Quick link to: EFIMAS home, WP4, CS9

Case study 9: Approach 2c: The TEMAS “Evaluation Frame” EF

The TEMAS “Evaluation Frame”(EF) of fisheries management regimes (Ulrich, Andersen, Sparre, and Nielsen, 2007) is presented and further developed through model application for the Baltic cod fisheries (Sparre, 2008b), together with the simulation model behind it (Sparre, 2008a,b). The EF executes two parallel simulations of the reference system and the alternative system. Each system comprises an “operational model” which simulates input data to a “management model”. The operational model system simulates the “true world”, whereas the management model simulates the advisory process of ICES combined with the management procedures of EU. The simulations may be executed in deterministic modes as well as in stochastic mode. The management regimes are compared by aid of a suite of “measures of performance”, which are defined by the various groups of stakeholders. Examples are the traditional measures of ICES, the Spawning Stock biomass and the average fishing mortality. Other measures can be bio-economic measures such as the net present value of cash flow, or employment measured in man-years. The EF is demonstrated for marine protected areas (MPA), closed seasons, and restrictions on maximum number of sea days (effort regulation) as management tools for the Baltic cod fisheries. The purpose of these MPAs and effort regulations is the recovery of the Baltic cod stock. This approach is in detail described in Sparre (2008a,b).

The principal components of TEMAS for one time period of a dynamic process
The principal components of TEMAS for one time period of a dynamic process

Data and parameters

Stock data

Population data originates from the ICES WGBFAS (2007) to fit stock-recruitment (SSB-R), maturity-at-age, and weight-at-age relationships. The eastern Baltic Cod decline is supposed to be the result of partly a change in the Baltic Sea abiotic conditions affecting the cod reproductive success and its growth rate. Accordingly, ‘good’ and ‘bad’ forcing environments were identified for favourable and unfavourable conditions for cod reproduction, respectively, and led to differentiate two set of values for population parameters depending on the years of inflow (‘good years’; Köster et al., 2005) from the North Sea to the Baltic Sea.

Fleet data

The exploitation part sub-model including the initial allocation of effort was conditioned from the monthly- and ICES square- disaggregated logbook data for the main participating countries in the Baltic cod fishery (Denmark, Sweden, Latvia, Poland, Germany), and the missing effort from other unreported countries was completed by calibrating on basis of the ICES WGBFAS (2007) eastern cod stock landings. The international fleet based landings and effort data dis-aggregated on month and ICES square was obtained under the cooperation between the EU-FP6 EFIMAS, PROTECT and UNCOVER projects.

A selection of input parameters of TEMAS have been made stochastic variable by multiplication with a “stochastic factor” with mean value 1.0 and a standard deviation, which is an input parameter to TEMAS (the blue cells in the input worksheet). This accounts for example for the stock-recruitment relationship, the von Bertalanffy growth parameter K, the fish condition factor, the fleet based catchability by year and stock, a environmental factor determining high recruitment (inflow years) or low recruitment (stagnant years).

Modelling and parameters

Scenarios evaluated

The scenario-evaluations under this approach mainly cover effort regulation - either direct or indirect effort regulation, where the latter is through introduction of different scenarios for closed seasons and areas in the Baltic Sea cod fishery.

The five pairs of regime comparisons of the current TEMAS program
The five pairs of regime comparisons of the current TEMAS program.

In the standard implementation of TEMAS, five pairs of alternative management regimes are considered. The five pairs or regime comparisons suggested here may not be the most relevant examples and should be considered illustrations of the concepts, rather than the only examples for TEMAS.

Dimensions and parameters in modelling

The main dimensions of the model are four stocks (the eastern and western Baltic cod, sprat, and an ”other stock” component), four areas (the eastern and western Baltic Sea, a given defined closed area for fishery, and an “other area”), 6 seasons of two-month-periods, and trawl and gillnet fleets (and an “other fleet” component) by country and vessel length group ( > 12 m, 12-24m, > 24 m). Additionally, the model is dimensioned to a low row of fisheries defined from the combination of fleet and gear mesh sizes. A number of riggings are defined according to the fisheries. The model include an “other” component for fleets, riggings and countries as well.

