© 2004 by ICES/CIEM International Council for the Exploration of the Sea/Conseil International pour l'Exploration de la Mer
From single-species advice to mixed-species management: taking the next step
a Danish Institute for Fishery Research, Department of Marine Fisheries Charlottenlund Slot, DK-2920 Charlottenlund, Denmark
b European Commission, Directorate-General Fisheries Rue de la Loi 200, B-1049 Bruxelles, Belgium
*Correspondence to S. A. Reeves: tel: +45 3396 3467; fax: +45 3396 3333. e-mail: sar{at}dfu.min.dk.
Fishery management advice has traditionally been given on a stock-by-stock basis. Recent problems in implementing this advice, particularly for the demersal fisheries of the North Sea, have highlighted the limitations of the approach. In the long term, it would be desirable to give advice that accounts for mixed-fishery effects, but in the short term there is a need for approaches to resolve the conflicting management advice for different species within the same fishery, and to generate catch or effort advice that accounts for the mixed-species nature of the fishery. This paper documents a recent approach used to address these problems. The approach takes the single-species advice for each species in the fishery as a starting point, then attempts to resolve it into consistent catch or effort advice using fleet-disaggregated catch forecasts in combination with explicitly stated management priorities for each stock. Results are presented for the groundfish fisheries of the North Sea, and these show that the development of such approaches will also require development of the ways in which catch data are collected and compiled.
Keywords: mixed fisheries, North Sea groundfish, technical interactions
Received 16 December 2003; accepted 19 August 2004.
 |
Introduction |
|---|
 Top Introduction Forecasts for mixed fisheriesCase study: North Sea...ResultsDiscussionReferences |
|---|
Most scientific agencies worldwide, including ICES, have traditionally given fishery management advice on a stock-by-stock basis. Such an approach has long been recognized as potentially problematic because it disregards technical interactions, i.e. cases where more than one species is caught in the same area, and different fleets catching differing proportions of the various species. Ignoring the mixed-species aspect of a fishery can mean, for instance, that the quota for one species is exhausted early in the season, but boats continue fishing and catching that species because there is still quota available for the other species in the fishery. As a result, the quota for the first species would not provide an effective constraint on fishing mortality on that species. Alternatively, the advice might be complete closure to protect a particular stock, which might also result in loss of fishing opportunities on other stocks that might be in a better state.
The mixed-species aspect of fisheries would be better addressed by giving fleet- or fishery-based advice instead of stock-based advice. Such approaches would require more explicit recognition of the fleet structure of fisheries, ideally to the extent that it was reflected at all stages of the sampling, data compilation, and stock assessment processes. The result would be assessments that described trends in exploitation at a fleet as well as at a stock level. As an aside, such an approach would also facilitate advice on ecosystem effects of fishing activity, because such issues are typically fleet- or area-specific. An approach of this nature would take time to implement, but an intermediate step would be one that takes the current, stock-based advice as a starting point, then uses additional fleet information to generate advice that accounts for the technical interactions. This paper documents one such approach. The work was prompted by the situation in the groundfish fisheries of the North Sea at the end of 2002, so that case study is used to illustrate the approach.
The main target species for the groundfish fisheries in the North Sea are the gadoids cod (Gadus morhua), haddock (Melanogrammus aeglefinus), whiting (Merlangius merlangus), and saithe (Pollachius virens), and the flatfish plaice (Pleuronectes platessa) and sole (Solea solea). The fishery can be very broadly characterized as a mixed cod/haddock/whiting fishery centred on the northern North Sea, a mixed flatfish fishery centred on the southern North Sea, and a saithe fishery along the northern margins of the North Sea. However, this is only a very broad generalization, cod, whiting, and plaice in particular having a wider distribution than implied by this picture, with the result that they also feature as by-catches in other fisheries.
At the end of 2002, the ICES advice for the separate groundfish stocks implied complete closure of the cod fishery, and substantial reductions in fishing mortality on most of the other stocks, though allowing for an increase in fishing mortality on saithe. The mixed-fisheries nature is neatly illustrated by the fact that some fleets catch both cod and saithe, so the advice implied that those fleets could both increase their effort (on saithe), and not fish at all (for cod). Although given on a single-stock basis, the ICES advice for North Sea groundfisheries for 2003 also clearly stated a need to account for the mixed-species nature of the fisheries, but the advice did not quantify the technical interactions in them. This led to a need for further advice and forecasts that accounted for the magnitude of these technical interactions, and hence the development of the approach documented here.
|
Forecasts for mixed fisheries |
|---|
|
TopIntroductionForecasts for mixed fisheriesCase study: North Sea...ResultsDiscussionReferences |
|---|
A conventional catch forecast is used as the basis of ICES advice for many of its stocks, and is computationally straightforward (Shepherd and Pope, 2002). The population numbers from a catch-at-age analysis are projected forward using a single vector of fishing mortality-at-age multiplied by a single overall F-multiplier. Using an appropriate F-multiplier, it is possible, in principle at least, to set a total allowable catch (TAC) to obtain the required F.
