Modelling Floodplain River Fisheries – an Introduction Training Workshop Materials · 2016. 8....

Modelling Floodplain River Modelling Floodplain River Fisheries Fisheries –– an Introductionan Introduction

Training Workshop MaterialsTraining Workshop MaterialsUK Department for International Development (DFID)UK Department for International Development (DFID)

Fisheries Management Science Programme (FMSP)Fisheries Management Science Programme (FMSP)

June 2005 June 2005 Ashley S. Halls Ashley S. Halls

Aquae Sulis Ltd (ASL)Aquae Sulis Ltd (ASL)www.aquaewww.aquae--sulissulis--ltd.co.ukltd.co.uk

BackgroundBackgroundThis presentation is one of a series of five presenting key outpThis presentation is one of a series of five presenting key outputs uts frfromom FMSP floodplain projects, carried out in the Asian region FMSP floodplain projects, carried out in the Asian region between 1992 and 2005. The five papers focus on:between 1992 and 2005. The five papers focus on:

–– General management guidelines for floodplain river fisheries (asGeneral management guidelines for floodplain river fisheries (aspublished in FAO Fisheries Technical Paper 384/1)published in FAO Fisheries Technical Paper 384/1)

–– Selection and management of harvest reserves (key messages)Selection and management of harvest reserves (key messages)–– Materials for a training course on harvest reservesMaterials for a training course on harvest reserves–– Flood Control Impacts on Fisheries: Guidelines for Mitigation Flood Control Impacts on Fisheries: Guidelines for Mitigation –– ModellingModelling floodplain river fisheriesfloodplain river fisheries

This presentation was prepared by FMSP Project R8486 This presentation was prepared by FMSP Project R8486 ––‘‘Promotion of FMSP guidelines for floodplain fisheries managementPromotion of FMSP guidelines for floodplain fisheries managementand sluice gate controland sluice gate control’’

IntroductionIntroduction

The following materials are provided for adaptation or use in The following materials are provided for adaptation or use in workshops aimed at building awareness of the range of models andworkshops aimed at building awareness of the range of models andapproaches that can be used to guide the management of floodplaiapproaches that can be used to guide the management of floodplainn--river fisheries resources. river fisheries resources.

The models and approaches described here were either developed oThe models and approaches described here were either developed or r applied under research projects funded under the Fisheries applied under research projects funded under the Fisheries Management Science Programme of the UK GovernmentManagement Science Programme of the UK Government’’s s Department for International Development (DfID).Department for International Development (DfID).

Further details of the empirical models and methodologies descriFurther details of the empirical models and methodologies described bed here can be found in Section 14 of Hoggarth et al (in press) whihere can be found in Section 14 of Hoggarth et al (in press) which ch may be provided as a handout. may be provided as a handout.

Other relevant papers, reports and sources of information are Other relevant papers, reports and sources of information are provided under each section.provided under each section.

Full references and URLs are provided at the end of this presentFull references and URLs are provided at the end of this presentation.ation.

IntroductionIntroduction–– What are modelsWhat are models

–– Types of modelsTypes of models

–– Purpose of modelsPurpose of models

–– Further ReadingFurther Reading

1. Empirical models1. Empirical models1.1. Linear models1.1. Linear models

1.1.1. Simple Linear Regression1.1.1. Simple Linear Regression

1.1.2. Multiple Linear Regression (MLR) and General Linear Model1.1.2. Multiple Linear Regression (MLR) and General Linear Models (GLM)s (GLM)

1.2. Non1.2. Non--linear modelslinear models1.2.1 Empirical surplus production models (Non1.2.1 Empirical surplus production models (Non--linear regression)linear regression)

1.2.2. Bayesian Networks (BNs)1.2.2. Bayesian Networks (BNs)

2. Population Dynamics Models2. Population Dynamics Models2.1. Age structured Populations Dynamics (ASPD) 2.1. Age structured Populations Dynamics (ASPD)

2.1.1 Dynamic Pool Model for Floodplain Fisheries2.1.1 Dynamic Pool Model for Floodplain Fisheries

2.1.2 BEAM 42.1.2 BEAM 4

ContentContent

What are models?What are models?Models are quantitative descriptions of processes or relationshiModels are quantitative descriptions of processes or relationships ps among variables.among variables.

Types of ModelTypes of ModelIn the context of fisheries management, models can be divided In the context of fisheries management, models can be divided into 2 main categories:into 2 main categories:

Empirical ModelsEmpirical Models. These are simply statistical descriptions of the . These are simply statistical descriptions of the observed relationships among two or more variables of interest.observed relationships among two or more variables of interest.

Population dynamics modelsPopulation dynamics models. These attempt to explicitly model . These attempt to explicitly model the dynamics of fish populations based upon established theoriesthe dynamics of fish populations based upon established theoriesof population behaviour, and biological and ecological processesof population behaviour, and biological and ecological processes..

IntroductionIntroduction

IntroductionIntroductionPurpose of ModelsPurpose of Models

•• Models are used to make predictions about, or improve Models are used to make predictions about, or improve understanding of the response of important understanding of the response of important dependentdependent variables to variables to changes in changes in independentindependent variables.variables.

•• DependentDependent variables are often referred to as variables are often referred to as responseresponse, , output output or or performanceperformance variables and typically include catch, indices of variables and typically include catch, indices of abundance, incomeabundance, income……etc.etc.

•• IndependentIndependent variables are often referred to as variables are often referred to as inputinput or or explanatoryexplanatoryvariables. Examples include fishing effort, stocking density, variables. Examples include fishing effort, stocking density, environmental variables (e.g. flood extent)environmental variables (e.g. flood extent)……etc.etc.

Further Reading: Haddon (2001).Further Reading: Haddon (2001).