TEMAS offers the opportunity to account for spatial aspects, in the sense that fish and fleets can be allocated to a number of areas in a given time period. TEMAS uses a simple “box-model” to handle spatial aspects. TEMAS however, is not suited for handling of a large number of areas. It is not anticipated that TEMAS applications will use more than, say, 10 divisions of the total area.

Operational model and parameters

The TEMAS operational model is a multi-species, multi-fleet dynamic software implementation of a bio-economic stochastic simulation model, which focus on the analysis of the effect of technical management measures. Technical management measures, however, cannot be analysed in isolation from other factors influencing the fisheries system. Therefore, TEMAS covers all essential components of the fisheries system, seen from the angle of a fisheries manager.

TEMAS focuses on the description of fishing fleets and their technical activities, rather than anything else of the ecosystem/fisheries/fisheries economics complex. The focal point in TEMAS are the vessels of the fleets. The idea is that the basic instrument for fisheries management is the capacity of fishing fleets, which in turn is controlled by controlling the number of vessels in each fleet. The number of vessels determines the upper limit of the effort that can be exerted, and the maximum effort determines the upper limit of the fishing mortalities, the effort can create. This way of thinking is somewhat different from the traditional ICES approach of evaluating fish stocks, where the system starts with the input F, without much consideration on what created the F and what controls the F. In this context, technical management measures become a detail, which can to certain degree modify the overall F created by fishing capacity.

Fishing Mortality = (Number of boats) * (Number of Sea-Days/period/vessel) * (Technical measures) * (Catchability Coefficient)

where fishing mortality is an age and species specific variable.

The “Number of Days/period” and “Technical measures” are the factors (partly) controlled by managers. The “Number of Days/period” can be the munber of sea day per year, which can be controlled by various management regulations, such as TAC (total allowable catch), maximum number of sea days per year or closed seasons.

The “catchability coefficient”, is a measure for the ability of the gear to catch a certain species.

The “number of sea-days/period/vessel” may be determined by several factors, of which fisheries regulations is one. Other factors determining the activities of fishing fleets are the choices made by the fishers, depending on economics and availability of resources. TEMAS attempts to describe the behaviour of fishing fleets, with respect of effort allocation in time and space.

By a technical management measure is meant a regulation measures which (1) Specify rules for gears (e.g. minimum mesh sizes) (2) specify on minimum landing size (3) specify areas closed for fishery (4) gives specifications for vessels (such as maximum engine power) (5) Specify limits for by-catch percentages and target species percentages (6) specify rules for equipment for handling of catch.

Bilogical model, parameters and data:

The biological model behind TEMAS, is the traditional stock based model by Thompson and Bell (1934), which has been discussed in many textbooks on dynamics of fish stocks (E.g. Ricker, 1975, Beverton & Holt, 1957, and with emphasis on tropical fisheries: Sparre & Venema, 1992). The major part of the biological model behind TEMAS is the traditional stock VPA model, or generalizations of the traditional model. TEMAS extends the traditional models with a spatial model, accounting for, e.g. migration using the approach of Quinn et al, 1990). All these models originally were thought of as “fish stock assessment model”, where parameters were estimated by (e.g.) VPA or “Cohort analysis” (Virtual Population Analysis, Derzhavin, 1922, Fry, 1949). A resent summary of the contemporary practice of VPA is given in Lassen & Medley, 2001. (See Sparre, 2008b).

The TEMAS model is a multi-stock model and can operate with several stocks and VPAs. In addition to cohort analysis, the traditional stock related analyses are: (a) Estimation of growth parameters (b) Estimation of spawning seasons © Maturity ogive (percentage mature as a function of age) (d) Estimation of natural mortality. Combined with spatial data, the above data may also be used for estimation of migration routes, spawning grounds, nursery grounds, distribution by depth zone, etc. The TEMAS model includes fish migration models between areas and seasons.

