Quick link to: EFIMAS home, WP4, CS1

Case study 1: Approach 2 - Accounting for fleet behaviour and effort allocation in the North Sea flatfish fisheries using the TEMAS software

Prior to FLR development, management scenarios for the North Sea flatfish fisheries were performed within the EU FP5 TECTAC project, using the existing TEMAS software. This work was pursued within EFIMAS. The main focus is on a fine description of fleets capacity and activity, including effort allocation across the various metiers (gear, mesh size, target species), and bioeconomic consequences of management scenarios. One main purpose of this approach is to understand to which extent mis-specifying or not taking into account fleets dynamics and adaptation to external factors (stock abundance, marked price, management actions) may affect the results of management scenarios.

Data and parameters

Stock data

Only sole and plaice are considered here. Other species are not explicitly modelled, but are accounted for through a constant economic parameter reflecting their value in the revenue per effort in each metier.

Fleet data

11 fleets are explicitly included, encompassing the most important Dutch, English and Danish flatfish fleets. The number and definition of the fleets to be explicitly included in the model is the result of an extensive trial and error process, to get the right balance between model tractability and transparency in one hand, and actual difference in behaviour in the other hand. All other national and international sources of fishing mortality on sole and plaice are aggregated into some “other fleets”. 25 national metiers were defined, based on gear and mesh size used, sometimes supplemented by the target species identified from national catch and effort data. Each fleet engage in 3 to 5 different metiers over time.

Fleet data were available from 1998 to 2003.

Modelling

TEMAS model

The TEMAS model is described in details in Ulrich et al. (2007). It combines a MSE (Operating Model plus Management Procedures) with a module for fleets behaviour. The model is run on a monthly time step over the period.

The cornerstone of the operating model is the effort module, which links the biological, economic and behaviour modules together. The fishing process is modelled through both the fleet level and the metier level. A fleet can engage in several metiers during each time period. The effort allocated to each metier is given by the product of the overall effort by fleet multiplied by the probability of choosing a given metier. In this case study, the overall effort is kept constant. The metier probability for a given fleet is derived from the Random Utility Model (RUM), which assumes that a fleet will choose the alternative that gives the highest utility among a finite number of alternatives.

Conditioning of Operating Model

Historical Estimates of Time Series
Biological Parameters

Von Bertalanffy growth, maturity ogive and recruitment models have been fitted to ICES WGNSSK (2005) data.

Stock Recruitment Relationships

Recruitment is assumed to take fully place at the beginning of the year, and is recalculated to age 0 assuming constant natural mortality pattern. As such, the whole cohort is modelled.

Fisheries & Fleets

The length-based selectivity logistic ogives were taken from the literature. Undersized discarding ogives were estimated empirically from onboard sampling or from the literature.
The Utility function for effort allocation is based on a reduced set of significant explanatory variables selected from empirical analysis of national catch and effort data. These includes the average value per unit-of-effort during the previous time step (as proxy for economic attractiveness of alternative choices), and information on the fleet’s fishing pattern during the previous time step (as proxy for recent knowledge) and one year earlier (as proxy for seasonality and tradition).

Economics

Fleets economic results are expressed by estimates of revenue and contribution margin. Prices by species were estimated separately for each country and are fixed for the whole simulation period. A constant parameter accounting for the revenue per unit of effort from species other than sole and plaice was estimated for each fleet and fishery, in order to include all sources of revenue for the fleets in the behaviour model. The contribution margin was calculated as the total revenue minus the variable costs. The variable cost are divided into i) variable costs proportional to effort (fuel, gear maintenance and other pre-split cost), ii) costs proportional to landings (e.g. various landing costs and fees) and iii) crew share as proportional to total revenue of the fleet. The costs data were provided by the national research institutes for fisheries economics.

Reference Points

Those from the ICES Precautionary Approach Harvest Control Rules (HCR).

Observation Error Model

No observation error is assumed

Management Procedure

the assessment is a simplified VPA model. The management module applies the ICES Precautionary Approach Harvest Control Rules (HCR) and the annual TAC’s are sat on the basis of the short-term forecast following the simulated assessment process. The TAC is initially splited between countries according to the principle of relative stability. We assume also a relative stability across fleets within each country, where fleets share are equal to those observed in 2003. The key issue of mixed-fisheries is the conflicting single-species management objectives and mismatching quota shares for the individual fleets, where the fleets cannot fully avoid harvesting species together when quota levels do not match. In accordance with the observed patterns in the flatfish fisheries (ICES, 2005), it is assumed that the fishermen will fish at the highest level authorized implying that any excess ‘over quota catch’ of some stocks is discarded (no revenue).