Empirical Models: Simple Linear ModelsEmpirical Models: Simple Linear Models

•• Simple empirical models are frequently used in floodplain river Simple empirical models are frequently used in floodplain river fisheries to describe the linear relationships between two variafisheries to describe the linear relationships between two variables bles of interest. of interest.

•• They are typically fitted to estimates of annual catch and a varThey are typically fitted to estimates of annual catch and a variety iety of different explanatory variables (e.g. resource area, fishing of different explanatory variables (e.g. resource area, fishing effort, hydrological indiceseffort, hydrological indices……etc) using linear regression methods.etc) using linear regression methods.

•• Variables are often first logVariables are often first logee transformed to ensure that the transformed to ensure that the normality assumptions behind the regression method are met.normality assumptions behind the regression method are met.

•• When few or no estimates of annual catch are available for a givWhen few or no estimates of annual catch are available for a given en fishery or management location, some estimate of potential yieldfishery or management location, some estimate of potential yieldmay be obtained by comparing estimates reported for other river may be obtained by comparing estimates reported for other river fisheries or management sites in relation to common explanatory fisheries or management sites in relation to common explanatory variables such as resource area.variables such as resource area.


•• Models based upon such Models based upon such among among fishery comparisonsfishery comparisons can provide can provide planners and policy makers with planners and policy makers with an an approximateapproximate indication of the indication of the potential yield from the river potential yield from the river fishery.fishery.

•• Figure 1 illustrates the Figure 1 illustrates the relationship between logrelationship between logee catch catch and floodplain area for Asian and floodplain area for Asian river systems reported by the river systems reported by the DfIDDfID--funded Project R5030 (see funded Project R5030 (see MRAG 1993; 1994 and Halls MRAG 1993; 1994 and Halls 1999). 1999).

0 5 10 15ln floodplain area (km 2)

0

5

10

15

ln c

atch

(t y

-1)

Figure 1. Potential yield from Asian floodplain rivers plotted as a function of floodplain area with fitted regression lines on loge transformed scales.


•• Details of this and other best fitting models for predicting annDetails of this and other best fitting models for predicting annual ual catches from tropical river fisheries developed under R5030 are catches from tropical river fisheries developed under R5030 are summarised in Table 1 below, together with guidance for estimatisummarised in Table 1 below, together with guidance for estimating ng 95% confidence intervals around the predictions. 95% confidence intervals around the predictions.

Table 1 Summary of the best fitting regression models for predicTable 1 Summary of the best fitting regression models for predicting ting multispecies potential yield from floodplainmultispecies potential yield from floodplain--river systems where river systems where aa and and bb are are the constant and slope parameters of the linear regression modelthe constant and slope parameters of the linear regression model: : Y = a + Y = a + bxbx, and where , and where n n is the number of observations, R is the correlation is the number of observations, R is the correlation coefficient, and P is the probability that the slope parameter, coefficient, and P is the probability that the slope parameter, bb = 0. = 0. SSbb is is the standard error of the estimate of the slope coefficient, the standard error of the estimate of the slope coefficient, bb, is the , is the residual mean square, and is the mean value of the observresidual mean square, and is the mean value of the observations of the ations of the explanatory variable.explanatory variable.

Relationship Continent a b Sb n X XYs .2 R P

ln catch vs ln FPA Asia 2.086 0.996 0.083 13 4.31 0.531 0.97 <0.001 ln catch vs ln length Asia -14.88 3.234 0.585 5 8.06 0.680 0.96 0.01 ln catch vs ln DBA S. America -3.60 0.936 0.218 15 12.87 1.457 0.77 0.001

X


Prediction intervals for yield corresponding to new observationsPrediction intervals for yield corresponding to new observations of of X, is given by:X, is given by:

where where is the standard error of the estimate given by:is the standard error of the estimate given by:

where is the residual mean square (the variance of Y afwhere is the residual mean square (the variance of Y after ter taking into account the dependence of Y on X), and Staking into account the dependence of Y on X), and Sbb is the is the standard error of the estimate of the slope coefficient, b (Zar standard error of the estimate of the slope coefficient, b (Zar 1984, 1984, p272p272--275).275).

}ˆ{)2-;2/-1(ˆnewnew YsntY ×± α

}ˆ{ newYs

2.

2

2

.2

)(

)-(11}ˆ{

b

XY

iXYnew

Ss

XXn

sYs +⎟⎠⎞

⎜⎝⎛ +=

XYs .2


ApplicationsApplications

•• Generally speaking, these types of models provide only very Generally speaking, these types of models provide only very impreciseimprecisepredictions because of the significant measurement error associapredictions because of the significant measurement error associated with ted with the potential yield estimates used to fit the models. the potential yield estimates used to fit the models.

•• Potential yields are often estimated using (i) the Generalised FPotential yields are often estimated using (i) the Generalised Fishery ishery Development Model (GFDM) approach described by Grainger and GarcDevelopment Model (GFDM) approach described by Grainger and Garcia ia (1996), (ii) as the average annual catch value, or worst (iii) f(1996), (ii) as the average annual catch value, or worst (iii) from a single rom a single observation, all of which are subject to potentially significantobservation, all of which are subject to potentially significant measurement measurement and estimation error (no account is taken of fishing effort). and estimation error (no account is taken of fishing effort).

•• The utility of these models is therefore restricted to providingThe utility of these models is therefore restricted to providing a rough a rough indication of the likely potential of the fishery for policy andindication of the likely potential of the fishery for policy and development development planning purposes.planning purposes.

