D4.1 Review of Environmental Models - SEAT Global

www.seatglobal.eu
Review of Environmental
Models
31 August 2010
Authors: L.I. Munro, L. Falconer, T.C. Telfer
and L.G. Ross
SEAT Deliverable Ref: D 4.1
Page 1

INTRODUCTION
The global demand for aquaculture products has risen in recent years due to an increased
need for food as a result of the growing global population. In order to meet this demand
countries in the EU are reliant on exports from other regions; particularly Asia which is
responsible for producing the largest amount of aquaculture products in the world (De Silva
& Davy 2009; Lymer et al, 2008). It is important to ensure that whilst the demand for
aquaculture products is being met the producers also aim for ethical and sustainable
products. Sustainability is a major challenge in aquaculture development as a sustainable
aquaculture system that does not provide a suitable standard of living for the producer will
not last long and likewise a highly profitable aquaculture system that heavily exploits and
degrades the natural resources and the environment will also have a limited lifespan (Leung
& El-Gayar, 1997). It is vital that aquaculture producers get an acceptable balance between a
profitable business and a sustainable business to ensure that aquaculture can continue to
grow and develop so as to meet the global demand.
The SEAT (Sustaining Ethical Aquaculture Trade) project is a large-scale collaborative EU
project with 14 participating organisations from Europe and Asia. The overall aim of the
project is to increase the sustainability and ethical factors (including food safety and quality,
animal welfare, environmental services, socioeconomic issues) of four major Asian
aquaculture products (tilapia, Pangasius catfish, peneid Shrimp and Macrobrachium Prawn)
that are imported into Europe . Trade has become an important factor in Asian aquaculture
where the industry has evolved in the last few decades from production of fish and shellfish
mainly for domestic consumption to the export of fresh and processed aquaculture produce
(Ahmed, Dey & Garcia, 2007). Through studying the production and trade of the case study
species from four producer countries in South-east Asia (Bangladesh, China, Thailand and
Vietnam) the SEAT project will obtain a detailed understanding of the food production and
market chains and this knowledge will be used to further enhance sustainability and ethical
trade between the Asian producers and the European market.
An important aspect of sustainability is environmental management as it is essential that any
activity (such as aquaculture) that takes place does not significantly alter or negatively
impact the environment or the commercial/social use of the environment by other users
both at present and in the future (Black, 2001). The impacts of aquaculture on the
environment are well documented; use of natural resources (space, seed, feed),
introduction of organic material into the environment (uneaten food, faeces etc), input of
chemicals into the environment (therapeutants, pesticides, disinfectants etc), introduction
of feral animals (escapees from aquaculture systems), disease and parasites and the
subsequent interaction with the natural animal populations (Beveridge, 2004; Black, 2001;
Frid & Dobson, 2002; Pillay, 2004). These issues are relevant to all aquaculture production
systems; however the degree of impact varies as there are many factors that influence the
type and scale of impact from an aquaculture system such as the species produced,
structure of system, stocking density, food supply and husbandry techniques (Cripps &
Kumar, 2003; Pillay, 2004).
It is essential when working towards sustainable aquaculture systems that these issues are
both monitored and managed to ensure that any potential impact is minimised. Variables
known as environmental indicators are biological, chemical, physical, social and/or economic
variables that signal change and can therefore be used to determine any impact (Frankic &
Hershner, 2003). There are many different indicators used throughout the industry; Kalantzi
& Karakassis (2006) reviewed 41 papers analysing the benthic effects of fish farming and
Page 2

found that over 120 biological and geochemical variables were monitored between the
studies. The use of different variables can make it difficult to compare impact between sites
and study areas, however Kalantzi & Karakassis (2006) also noted that it would be difficult to
develop a uniform suite of indicators across a large geographical area that would be robust
enough to take into account all of the complex interactions of variables and geographical
variations such as sediment type, depth and latitude. A good monitoring programme should
therefore be designed so that selected indicators are simple to understand, practical, costeffective,
easily recorded and sampled at an appropriate frequency so that a signal can be
detected above the natural background environment that would indicate a change in the
environmental status (Southall, Telfer & Hambrey, 2004; Fernandes et al, 2001; Hansen et al,
2001).
Monitoring programmes will often specify the variable(s) to be measured (e.g. dissolved
oxygen, suspended solids, redox potential ) without defining the threshold value beyond
which mitigation is required (GESAMP, 1996). Some countries use environmental quality
objectives (EQO) and environmental quality standards (EQS) as a method of managing
environmental impact (Fernandes et al, 2001; Beveridge, 2004). EQS are the threshold value
of an environmental indicator and they are usually established by regulatory authorities
(GESAMP, 1996). EQOs are developed to manage the environment to ensure minimal
impact (e.g. an EQO could be protection of aquatic life) and the EQS are then set to achieve
the objectives (e.g. an EQS could be maximum level of a pollutant) (Fernandes et al, 2001). It
can be expensive and time consuming to define and set EQOs and EQS and many countries
do not have regulation systems in place to facilitate the process (Beveridge, 2004). In order
to help countries develop standards for certifying aquaculture the World Wildlife Fund
(WWF) developed several species-specific criteria resulting from a series of “dialogues”
(consisting of producers, buyers, nongovernmental organisations (NGO’s) and other
stakeholders) which aim to minimise or eradicate many environmental and social impacts of
aquaculture (WWF, 2010). A monitoring programme does not need to be complicated and,
as many aquaculture production systems are limited by time and finances, it is more
efficient to develop a simple programme that is easily understood, can be sampled easily
and thus likely to be adhered to rather than a complicated system that may have detailed,
scientific measurements but is at risk of being abandoned due to impracticality of day to day
management. This is particularly relevant to many aquaculture systems within Asia.
It can be expensive both in terms of time and money to collect the large amount of data
required for a full environmental assessment; particularly if the study area has a substantial
size or consists of multiple locations. Conversely, baseline data can often be used more
effectively and efficiently in computer models developed to predict changing environmental
conditions. The advantage of modelling is that it can be used to predict environmental
impact where as measurements from monitoring studies are an indication of what has
already happened. Environmental models can be thought of as simulations of reality
incorporating knowledge and experience to simplify complex situations, and simulate
complicated operations or situations that may be difficult or unsafe (worst case scenario
modelling) to manipulate in the natural environment (Leung & El-Gayar, 1997;
Nirmalakhandan, 2002; Mulligan & Wainwright, 2004).
Environmental modelling has become increasingly popular in the sciences in recent years
(Smith & Smith, 2007), and models are used widely for regulatory purposes by organisations
such as SEPA (Scottish Environment Protection Agency) and US EPA (US Environment
Protection Agency), which are involved in licensing of environmental activities (SEPA, 2010;
USEPA, 2010). The demand for environmental regulation and predictive analysis has been
Page 3