The implicit assumption of that approach is that there is one fleet targeting one species. If instead there are multiple fleets targeting one species, then things become more complicated as it becomes necessary to consider how the required change in fishing mortality should be distributed across the different fleets. One option is that the same F-multiplier be applied to all fleets. This would be the simplest option, particularly if all fleets only target one species. However, if different fleets target different mixes of species, then it may be considered appropriate to reflect this in the allocation of F-multipliers to fleets. Even if all fleets were targeting the same single species, then the same TAC could also be obtained by applying different F-multipliers to different fleets if there were policy reasons for doing so. Note, this differential allocation of F-multiplier would have no conservation basis (as long as F is reduced, it doesn't matter how), but it would be entirely a policy decision.
Another hypothetical example would be of one fleet targeting multiple species in a mixed fishery. In that case, it would be possible to set individual TACs for each species on the basis given above. The problem comes when conflicting advice leads to different implied F-multipliers in the different TACs for the same fleet. This requires some compromise across the different species, and again it requires some form of external input, such as a policy decision on the relative importance assigned to the different species.
From these examples, it appears possible to overcome the problems attributable to multi-fleet and multispecies aspects of fisheries through assigning priorities to the species and fleets in the fishery, and running forecasts that account for those priorities. In the current context, our starting point is single-species stock assessments and advice, and what is required is to generate advice that accounts for the mixed-species nature of the fishery. Therefore, in addition to the data already in the assessments, we require a relative weighting for each species to reflect the priority assigned to it, together with information on the fleets within the fishery, and the catches by those fleets broken down by species.
To consider first the species priorities, it is possible to envisage a means by which these might be derived objectively, using for example the state of each stock relative to a specific reference point, or the relative economic value of each species. However, many other methods could also be used to derive such weightings, and it must be stressed that there is no scientific basis for choosing between them. The selection of these weightings should be regarded as a policy decision, not a scientific one.
The use of fleet data within a forecast raises questions about how the change in fishing mortality required for a given species will be allocated across the different fleets. As with the species weightings, this is largely a question of priority, and hence represents a policy input rather than a scientific decision. However, the catch compositions of the fleets offer an obvious source of information that can be used to guide such decisions, even in the absence of policy input.
To formalize the problem, the initial stage is to introduce additional subscripts into the conventional catch equation. If k fishing fleets exploit j species at age a, each with partial fishing mortality Fk,j,a for each fleet, then the total annual catch numbers (C) and catch weight (CW) are the sum of catches from each fleet, such that
|
| (1) |
|
| (2) |
To perform a catch forecast for the purpose of calculating a TAC, we require assumed, status quo values for fishing mortality and weight at age, etc., which would typically be estimated using recent average values. These are here indicated by a "prime" superscript (e.g. F'). Other variables refer to quantities for the forecast period.
The forecast also requires the implicit assumption that fishing practices will remain unchanged. The TAC can thus be estimated from
|
| (3) |
|
| (4) |
Now, assume that it is desirable to alter fishing mortality for one species independently of another. A variety of different alterations in fack can be used interchangeably to effect modifications in Fj, and the problem is underdetermined. In general terms, for a single species in a given year, fisheries management will aim to result in a specified change in the absolute fishing mortality on species j. This can be achieved through a variety of means, including closures or effort control, but we will here assume that a TAC is required for the purpose. Assume that Fj for TAC estimation is determined from F'j and a factor, 
Fj, such that Fj = F'j x Fj.
The relative change in each fleet's partial fishing mortality, F'j,k, can be defined by
|
| (5) |
j, is a scaling factor applied for all fleets catching species j, and pk,j represents how the overall effort reduction is allocated across fleets in order to achieve the desired change in fishing mortality for species j. The value pk,j is a type of "weighting factor" for effort reduction which could be supplied externally to reflect policy considerations, such as the ecosystem effects of fishing with a particular gear, or could be estimated on the basis of the catch compositions of individual fleets. As a fleet cannot reduce F to less than zero, the product of alpha and p cannot exceed 1, so the minimum of the product and 1 must be used.