•• Note that whilst the examples illustrated above are based upon Note that whilst the examples illustrated above are based upon comparisons across wide geographical scales, this modeling approcomparisons across wide geographical scales, this modeling approach may ach may be equally, if not more, relevant on a more be equally, if not more, relevant on a more local scalelocal scale, particularly in the , particularly in the context of context of adaptive coadaptive co--managementmanagement. . (see Co(see Co--management guidelines management guidelines presentation)presentation)

Database ResourceDatabase Resource

•• Estimates of potential yield for lakes and rivers, and a wide raEstimates of potential yield for lakes and rivers, and a wide range of nge of corresponding habitat variables (e.g. resource area, indices of corresponding habitat variables (e.g. resource area, indices of primary productivity and hydrological variables) have been compiprimary productivity and hydrological variables) have been compiled led by Project R5030 from the literature and entered into a by Project R5030 from the literature and entered into a ““Lakes and Lakes and Rivers DatabaseRivers Database””. This database resource will shortly become . This database resource will shortly become available on a CD published by FAO (see Dooley at al. in press).available on a CD published by FAO (see Dooley at al. in press).

Other ModelsOther Models

•• Welcomme (1985; 2001) contain other examples of linear empiricalWelcomme (1985; 2001) contain other examples of linear empiricalmodels for predicting fish yields and species richness in tropicmodels for predicting fish yields and species richness in tropical river al river basins.basins.


•• When the response of a variable (e.g. annual catch) to two or moWhen the response of a variable (e.g. annual catch) to two or more re independent variables (covariates) is of interest (e.g. floodplaindependent variables (covariates) is of interest (e.g. floodplain area and in area and annual rainfall), then multiple linear regression (MLR) methods annual rainfall), then multiple linear regression (MLR) methods would be would be applicable. applicable.

•• When using MLR methods it is important to ensure that the explanWhen using MLR methods it is important to ensure that the explanatory atory variables included in the model are indeed independent to avoid variables included in the model are indeed independent to avoid spurious spurious results (e.g. rainfall and flooded area are unlikely to be indepresults (e.g. rainfall and flooded area are unlikely to be independent).endent).

•• It is generally recommend that you should have at least 10 to 20It is generally recommend that you should have at least 10 to 20 times as times as many observations (cases, respondents) as you have variables, otmany observations (cases, respondents) as you have variables, otherwise herwise the estimates of the regression line are probably very unstable the estimates of the regression line are probably very unstable and unlikely and unlikely to replicate if you were to repeat the study. to replicate if you were to repeat the study.

•• Beware of automatic (forward and backward) stepwise fitting methBeware of automatic (forward and backward) stepwise fitting methods. It ods. It is often safer to employ a manual backward stepwise fit, startinis often safer to employ a manual backward stepwise fit, starting with all g with all the variables in the model and then dropping the least significathe variables in the model and then dropping the least significant variables nt variables in turn. For unbalanced designs, it is often necessary to returin turn. For unbalanced designs, it is often necessary to return dropped n dropped variables to the model to determine the effect of different combvariables to the model to determine the effect of different combinations of inations of variables. variables.

•• Further useful guidance on fitting Further useful guidance on fitting MLRsMLRs can be found at can be found at http://http://www.statsoft.com/textbook/stmulreg.htmlwww.statsoft.com/textbook/stmulreg.html

Empirical Models: Multiple Linear Regression (MLR)Empirical Models: Multiple Linear Regression (MLR)

Empirical Models: General Linear Models (GLM)Empirical Models: General Linear Models (GLM)

•• Sometimes researchers are interested in understanding the effectSometimes researchers are interested in understanding the effects of s of both both factorsfactors (categorical variables) and (categorical variables) and covariatescovariates (scale variable) on (scale variable) on dependent variables such as catch or CPUE.dependent variables such as catch or CPUE.

•• This is often the case in the context of This is often the case in the context of adaptive or coadaptive or co--managementmanagementwhen opportunities frequently exist to compare the outcomes when opportunities frequently exist to compare the outcomes (performance) of local management activities among sites. These(performance) of local management activities among sites. Thesecomparisons can generate comparisons can generate lessons of success and failurelessons of success and failure which can which can be used to adapt management plans accordingly (see Halls et al 2be used to adapt management plans accordingly (see Halls et al 2002 002 and the accompanying coand the accompanying co--management guidelines presentation). management guidelines presentation).

•• Examples of important factors of interest might include:Examples of important factors of interest might include:–– Community based management: Present (1), absent(0).Community based management: Present (1), absent(0).–– Management: Gear bans (1); closed seasons (2); reserves (3).Management: Gear bans (1); closed seasons (2); reserves (3).–– Extent of poaching: low (0); medium (1); high (2).Extent of poaching: low (0); medium (1); high (2).

•• Examples of covariates might include:Examples of covariates might include:–– Fishing intensityFishing intensity–– Ratios describing the morphological characteristics of waterbodiRatios describing the morphological characteristics of waterbodies es

(eg dry season area: flood season area)(eg dry season area: flood season area)–– Indices of flooding extent and duration.Indices of flooding extent and duration.

•• In this case, the use of the In this case, the use of the General Linear Models (GLM)General Linear Models (GLM) approach approach would be applicable.would be applicable.

•• GLMs are similar to regression models but can deal with both facGLMs are similar to regression models but can deal with both factors tors (fixed and random) and covariates. (fixed and random) and covariates. The factor variables effectively The factor variables effectively divide the population into groups.divide the population into groups.

•• Detailed guidelines for building interdisciplinary GLMs for smalDetailed guidelines for building interdisciplinary GLMs for small scale l scale fisheries have been developed by R7834 (see Halls et al 2002 andfisheries have been developed by R7834 (see Halls et al 2002 andHoggarth et al. in press). Hoggarth et al. in press).

•• These include examples of models fitted to data compiled from coThese include examples of models fitted to data compiled from co--management projects worldwide, as well as guidance on management projects worldwide, as well as guidance on identifying identifying sampling units, important variables, data levels and cleaning, sampling units, important variables, data levels and cleaning, exploratory analysis, sample sizes, sensitivity analysisexploratory analysis, sample sizes, sensitivity analysis……etc. etc.