increasing and therefore models have been consistently created, developed and adapted
since the 1970s (Ford, 1999) The first models were basic and based on mathematical
equations developed from data collected over a period of time (Hall & Day JR. 1977). System
dynamics models were first developed by Forrester in the 1960s using mathematical
equations, which formed the basis of future models (Vlachos et al., 2007).
Environmental models are essentially tools, based on mathematical algorithms, which
enable predictions of environmental change and their consequences (Ford, 1999) using
baseline and subsequent monitoring data. To establish the usefulness of a model a
sensitivity analysis must be carried out to determine if a model remains accurate when
forcing functions are altered. System dynamics models tend to have a causal loop interface
to allow the user to recognise major feedback mechanisms be they positive or negative
(Ford, 2009), to better understand the interaction of systems within the final models
developed (Vlachos et al., 2007).
Environmental models are also used in aquaculture for farm management to simulate the
quality of the water within the farming system to help minimise fish deaths and predict
profitability (Beveridge, 1996). A common method used for basic modelling is called the
Mass Balance equation this can be utilised for many different parameters but is most widely
used in a water quality context to model nitrogen and phosphorus concentrations in and
from aquaculture systems. When using such models there has been a “blanket” approach to
their implementation through application of general guidelines. However, it is now clear that
these general guidelines are not relevant for every system (Panchang et al, 1997), for
example site suitability for net pen culture should be modelled and considered on a site by
site basis as environmental variability make a general approach immaterial (Dudley et al,
2000). Consequently, it is important that data available are typical to the system selected to
prevent any restrictions on the model’s usefulness (Cromey et al, 2002).
Dynamic models are normally applied through computer software either coded as specific
programs, or formulated through modelling framework software, such as STELLA TM or
VENSIM. The latter offers a flexible and consistent approach to modelled giving the
opportunity to develop a range of models which can be easily disseminated and used, while
allowing further model development and adaptation by other users. For example IAAS and
MMFA (see Section 1) were both created using the software STELLA TM , but these models
have independent applications within different systems (Jamu et al., 2002a; Schaffner et al.,
2009).
Environmental modelling is done at different spatial and temporal scales, even though it is
mostly thought of as investigating local dynamic changes at farm level. Larger spatial scale
modelling is often required when investigating aquaculture development and its impacts at a
regional level where, for example, impacts from individual farms can be additive and
combined with other inputs from agricultural or urban run-off (Midlen & Redding, 1998).
Geographical Information Systems (GIS) are modelling frameworks designed for use
different spatial levels, as they can provide both general and site specific information and
investigate issues at both local and regional scale (Silvert & Cromey, 2001). GIS is particularly
useful as an environmental management tool as it organises, analyses and presents
geographical data in a useful and efficient manner, using standard data formats. In terms of
aquaculture development, the advantage of GIS is that the impact from several farms could
be analysed on a regional scale as well as taking into account inputs from other sources,
therefore the results are truly representative of the activities taking place in the area and
the subsequent environmental conditions.
Page 4

Work Package 4 (WP4) of the EC 7FP SEAT (Sustainable Ethical Aquaculture Trade) research
and development project will use environmental models to predict the fate and distribution
of major aquaculture wastes; assessing environmental impact of waste nutrient inputs and
chemical contamination (particularly linked with Work Packages 6 and 7). System dynamics
modelling will look at farm level inputs such as feed and chemicals and then spatial
modelling will incorporate the wider field inputs taking into account nutrient and
contaminant inputs from external sources such as agriculture resulting in a more holistic and
representative indication of the environmental situation in each case study area. Preliminary
dynamic and spatial models will be developed initially and then these models will be further
developed to establish the environmental models that will be used within the SEAT project
meeting the requirements of Deliverable 4.2b. The models developed and their outputs will
have the potential to be used in wider systems modelling.
Page 5

SECTION 1 – REVIEW OF ENVIRONMENTAL MODELS USED IN AQUACULTURE
DEVELOPMENT AND REGULATION
1.1 Introduction
System dynamic models have become more widely used for investigation of environmental
and ecosystem change in recent years. Of particular relevance and those created for use in
aquaculture development and sustainability; most of which are at Farm process level.
Mechanistic models are considered most suitable for simulating nutrient flow (i.e. nitrogen
and organic matter) in relation to aquaculture, whereas stochastic models are most suited to
the simulation of fish growth, and temperature and dissolved oxygen changes (Ford, 1999).
Modern models tend to have both mechanistic and stochastic components, to enable a
more robust systems simulation. SEAT Work Package 4 aims to produce system dynamics
models to simulate the fate of nutrients and chemicals at local area and farm level for
selected species and localities in each study country. These models will be used as tools to
aid the move towards sustainability of the aquaculture trade in SE Asia. A number of models
have been developed for, or can be used in, sustainable aquaculture development and
management, and will be reviewed here. However, many dynamic models are developed for
specific purposes and are compiled within a variety of modelling software packages. As it is
likely specific models will be developed as part of WP4 the relevant modelling packages are
also reviewed. Conclusion and discussion of the applicability of the models and software to
the SEAT WP4 will be made.
1.2 Modelling Packages and Frameworks
1.2.1 DYNAMO,
DYNAMO is one of the original system dynamic modelling software packages using objectoriented
modelling (Ford, 2009). It was used widely from the 1960s-1980s and has produced
one of the most well known models; World3 (Cellier, 2008), which simulated the interactions
between population, industrial growth, food production in relation to ecosystem limitations.
Though DYNAMO was originally written in an assembly language for mainframe computers
in 1959 (Sammet, 1969), it now uses Mathematical Formula Translating (FORTRAN)
computer language for PC computers (Cellier, 2008), though more recently DYNAMO has
been rebuilt Lisp for Windows for greater PC applicability (Normark, 1997), but has largely
been superseded for building dynamic models by integrated GUI-based systems.
DYNAMO paved the way for the newer more up to date modelling packages such as STELLA,
Vensim and Powersim (see below), all of which have kept the original format of Forrester’s
(1961) object oriented modelling approach (Ford, 2009).
1.2.2 STELLA TM ,
STELLA was one of the first object-oriented modelling software developed after DYNAMO
(Ford, 1999). Originally developed for used on Apple Macintosh computers, it has been
redesigned for PCs using MS Windows operating systems as these are more widely used
(Losordo and Allan, 2005). STELLA is object oriented modelling software, which allows
efficient building of models through a “drag and drop” graphic interface using stocks and
Page 6