The Fj,k represent the effort modification factor which would be applied to fleet k if management was intended only to apply to species j in isolation of the other species caught in the fishery. For this reason, applying Fj,k as the effort modification factor, fack, in Equation (3) will lead to a unique solution. However, this only applies where management decisions in a multi-fleet fishery are driven by the conservation needs of a single species. Often this will not be the case, and conservation needs for each species give different effort-reduction scenarios. This is reflected in the different Fj,k estimated for each species within a fleet, and it is necessary to find a way of combining them to produce a single fleet effort modifier for each fleet in the fishery. In addition, the derivation of Fj,k does not reflect the extent to which different species are caught by the same fleet, so it is also necessary to account for these technical interactions in some way.
Multicriterion forecasting can only be developed if an appropriate weight is assigned to each conservation criterion. Assume that a decision weight 
j can be developed, representing the importance of the conservation criteria for each of the j species. It would be desirable for these weights to be based primarily on the state of the individual stocks with respect to their precautionary reference points. However, as economic or other criteria are also likely to influence the overall management objectives for the fisheries, the j could also be supplied externally to reflect policy considerations.
In addition to a weighting to reflect the relative importance assigned to each species, it is necessary to reflect the extent to which the different species are caught together. To represent this, we define gk,j as a fleet target factor, describing the relative importance of species j for fleet k. If we also specify that the fleet target factors for a given fleet sum to 1, i.e.
|
| (6) |
The overall fleet effort modifier, fack, as defined in Equation (7), can then be derived as the sum of the Fj,k weighted by the decision weights j, and the fleet target factors gk,j. The overall fleet modifier would thus be:
|
| (7) |
Having specified such decision weights, the calculation for multicriterion TAC setting could then be as follows:
- Specify all relevant decision weights, j, where the sum of j = 1;
- Specify the effort-reduction rates pk,j;
- Specify the desired adjustment factor Fj to status quo fishing mortality for each species, Fj,a = F'j,a x Fj;
- Calculate individual "single species" TACs (STACs) from Fj,a;
- Calculate all Fj,k, the species- and fleet-specific effort factors required to achieve each STACj:
where
j is estimated by minimization of 
- Calculate weighted fleet effort-reduction factors, fack, from Equation (7);
- Calculate each mixed-species TAC from Equation (3).
|
Case study: North Sea groundfish fisheries |
|---|
|
TopIntroductionForecasts for mixed fisheriesCase study: North Sea...ResultsDiscussionReferences |
|---|
Stock status and advice
The starting point for this case study is the assessments and advice for the North Sea groundfish in late 2002 (ICES, 2002; ICES, 2003a). For background purposes, we first summarize the states of the relevant individual stocks, and the advice given for each. The ICES advice for North Sea cod included the statement "ICES recommends a closure of all fisheries for cod as a targeted species or by-catch" (ICES, 2002). For haddock, the advice was "Since haddock is mostly taken in demersal fisheries with cod and whiting, the advice for cod determines the advice for haddock. Unless ways to harvest haddock without by-catch or discards of cod can be demonstrated, fishing for haddock should not be permitted." A similar statement was made for whiting. The advice for plaice included the statement "Management of fisheries taking plaice must respect the stringent restrictions on the catch and discard rates advised for cod, with effective monitoring of compliance with those restrictions". The advice for sole included a similar statement. The advice for the saithe stock did not include any specific reference to cod.
While the advice for most species emphasized the need to protect cod, comments were also made about the management of the stock, assuming that it was possible to manage it independently of the cod stock. Table 1 summarizes this advice, as well as the current state of exploitation of the stock (i.e. current F), and the change in F implied by the advice. The final column in Table 1 gives the required overall fishing mortality multiplier advised for each fleet. These are Fj expressed in percentage terms. The advice implies substantial reductions in fishing mortality for all species except saithe, for which an increase of up to 38% would be possible. However, as the advice for saithe was "Fishing mortality in 2003 should be less than Fpa" (where Fpa for the stock is 0.4), one possible interpretation is that F should have been maintained at the current level, an interpretation adopted for the scenarios run here.
|
Fleet data
Data on species catches by fleet in the North Sea groundfish fisheries are given in ICES (2003a). For the current purposes, the data have been reorganized and edited to facilitate more comprehensive investigation of the modelling approach. Although the resulting data are similar to those given in ICES (2003a), for use here the nation associated with each fleet has been obscured by the use of a code letter, because the intention is merely to illustrate the approach rather than to provide management options for each fleet. The data are total catches of each species for the three most recent years (1999–2001), in weight only; they are not disaggregated by age.