•• More general guidance on More general guidance on GLMsGLMs can be found in can be found in McCullaghMcCullagh & & NelderNelder(1989).(1989).

Empirical Models: General Linear Models (GLMs)Empirical Models: General Linear Models (GLMs)

Empirical Models: Empirical Models: NonNon--linear modelslinear models•• NonNon--linear models are fitted when linear models are fitted when

relationship between two variables is relationship between two variables is not linear, or cannot be not linear, or cannot be ‘‘linearisedlinearised’’ by by means of data transformations.means of data transformations.

•• Typically, models are fitted using Typically, models are fitted using nonnon--linear least squares methodslinear least squares methods or other or other more sophisticated methods e.g. more sophisticated methods e.g. maximum likelihood methods and maximum likelihood methods and Bayesian estimation.Bayesian estimation.

•• For example, Halls et al (2002) fitted For example, Halls et al (2002) fitted a nona non--linear modified Fox surplus linear modified Fox surplus production model using nonproduction model using non--linear linear least squares to estimates of catch least squares to estimates of catch per unit area (CPUA) and fisherman per unit area (CPUA) and fisherman density assembled from a number of density assembled from a number of floodplain rivers to provide some floodplain rivers to provide some estimate of fishing intensity estimate of fishing intensity corresponding to maximum yield. corresponding to maximum yield. (see Figure 2). (see Figure 2).

0 2 4 6 8Square root number of fishers km -2

-1

0

1

2

3

4

Ln C

PUA

(ton

nes

km-2

yr-1

)

0 2 4 6 8-1

0

1

2

3

4

0 2 4 6 8-1

0

1

2

3

4

0 2 4 6 8-1

0

1

2

3

4

Figure 2 CPUA vs. fisher density for floodplain rivers with fitted Fox model. Africa ( • ); Asia ( ▲ ); and South America ( ■ ). R2 = 0.80. Note axis scaling. Source: Halls et al (2002)

The model predicts a maximum yield The model predicts a maximum yield of 132 kg haof 132 kg ha--11 yryr--11 at a fisher density at a fisher density of about 12 fishers kmof about 12 fishers km--22..

Empirical Models: Empirical Models: NonNon--linear modelslinear models

•• Further details of this and other related models can be found inFurther details of this and other related models can be found inHalls et al (2002) and Hoggarth et al (in press)Halls et al (2002) and Hoggarth et al (in press)..

•• Further advice on fitting nonFurther advice on fitting non--linear models for fisheries applications linear models for fisheries applications may be found in Chapter 6 of Hilborn & Walters (1992) and Sectiomay be found in Chapter 6 of Hilborn & Walters (1992) and Section n 3.3 of Haddon (2001).3.3 of Haddon (2001).

Empirical Models: Empirical Models: :: Bayesian networks (BNs)Bayesian networks (BNs)

•• Unlike GLMs that deal with quantitative dependent (response) varUnlike GLMs that deal with quantitative dependent (response) variables, iables, Bayesian networks (BNs)Bayesian networks (BNs) provide opportunities to model more qualitative provide opportunities to model more qualitative (categorical) response variables such as equity, compliance, emp(categorical) response variables such as equity, compliance, empowerment owerment etc. etc.

•• BNs comprise nodes (random variables) connected by directed linkBNs comprise nodes (random variables) connected by directed links. Prior s. Prior probabilities assigned to each link (established via tables of cprobabilities assigned to each link (established via tables of conditional onditional probabilities) determine the status of each node.probabilities) determine the status of each node.

•• Conditional probabilities can be generated from crossConditional probabilities can be generated from cross--tabulations of the tabulations of the data data oror by using subjective probabilities encoded from expert opinions.by using subjective probabilities encoded from expert opinions.

•• BNs are able to model complex and intermediate pathways of causaBNs are able to model complex and intermediate pathways of causality in a lity in a very visual and interactive manner to improve understanding of cvery visual and interactive manner to improve understanding of coo--management systems and fisher behavior. management systems and fisher behavior.

•• BNs can also be used as a management tool or expert system for BNs can also be used as a management tool or expert system for diagnosing strengths and weaknesses among codiagnosing strengths and weaknesses among co--management units and for management units and for exploring exploring ‘‘what ifwhat if’’ scenarios. scenarios.

Empirical Models: Empirical Models: :: Bayesian networks (BNs)Bayesian networks (BNs)

•• Netica software for constructing Netica software for constructing BNsBNs is useris user--friendly, inexpensive, friendly, inexpensive, and easy to learn. and easy to learn. http://http://www.norsys.comwww.norsys.com//

•• Further information about BNs together with detailed guidelines Further information about BNs together with detailed guidelines for for their construction (with examples) can be found in Halls et al (their construction (with examples) can be found in Halls et al (2002) 2002) and Chapter 14 of Hoggarth et al (in press)and Chapter 14 of Hoggarth et al (in press)..

Further Reading: Further Reading: CowellCowell et al (1999).et al (1999).

Web resources:Web resources:

http://en.wikipedia.org/wiki/Bayes'_theoremhttp://en.wikipedia.org/wiki/Bayes'_theorem

http://en.wikipedia.org/wiki/Bayesian_inferencehttp://en.wikipedia.org/wiki/Bayesian_inference

•• AgeAge--structured population dynamics (ASPD) models apply growth and structured population dynamics (ASPD) models apply growth and mortality rates to individual cohorts (agemortality rates to individual cohorts (age--groups) recruited to the fishery in groups) recruited to the fishery in order to determine how the overall population number or biomass order to determine how the overall population number or biomass will will respond to agerespond to age-- or sizeor size--dependent rates of exploitation or management dependent rates of exploitation or management interventions. interventions.