flows to display the route of conceptual data flow through the model (Costanza & Gottlieb,
1998). By defining rules and algorithms to control this data transfer between objects, system
dynamic models can be built.
STELLA is a visual modelling tool, with a User-Friendly interface, which has been utilised in
both business and academia (Isee systems, 2010). This interface has allowed STELLA to be
one of the most widely used of the modelling software packages available to date (Costanza
& Voinov, 2001). Models are constantly being developed for different systems and areas of
research and circulated through application of a VLE version of the software. This not only
allows re-use and flexible programming of models by many users but the estabilishment of a
library of routines which can be incorporated into a variety of models (Isee systems. 2010).
STELLA has been widely used for management in environmental systems, for example for
use in the US to model changes in culture of sugarcane in Brazil for the production of
ethanol for use as a biofuel (Dias De Oliveira et al., 2005). It has also been used extensively
to develop models for prediction of climate change on environmental systems. More
recently STELLA has being used increasingly to build models of aquatic systems including
sites for aquaculture production and management (Grant et al., 1997), especially concerning
nutrient cycling and material flow; common concerns for sustainable aquaculture (Pillay and
Kutty, 2005).
An example of models of systems dynamics developed in STELLA include a wetland model
for the Olentangy River Wetland Research Park (ORWRP) (Zhang & Mitsch, 2005)
Simulations were created of the hydrological processes and nutrient flow interactions
between multi-wetland areas (Zhang & Mitsch, 2005). These models, which we based on a
mass balance approach for nutrient flow, included evaporation and retention from the river
as well as outside influences such as storm water additions. Subsequent field verification of
the models showed that they were a very accurate simulation of actual systems interactions.
STELLA models of dynamic fluctuations within systems can be readily transferred to spatial
modelling frameworks, such as GIS (Mazzoleni et al., 2003). Coupling a model developed
through STELLA with GIS can provide the user more visually accessible information of what is
happening within a system at different spatial scales, thus making it easier to convey the
results from any given simulation to a non-specialist end-user group (Elshorbagy et al., 2005)
1.2.3 Powersim TM
Powersim is graphical interface object-oriented modelling software, originally created for
modelling business strategies (Ford, 2009; Powersim, 2010). However, this software has
been shown to provide an excellent modelling framework for dynamic modelling of
biological systems (Franco et al, 2006).
Similar in nature to STELLA, Powersim has greater functionality and the ability to produce
self standing, compiled models for easy dissemination and integration into other modelling
frameworks such as GIS. The software is also easily accessible to non-mathematicians and is
particularly well designed for modelling spatial dynamics (Maxwell, 1999).
Powersim has been used to model the uptake of nitrogenous substances by marine
phytoplankton, namely the interactions between nitrate and ammonium within the
environment and organism (Flynn et al, 1997). The model took into account transport of
ammonia and nitrate into the cell and its subsequent cellular level biochemical reactions.
Page 7

Influence of the photosynthesis and respiration were assessed by including biochemical
effects of photoperiod into the model (Flynn et al., 1997), to give good estimates of the
relationship between phytoplankton productivity and nitrogenous nutrients within the
system. The outcome, therefore, simplified a complex system to give good estimates
appropriate to effective water quality management.
The Powersim software has been used to simulate sustainable shellfish culture in China
(Nunes et al., 2003). Powersim was used specifically for this task as, not only is it a powerful
modelling framework, but the graphic use of object-oriented stocks and flows allowed
stakeholders within the research project to see the modelling processes and the steps
involved in providing the outcomes (Nunes et al, 2003). Implementation of the outcomes of
the model into actual aquaculture practice was found to be easier as there was considerable
stakeholder “buy-in” and understanding at each stage of the modelling and management
process.
A further example of the use of Powersim for aquaculture is the development of a growth
model for penaeid shrimp (Franco et al, 2006). The model simulated the growth of the
shrimp, P. indicus, from early life stages, not including any larval stages, due to their short
duration in relation to the complete life cycle (FAO.1980). The model has applications as a
feed management tool, as it took into account physiological processes including ingestion,
which has wider aquaculture development implications.
1.2.4 VENSIM
Like STELLA and Powersim, Vensim is an object-oriented modelling framework with a graph
front end interface to create and adapt systems dynamics models. It was created by Ventana
systems as an in-house model for consultation on government and business projects
(Ventana, 2010). It is a powerful but highly complex framework (Ford, 2009), though a
personal learning environment (PLE) version is available as freeware
Vensim PLE software has been used to model water management issues in Las Vegas,
Nevada (Stave, 2002) to provide the public and water users with a visual concept of the
water levels available in Lake Mead, where the water is going and its primary uses in the city
(Stave, 2002). Vensim is useful for conveying such parameters as it can formulate causal loop
diagrams for visual representation of model outputs, whilst employing simplified conceptual
diagrams containing standard stocks and flow notation (Ford, 2009).
Likewise, Vensim has been used to model the dynamics of the Cannonville reservoir in New
York State, which provides New York with water. The reservoir suffers from significant
annual euthrophication and required effective predictive dynamic models for water resource
management (Schneiderman et al, 2002). Vensim was used to develop the GWLF
(Generalised Watershed Loading Functions) to model the point and non-point sources of
pollution including that from agriculture and surrounding sewage treatment plants. The
model was developed using four sub-modules to create the overall model, containing data
for sediment and particulate nutrients, hydrologic water balance, dissolved nutrients and
septic systems nutrients (Schneiderman et al., 2002). These four sub-modules were then
modelled by defining the point and non-point inputs to the system. (Schneiderman et al,
2002). The results were found to be effective in predicting the monthly stream flow,
sediment yield, particulate P and dissolved nutrients, though it did not produce effective
simulations for nutrient fluxes as it failed to simulate an extreme event which occurred in
Page 8