The fleets used are defined on the basis of a combination of nation, gear type, and vessel size. Each fleet is identified by an initial code letter corresponding to a country, together with a three-letter code indicating the gear type, and for some gears a number indicating a nominal size category for the vessels involved. These range from 1 for the smallest vessels, to 3 for the largest. The gear codes are identified in Table 2, the fleets in Table 3.
|
|
Eight different countries are represented in the data set. Each has a subset of the different gears and size categories, leading to a total of 52 different fleets. The species compositions of the fleet catches are illustrated in Figures 1 and 2, which provide an effective illustration of the complexity of the North Sea groundfish fisheries.
|
|
Scenarios
To illustrate the modelling approach documented here, three scenarios (i.e. sets of species decision weights,
j) have been run. They are assumed to be specified externally on the basis of a policy decision, and are given in Table 4.
|
In addition to the
j, the method also requires that fleet effort-reduction rates, pk,j, be specified. As with the j, this is to some extent a policy decision. However, the fleet catch data do give a basis by which fleets can be prioritized for the purpose. The pk,j are used to allocate how the required effort reduction for a given species should be allocated across fleets, and this can be done using the species compositions of the fleet catches in a number of ways. The basic options are:- ignore this information, and treat all fleets equally, i.e.
(8)
- consider catches within the fleet, such that the effort reduction allocated to a fleet reflects the proportion of the species within the fleet's catch, i.e.
(9)
- consider catches within species, such that the effort reduction allocated to a fleet reflects the proportions of the total international catch of the species as a proxy for the partial F exerted by that fleet on that species, i.e.
(10)
It is also necessary to define the fleet target factors, gk,j, that describe the relative importance of species j for fleet k. The simplest way of defining gk,j is to use the mean proportion of species j in the catch of fleet k, i.e.
|
| (11) |
In defining pj,k, the fleet catch information is only used to allocate the effort reduction for a given species across fleets within that species. As such, this is distinct from the use of information on the species catch within fleet to define the fleet targeting factor, gj,k.
|
Results |
|---|
|
TopIntroductionForecasts for mixed fisheriesCase study: North Sea...ResultsDiscussionReferences |
|---|
The results from the three scenarios run with the full data set are summarized in the form of the fleet effort factors, fack, in Figure 3, and by comparing the required single-species F-multipliers with the resultant mixed-species F-multipliers in Figure 4. In scenario A, where highest priority is given to cod, the results imply a substantial reduction in effort for all fleets except those fishing for saithe and, to a lesser extent, flatfish. It also results in overall F-multipliers that are well below the single-species advice, except for that on cod. With scenario B, which essentially disregards the advice for cod, the effort reductions implied are less severe and apply more evenly across the fleets. As mentioned previously, saithe fleets are least affected. The overall F-multipliers for each species resulting from scenario B are close to the single-species advice, except for cod, where the result is an F-multiplier greatly in excess of the advice. The results for scenario C are intermediate between those for scenarios A and B.
|
|
In the results presented in Figure 3, it is clear that in all cases the fleet effort multiplier for fleet 18 (labelled "D_OTB1") is very low. This fleet is essentially a targeted whiting fleet, with some by-catch of cod and plaice. Whiting constitutes 72% of that fleet's total catch, but no more than 32% of the total catch of any other fleet. In that example, the required effort reduction for whiting is allocated according to the proportion of whiting in each fleet's catch. As a result, the fleet receives a very high proportion of the total effort reduction required for whiting, leading to an individual fleet multiplier for whiting of zero in all scenarios. The very low fleet effort multiplier for that fleet in all cases is therefore a consequence of allocating effort reduction according to the proportion of species in the fleet catch in a case where one fleet has an exceptionally high proportion of one species in its catch.