•• These types of models are often referred to as These types of models are often referred to as Dynamic Pool Models.Dynamic Pool Models.

•• Project R5953 modified a basic ASPD to include densityProject R5953 modified a basic ASPD to include density--dependent growth, dependent growth, mortality and recruitment to explore the effects of hydrologicalmortality and recruitment to explore the effects of hydrological modification modification (flood control) and management interventions of floodplain fishe(flood control) and management interventions of floodplain fishery yieldsry yields. . This This Dynamic Pool Model for Floodplain FisheriesDynamic Pool Model for Floodplain Fisheries is described in detail by is described in detail by (Halls et al 2001).(Halls et al 2001).

•• A combination of hydrological conditions and ageA combination of hydrological conditions and age--dependent fishing dependent fishing mortality rates drives changes to numerical and biomass density.mortality rates drives changes to numerical and biomass density. These in These in turn effect rates of recruitment, growth and natural mortality (turn effect rates of recruitment, growth and natural mortality (Figure 3).Figure 3).

Population Dynamics Models: ASPD ModelsPopulation Dynamics Models: ASPD Models

Hydrology

Population Density w

GrowthNatural Mortality

Biomass w

Recruitment

Natural Losses

Biomass w +1

Fishing Mortality Yield

Fishing Effort Catchability

Hydrology

Population Density w

GrowthNatural Mortality

Biomass w

Recruitment

Natural Losses

Biomass w +1

Fishing Mortality Yield

Fishing Effort Catchability

Figure 3 Figure 3 Schematic representation of the population model illustrating thSchematic representation of the population model illustrating the processes by which e processes by which the biomass in week w becomes the biomass in the following week,the biomass in week w becomes the biomass in the following week, w+1. The weekly process is w+1. The weekly process is repeated for the 52 weeks of the year, after which recruitment, repeated for the 52 weeks of the year, after which recruitment, determined by the surviving determined by the surviving spawning stock biomass, is added at the end of week 52. Solid lispawning stock biomass, is added at the end of week 52. Solid lines indicate direct influences or nes indicate direct influences or operations and broken lines indirect influences or occasional opoperations and broken lines indirect influences or occasional operations. Source: Halls et al erations. Source: Halls et al (2001)(2001)


•• The model has been fitted to landings of a small but widely The model has been fitted to landings of a small but widely abundant cyprinid, abundant cyprinid, Puntius sophorePuntius sophore in Northwest Bangladesh in Northwest Bangladesh (see Halls et al 2001). This species is abundant throughout (see Halls et al 2001). This species is abundant throughout Bangladesh and southern Asia, and shares similar life history Bangladesh and southern Asia, and shares similar life history characteristics with characteristics with HenicorhynchusHenicorhynchus species that dominate species that dominate catches in the Tonle Sap and Lower Mekong rivers.catches in the Tonle Sap and Lower Mekong rivers.

•• A simple hydrological model was used to generate weekly A simple hydrological model was used to generate weekly estimates of flooded area and volume required as an input to theestimates of flooded area and volume required as an input to themodel.model.

•• The model was used to explore the potential effects (benefits) oThe model was used to explore the potential effects (benefits) of f water level management within a flood control scheme and water level management within a flood control scheme and introducing closed seasons (to reduce overall effort) on fisheriintroducing closed seasons (to reduce overall effort) on fisheries es yield. yield.

•• The results indicated that beyond a The results indicated that beyond a flood water height (~ 9m at the flood water height (~ 9m at the study site) fish production is study site) fish production is determined mostly by dry season determined mostly by dry season water levels with production water levels with production increasing almost linearly with increasing almost linearly with increasing mean dry season water increasing mean dry season water levels (Figure 4).levels (Figure 4).

•• The model predicted that yield can The model predicted that yield can be improved by be improved by retaining moreretaining morewater during the water during the dry seasondry season. .

•• Lost yield arising from the Lost yield arising from the reductions in flood season water reductions in flood season water heights caused by flood control heights caused by flood control embankments could be embankments could be compensated by compensated by increasing the increasing the dry season water levelsdry season water levels(volumes) on modified floodplains. (volumes) on modified floodplains.

Figure 4. Isopleths of yield kg haFigure 4. Isopleths of yield kg ha--11yy--11

for for P. sophoreP. sophore in response to different in response to different combinations of dry and flood season combinations of dry and flood season water levels. Source: Halls et al (2001).water levels. Source: Halls et al (2001).



•• Closing the fishery during any month of the year was predicted tClosing the fishery during any month of the year was predicted to increase o increase production by at least 30% (fishery was heavily overproduction by at least 30% (fishery was heavily over--exploited). exploited).

•• Annual production was found to be maximized by removing ~ 85% ofAnnual production was found to be maximized by removing ~ 85% of the the fish biomass during October (and closing the fishery for the remfish biomass during October (and closing the fishery for the remaining 11 aining 11 months of the year) just prior to the drawdown (ebb flood) when months of the year) just prior to the drawdown (ebb flood) when fish have fish have achieved the majority of their yearachieved the majority of their year’’s growth and before losses due to s growth and before losses due to densitydensity--dependent mortality become significant. The surviving fraction dependent mortality become significant. The surviving fraction of of the spawning stock maximizes next yearthe spawning stock maximizes next year’’s densitys density--dependent recruitment. dependent recruitment.

•• Such a highly seasonal fishery is unlikely to be practicable or Such a highly seasonal fishery is unlikely to be practicable or equitable equitable given the prevailing access rules, particularly in Bangladesh. given the prevailing access rules, particularly in Bangladesh.

•• The greatest gains for the smallest initial sacrifices were predThe greatest gains for the smallest initial sacrifices were predicted to be icted to be achieved by closing the fishery during the dry season (Januaryachieved by closing the fishery during the dry season (January--April) when April) when small catches comprise the few remaining spawning individuals small catches comprise the few remaining spawning individuals experiencing low rates of growth and natural mortality. experiencing low rates of growth and natural mortality.