1.2.5 WASP
January 1996 (Schneiderman et al., 2002). However, further development of GWLF has lead
the USEPA to consider integrating this model within their BASIN Watershed Assessment
Tools. (Schneiderman et al, 2002)
The Vensim modelling package has been successfully linked with ArcGIS via Excel
spreadsheets to model soil erosion and the related nutrient impact on the Keelung
watershed in Taiwan (Yeh et al., 2006). Here five parameters were considered including soil
erosion, sediment transport, runoff, nutrient flow and economic impacts (Yeh et al., 2006).
The basic watershed characteristics were measured through ArcGIS and the system
dynamics were carried out in Vensim. Excel spreadsheets were used to convert results
produced from ArcGIS to inputs for the Vensim model and for storage of the resulting
Vensim outputs (Yeh et al., 2006).
WASP is an environmental modelling tool developed by the USEPA for water bodies (USEPA
2010). The hydrodynamic aspect of the model utilises the Saint-Venant equations for
variable flow (Vuksanovic et al., 1996). It is robust modelling system which can be used for
simulations in water bodies as diverse as coastal and estuarine environments (USEPA, 2010).
It simulates the transport of contaminants through surface waters (US EPA 2010) and while
the results can be useful they are general and often not specific enough to make detailed,
system specific management decisions (Schaffner et al, 2009).
The WASP program has been widely used for estuarine systems throughout the world
(USEPA, 2010). The program is simple to use and can easily integrate data from
hydrodynamic and sediment transport models (USEPA, 2010). The post-processing facility
provides two data displays: the first giving a spatial modelled outputs for each parameter,
and the second graphical showing concentrations over time to illustrate general trends for
comparison with empirical data (USEPA, 2010). WASP has been used effectively for “realtime”
development of management strategies for the ‘Little River embayment in Lake
Allatoona, Georgia’ (US EPA, 2010). The model was tested in an estuary in Belgium to
simulate the flow of PCBs at that particular time in the estuary. The WASP has also been
used to simulate the flow and spatial distribution patterns of PCBs in a Belgian estuary. The
simulation included transport and sorption rates of the isomers studied to sediment
particles (Vuksanovic et al., 1996). Comparison to measured levels of PCBs showed a clear
correlation between high of PCBs and areas of enhanced turbidity, agreeing with the
hypothesis that sorption of PCBs correspond to levels of particulate materials and resulting
in an effective simulation model (Vuksanovic et al., 1996). The eutrophication function of
WASP was used to determine nutrient transport and cycling to attempt to determine how to
reduce the nutrient loading on the estuary (Wool et al., 2003). The performance statistics
found the model to simulate the system nutrient dynamics and predict the water quality
accurately (Wool et al., 2003).
The WASP model was designed to be highly adaptable where users have the opportunity to
customize sub routines for different aquatic systems (USEPA, 2010). Parameters and state
variables can be added, removed and modified to fit to a system where the user has a
detailed knowledge of environmental interactions (Ward Jr & Benaman, 1999).
When using WASP, external hydrodynamic tools can be avoided by incorporating the flow
data directly in the program using a link to other models such as the companion
Page 9

hydrodynamic model DYNHYD for simulating complex hydrodynamics in the system (Ward Jr
& Benaman, 1999). WASP also has the facility to be directly integrated in to GIS for spatial
modelling, using a series of subroutines, called WISP, for data format interchange (Ward Jr &
Benaman, 1999).
1.2.6 ECOPATH
The Ecopath model software operates through a mass balance approach to estimate energy
transfer through different trophic levels within ecosystems to investigate change in
population dynamics over a predefined time period, usually 1 year (Christensen and Walters,
2004). The software was developed in the early 1980s and has been developed and
improved over a series of versions (Ecopath. 2010). All have been freely available online for
research and educational use. Due to its adaptability Ecopath has been used in ecosystems
ranging from temperate to the tropical climates. It has also been used to investigate
ecological pathways and energy balances in pond to open sea systems. In 2004 the Ecopath
package was used for ecosystem studies in 120 countries with over 2400 registered users
(Christensen and Walters, 2004).
Most studies involving Ecopath are based on management of fisheries stocks. The software
enables the user to treat the different trophic levels as functional entities within and
ecosystem (Pauly et al, 2000). Although this can be useful on some levels this is not an
effective way to model a full system dynamic as Ecopath fails to account for interactions of
trophic flows in a system, e.g. it is not suitable for modelling feeding patterns for multispecies
interactions (Christensen & Walters, 2004). In consequence, Ecopath now includes
two additional model components; Ecosim and Ecospace (Ecopath. 2010; Pauly et al, 2000).
Thus, later versions of the software allow the user to fit their models to the real time data
and visually display spatial data (Christensen and Walters, 2004).
1.2.7 ECOSIM
Ecosim is an “add-on” component to the Ecopath simulation programme (Ecopath, 2010) to
further investigate the the dynamics of system interactions. It focuses mainly on biomass
dynamics and consequences on this of multispecies interactions within the ecosystem
(Christensen and Walters, 2004).
The Ecosim package models trophic flows and their dynamic interactions within the Ecopath
modelling software. Ecopath with Ecosim (EwE) has been used to model multispecies culture
as well as investigate interactions between fisheries stocks and their predator species
(predator–prey interactions). One particular use of this in fisheries has been to look at the
relation between fisheries and upwelling to develop fisheries management policies in these
productive areas and changing environments (Pauly et al, 2000).
1.2.8 ECOSPACE
Ecospace is the third component of the Ecopath package. This incorporates all of the key
elements of the Ecosim software to model the dynamic flows within a spatial framework
Page 10