There are four options for stock and recruitment model in TEMAS: (1) Beverton and Holt model (Beverton & Holt, 1957) (2) “Hockey stick” model (Barrowman & Meyers, 1999), (3) Ricker Model (Ricker, 1954) (4) the general Deriso-Schnute Model (Deriso 1980, Schnute, 1985). The deterministic recruitment model in TEMAS is a function of spawning stock, SSB, only. Further¬more, TEMAS has the option to let recruitment becomes a stochastic variable. (See Sparre, 2008b).

Technical and economical model, parameters and data:

Fleet and fishery:

The technical units of TEMAS are the “fleets”. A formal definition of fleet is: A “fleet” is a group of uniform vessels, which have approximately the same size and the same construction. The vessels should use the same type of gear and fishing techniques and most often, they share fishing grounds.

The TEMAS model includes number of vessels and effort and capacity by fleet, and keeps track of the age distributions of vessels in a fleet as the ICES model keeps track of the age distributions of fish. Fishing mortality is calculated from catch, landings, discards and stock numbers. TEMAS calculates partial F by fleet and stock which sums up to total stock based F. In addition, the partial F’s by fleet is divided into area and season specific mortality. The division into partial F’s is based on the catch and effort allocation and relationship between fishing mortality, effort and catchability, and the simulated area-fishing mortality is derived from the effort and the selection ogive. Advanced models and analyses for effort allocation and catchability patterns as well as effort re-allocation patterns is included in TEMAS. The effort is estimated as number of fishing days or days at sea. The capacity is the maximum number of fishing effort units (fishing days or sea days) that a fleet can exert in a time-period.

Included in the model are models or estimates for discard and selection ogives by fleet, minimum landing sizes by species,

Behaviour models:

TEMAS contains several options to model the behaviour of fishing firms during the fishing season and from year to year. The inspiration for this comes from the textbook by Vani K. Borooah (2002), as a general reference in behaviour theory, and also for certain elements from various papers dealing with fishers behaviour (e.g. Mistiaen & Strand, 2000, Wilen et al, 2002, Bockstael & Opaluch, 1983, Dupont, 1993). The types of behaviour models used by TEMAS is discrete choice models and relative utility models (RUM). (See Sparre 2008b).

With respect to fleet dynamics and structural behaviour TEMAS contains two options to model the behaviour of fishing firms during the fishing season and from year to year: i) Random Utility Model (RUM), or ii) Ad hoc behaviour rules. Effort re-allocation can be controlled in TEMAS in two ways: i) Giving effort as input (e.g. historical effort allocation patterns), or ii) Let the “Effort-rule” decide the effort. The statistical model and theory behind the RUM is comprehensive. However, it is also complicated and data demanding. The “Ad hoc” approach is kind of a short cut method, which indeed can be questioned and is not supported by a huge literature as the RUM is.

There are short-term and long term behaviour rules in TEMAS. The short term (trip) behaviour is a model by which we can predict the probabilities of the different choices a fisher makes on the trip-level. A similar model is used for the long-term behaviour. The five rules currently in the TEMAS package are:

Fishing effort rule: This is a rule for where to fish at which time with which gear. So far, the rule implemented only decides whether to fish or not to fish. The remaining behaviour is fixed by input parameters. The model for the “trip-related” behaviour is based on a mixture of “tradition” and “recent experience”. “Tradition” here means what was done last year (at the same time) and recent experience means the value of landings in the foregoing period relative to the costs of fishing.

Decommission (Rule). This (and the three following rules) are the so-called long term rules, which determines the capacity of the fishing fleets. The decommission rules takes the decision on accept of a decommission compensation based on the recent economic performance of the fleet and the age structure of the fleet.

Dis-investment rule. This rule decides on the bankruptcy of a vessel based on the recent economic performance of the fleet. Attrition rule: The attrition rule takes the decision on scrapping a vessel due to old age based on the age structure of the fleet. Investment rule: This rule decides on the investment in a new vessel based on the recent economic performance of the fleet.

When predicting the effect of management measures, it is obviously very interesting to predict both the short-term and the long-term reaction of fishers.