Sensitivity of results to choice of decision weightings
To extend the analyses summarized above, and to illustrate the sensitivity of the results to the choice of the species decision weights, j, a series of runs was made with the decision weight on cod varied continuously between 0 (scenario B) and 1 (scenario A), with the remaining weighting equally distributed across the other species. The results from those runs are summarized by species in Figure 5, and by fleet in Figure 6. Figure 5 shows the F-multiplier that would result for each species under each combination of decision weights compared with the F-multiplier that would result for cod. The species curves converge towards the cod curve as the weighting applied to cod is increased, but the rate at which they converge provides an indication of the extent to which the fishery on the species can be managed separately from that on cod. The curves for haddock and whiting lie very close to the cod curve, reflecting the fact that any restriction on fishing effort on cod will lead to a similar reduction in effort on haddock and whiting. In contrast, the effect on the overall fishing multiplier for saithe is relatively slight for scenarios in which the decision weight applied to cod is 0.8 or less. This indicates scope for allowing a relatively unrestricted saithe fishery, while still implementing substantial effort reduction for cod. The curves for plaice and sole are similar to each other, and are intermediate in form between those for haddock and whiting and that for saithe. This indicates that the fisheries for plaice and sole are less associated with cod than the haddock and whiting fisheries, but are not as independent of cod as is the saithe fishery.
|
|
Figure 6 gives the overall fleet effort factors that would result for each fleet, given each set of species decision weightings. The form of the fleet curves reflects the species composition of the fleet catches, such that fleets that target cod are subject to substantial reductions in effort even when only a small proportion of the decision weight is applied to cod. This leads to fleet curves that are concave. At the other extreme, those fleets with little or no cod catch (typically fleets targeting saithe) have highly convex curves, because they are only subject to significant effort reduction once the proportion of decision weighting applied to cod approaches 100%.
Sensitivity of results to fleet definitions
In the example data set used here, the distribution of catch weight across fleets is highly skewed, with relatively few fleets accounting for a very high proportion of the total catch (Figure 1). If the 52 fleets are ranked in order of decreasing total catch weight, then the top five account for >50% of the total catch, and the top 20 for almost 90%. To investigate the effect that this may have on the results, runs were made using subsets of the data in which the less important fleets were excluded. For this purpose, "importance" was defined by determining the proportions of the total catch of each species accounted for by each fleet, then ranking the fleets by the highest of these. The subsets of data used consisted of the most important 5, 10, 20, 30, and 40 fleets. As a control on the possible effects of reducing fleet numbers on the results, runs were also made with five randomly chosen subsets of the same sample size, to allow comparison. As a performance measure to allow the results of each run to be summarized as a single value, the sum of squares of the differences between the single- and mixed-species F-multipliers was used. By this measure, a smaller value indicates improved performance. The runs used scenario C, with other options as above.
The results from these runs are given in Figure 7. By selecting only the most important fleets, there is very little change in the overall result, even if only the five most important fleets are used. In contrast, with the random subsets, performance deteriorates once the number of fleets decreases below 30.
|
|
Discussion |
|---|
|
TopIntroductionForecasts for mixed fisheriesCase study: North Sea...ResultsDiscussionReferences |
|---|
The basic problem addressed here is how to estimate the fleet effort factors (fack), given the specified constraints on stock-specific fishing mortality and the weighting applied to each fleet and species. As such, the approach is an application of linear programming (LP), a technique previously applied in other fisheries contexts, particularly in relation to mixed-species fisheries. Gundermann et al. (1974) used an LP approach to investigate how to allocate quotas for different species across different fisheries within the North Sea. Brown et al. (1979) considered the same problem in relation to the fisheries off the northeastern USA, and Fukuda (1976) considered theoretical aspects of quota allocation in mixed-species fisheries. Overholtz (1985) considered the potential yield of several demersal fish assemblages on Georges Bank. Murawski and Finn (1986) used an LP approach to investigate optimal allocation of effort across six different mixed-species trawl fisheries in order to meet single-species fishing mortality constraints, and Spencer et al. (2002) used LP to investigate the effect of halibut by-catch constraints on the potential yield of flatfish in the shelf fisheries of the eastern Bering Sea.
The approach described here differs from those described above primarily in the use of explicit weightings by fleet (pj,k) and species (j). This means that there is no overall objective function optimized in the present approach. The minimization done at stage E of the algorithm is used only to estimate the common scaling factor j. The fleet weightings, pj,k, constrain the means whereby changes in fishing mortality will be distributed across fleets, and hence reduce the tendency towards "solutions which are extreme, sparse and ruthless", a comment of Shepherd and Garrod (1982). This tendency is apparent in the results of Murawski and Finn (1986) where, in most of the scenarios investigated, two fisheries were "switched-off" completely. The use of the fleet and species weightings removes the need to categorize fisheries as either targeted or by-catch, as in Brown et al. (1979).