1.2.9 SIMILE
(Pauley et al., 2000). It utilises a grid system and maps to model spatial dynamics as a series
of box models (Ecopath, 2010).
Simile is a system dynamics modelling package with the same “stocks and flows” objectorientated
interface as programs such as STELLA, Powersim and Vensim (Ford, 2009). It was
originally derived from a model known as “Agroforrestry Modelling Environment”
(Muetzelfeldt & Taylor, 1997). The Simile package incorporates system dynamics modelling
with spatial modelling using grid reference (Muetzelfeldt et al, 2003)
The Simulink package has been used in Quebec, Canada, to determine the carrying capacity
of the Lagune de la grande-Entrée for suspended mussel culture (Grant et al,. 2007) the
model is a simple set up which allows the user to determine the effects of stocking densities
on the surrounding environment and enable an informed decision to be made as to whether
the lagoon can accept the growth of aquaculture in the area or not. The model consisted of
five mass balance equations which included mussel mass growth, the rate of change in
detrital matter concentration, rate of change in phytoplankton concentration, the rate of
change in zooplankton concentration and the rate of change in nitrate and ammonia
concentration (Grant et al., 2007). In this example the model suggested that the lagoon was
robust enough to take on more aquaculture after having the model validate and run through
a sensitivity analysis. (Grant et al, 2007)
1.2.10 Spreadsheet-based dynamic models
Spreadsheets can produce both simple and complex models using in build numerical
functions or programmed macros to manipulate data and perform multiple calculations
(Ford, 2009). Macro codes are often available as proprietary add-ons to the spreadsheet
software. Spreadsheets are designed to produce answers to complex calculations mainly for
a single time point but can be used repeated to reproduce a dynamic time set. However, if
this is done within the confines of the workbooks it is both time consuming and results in
the production of a large number of sheets with potentially hundreds of linked cells of data
making the whole model cumbersome and difficult to understand (Ford, 1999).
Spreadsheets do offer a modelling framework which is flexible and can be easily used to
build and test preliminary models, which can later be refined in modelling software
mentioned above or in native program code. It provides a clear and organised way of
showing calculations and their numerical results (Ford, 2009). The final spreadsheets can
also be linked to other dynamic modelling packages to supplement spatial modelling using a
grid search method for multiple parameters, such as Powersim (Vlachos et al., 2007)
MS Excel is probably the most commonly used spreadsheet for complex calculations and
modelling. It has the advantages of being almost universally used, so outputs are easily
transferable, and it uses VBA as its macro script, which is a standard and widely used and
understood. Spreadsheets are used for completely developed models or for formatting and
storage of data for entry into other modelling packages, for example Vensim and ArcGIS (Yeh
et al, 2006), and IDRISI (Perez et al, 2002). In the former case, the spreadsheet provided a
bridge allowing the conversion of outputs from ArcGIS into a form suitable for use in Vensim
for dynamic modelling of the Keelung watershed in Taiwan (Yeh et al, 2006). Spreadsheets
Page 11

have been used widely in aquaculture to growth and food requirement modelling, feed
formulation modelling and economic analysis of aquaculture business. Spreadsheets have
also been used in modelling environmental impact and carrying capacity both directly, by
modelling the distribution of particulate (Telfer, unpublished, reported in Ferreira et al 2008)
and chemical (SEPA, 2007) wastes, and indirectly by data input and pre-processing of data
for entry into IDRISI GIS system (Perez et al, 2002). Both models, designed for coastal cage
aquaculture incorporates bathymetry, hydrography, production data and settling velocities
of various feeds and resultant faecal materials, to simulate the dispersion of waste. An
example of a dispersion model for fish cages containing different species of fish in
Huangdung Bay, China, is given in Figure 1.1.
Northings (m)
Particulate Carbon input per year from the Demonstration Site (15m)
120
110
100
90
80
70
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Eastings (m)
Figure 1.1 CAPOT (Cage Aquaculture Particulate Output and Transport) model output
for coastal cages in Huangdung Bay, China. Four species of fish grown at
different production regimen are interspersed between 120 cages. (Telfer,
presented in Ferreira et al, 2008).
A summary of dynamic models mentioned and links to further information about these
models are given chronologically in Table 1.1. This shows progression from coded models to
object-oriented modelling frameworks seems to have been a consistent trend, where today
many models are developed using an interface such as Vensim, which can build and test
models and once finalised can compile for external use.
9000
7000
5000
3000
1000
750
100
50
10
5
0
gC/m2/y
Page 12

Table 1.1: Table showing modelling software developed, chronologically.
Model
Package
Date Type Website
Dynamo 1960 Stock and
flow
(Originally
developed
for business)
No longer in use
WASP 1983 Dynamic US Environment Protection Agency
compartment
modelling
system
http://www.epa.gov/athens/wwqtsc/html/wasp.html
Spreadsheets 1985 Cell based MS Excel TM
Stella 1985 Stock and Isee Systems
Flow
http://www.iseesystems.com
Ecopath with 1990 Static and NOAA
Ecosim (EwE) dynamic
modelling
with a spatial
aspect
http://www.ecopath.org
Vensim 1991 Stock and Ventana Systems Inc.
flow
http://www.vensim.com
Simile 2002 Stock and Simulistics
flow
http://www.simulistics.com
Powersim 2002 Stock and Powersim
flow business
simulation
http://www.powersim.com
1.3 Dynamic models developed to date specific to aquaculture
management and development
1.3.1 DEPOMOD series of waste dispersion models
Marine aquaculture waste deposition and dispersal has become recognised as an
increasingly important factor in developing sustainable aquaculture practices (Chamberlain
& Stucchi, 2007). A number of models have been developed which have been particularly
relevant to aquaculture practice in coastal areas. Of particular relevance in regulation of
coastal aquaculture are the DEPOMOD series of particulate waste dispersion models
(Cromey et al, 2002), which have “spawned” sibling models for other species and locations.
DEPOMOD
DEPOMOD is a waste dispersion model developed for Atlantic Salmon Aquaculture farms
(Cromey et al, 2002) from an earlier sewage waste dispersion model BenOss. DEPOMOD is
used extensively by the Scottish Environment Protection Agency (SEPA) for designating
license and discharge consent status of coastal marine fish farms (SEPA, 2007). According to
Cromey et al (2002), it has been validated for numerous scenarios including varying
Page 13