Example of decision tree for trip related behaviour

Example of decision tree for trip related behaviour.

Economic models:

The economic model in TEMAS serves two purposes: i) Modelling of fishers behaviour, and ii) Provision of measures of system performance.

Economics plays an important role in the evaluation of fisher’s reaction to the introduction of regulations. In the context of the Baltic case study, the important regulations under study are the MPAs in time and space. How fishers reallocate or moderate their effort in reaction to technical regulations, (like MPAs) is in the TEMAS model dependent on three factors:

1) Economy of fishing operations
2) Tradition (Whish fishing operations were made in the past)
3) The regulation (e.g. MPA in space and time)

Economy in the context of TEMAS is similar to an examination of accounts. The key issue in the TEMAS economic model is the cash flow, the difference between income and costs. Income, costs and cash flow are key issues in choice making of fishers in the TEMAS model. In Sparre (2008a,b) is the economic model and its linking to the behaviour models described. The income (the value of the landings) links the economic model to the “production model”, the technical/biological model of TEMAS.

Like the biological models has a suite of measures of performance, such as SSB, fishing mortality, landings, value of landings etc., the economic model can provide overall measures for the performance of the system. These measures are stakeholder specific, as the evaluation of fisheries depends on who is evaluator. For example, Fishing industry, Government Treasury, Society (in general) do usually not evaluate the same way. TEMAS allows for an optional number of economic models, each of which reflects the view of a stakeholder group. The outputs are a suite of measures of performance of the fishing industry or individual fleets.

There are no fixed economic models in TEMAS, but there is a frame by which the user can select the desired model(s) from a family of economic models. It has been attempted to make the family of economic models as wide as possible. A common feature is that the models are all dynamic models, as is the biological model of TEMAS.

There are 3 economic models in the current version of TEMAS, reflecting the views of three groups of stakeholders

1) FINANCIAL ANALYSIS OF FLEETS: From the point of view of vessel owners.
2) GOVERNMENT BUDGET: The impact of the fleets on the government budget.
3) ECONOMIC ANALYSIS: The economic performance from of the economy as a whole.

All three models operate with the same concepts of costs, earnings and investments, but (possibly) with different parameters.

The economic model calculates the cash flow (Revenue – costs) for each time period and eventual it computes the net present value over the time horizon simulated. The economic model was designed by Mr. Rolf Willmann, of the fisheries department of FAO, Rome (Sparre and Willmann, 1993).

The economic part of TEMAS uses the concepts developed for project analysis to evaluate the financial and economic performance of the fishery during the project horizon (i.e. simulation life span) given different fisheries management measures, government financial transfers, and assumptions about the investment and operational behaviour of fishing firms. The financial performance is assessed from the point of view of both the fishing firms and the government treasury (Gittinger, 1984, Little & Mirrlees, 1974, Squire & Tak , 1975 and Dasgupta et al, 1972). (See Sparre, 2008b).

The key performance measures of project analysis are the net present value (NPV), equal to the discounted net cash flow. Other economical measures of performance are given in Sparre (2008a,b).

TEMAS includes both economical and financial analysing and consider dynamics in values, prices, different types of costs, vessel/owner and crew share, taxes and subsidies and license fees, decommission payments, dis-investment, attrition, investment, government treasury, etc. (see Sparre, 2008a,b).

The economic analysis includes certain costs that are usually not paid for by the fishing firms and are thus excluded from their financial calculus. These include fisheries management costs such as research, administration and surveillance and enforcement. These costs lead to a cash outflow from the government budget or treasury. This cash outflow, however, might not be equal to their true costs to society to be accounted for in the economic analysis as is further explained below.

The economic analysis uses shadow prices of inputs whenever there is a discrepancy between the prices paid by fishing firms or the government and the economy wide opportunity costs of such inputs. For example, where fuel prices are subsidised, thus lowering fuel expenditures incurred by fishing firms, the economic analysis will be based on fuel prices net of such subsidies.

The financial performance of fishing firms will be affected by the way investments into fishing craft and gear have been financed, i.e. own savings or loans, and by the capital servicing terms of any loans taken in the past or in future years.

