Quick link to: EFIMAS home, WP4, CS9

Case study 9: Approach 2b: A model-based evaluation of Marine Protected Areas for fishery management in the case of strong environmental forcing - the example of Eastern Baltic cod

Data and parameters

Stock data

the long-term average distribution pattern of the youngest egg stage (Bagge et al., 1994; Hinrichsen et al., 2007)

hydrodynamic model developed by Lehmann (1995), which provides current velocities and environmental variables (temperature, salinity, oxygen)

ICES Baltic International Trawl Survey database.

Fleet data

The exploitation model was parameterised from logbook data from the five countries (83% of total catches) Poland, Sweden, Denmark, Germany and Latvia for which data were available. The data used was the same as used in the FLR-modelling under CS9, Approach 2.

Modelling

ISIS-Fish (Mahévas and Pelletier, 2004; Pelletier and Mahévas, 2005) is a model of fisheries dynamics based on three sub-models: a population model, a fishing activity model and a management model. Each sub-model is spatially explicit and operates on a monthly time step. Within the fishery region, zones are independently defined for each population area (spawning grounds, nursery areas, feeding grounds), each fishing activity and each management measure. Seasons (i.e., sets of successive months) are also defined for each population group (age classes), each fishing activity and each management measure. Within each zone and season, fishing effort and population abundance are assumed to be homogeneously distributed. Seasonal migrations between population zones are considered and zone-specific catchability depends on seasons. The exploitation model calculates the standardised effort per fishing activity affecting the population in each zone and month. In ISIS-Fish, fishing units are not individually identified but grouped into fleets and exploitation is described by métiers and strategies. A métier is characterised by a combination of gear, target species, zone and season. Fishing effort is standardized between gears, and a selectivity model is defined for each combination of gear and species, with a parameter that can possibly be modified through management measures, e.g. mesh size. Vessels that practise a similar sequence of métiers during the year constitute a fishing strategy that is characterised by a seasonal allocation of fishing effort between métiers. The management model describes the management scenario considered, its impact on the fishing activity in particular due to fisher’s response to management. At each time step the model calculates changes in the distribution of fishing effort among métiers of a strategy and generates the corresponding catch and abundance estimates for each zone. Further detail and the equations are given in Pelletier and Mahévas (2005).

Conditioning of Operating Model
Biological Parameters

The cod population was structured into eight age classes the last one being a plus-group, i.e. all fish older than 8 years accumulate in this class. Von Bertalanffy growth curves were defined for each environmental scenario based on observed weight-at-age data. Similarly, mean weight–at-age for each environmental scenario was obtained by fitting exponential weight-at-age curves. Natural mortality was assumed to be 0.2 for all age classes 2-8. Spawning stock biomass was calculated from stock numbers and scenario-specific observed maturity ogives for each spawning ground, summed over all areas and fed into scenario-specific Beverton-Holt stock-recruit relationships. Age 2 recruits were then redistributed to nursery areas using observed distribution patterns from the ICES Baltic International Trawl Survey database.

Fisheries & Fleets

Data were available per month, ICES statistical rectangle, vessel size group, gear type and country. Three vessel size groups were considered: <12m, 12-24m, >24m. An average trip duration was assigned to each of these groups, and three main gears were considered, namely trawl, gillnet and “other gears”, the latter mainly comprising longlines. Selectivity curves for the main gears (trawls and gillnets) were taken from Nielsen et al. (2008, in revision).

The standardisation factor of a gear Fstd quantifies the ratio in overall catch between each gear and a reference gear (i.e. the difference in efficiency between gears, all other things being equal). Standardisation factors were estimated for each gear by fitting a Generalized Linear Model (GLM) to log-book Catch Per Unit Effort (CPUE) data. The model is loglinear with factors gear, month and zone, including an interaction between month and zone. As the only target species in the model was cod, metiers were defined based on fishing zones, gears used and fishing seasons for that species. In total 23 metiers and 19 corresponding fishing zones were identified. A fishing zone was defined as a group of contiguous statistical rectangles comprising at least 80% of the fishing effort of that métier.

Catchability in ISIS-Fish is defined as the probability that a fish present in a specific zone during a season is caught by a standardised effort unit from a non-selective vessel (see Pelletier and Mahévas, 2005 for a discussion). Catchability coefficients were fitted by calibrating the model against total quarterly catches over an arbitrarily chosen period of two years (2002-2003; Figure 3)

Management

Management options considered in the model comprised the exclusion of fishing effort at different temporal and spatial scales. Four scenarios were simulated over periods of 20 years, each under favourable and unfavourable conditions for cod reproduction - (i) no closures, (ii) spawning closure in the Bornholm basin in combination with a closed season based on the IBSFC management plan for 1995, (iii) three small spawning closures plus a closed season based on the management plan proposed by the EU Commission for 2007, (iv) year-round spawning ground closures in the Bornholm Basin and Gdansk Deep.