environmental conditions and farm sizes. It is considered a good compromising modelling
system to use, as it only requires data that is readily available for the systems scrutinised
making modelling a much simpler task (Cromey et al, 2002), while not having the reliability
and accuracy of 3D hydrodynamic models which require a far more complex and complete
data set. DEPOMOD has been modified for use with other species of fish, for example
Atlantic cod and Atlantic halibut, and for a number of localities throughout the world. At
present the DEPOMOD model is being tested in Canada in order to validate it for use with
finfish farms in British Columbia (Weise et al, 2009). DEPOMOD has also been programmed
to allow for adjustment to suit differing environments, however the resuspension velocity is
generalised at 9.5 cm s -1 , it can however be switched off (Chamberlain & Stucchi, 2007).
DEPOMOD is used at the pre-development stage for fish farm sites to estimate the
maximum production biomass which can be supported by local environmental services
(Cromey et al, 2002). DEPOMOD has also been use within the EU project ECASA (Ecosystem
Approach for Sustainable Aquaculture), which promoted sustainable development of
aquaculture using an Ecosystem approach (SAMS. 2004).
DEPOMOD has recently undergone testing to determine whether it can be developed for
shellfish biodeposition modelling in Quebec, Canada. (Weise et al, 2009) Due to the
difference in the two types of farming key changes were considered including farm
structure, waste values and sinking velocities (Weise et al, 2009). In comparison to finfish
DEPOMOD the shellfish version was developed to cover a significantly larger area due to the
farms covering a much larger area. The shellfish version has been found to be effective in
predicting dispersal and resuspension however is not yet capable of simulating advection of
resuspended material that is not farm-derived (Weise et al., 2009). The shellfish version uses
2-dimensional hydrographic information which allows considerable area coverage but
decreases the accuracy of the prediction. The lack of models for estimation of shellfish waste
distribution means that Shellfish DEPOMOD is likely to be used widely. DEPOMOD has also
been developed for Atlantic cod as COD –MOD (Cromey et al., 2009) and for sea bream in
the Mediterranean as MERAMOD (SAMS, 2004).
COD-MOD
The COD-MOD model is derived from DEPOMOD but for use for Atlantic cod grown in
marine cages (Pillay & Kutty, 2005). It uses 5 modules: grid generation, fish bioenergetics,
particle tracking, re-suspension and benthic faunal response (Cromey et al., 2009). Grid
generation is developed, by considering physical environmental factors such as depth, cage
and sampling station positions and hydrography. The model was re-parameterised from
DEPOMOD by analysing the digestibility of feed pellets suitable for cod and the volume and
nutrient content of faecal material deposited in a system and the potential effect on
macrofauna in the surrounding region (Cromey et al., 2009). Though COD-MOD seems to
work well when validated using field data, as little is still known regarding the environmental
aspects of cod grown in marine cages, considerable amounts of data are required for its
operation rather than the general derived assumptions used for salmonids.
MERAMOD
MERAMOD is also a development from the DEPOMOD model but has been adapted for sea
bream in the Mediterranean. It is also a waste dispersion model which takes into account
total solids and other particulates as well as accounting for macrobenthic indicator species,
and estimates for the effects by consumption of released wastes by indigenous fish species
attracted to the culture cages (SAMS. 2004). Unlike its predecessors MERAMOD is
Page 14

programmed in Borland Delphi 7 (SAMS, 2004). The model was used as part of the ECASA
project as it user friendly and has been validated. It is also easily adaptable for different
species and environments (SAMS, 2004).
1.3.2 FARM (Farm Aquaculture Resources Management)
The Farm Aquaculture Resources Management Model is based on shellfish cultivation, of
which there are relatively few models (Ferreira et al, 2007). It has been condensed from
more complex models and utilises parameters for which data is readily available. The
program was developed in order to allow an aquaculturist to determine the maximum
shellfish density that would provide a sustainable carrying capacity to the surrounding area
(Ferreira et al., 2007). It takes into account a variety of factors including farm size, food
supply and any environmental parameters to calculate the yield of the farm (FARM, 2010).
The model involves the integration of a number of models including; growth models for both
finfish and shellfish, production models, economic models and nutrient and eutrophication
models (Ferreira et al, 2009). It has been applied in various locations in the EU from Scotland
to Italy (Ferriera et al., 2009). FARM is not only an environmental monitoring tool but also a
tool, which can take into account the economics of the mariculture practice in question in
turn assisting in developing in better management for the farm concerned. (Ferreira et al,
2007) FARM is currently undergoing more developments to allow it to model for other forms
of culture such as seaweeds and fish cages (Ferreira et al, 2007).
1.3.3 APEM (Aquaculture Pond Ecosystem Management)
Many models have been developed for pond aquaculture, as this is a method responsible for
culture of many species (Boyd & Tucker, 1998). These models are often site specific and data
intensive rendering them too specific to be used on other systems (Losordo & Piedrahita,
1991). The Aquaculture Pond Ecosystem Model was developed in order to provide a simpler
more general, user friendly, method for modelling these systems, enabling application to a
number of ponds types (Culberson and Piedrahita, 1996). The model looks at the biological,
physical and chemical aspects of the pond but simplifies the processing. In doing this
Culberson and Piedrahita (1996) have produced a model which covers a larger range of
ponds and allows for ponds with limited data sets to be simulated. The model produced
accounts for Temperature and thermal energy dynamics, simulating the vertical heat
transfer within the pond as well as the Dissolved Oxygen dynamics, simulating plankton
photorespiration and sediment and water column respiration. The model was developed in
the STELLA framework and consisted of stocks and flows to simulate an outcome.
To determine the range of use of the model, simulations were carried out for three areas,
using data from the Pond Dynamics/Aquaculture Collaborative Research Support Program
database including; Honduras, Rwanda and Thailand (Culberson and Piedrahita, 1996). The
ponds covered a range of areas; coastal, inland and elevated from sea to mountains in
Africa.
The model was found to simulate erroneous results for the temperature model in Honduras,
resulting in underestimations, however consistently the errors showed only a 1 o C difference
for the entire model run. A similar issue arose when the model was used for pond culture in
Rwanda; this time the difference was consistently 1.5 o C different.
Page 15