The financial performance of the government treasury depends on the cash inflows from the fishery through taxes, licensing fees, fines etc. and cash outflows for fisheries management expenditures, subsidies, etc. during the project horizon.

The economic analysis applies opportunity costs of capital to reflect the real social cost of using capital in fisheries rather than elsewhere in the economy. The opportunity cost concept is only applied to new investments. Past investments are sunk costs to the extent that they have no alternative economic use outside of fisheries.

In the financial analyses, labour costs are based on observed payments made to the fishing crew or government employees. In the economic analysis, opportunity cost of labour is applied to reflect the real social cost of employing people in fishing or government rather than elsewhere in the economy.

In the financial analysis, payments made to fishing firms to decommission excess fishing capacity increase their net cash flows. Some firms may exit the fishery altogether and may invest decommissioning payments into other economic activities. If so, these firms would not be further considered in the simulation model of the fishery.

Decommissioning payments (i.e. compensations to fishing firms and to displaced fishing crews) are considered as transfer payments, i.e. a cash outflow from the government treasury. These payments are not considered a cost in the economic analysis.

No adjustments are made to fish prices observed in the market which are assumed to accurately reflect social values. However, a simple function has been included to model changes in fish prices as a result of changes in fish landings.

Uncertainty, error and sensitivity assessment and estimation by EF/TEMAS:

The TEMAS model can do single deterministic simulations or multiple stochastic simulations. The multiple stochastic simulations executes a number of single deterministic simulations (say 1000 simulations), each of which based on parameters drawn by a random number generator. We shall forget about multiple stochastic simulations for the time being, and concentrate on single deterministic simulations.

Simulation procedure

TEMAS accounts for a number of different types of “errors” in the system. An error means a “deviation from the model”, or “something that can go wrong” .

1. Measurement error. Errors in input data, such as catch at age data, caused by data being estimated from samples, and not from complete enumeration.

2. Estimation error. Errors caused by the method used to estimate parameters, or erroneous assumption about the data.

3. Model misspecification error. Errors caused by incomplete or wrong understandings of the mechanism behind the system dynamics. The assumed Stock/recruitment relationships may be candidates for model misspecifications.

4. Implementation error. The errors caused by regulations not being reacted to as assumed. The fishers may find ways to implement regulations, which do not lead to the achievements of the intensions of regulations.

The software will be able to simulate the effect of errors and bias, by stochastic simulations. Stochastic simulation is simple to repeat the same calculations a large number of times, each time with new parameter-values drawn by a random number generator. The stochastic simulation requires specifications of probability distributions of those parameters which are considered stochastic variables.

The stochastic simulation module simply executes TEMAS a large number of times (say, 1000 times), and each time it draws parameters and initial condition variables by random number generators, executes a simulation over a series of years. The parameters of the probability distributions of parameters are given as input. At the end it retrieves the results of all 1000 simulations and converts them into, for example, frequency diagrams. TEMAS offers (in its present version) two probability distributions: (1) Normal distribution (2) Log normal distribution.

We will simply not be in a position to say anything about the prediction power. The output of the model is in the best case of the nature: “It is likely that management regime A gives a better performance than management regime B” with respect of a selected measure of performance. TEMAS should not be used to quantify, for example, the expected spawning stock biomasses.

Calibration - Conditioning

The statistical estimation of parameters in TEMAS, is more or less assumed to be a problem isolated from the simulations with TEMAS. Somehow, we assume that parameters are available from various (not specified, by “reliable” sources). This will never strictly be the case in any application of TEMAS. Actually, many of the crucial parameters of TEMAS cannot be estimated by robust statistical methods, involving estimation of variance and co-variances and all their derivatives in the form of statistical diagnostics. Many (most) parameters of TEMAS are “guesstimates” rather than “estimates” (as defined in standard textbooks of statistical inference). The reason for this is not that parameter estimation methodology is not available, but that available data are of a poor quality, but perhaps more important is, that the basic mechanism behind the system dynamics is not understood. The so-called “process errors” of TEMAS are not known. Thus, it is not possible to separate “process errors” and “measurement errors”, but both are probably big