Results

Some single deterministic simulations were run, investigating the effects of using constant effort allocation versus RUM-based dynamics allocation. Scenario used a unique random recruitment draw simulating a decrease of the most valuable species (sole), in order to evaluate the potential reallocation of effort towards plaice. Furthermore, robusteness trials were conducted on the sensitivity of results to the estimated coefficient of economic attractiveness β in the utility function, i.e. what would happen if the fleets were much more flexible to switch among metiers and would chose every month the metiers giving the highest value instead of being mainly driven by tradition and seasonality.

modelfit_galwayppt.jpg fleeteffect_galwayppt.jpg

Figure 1. Observed and predicted stock dynamics over 1998-2003. Figure 2. Partial F by stock (top) and % of difference in partial F when applying behaviour model



Over the period 1998-2003 (figure 1), the model is able to reproduce main trends of the historical ICES data. The behaviour model affect marginally the results of the simulationsat at the stock level, estimating slightly higher catches and F and lower SSB for plaice. However, differences occur at the fleet level (figure 2). The main fleets (Dutch beam trawl) are little affected by the behaviour model, but some minor fleets show some significant changes in their dynamics, increasing their exploitation on plaice stock when sole stock decreases. These differences are explained by the overall flexibility of the fleets considered. The Dutch beam trawl fleets have no other alternatives than these two stocks and metiers, and cannot therefore show major variations in their dynamics. On the contrary, minor fleets such as danish gillnetters are less dependent on these only two flatfish stocks, and can allocate their effort to a wider range of choices, explaining thus a greater response to changes in sole biomass.


increasedbeta_galwayppt.jpg Figure 3. Effect on increased economic attractiveness on stock dynamics. Left:plaice, right:sole. Top:SSB, bottom:F

Robustness trials were conducted by increasing the parameter of economic attractiveness (β factor of VPUE) in the utility function by a factor 5, 10, 20 and 40, corresponding to a higher response of fishermen towards most profitable metiers and less weight on tradition. It appears clearly (figure 3) that such higher flexibility would imply an increased fishing pressure on plaice when sole biomass decreases. The main change in the dynamics is observed when changing from factor 10 to 20, but not when changings to values lower than 10 or higher than 20. That means that there is a clear need to identify whether the fleets are considered as mainly flexible and revenue-oriented, or mainly conservative and tradition-oriented. This qualitative categorisation matters strongly on simulation results, but the exact quantifcation of the β parameter within each category is not very important.

Dissemination

This work was preented to the ICES symposium on Fisheries Management Strategies, 27-30 june 2006. See the presentation slides.ppt. An article is soon to be submitted.

Ulrich, C., Andersen, B.S., Sparre, P.J., and Nielsen, J.R. 2007. TEMAS: fleet-based bioeconomic simulation software to evaluate management strategies accounting for fleet behaviour. ICES J. Mar. Sci., 64: 647-651. Ulrich et al. (2007).

Links to Other Work

This work has been initiated under and conducted as a cooperation with the EU FP5 TECTAC Project.

References

Ulrich, C., Andersen, B.S., Sparre, P.J., and Nielsen, J.R. 2007. TEMAS: fleet-based bioeconomic simulation software to evaluate management strategies accounting for fleet behaviour. ICES J. Mar. Sci., 64: 647-651. Ulrich et al. (2007).

Acknowledgements

EFIMAS Contribution to the work

This work has been initiated under and conducted as a cooperation with the EU FP5 TECTAC Project.

Participants: Bo Sølgaard Andersen, Youen Vermard, Clara Ulrich, Holger Hovgaard, Dave Bromley, Simon Mardle, Jan-Jaap Poos, Per Sparre, Hans von Oostenbrugge

  1. Clara Ulrich 2007/02/08 11:02

J. Rasmus Nielsen 2008/07/30 14:29

 
efimas1/wp4/cs1/appr2/main.txt · Last modified: 2008/11/16 00:27 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