By considering data at a fleet level, the approach documented here provides a natural basis for advice on fleet-based effort management, although it is equally straightforward to translate this into TAC advice. Effort-based management approaches potentially provide a more robust approach for dealing with mixed-species fisheries than quota-based approaches (Murawski and Finn, 1986). However, the North Sea stocks considered here fall within the remit of the EU's Common Fisheries Policy (CFP). A key component of this is the principle of relative stability (Holden, 1994), which ensures equality of fishing opportunities across nations. This requires that some measure of fishing opportunities is required each year in order that this can then be divided between participating nations. In practice this means that TAC-based advice, rather than effort-based advice, is required for each stock. The resultant TACs are then divided between nations according to fixed allocation keys. This imposes limits on the extent to which fleets can be managed separately, although Shepherd (2003) recently outlined a scheme that could allow relative stability to be expressed in terms of effort, rather than total catch.
A TAC has two interlinked roles. It provides a measure of the overall fishing opportunity, which can then be allocated to interested parties, and it can serve to limit fishing activity through providing an upper limit on fishing opportunities. Many of the earlier applications of LP (Gundermann et al., 1974; Fukuda, 1976; Brown et al., 1979) were concerned with the allocation aspect, i.e. establishing how a pre-determined TAC should be allocated across fleets. In the current context the allocation is pre-determined, at the national level at least, and we are concerned with setting TACs that meet the conservation needs of the different stocks.
In the example application made in this paper, the decision weights (j) have been taken as an external policy input. This reflects the fact that the TAC-setting process for North Sea stocks includes a negotiation stage (Holden, 1994; ICES, 2001), which means that policy input has a role. With the wider adoption of harvest rules as the basis for advice for European fisheries, something anticipated under the revised CFP (European Commission, 2002)1, it will become appropriate to derive more objective ways of defining the weightings. One approach might be to derive an overall stock weighting as the product of a species weighting (reflecting the "value" of that species in, for example, economic or ecological terms), and a conservation weighting (reflecting the current state of the stock in relation to conservation reference points).
The mixture of species caught by a fishing vessel will depend on the area and season in which it is fishing, and on the gear it is using. These may change in response to changes in the relative availability of a target species, where this in turn could reflect stock abundance or availability of fishing opportunities (i.e. quota) relative to that abundance. The approach summarized here does not capture the potential for the fleet to change its fishing practice in response to changing circumstances. Instead, it uses the assumption of an unchanged exploitation pattern, which is typically also used in a conventional forecast (Rivard and Foy, 1987). In this respect, this limitation reflects the need to perform a catch forecast in order to set a TAC, so it is a problem associated with management by TAC rather than a specific problem of the approach.
If there are strong spatial and seasonal influences on a fleet's catch composition, then these may offer alternative ways of addressing the mixed-fishery problem, i.e. through closure of areas/seasons which have relatively high catches of the species of most conservation concern. An extension of this would be the complete closure of all fisheries in an area once the quota for one target species is exhausted. This corresponds to the "weak stock management" approach in use for fisheries off the west coast of the USA, which means that harvest of healthier stocks must sometimes be curtailed to prevent overfishing or to rebuild overfished stocks. Another attempt to address mixed-fisheries issues in quota management situations was the "two-tier" approach adopted by ICNAF (Brown et al., 1979; Anthony and Murawski, 1986). This involved overall quotas designed to be 15–20% less than the sum of the individual species TACs, with the intention of making some allowance for by-catch mortality and species interactions, as well as permitting some stock recovery (Anthony and Murawski, 1986).
For all scenarios of the North Sea case study presented here, the mixed-species F-multipliers that result for cod are close to those for haddock and whiting, and the values for plaice and sole are similarly linked, whereas those for saithe are largely independent of those for other species. These results reflect the broad categorization of the North Sea fisheries apparent in the data used here, i.e. a mixed cod/haddock/whiting fishery, a flatfish fishery, and a fishery for saithe. This may be an over-simplification of the actual fisheries, but it is a reflection of the fact that the results of the scenarios are largely dependent upon a few major fleets, with most fleets having little influence. This is an issue relating to the data available for the particular example investigated here. It suggests that it may be more desirable for fleets to be defined such that they are generally all rather similar in "size", i.e. total catch or proportion of total species catch. However, it would be desirable to investigate this with other data sets before making more general conclusions, not least because the domination of the fisheries by a small number of major fleets may be a reasonable picture for the North Sea. Further, it is clear that TAC management alone would not provide any significant protection for cod unless there were also substantial reductions of the TACs for associated species, primarily haddock and whiting.