However, it is not satisfactory to make a complete separation between the “real world” and the simulations by TEMAS. One would like to maintain the humble illusion that TEMAS does indeed resemble to the real world, although we do not dare make statements about the “prediction power” of TEMAS. The calibration of TEMAS is a rather ad hoc attempt to make TEMAS not deviate “too much” from the reality. TEMAS calibrates some of its parameters by aid of the so-called modified 2-criterion (Sokal and Rohlf, 1981)

Calibration procedure

where “Xcalculated” symbolises a prediction-variable of the model. “Xobserved” indicates the value of X observed from a historical period. The same model is used for both prediction and estimation. Xcalculated depends on the indigenous parameters, and 2 is minimised with respect of the indigenous parameters. “Indices” is a subset of the indices available in TEMAS.

Management

TEMAS contains a suite of options for pre-prepared pairs of management regimes. The natural reference for comparison is the “traditional management regime”, based on the total TAC. EFIMAS can simulate an ICES procedure, and transformation of a TAC into effort by fleet. The TAC is based on scientific advice given by the ACFM of ICES. The advice of ACFM is based on the fish stock assessment executed by the assessment working groups of ICES. TEMAS can simulate the entire process from assessment to implementation of regulation. TEMAS also simulates the most recent adaptive approach of the EU commission, where effort and TACs are reduced/increased with a maximum percentage per year, until a desired goal is achieved. As the principal input to the operational model of TEMAS is effort by fleet (derived from fleet capacity), TAC must be converted into effort, in order to establish the feed back from the management to the operational model of TEMAS. Furthermore, TEMAS contains a module in relation to mixed fisheries harvest rules. TEMAS include harvest and management rules in relation to current TAC management regime including simulation of a ICES stock assessment and forecast, effort management by capacity or days at sea reduction, indirect effort management through closed seasons and areas (MPA), the adaptive approach, effort management by maximum sea days combined with MPA and TAC, and take also into account the relative stability and historical rights. (See Sparrre, 2008a,b and Ulrich et al., 2007).

The evaluation of TAC/effort management is connected to the evaluation of simultaneously implemented technical management measures, such as gear regulation, closed areas (MPAs) and closed seasons. Essentially, these measures also aim at reducing or redistribute effort.

Tuning of TEMAS: By “tuning” is meant the processes of finding the reference simulation of TEMAS. The reference simulation is the situation (scenario) relative to which all the other simulations are made, and are compared to in the evaluation frame. Tuning involves the calculation of certain parameters. It should be noted that tuning does not involve a proper statistical estimation of parameters.

TEMAS Tuning

The reference simulation will usually be chosen to be a simulation in equilibrium, that is, a simulation where all results are equal in all years of the time series under study. Furthermore, the reference simulation will usually be chosen to be the fisheries situation of the current situation (current year). TEMAS is said to reproduce the current situation when it can reproduce the landings (in weight) observed the last data year for each combination of fleet, stock, time period and area. To achieve this goal completely is usually impossible, so one can only hope for a reasonable approximation. Taking in to account all the sources of uncertainties involved in TEMAS, there is no reason to make too much effort in achieving a complete reproduction of observed catches

The five types of tuning offered by TEMAS is:

1) N(first year) = N(last year). To achieve equilibrium
2) BH(New) = BH(old)*Land(Obs)/Land(Calc), or the similar parameter in an alternative S/R-model. . Tune recruitment to observed landings
3) Q(New) = Q(old )*Land(Obs)/Land(Calc). Tune catchability to observed landings
4) Q(New) = Q(old )*F(Obs)/F(Calc). Tune catchability to observed total fishing mortality
5) Q = F/Effort by area and fleet. Compute individual catchabilities to observed area fishing mortalities

Results and status

TEMAS Evaluation Frame Baltic Application

The ultimate objective of TEMAS is to compare two alternative management regimes, by simulating the fisheries system over a series of years for both regimes, and eventually it compare the performance of the two regimes during the time period (as explained in the introduction). The operating system generates (“fake” or “hypothetical”) input data to the alternative management models, and it predicts the effect of the alternative management regulations on the eco-system and the fisheries. Eventually it compares the two alternatives by comparison of selected measures of performance.