•• A closed season toward the end of the dry season could alternatiA closed season toward the end of the dry season could alternatively take vely take the form of dry season reserves (see accompanying presentations the form of dry season reserves (see accompanying presentations on on harvest reserves).harvest reserves).

•• Full details of the model algorithms and results can be found inFull details of the model algorithms and results can be found in Halls et al Halls et al (2001).(2001).

Other applicationsOther applications

•• The model has also been used to explore how water within flood The model has also been used to explore how water within flood control schemes (compartments) can best be managed for the control schemes (compartments) can best be managed for the benefit of benefit of both agriculture and fisheriesboth agriculture and fisheries (see Shanker et al 2004; (see Shanker et al 2004; 2005). The results of these investigations have been summarised2005). The results of these investigations have been summarised in in the accompanying presentation on the accompanying presentation on sluice gate managementsluice gate management..

•• The model has also been used to examine the effects of The model has also been used to examine the effects of dam dam releasesreleases of different depth and duration on downstream resident of different depth and duration on downstream resident fish populations (see Halls & Welcomme 2004). fish populations (see Halls & Welcomme 2004).

BEAM 4 BEAM 4 -- BioBio--Economic Analytical Fisheries ModelEconomic Analytical Fisheries Model

•• BEAM4 is a BEAM4 is a multispeciesmultispecies, , multigearmultigear yieldyield--perper--recruit simulation recruit simulation modelmodel

•• It can be used to assess the potential impacts of different fishIt can be used to assess the potential impacts of different fishery ery management measures (effort controls, closed seasons, minimum management measures (effort controls, closed seasons, minimum size limits etc) on fishery yieldssize limits etc) on fishery yields

•• Software originally published by FAO as 'BEAM4' (Software originally published by FAO as 'BEAM4' (SparreSparre & & WillmannWillmann, 1991). Now available as the general analytical YPR , 1991). Now available as the general analytical YPR model in model in FiSATFiSAT software suite (downloadable from FAO web site)software suite (downloadable from FAO web site)

•• Model applied by DFID project R4791 to floodplain river fishery Model applied by DFID project R4791 to floodplain river fishery data data from Bangladesh, Indonesia and Thailand (see Hoggarth & from Bangladesh, Indonesia and Thailand (see Hoggarth & Kirkwood, 1996)Kirkwood, 1996)

BEAM 4 BEAM 4 -- Model inputsModel inputs

•• BEAM 4 has high data requirements, but approximate inputs can beBEAM 4 has high data requirements, but approximate inputs can beestimated from a short time series sample of length frequency daestimated from a short time series sample of length frequency data ta (or from a sample of aged fish, for species where ageing is poss(or from a sample of aged fish, for species where ageing is possible)ible)

•• Data inputs:Data inputs:

–– Biological parameters for each species in the model Biological parameters for each species in the model -- growth rates (K, growth rates (K, LLinfinf, t, t00) and mortality rates (Z and M) ) and mortality rates (Z and M) -- estimated in this analysis from a estimated in this analysis from a 9 month time series of length frequency samples (ELEFAN method)9 month time series of length frequency samples (ELEFAN method)

–– Size selectivity of each gear type for each species, determined Size selectivity of each gear type for each species, determined approximately from the length frequency dataapproximately from the length frequency data

–– Seasonality of each gear (modelled by entering the actual monthlSeasonality of each gear (modelled by entering the actual monthly y fishing efforts of each gear)fishing efforts of each gear)

•• For the R4791 analysis, the model was fitted for up to five specFor the R4791 analysis, the model was fitted for up to five species ies guilds in each fishery (each country study site) and up to ten fguilds in each fishery (each country study site) and up to ten fishing ishing gear typesgear types

Study SitesFish Species Guilds

Von Bertalanffy Growth Rates

Mortality Rates(yr-1)

K (yr-1) Linf (cm) Natural, M Fishing, F1

BangladeshLarge CarpsLarge PredatorsMedium-sized BlackfishSmall WhitefishSmall Shrimps

0.30.70.70.7'1.0

75703015'4

0.60.92.43.0'3.0

4.12.12.52.4'1.9

IndonesiaFloodplain PredatorsRiver PredatorsMedium-sized BlackfishMedium/small whitefishLarge Shrimps

0.70.6

0.65'0.7'1.0

655828'20'30

0.90.91.2'1.5'2.0

2.02.02.0'2.0'2.0

ThailandSnakehead PredatorsMedium-sized BlackfishMedium-sized Whitefish

0.50.6'0.6

6530'30

0.91.2'1.2

2.02.0'2.0

Growth and mortality rates used in R4791 BEAM 4 analysisGrowth and mortality rates used in R4791 BEAM 4 analysis

11 Maximum fishing mortality rate, for fish at lengths fully selectMaximum fishing mortality rate, for fish at lengths fully selected by all gear types.ed by all gear types.