Technical interactions have been accounted for in a large number of modelling approaches (see, for example, the brief review in ICES, 2003b). In relation to previous work, the current approach represents a generalization of an age-structured catch forecast that allows for explicit inclusion of policy priorities for species and fleets. It is this aspect that makes the approach applicable in cases where advice for different species in mixed fisheries is given on a stock-by-stock basis, with no accounting for technical interactions. It should, however, be considered as an interim approach, pending the development of a more integrated, fishery-based approach to the assessment and advisory process.
The selection of fleets for use in the forecasts is a complex question that requires some trade-off between the need to classify vessels according to similarity of fishing practice (e.g. a posteriori multivariate definition of métiers according to catch composition) and the need to be able to assign vessels to categories a priori for management purposes. The latter requirement is strong, and perhaps the best approach would be to use relatively coarse fleet definitions based, for example, on vessel size and gear for forecast purposes, but to use more detailed analyses to identify spatial and seasonal aspects of mixed fisheries that might facilitate their management (Murawski et al., 1983; Murawski and Finn, 1988). One recent study (Ulrich and Andersen, 2004) demonstrated considerable flexibility of activity among Danish fishing vessels, and concluded that separation into distinct management units on the basis of catch compositions was difficult. Such a conclusion implies that fleet definitions based on vessel and gear characteristics might be of more practical use than definitions based on catch information. The question of fleet definition for forecasts is considered further in ICES (2003b, 2004).
The definition of fleets involves practical as well as scientific considerations concerning, for instance, the design of national schemes for data collection. These may impose limits on how fleets may be defined for individual nations, and in particular on whether age-disaggregated data are available for all fleets. In the short term, such practical considerations are likely to be at least as important as more scientific considerations in the question of fleet definitions.
|
Acknowledgements |
|---|
Much of the development of the approach documented here took place during the STECF SGRST meetings on mixed fisheries in Brussels in October 2002 (STECF, 2002)2 and October 2003 (STECF, 2003)3, and development was also assisted by discussion in relation to the ICES Study Group on the Development of Fishery-based Forecasts. The authors thank participants in those meetings for their contributions, and also acknowledge the contributions of two anonymous referees, whose comments led to substantial improvements to the manuscript. The results and conclusions drawn do not necessarily reflect the views of the authors' employers. The method described here corresponds to method 2 of STECF (2002). Further background to the work is given in ICES paper CM 2003/V: 01, which the current paper supersedes. The model is implemented in the R programming language, and is called MTAC. It can be found at http://www.ices.dk/committe/acfm/wg/asoft/MTAC/, along with a test data set.
|
Footnotes |
|---|
1 At: http://europa.eu.int/comm/fisheries/doc_et_publ/liste_publi/facts/reform_en.pdf.
2 At: http://europa.eu.int/comm/fisheries/doc_et_publ/factsheets/legal_texts/docscom/en/sec_2002_1373_en.pdf.
3 At: http://europa.eu.int/comm/fisheries/doc_et_publ/factsheets/legal_texts/docscom/en/sec_2003_1428_en.pdf.
|
References |
|---|
|
TopIntroductionForecasts for mixed fisheriesCase study: North Sea...ResultsDiscussionReferences |
|---|
-
Anthony V.C. and Murawski S.A. (1986) Managing multispecies fisheries with catch quota regulations – the ICNAF experience. ICES Cooperative Research Report 139:42–57.
Brown B.E., Brennan J.A., Palmer J.E. (1979) Linear programming simulations of the effects of bycatch on the management of mixed species fisheries off the northeastern United States. Fishery Bulletin, US 76:851–860.
European Commission. (2002) Communication from the Commission on the Reform of the Common Fisheries Policy – Roadmap.
Fukuda Y. (1976) A note on yield allocation in multi-species fisheries. ICNAF Research Bulletin 12:83–87.
Gundermann J., Lassen H., Nielsen E. (1974) Splitting catch quotas of several species on a number of fisheries using linear programming. ICES Document, CM 1974/F: 46.
Holden M. (1994) The Common Fisheries Policy. Origin, Evaluation and Future(Fishing News Books, Oxford) 288 pp.
ICES. (2001) Report of the Working Group on Fishery Systems. ICES Document, CM 2001/D: 06.
ICES. (2002) Report of the Advisory Committee for Fishery Management, October 2002. ICES Cooperative Research Report, 255.