Summary of the TEMAS Evaluation Framework

Summary of the Evaluation Framework, as implemented by the TEMAS software.

The figure above attempts to summarize the basic features of TEMAS in a single graph. The figure operates with four types of errors:

1. Measurement error. Errors in input data, such as catch at age data, caused by data being estimated from samples, and not from complete enumeration.
2. Estimation error. Errors caused by the method used to estimate parameters, or erroneous assumption about the data.
3. Model misspecification error. Errors caused by incomplete or wrong understandings of the mechanism behind the system dynamics. The assumed Stock/recruitment relationships may be candidates for model misspecifications.
4. Implementation error. The errors caused by regulations not being reacted to as assumed. The fishers may find ways to implement regulations, which do not lead to the achievements of the intensions of regulations.

The software can simulate the effect of errors and bias, by stochastic simulations. The stochastic simulation requires specifications of probability distributions of those parameters which are considered stochastic variables. The figure should be considered as an illustration of the calculations for one time period. These calculations are repeated for as many time periods (and years) as chosen the user of the software. Chronologically, the events taking place are:

1. The operating model produces input to the management model for year “y”
2. The management produces management regulations for year “y+1”
3. The management regulations for year “y+1” is used as input to the operational model, to produce input to the management model in year y+1,
Etc.

The stochastic simulation module simply executes TEMAS a large number of times (say, 1000 times), and each time it draws parameters and initial condition variables by random number generators, executes a simulation over a series of years. At the end it retrieves the results of all 1000 simulations and converts them into, for example, frequency diagrams. Below is shown an example of output from stochastic simulation with TEMAS, namely a time series of total Revenue with indication of the stochastic variation, in the form of SD (Standard deviation) and maximum/ minimum values.

Example

It should be noted that the operational model of TEMAS contains many parameters which cannot be estimated by the data currently available. Therefore a large number of parameters will have be assigned “plausible” values, that is, values not estimated by statistical methods and observations but values which are believed to be “reasonable”. Likewise, TEMAS will contain a number of sub-models which has not been verified by recognized statistical tests. Therefore, the concept of “prediction power” may not be applicable to TEMAS.

There is no alternative to this approach, when it comes to test alternative management regimes, which has not been tested earlier. A real statistical experimental design would require that the two alternative management regimes were tested on two identical ecosystems, and such an experiment will never become possible in practice.

The links between the components of TEMAS can be summarised as follows:

The operational model of TEMAS integrates the biology, technical features, economy and behavioural features as illustrated in the figure below. TEMAS integrates seven components:

• Management model.
• Generation of stochastic input from ecosystem.
• Biological/technical model.
• Short term behaviour model (trip related behaviour of fishers).
• Economic model (costs and earnings).
• Long term behaviour model (investment/disinvestment related behaviour of fishers).
• Evaluation of system performance.

The focal point in TEMAS is the capacity (the number of vessels by fleet). The capacity is determined by the long term behaviour model, which predicts the number of investments in new vessels, the number of disinvestments, the number of attritions (vessels “dying” from old age) and removals (scrapings) of vessels due to decommission. The long term behaviour is determined by the economic model, which predicts costs and earnings. Costs are variable and fixed costs. The variable costs are derived from the effort, and the earnings from the value of the catch.

Operational model of TEMAS Evaluation Frame

The complete operational model, combining biology, technical features, economy and behavioural features, together with it’s links to the management model and the evaluation.

The effort is derived from the number of sea-days, which in turn is determined by capacity and the short term behaviour model. Both the short term and the long term behaviour are influenced by the management regulations. The management model simulates the bodies that give advice (e.g. ICES) and which decide on the management measures (e.g. the EU fisheries commission). The effort produces the fishing mortality, which is input to the biological model together with stochastically generated input.