1.3.4 IAAS (Integrated Aquaculture/Agriculture System)
Aquaculture systems in SE Asia often integrate aquaculture and agriculture, resulting in a
shared movement of material between the 2 systems (Pullin et al, 1993). Although
integration of these systems is well known, there are few models, which have been
developed specifically for this purpose.
The IAAS (Integrated Aquaculture/Agriculture System) model is a relatively new model
developed by Jamu and Piedrahita (2002a). Its purpose is to simulate the movement of
organic matter between the two systems. The model simulates fish and crop production,
organic matter, phytoplankton, dissolved oxygen and the nitrogen dynamics in the system
(Jamu & Piedrahita, 2002a). The IAAS model is made of 2 modules - Aquaculture and
Agriculture - which are then divided into four sub modules. The Aquaculture module is split
into fish pond water column and fish pond sediment, whereas the Agriculture module is
separated into terrestrial soil and terrestrial crop. Within these sub-modules there are 19
state variables each with a separate mass balance simulations (Jamu & Piedrahita, 2002a).
Model simulations use time steps of 0.125 days, enabling simulation of multiple production
cycles and potentially over longer periods of time (Jamu & Piedrahita, 2002b).
The model was developed through the software package STELLA TM (Jamu & Piedrahita,
2002b) and calibrated using data from the PD/A CRSP for a site in Rwanda cultivating tilapia
in a fertilised pond (Jamu & Piedrahita, 2002a). Having undergone validation it was found
that the model performed well for the fish and crop biomass production, organic matter and
nitrogen. It did however demonstrate errors when simulating phytoplankton production,
probably through not considering zooplankton grazing. If this was added the model would
be considerably more robust (Jamu & Piedrahita, 2002b).
1.3.5 AWATS (Aquaculture Waste Transport Simulator)
This is a waste transport model developed using a series of existing validated models. The
overall model simulates the transport of wastes from finfish net pen aquaculture systems
and is intended to be used for regulatory purposes (Dudley et al, 2000). The model package
includes the finite difference model, DUCHESS, integrated with the software SMS (Surfacewater
Modelling Systems) to provide a user friendly interface with a graphical output. The
package also includes the transport model TRANS which simulates the advection and resuspension
of particulate wastes (Dudley et al, 2000). The final component of the AWATS
package is the Automated Coastal Engineering System (ACES) to model wave velocity which
effects waste settling rates.
The AWATS modelling package has been tested in the USA in 3 different Bays (Machias Bay,
Blue Hill Bay and Cutler Harbor). The results from the simulations proved to be “relatively”
robust, showing its strength for environmental regulatory purposes although more data are
required to calibrate the model (Dudley et al, 2000).
1.3.6 MMFA (Mathematical Material Flow Analysis)
The MMFA model was originally developed for economic modelling in the 1930s (Hall & Day
Jr, 1977). In recent years the model has been developed for use in pollution monitoring and
Page 16

now follows the material flows through a system. It is now a stock and flow model as it is
built in STELLA and can provide a visual aid for pollution flows (Schaffner et al, 2009).
The MMFA has been used in conjunction with WASP in Thailand on a catchment known as
the Thachin River Basin (Schaffner et al., 2007), which is perceived as the most polluted river
in Thailand (Simachaya, 2003). The MMFA was used to quantify the nutrient loading and
consequent carrying capacity of the river, by taking into account both agriculture and
aquaculture activities simulating the loading from each to determine the largest polluter in
terms of nitrogen and phosphorus (Schaffner et al., 2009). The model confirmed that
aquaculture was the most significant nutrient polluter in the Thachin River (Schaffner et al,
2009). The benefit of this mode was simulations could be carried out using whatever data
was available (Schaffner et al., 2007). This is a clear advantage as there is often a paucity of
data in such systems. It is also advantageous as the model can be easily adapted for use in
other similar aquatic systems and where data is limited.
It is unadvisable to use the MMFA model on its own for any decision on catchment
management. It is desirable to use it in combination or to complement a system dynamics
model to provide a wider understanding of the workings and carrying capacity for any
aquatic system (Schaffner et al., 2007).
1.3.7 SWAT (Soil and Water Assessment Tool)
SWAT is a long-term assessment tool for simulations of watershed dynamics (Ullrich & Volk,
2009). The basic design of SWAT is in three layers: the sub-basin, reservoir routing and
channel routing (Spruill et al., 2000). SWAT was originally developed for use with nutrient
loading of water sources from land based activities (Huang et al, 2009), and has been used
for estimating nutrient loads to local watersheds in relation to salmon farming.
SWAT2000 was used in the Cannonsville reservoir in New York to simulate nutrient loading
of agricultural practices on the surrounding land, taking into consideration sediment flows
(Tolson & Shoemaker, 2004). Although the model was effective in showing some useful
findings, it was not effective in showing phosphorus loading as a validation method. There
were significant errors between empirical data and simulations (Tolson and Shoemaker,
2004). However, the model demonstrated that agricultural land used for corn was the main
contributor to phosphorus loading to the reservoir, even though the land used for this
purpose only took up approximately 1.2% of the survey area (Tolson & Shoemaker, 2004).
SWAT is a popular model as it has been used in a number of further studies:
In Kentucky to determine monthly runoff in karst areas and was found to have
effective analysis and simulation results for the system in this respect (Spruill et al.,
2000)
In the Vantaanjoki basin in Finland (Barlund & Kirkkala, 2008) to determine
management efficiency in simulating the nitrogen and phosphorus flow in
catchment and water bodies under the Water Framework Directive (WFD) (Barlund
& Kirkkala, 2008). In this case the model did not allow for nitrogen input from
forestry, which is significant and thus limited the model’s use for the present.
However, this can be adapted and used more effectively for management under the
WFD in future (Barlund et al, 2007).
Page 17