Example results from R4791 BEAM4 AnalysisExample results from R4791 BEAM4 Analysis

•• Figure shows % Figure shows % change to catch for change to catch for each gear type each gear type (listed on x(listed on x--axis) for axis) for four alternative four alternative management management measures (shown by measures (shown by symbols)symbols)

•• Note variation in Note variation in effects of different effects of different measures on each measures on each gear, but limited gear, but limited overall benefits of overall benefits of any measure, shown any measure, shown as TOTAL, averaged as TOTAL, averaged across all gearsacross all gears

•• These measures These measures would change the would change the allocation of catch, allocation of catch, but not the totalbut not the total

ReferencesReferences

Dooley, J., Jenness, J., AguilarDooley, J., Jenness, J., Aguilar--Manjarrez, J. & Riva, C. (in press.) African Water Manjarrez, J. & Riva, C. (in press.) African Water Resource Database (AWRD). GIS based tools for aquatic resource mResource Database (AWRD). GIS based tools for aquatic resource management. anagement. CIFA Technical PaperCIFA Technical Paper. No. 33. Rome, FAO, 2005. . No. 33. Rome, FAO, 2005. http://http://www.fao.org/fi/eims_search/publications_form.asp?langwww.fao.org/fi/eims_search/publications_form.asp?lang=en=en

Grainger, R.J.R. & Garcia, S.M. (1996). Chronicles of marine fGrainger, R.J.R. & Garcia, S.M. (1996). Chronicles of marine fishery landings (1950ishery landings (1950--1994): Trend analysis and fisheries potential. 1994): Trend analysis and fisheries potential. FAO Fisheries Technical PaperFAO Fisheries Technical Paper. . 359359. . Rome, FAO. 51pp. Rome, FAO. 51pp. http://http://www.fao.org/fi/eims_search/publications_form.asp?langwww.fao.org/fi/eims_search/publications_form.asp?lang=en=en

Haddon, M. (2001). Haddon, M. (2001). ModellingModelling and quantitative methods in fisheries, Chapman Hall, and quantitative methods in fisheries, Chapman Hall, London, 406pp.London, 406pp.

Halls, A.S. & Welcomme, R.L. (2004). Dynamics of river fish popHalls, A.S. & Welcomme, R.L. (2004). Dynamics of river fish populations in response to ulations in response to hydrological conditions: A simulation study. hydrological conditions: A simulation study. River Research and ApplicationsRiver Research and Applications. . 2020: : 985985--1000. 1000. http://www3.interscience.wiley.com/cgihttp://www3.interscience.wiley.com/cgi--bin/jissue/109857602bin/jissue/109857602

Halls, A.S., Burn, R.W., & Abeyasekera, S. (2002) InterdiscipliHalls, A.S., Burn, R.W., & Abeyasekera, S. (2002) Interdisciplinary Multivariate Analysis nary Multivariate Analysis for Adaptive Cofor Adaptive Co--Management. Management. Final Technical Report to the UK Department for Final Technical Report to the UK Department for International Development, MRAG Ltd, London, January 2002, 125ppInternational Development, MRAG Ltd, London, January 2002, 125pp. . http://p15166578.pureserver.info/fmsp/Home.htmhttp://p15166578.pureserver.info/fmsp/Home.htm

Halls, A.S., Kirkwood, G.P. and Payne, A.I. (2001). A dynamic pHalls, A.S., Kirkwood, G.P. and Payne, A.I. (2001). A dynamic pool model for floodplainool model for floodplain--river fisheries. river fisheries. Ecohydrology and HydrobiologyEcohydrology and Hydrobiology , , 11 (3): 323(3): 323--339. 339. http://http://www.ecohydro.pl/index.phpwww.ecohydro.pl/index.php

Halls, A.S. 1999. Halls, A.S. 1999. Spatial models for the evaluation and management of Inland FisheSpatial models for the evaluation and management of Inland Fisheriesries. . Final Report prepared for the Food and Agriculture Final Report prepared for the Food and Agriculture OrganisationOrganisation of the United of the United Nations (FIR 1998 Nations (FIR 1998 PlansysPlansys 232200120).232200120).

HilbornHilborn, R. & C.J. Walters (1992). Quantitative Fisheries Stock Assess, R. & C.J. Walters (1992). Quantitative Fisheries Stock Assessment. Choice, ment. Choice, Dynamics and Uncertainty. London, Chapman Hall.Dynamics and Uncertainty. London, Chapman Hall.

Hoggarth, D.D., Abeyasekera, S., Arthur, R., Beddington, J.R., BHoggarth, D.D., Abeyasekera, S., Arthur, R., Beddington, J.R., Burn, R.W., Halls, A.S., urn, R.W., Halls, A.S., Kirkwood, G.P., McAllister, M., Medley, P., Mees, C.C., Pilling,Kirkwood, G.P., McAllister, M., Medley, P., Mees, C.C., Pilling, G.M., Wakeford, R., G.M., Wakeford, R., and Welcomme, R.L. (in press). Stock Assessment for Fishery Maand Welcomme, R.L. (in press). Stock Assessment for Fishery Management nagement –– A A Framework Guide to the use of the FMSP Fish Stock Assessment TooFramework Guide to the use of the FMSP Fish Stock Assessment Tools. ls. FAO FAO Fisheries Technical PaperFisheries Technical Paper No. XXX. Rome, FAO. 2005. XXX pp. No. XXX. Rome, FAO. 2005. XXX pp. http://http://www.fao.org/fi/eims_search/publications_form.asp?langwww.fao.org/fi/eims_search/publications_form.asp?lang=en=en

McCullagh, P. & Nelder, J.A. (1989). Generalized Linear Models.McCullagh, P. & Nelder, J.A. (1989). Generalized Linear Models. London, Chapman & London, Chapman & Hall.Hall.

MRAG (1993) Synthesis of Simple Predictive Models for Tropical MRAG (1993) Synthesis of Simple Predictive Models for Tropical River Fisheries. River Fisheries. Report to the Overseas Development Administration. 85 pp. Report to the Overseas Development Administration. 85 pp. http://p15166578.pureserver.info/fmsp/Home.htmhttp://p15166578.pureserver.info/fmsp/Home.htm

MRAG (1994) Synthesis of Simple Predictive Models for Tropical MRAG (1994) Synthesis of Simple Predictive Models for Tropical River Fisheries River Fisheries --Supplementary Report. Report to the Overseas Development AdminiSupplementary Report. Report to the Overseas Development Administration. 29 pp. stration. 29 pp. http://p15166578.pureserver.info/fmsp/Home.htmhttp://p15166578.pureserver.info/fmsp/Home.htm

Shankar, B., Halls, A.S., & Barr, J. (2005). The effects of surShankar, B., Halls, A.S., & Barr, J. (2005). The effects of surface water abstraction for face water abstraction for rice irrigation on floodplain fish production in Bangladesh. Intrice irrigation on floodplain fish production in Bangladesh. Int. J. Water, Vol. 3, No. 1, . J. Water, Vol. 3, No. 1, 2005.2005.