Results

Under favorable environmental conditions a simulation without closures showed a stock recovery to levels around Bpa after 18 years, even when the effort was increased to account for illegal landings and discarding. This indicates that the present total effort would be sustainable on the long run under such conditions.

On the contrary, under unfavourable environmental conditions, none of the proposed or implemented closure scenarios was able to recover the stock even to Blim. Such a scenario of consistently low recruitment might be overly pessimistic as even during long stagnation periods, infrequent inflows were observed.

As both population and exploitation models are subject to some uncertainties, the interpretation of SSB and yield should be cautious. Knowing that, our results demonstrated that closed seasons of the entire fishing area had a much larger impact on recovery rates, final stock sizes and yield compared to regionally restricted spawning area closures. A possible reason for the limited impact of spawning closures might reside in the effort reallocation rule implemented in our model.

In summary, our results showed that the ISIS-Fish model for Baltic cod is producing scenarios of stock and yield development in a realistic order of magnitude and comparable to past projections (ICES, 2005). The 1995 closure scenario provided an option to compare simulation results to data from stock assessment during years where the simulation period overlapped with the assessment (ICES, 2005). The comparison yielded a 50% higher SSB of the simulation at the end of the overlap period. The difference might be explained by the lower exploitation rates in the model due to a longer closed season in the simulation compared to the closed seasons implemented in the real world during most of the overlap years. Considering this difference, both estimates correspond well and proof the validity of our model.

Despite the strong and obvious influence of environmental conditions, we further conclude that conditioned on model assumptions for effort reallocation, the reduction of effort and thus fishing mortality as imposed by closed seasons is more efficient at stock recovery rather than reduction of spawner disturbances through the implementation of spatially restricted spawning closures. An effective, traditional management regime may thus be a viable alternative to the MPA design currently implemented in the Baltic Sea

Dissemination

Gerd Kraus, Dominique Pelletier, Julien Dubreuil, Christian Moellmann, Hans-Harald Hinrichsen, Francois Bastardie, Youen Vermard and Stéphanie Mahevas. 2008 (Accepted). A model-based evaluation of Marine Protected Areas for fishery management in the case of strong environmental forcing - the example of Eastern Baltic cod (Gadus morhua callarias L.). Accepted, Symp. Proc. ICES J. Mar. Sci.

The accepted paper can be viewed here:
Paper with results

Links to Other Work

The present study links up to the ISIS-Fish applications under EFIMAS WP4 Case Study 6 and Case Study 4. The work has been done under the EU FP6 PROTECT project in cooperation with the EU-FP6-EFIMAS and EU-FP6-UNCOVER Projects as well as partly in cooperation with a Danish national government project. (See also under acknowledgements).

References

Mahévas, S., and Pelletier, D. 2004. ISIS-Fish, a generic and spatially explicit simulation tool for evaluating the impact of management measures on fisheries dynamics. Ecological Modelling, 171: 65-84.

Acknowledgements

The present study evaluates aspects of the Baltic cod fishery and population biology mostly based on results from the Commission of the European Communities Specific Targeted Research Project (STREP) EU-FP6-513670 (PROTECT). Additional results were obtained through STREP projects EU-FP6-502516 (EFIMAS), and EU-FP6-022717 (UNCOVER). We are especially indebted to project participants providing data sets, to Benjamin Poussin for immediate online software support and to J. R. Nielsen, P. J. Sparre and S. Lehuta for valuable discussions while building the model. Furthermore, we would like to thank for the contribution of fleet based catch and effort data for all Baltic institutes participating in EFIMAS.

EFIMAS Contribution to the work

About 20% of this work is done under EFIMAS.

Participants:

Gerd Kraus (DTU-Aqua), Dominique Pelletier (IFREMER), Julien Dubreuil (IFREMER), Christian Moellmann (Hamburg University), Hans-Harald Hinrichsen (IFM-GEOMAR), Francois Bastardie (DTU-Aqua), Youen Vermard (IFREMER) and Stéphanie Mahevas (IFREMER)

 
efimas1/wp4/cs9/appr2b/main.txt · Last modified: 2008/07/07 23:15 by admin
 
Except where otherwise noted, content on this wiki is licensed under the following license:CC Attribution-Noncommercial-Share Alike 3.0 Unported
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki