for fisheries reference points such as Fmax. The traditional Beverton and Holt yield-model should be replaced by our model; for overfished stocks it delivers identical output values but, if dietary deficiency is detected in overstocked populations then the results correspond to the measured response of the ecosystem.

Beaked redfish inhabits North Atlantic waters in the depth range 100-950 m, over the continental shelf, slope and the open ocean. Individuals can live demersal or pelagic, at various stages of their life cycle. The geographical distribution of the species extends to most of the Atlantic waters from Newfoundland and the Labrador basin in the west to the Barents Sea in the east. Monitoring beaked redfish is challenging because of the species wide geographical distribution and large scale migrations; deep distribution, which complicates trawling and hydroacoustic surveys; difficulties with tagging; and persistent difficulties in taxonomic identification. These challenges make it a particularly problematic species to observe with conventional research methods. We review these key challenges and provide recommendations for the coordinated observation of Sebastes mentella in the North Atlantic that would best contribute to the assessment and ecological research on this species.

The aim of this paper is the provision of a set of simple equations, which can serve as a basic module of an ecosystem and can be used to describe the dynamics of cohort-biomass to exploitation sequentially. The approach is based on two growth equations, which are commonly used to describe individual mass of fish as a function of time. In the management of wild fish stocks, the von Bertalanffy-growth function is applied, while research in aquaculture primarily refers to a simple exponential growth model. In this study, both models are merged into a single model and a relationship between growth and feed intake was added. The resulting equation in conjunction with the traditional abundance equation offers advantages for modeling of the dynamics of the biomass of a single cohort of a fish stock as a function of assimilation rate, catabolism rate, natural mortality, fishing mortality and time. The approach provides also a simple link between ecology and economy. Model validation was carried out by computing the annual yield per recruit applying the Beverton and Holt yield model. A simple example shows that overstocking generates an economic loss in the same order of magnitude as overfishing. Obviously it is not sufficient only to protect the fish stocks. Fish needs also feed. This work aims to move this fact more into the focus of fisheries research.

Feed conversion ratio (FCR), which is used as a description of growth as a function of feed intake, depends mainly on the feeding rate but also on temperature. The estimation of the FCR is simple; however, the practical application of this ratio is limited to cases in which fish are fed beyond maintenance rate. In addition, FCRs are not comparable if they are measured at different temperatures. Ivlev (1939) proposed the more universal temperature independent metric K3 -- also known as feed efficiency. Even if the feed efficiency is a measure of the usable percentage of the food it has not been in common use in research and practice. This is most likely due to difficulties in determining this parameter in conjunction with temperature dependency of the maintenance rate and its temperature constant in practice. In this work we present an estimation method, which allows the determination of these parameters indirectly by the “system response” of fish. The parameter estimation is performed by least square fitting, comparing the food intake of the model with the food intake observed in feeding experiments with juvenile cod (Gadus morhua), carried out by Soofiani (1983) and Hawkins et al. (1985) over a wide range of feeding rates and temperatures. For individual fish, the model works not very accurately, while a comparison of the computed model results with the de facto consumed amounts (mean feed intake per fish and size group) demonstrates a high level of agreement. The relative estimation error is generally highest at the lowest consumption rates. The developed model is particularly suitable for the optimization of the feed ration at different temperatures in aquaculture systems.

Summary Estimates of fish biomass, derived from hydroacoustic registrations are generally dependent on biological information from trawl catches. Catch composition is necessary for accurate quantitative allocation of acoustic energy to the different species present in the water column. For this purpose, the species composition in the trawl catch must be representative of the true species composition in the ocean, which requires that the catchability of the trawl for individual species is known. The catchability is expected to be trawl- and species-specific as a result of variations in fish behaviour in front of the trawl (fright reactions, swimming speeds and other factors) which can affect the chance of fish being caught or escaping. We show that catchability estimates for several species can be simultaneously derived from combined trawl-acoustic samples through a simple and straight- forward optimization method. We apply the method to estimate the catchability of two important pelagic species in the Norwegian Sea, beaked redfish (Sebastes mentella) and blue whiting (Micromesistius poutassou) when fished with a large pelagic trawl, the Gloria 2048.

An important aim of fishery management is the conservation and sustainable exploitation of fish stocks whereas maximization of profit is the aim of fishermen. At first view this seems to be contradictory. The main management tools for stock exploitation are fishing effort (summarizing all effort limitations) and the so called technical measures (mesh size). The presented new model, using these control variables, is intended to replace the Gordon Schaefer model and based on the Beverton and Holt yield-per-recruit model. The Beverton and Holt model, however, do not take into account the feedback of stock-size changes on recruitment and therefore also on yield. Overcoming this disadvantage is one finding in this paper. For the computation of optimal exploitation pattern the Shepherd approach for construction of sustainable yield curves was extended by incorporating the control of mesh size along with the current exclusive control of fishing effort. This extension turns all fishing mortality reference points into reference functions of the age at first capture. The sole economic important demersal stock in the western Baltic Sea is the cod stock of that area. This paper describes an equilibrium model for the computation of long-term harvesting strategies for that stock including the control of mesh size into the well-established exploitation models. It was shown that the stock is more reliably protected against collapsing by economic considerations rather than by measures chosen according to the precautionary approach. If the Baltic cod stock is exploited according to the suggested approach with an optimized, much larger mesh size in the steady-state, a larger profit is ensured simultaneously with improved stock conservation. This model can also be applied for the maximization of the overall profit within mixed-species management by overlaying the single-species results. Here, however, it is only possible to find an overall optimum which generally differs from the single species optimum. An economic reasonable limit of exploitation is given by the FMSY-curve – the fishing mortality resulting for a given age at first capture in the maximum yield for this age while the Fpa-curve according to precautionary approach has to be considered as the final line of defense against stock collapse. The precautionary approach only calls for changes in effort limitations or technical measures if a violation of its principles has occurred in a multi-species fishery due to ecological reasons. Apart from stock protection the main objective of this paper is to find at the same time appropriate values for the management control variables defining an equilibrium maximizing the profit (maximum economic yield – MEY) of the fishery.