Shankar, B., Halls, A.S., & Barr, J. (2004). Rice versus fish Shankar, B., Halls, A.S., & Barr, J. (2004). Rice versus fish revisited: on the integrated revisited: on the integrated management of floodplain resources in Bangladesh.management of floodplain resources in Bangladesh. Natural Resources ForumNatural Resources Forum, , 2828: : 9191--101. 101. http://www.blackwellhttp://www.blackwell--synergy.com/toc/narf/28/2synergy.com/toc/narf/28/2



WelcommeWelcomme, R.L (1985). River Fisheries. FAO Fisheries Technical Paper , R.L (1985). River Fisheries. FAO Fisheries Technical Paper 262262, , FAO,RomeFAO,Rome. . http://http://www.fao.org/fi/eims_search/publications_form.asp?langwww.fao.org/fi/eims_search/publications_form.asp?lang=en=en

Welcomme, R.L. (2001)Welcomme, R.L. (2001) Inland Fisheries: Ecology and ManagementInland Fisheries: Ecology and Management. Fishing News . Fishing News Books, Blackwell Scientific, Oxford, 358pp.Books, Blackwell Scientific, Oxford, 358pp.

Zar, J. H. (1984). Zar, J. H. (1984). Biostatistical AnalysisBiostatistical Analysis. New Jersey, Prentice Hall. 718 pp.. New Jersey, Prentice Hall. 718 pp.

This presentation is an output from a project funded by the UK DThis presentation is an output from a project funded by the UK Department for epartment for International Development (DFID) for the benefit of developing cInternational Development (DFID) for the benefit of developing countries. The ountries. The

views expressed are not necessarily those of the DFID.views expressed are not necessarily those of the DFID.

This project was funded through This project was funded through DFID'sDFID's Fisheries Management Science Fisheries Management Science Programme (FMSP). For more information on the FMSP and other proProgramme (FMSP). For more information on the FMSP and other projects jects

funded through the Programme visit funded through the Programme visit http://http://www.fmsp.org.ukwww.fmsp.org.uk

Project details and creditsProject details and credits

FMSP Project R5953 FMSP Project R5953 –– Fisheries dynamics of modified Fisheries dynamics of modified floodplains in southern Asiafloodplains in southern Asia

•• Start Date: Start Date: 03/1994 03/1994 •• End Date: End Date: 03/199703/1997

•• Project Collaborators: Project Collaborators: –– MRAG (Dan Hoggarth, Ashley Halls); MRAG (Dan Hoggarth, Ashley Halls); –– CRIFI, Indonesia (CRIFI, Indonesia (FuadFuad CholikCholik, , AgusAgus UtomoUtomo, , OndaraOndara); ); –– BAU BAU MymensinghMymensingh (M.A. (M.A. WahabWahab, Kanailal Debnath, , Kanailal Debnath, RanjanRanjan Kumar Kumar

Dam)Dam)

•• Key References: MRAG (1997); Halls et al (1998); Hoggarth et al Key References: MRAG (1997); Halls et al (1998); Hoggarth et al (1999); Hoggarth et al (1999b). (1999); Hoggarth et al (1999b).

•• Project web page: Project web page: http://http://www.fmsp.org.ukwww.fmsp.org.uk/FTRs/r5953//FTRs/r5953/.htm.htm

FMSP Project R5030 FMSP Project R5030 –– Synthesis of simple predictive Synthesis of simple predictive models for river fish yields in major tropical riversmodels for river fish yields in major tropical rivers

•• Start Date: 04/1993Start Date: 04/1993•• End Date: 07/1993End Date: 07/1993

•• Project Collaborators: Project Collaborators: –– MRAG (Ashley Halls)MRAG (Ashley Halls)–– FAO (Jim FAO (Jim KapetskyKapetsky))

•• Key References: MRAG (1993; 1994); Halls (1999); Hoggarth et al Key References: MRAG (1993; 1994); Halls (1999); Hoggarth et al (in press); Dooley et al (in press).(in press); Dooley et al (in press).

•• Project web page: Project web page: http://http://www.fmsp.org.ukwww.fmsp.org.uk/FTRs/r5030//FTRs/r5030/.htm.htm

FMSP Project R7834 FMSP Project R7834 –– Interdisciplinary multivariate Interdisciplinary multivariate analysis for adaptive coanalysis for adaptive co--managementmanagement

•• Start Date: 01/10/2000 Start Date: 01/10/2000 •• End Date: 31/01/2002End Date: 31/01/2002

•• Collaborators: Collaborators: –– MRAG (Ashley Halls) MRAG (Ashley Halls) –– Reading University SSC (Bob Burn, Reading University SSC (Bob Burn, SavitriSavitri AbeyasekeraAbeyasekera) ) –– WorldFish Centre (WorldFish Centre (KuperanKuperan ViswanathanViswanathan) ) –– IFM (Doug Wilson, IFM (Doug Wilson, JesperJesper NeilsenNeilsen). ).

•• Key ReferencesKey References: : Halls et al (2002).Halls et al (2002).

Modelling Floodplain River Fisheries – an Introduction Training Workshop Materials · 2016. 8....

Documents

Transcript of Modelling Floodplain River Fisheries – an Introduction Training Workshop Materials · 2016. 8....