The stochastic input represents the “unpredictable ecosystem”. The main stochastic component is the unpredictable recruitment. The mechanisms that determines the recruitment of fish stocks is highly variable for most fish stocks. The relationship between stock and recruitment is not understood, and the only knowledge currently available, is the series of historical observations of recruitment. Other parameters in TEMAS can be treated as stochastic variables, in principle, any parameter. Output from the biological system is the yield (the catch in weight). The yield combined with the price/kg determines the revenue from fishing which is input to the economic model together with the costs.

The political “evaluation” is not a part of TEMAS. TEMAS attempts to create a suite of useful measures of performance, that can be used in a political evaluation of the system performance. TEMAS thus does not range the alternatives amongst management strategies. It does not attempt to give the optimum to a maximization problem. TEMAS does not contain a goal function. What is “best” is a decision left to the users of TEMAS.

Regrettably, the only obtainable international fishery data for different international fleets in the Baltic Sea to be used in the model (2003 data applied – see Approach 2a), has been for cod fishery leading to only Danish data for Baltic herring and sprat fisheries to be found. Furthermore, as for fisheries outside/beyond the Baltic Sea, the only obtainable data has been for the vessels partly operating in the Baltic Sea for the Danish fisheries and for the Danish fishing fleets and not for the international fisheries. This has made it difficult to obtain meaningful output from the model in the present implementation.

Dissemination

Sparre, P. J. 2008a. User’s Manual for the EXCEL Application “TEMAS” or “Evaluation Frame”. DTU-Aqua Report 190-08: 182 pp. ISBN 978-87-7481-077-3.
Users Manual for the TEMAS Evaluation Frame

Sparre, P. J. 2008b. Evaluation Frame for comparison of alternative management regimes using MPA and closed seasons applied to Baltic cod. DTU-Aqua Report 191-08: 298 pp. ISBN 978-87-7481-079-7.
TEMAS Evaluation Frame Baltic Application

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.

EFIMAS ECONOWS Report, Final Version 2008.
Final report ECONOWS 26/06/2008

Links to Other Work

This further development work of TEMAS Evaluation Frame has been done in a cooperation between the EU-FP6-EFIMAS and EU-FP6-PROTECT projects as well as in cooperation with a Danish national project.

The TEMAS model has been applied in the North Sea Flatfish case study as well under EFIMAS.

The TEMAS Model was in its initial form developed as a cooperation between a Danish national government project and the EU-FP5-TECTAC project.

References

See first of all comprehensive list of references in:

Sparre, P. J. 2008b. Evaluation Frame for comparison of alternative management regimes using MPA and closed seasons applied to Baltic cod. DTU-Aqua Report 191-08: 298 pp. ISBN 978-87-7481-079-7.

as well as under:

Sparre, P. J. 2008a. User’s Manual for the EXCEL Application “TEMAS” or “Evaluation Frame”. DTU-Aqua Report 190-08: 182 pp. ISBN 978-87-7481-077-3.

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.

EFIMAS ECONOWS Report, Final Version 2008.
Final report ECONOWS 26/06/2008

Acknowledgements

This work has been developed as a collaboration between the EU-FP6-EFIMAS and EU-FP6-PROTECT Projects as well as with a Danish national government project. The collaboration has resulted in developing and further development of bio-economic models and tools for management evaluation done under among other the EU-FP5-TECTAC Project. The TEMAS model has been applied in the North Sea Flatfish case study as well under EFIMAS.

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

The Baltic application of the TEMAS Evaluation Frame has mainly been financed by EFIMAS with significant contributions from the two other mentioned projects as well.

Participants:

Per J. Sparre, J. Rasmus Nielsen, Francois Bastardie, Eva Maria Pedersen, Jacob Nabe Nielsen, Ole Vestergaard, Bo Sølgaard Andersen, Clara Ulrich-Rescan (DTU-Aqua), Ayoe Hoff and Hans Frost (UNCPH-FOI).

 
efimas1/wp4/cs9/appr2c/main.txt · Last modified: 2008/07/08 09:58 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