ICES. (2003a) Report of the Working Group on Demersal Stocks in the North Sea and Skagerrak. ICES Document, CM 2003/ACFM: 02.
ICES. (2003b) Report of the Study Group on the Development of Fishery-based Forecasts. ICES Document, CM 2003/ACFM: 08.
ICES. (2004) Report of the Working Group on Demersal Stocks in the North Sea and Skagerrak. ICES Document, CM 2004/ACFM: 07.
Murawski S.A. and Finn J.T. (1986) Optimal effort allocation among competing mixed-species fisheries, subject to fishing mortality constraints. Canadian Journal of Fisheries and Aquatic Sciences 43:90–110.
Murawski S.A. and Finn J.T. (1988) Biological bases for mixed-species fisheries: species co-distribution in relation to environmental and biotic variables. Canadian Journal of Fisheries and Aquatic Sciences 45:1720–1735.
Murawski S.A., Lange A.M., Sissenwine M.P., Mayo R.K. (1983) Definition and analysis of multispecies otter-trawl fisheries off the northeast coast of the United States. Journal du Conseil International pour l'Exploration de la Mer 41:13–27.
Overholtz W.J. (1985) Managing the multispecies otter trawl fisheries of Georges Bank with catch optimization methods. North American Journal of Fisheries Management 5:252–260.[CrossRef]
Rivard D. and Foy M.G. (1987) An analysis of errors in catch projections for Canadian Atlantic fish stocks. Canadian Journal of Fisheries and Aquatic Sciences 44:967–981.
Shepherd J.G. (2003) Fishing effort control: could it work under the Common Fisheries Policy? Fisheries Research 63:149–153.[CrossRef][Web of Science]
Shepherd J.G. and Garrod D.J. (1982) Modelling the response of a fishing fleet to changing circumstances, using cautious nonlinear optimization. Journal du Conseil International pour l'Exploration de la Mer 39:231–239.
Shepherd J.G. and Pope J.G. (2002) Dynamic pool models. 2. Short-term and long-term forecasts of catch and biomass. In Hart P.J.B. and Reynolds J.D. (Eds.). Handbook of Fish Biology and Fisheries. 2. Fisheries(Blackwell, Oxford) pp. 164–188 410 pp.
Spencer P.D., Wildebuer T.K., Zhang C.I. (2002) A mixed-species yield model for eastern Bering Sea shelf flatfish fisheries. Canadian Journal of Fisheries and Aquatic Sciences 59:291–302.
STECF. (2002) Report of the Subgroup on Resource Status (SGRST) of the Scientific, Technical and Economic Committee for Fisheries (STECF). Mixed Fisheries. Brussels, October 2002.
STECF. (2003) Report of the Ad Hoc Working Group on Mixed Fisheries, Brussels, October, 2003.
Ulrich C. and Andersen B.S. (2004) Dynamics of fisheries, and the flexibility of vessel activity in Denmark between 1989 and 2001. ICES Journal of Marine Science 61:308–322.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Duarte, M. Azevedo, and M. Afonso-Dias Segmentation and fishery characteristics of the mixed-species multi-gear Portuguese fleet ICES J. Mar. Sci., April 1, 2009; 66(3): 594 - 606. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Pilling, L. T. Kell, T. Hutton, P. J. Bromley, A. N. Tidd, and L. J. Bolle Can economic and biological management objectives be achieved by the use of MSY-based reference points? A North Sea plaice (Pleuronectes platessa) and sole (Solea solea) case study ICES J. Mar. Sci., September 1, 2008; 65(6): 1069 - 1080. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Marchal A comparative analysis of metiers and catch profiles for some French demersal and pelagic fleets ICES J. Mar. Sci., May 1, 2008; 65(4): 674 - 686. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Ulrich, B. S. Andersen, P. J. Sparre, and J. R. Nielsen TEMAS: fleet-based bio-economic simulation software to evaluate management strategies accounting for fleet behaviour ICES J. Mar. Sci., May 15, 2007; (2007) fsm044v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Reeves and M. A. Pastoors Evaluating the science behind the management advice for North Sea cod ICES J. Mar. Sci., May 1, 2007; 64(4): 671 - 678. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-J. Ratz, E. Bethke, H. Dorner, D. Beare, and J. Groger Sustainable management of mixed demersal fisheries in the North Sea through fleet-based management a proposal from a biological perspective ICES J. Mar. Sci., May 1, 2007; 64(4): 652 - 660. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||