1.3.8 EcoWin2000
The development of this model began in 1991 and has continued since (Ferreira, 1995). It is
used for modelling of aquatic systems through an object-oriented approach in which a
‘series of self-contained objects’ are coordinated by a shell module which communicates and
transfers information between the individual objects (EcoWin2000, 2010). EcoWin2000 is
mostly used for analysis of large coastal systems by breaking the extended areas into boxes
which are dealt with individually (Ferreira, 1995). The outputs from each box can be used by
other modelling packages to provide information at a considerably smaller scale, for
example FARM which models farm scale events (see Section 1.3.2).
The EcoWin2000 package has been used in China to model shellfish polyculture and
determine the sustainability of current culture practices (Nunes et al., 2003). The model
known as GAMBEY (General Aquaculture Model for Bivalve Equilibrium Yield) was built and
tested in the Powersim package to allow the different steps to be shown, but the final model
was created and run in the EcoWin200 package (Nunes et al., 2003). The model is made up
of three sub modules; the human module, which simulates effects of human activities; the
shellfish module, focused on individual growth and population dynamics and the ecosystem
module, which contains 4 sub models including detritus, kelp, phytoplankton and nutrients
(Nunes et al, 2003). These modules are coupled to provide a fully functioning model of all of
the interactions taking place during the culture of shellfish. The model has been successfully
validated for Sanggou bay in China (Nunes et al, 2003).
The EcoWin package has also been used to determine phytoplankton productivity in marine
and estuarine systems (Duarte & Ferreira., 1997; Macedo & Duarte, 2006). This involved a
model developed for both the static and dynamic interactions between the phytoplankton
and the mixing layers (Macedo & Duarte, 2006) constructed using EcoWin software as it
enables the user to modify the coding for the model, unlike other packages such as STELLA
(Duarte & Ferreira, 1997) This model was simplified to allow for small changes in the
parameters to be visualised, thus nutrient limitation, phytoplankton sinking and zooplankton
grazing were all omitted from the model. Larger scale models for marine and estuarine
environments tend not to take phytoplankton productivity into consideration as they focus
on the larger processes taking place (Duarte & Ferreira, 1997). This model has also been
used as a component in a larger model due to its simplicity (Macedo & Duarte, 2006) thus
filling a previously neglected niche.
1.3.9 MOM (Modelling- Ongrowing fish farm- Monitoring)
MOM (Modelling- Ongrowing fish farm- Monitoring) is a Norwegian based model to simulate
the carrying capacity of a site for Aquaculture (SAMS, 2004). It is a robust model that has
been used for over 10 years by the Norwegian Environment regulatory body and was
originally developed for Salmon, which is a large market in the country (Hansen et al., 2001)
The model is made up of 4 sub models which all link to form the final model for aquaculture
site impact determination (Hansen et al., 2001). The 4 sub models include a fish model,
which models the uptake and retention of feed particles as well as excretion in relation to
temperature and fish mass. The second is a particle dispersion model which simulates the
“dispersion and sedimentation rates of feed and faecal pellets” The third model, the
sediment model, simulates the effect of differing volumes of organic matter deposited on
Page 18

1.3.10 KK3D
the sea bed and the effect it has upon the benthic fauna. The fourth and final model is the
Water Quality model which is split into two parts; 1) water quality in the Net pen cages
which determines the concentrations of O2 and ammonium produced, 2) water quality
discharged to surrounding bodies which is modelled based on the change in turbidity and
oxygen levels due to eutrophication (Stigebrandt et al., 2004).
The MOM model was implemented as part of the ECASA project as well as SPEAR as it is
robust and simulates most of the factors that can be thought of for net pen cage culture
(Stigebrandt et al., 2004). This model is data intensive and can take a long time to develop
for other culture methods and species if data is not available. The model has now been
developed to include shellfish farming (Zhang et al., 2009b).
KK3D is a waste deposition model for use in mariculture, which can be applied to both local
scale and water body scale (SAMS. 2004) The model is object-oriented and developed in
C++. It uses input data for feed composition, geometry of net-pens, variation in emissions
over a given time scale, faecal pellet solubility, current velocity, settling rate of faecal pellets
and realistic bathymetry data (Jusup et al, 2007) Unlike the deposition model DEPOMOD,
KK3D does not have a resuspension model connected to it (Jusup et al., 2007) It does
however contain a particulate organic matter model which is another particle tracking
module, which determines whether the particle will remain motionless on the seabed or
move again (Jusup et al., 2009). Data required in order to run the model is relatively easily
obtainable at minimal cost as it requires only daily datasets (Jusup et al., 2007).
Table 1.2: Summary description of models and model systems relevant to aquaculture
Model Type Package Reference
DEPOMOD Waste Dispersion Visual Basic
Cromey et al., 2002
(Salmon)
(Computer Language)
COD-MOD Waste Dispersion Visual Basic
Cromey et al., 2002
(Cod)
(Computer Language)
MERAMOD Waste Dispersion
(Med)
Borland Delphi 7 SAMS. 2004
FARM Resource management
for shellfish
STELLA Ferreira et al., 2007
APEM Environmental ecosystem STELLA Culberson &
dynamics
Piedrahita, 1996
IAAS Evironmental ecosystem STELLA Jamu & Piedrahita,
dynamics
2002 (a&b)
AWATS Waste transport Various Dudley et al., 2000
MMFA Material flow Spreadsheet Schaffner et al.,
2009
SWAT Water quality/
Visual Basic
Spruill et al., 2000
Groundwater modelling (Computer Language)
EcoWin2000 Ecosystem model Ecowin2000 software EcoWin website,
2010
MOM Environmental impact Hansen et al., 2001
Page 19

model
KK3D deposition C++ SAMS, 2004 (ECASA
website)
The model was implemented in the ECASA (Ecosystem Approach for Sustainable
Aquaculture) project “for the impact of finfish development on the state of the
environment.” (SAMS. 2004). The KK3D model has been used in a variety of Environmental
Impact Assessments in the Adriatic Sea off the coast of Italy (Brigolin et al., 2010) with
relatively accurate results, although it would benefit greatly from the incorporation of a
resuspension model.
1.4 Models advantageous to SEAT
The SEAT (Sustainable Ethical Aquaculture Trade) project is an EU project aimed at
developing sustainable Aquaculture for global trade. This is an important project as the EU
imports almost half of their aquatic food products, the vast majority from south East Asia.
The project aims to produce an aquatic food index to allow consumers to make better
informed decisions when choosing sustainable seafood. The project is aimed at the four
biggest imports pangasius, tilapia, shrimp and prawn. One of the outcomes for the project is
the production of environmental models for the fate and distribution of aquaculture wastes.
Some of the models in this report would be beneficial to the project and could provide a
basis for the work to be carried out. The Aquaculture Pond Ecosystem Management model
has potential to be of considerable use for the project as all of the species investigated are
cultivated in ponds in South East Asia. The model is considered a more general model than
most and would therefore potentially be suited to all of the methods of culture without
much adaptation.
The IAAS (Integrated Aquaculture/Agriculture Systems) model also has potential for use. It is
not unusual in South East Asia to have aquaculture farms in amongst agricultural land (Pillay
& Kutty, 2005). It is known that many aquaculture farms’ discharge water into channels
which run into irrigation systems for crops, thus not only irrigating crops but also fertilising
with any excess nutrients produced from aquaculture practices. The model has been
effective in its simulations but would need the addition of a zooplankton model to
accurately predict the phytoplankton productivity.
The MOM model, although developed for Norwegian aquaculture, has the potential for
development for South East Asia. It accounts for the uptake and retention of feed particles
as well as excretion in relation to temperature and fish mass, which for the species
investigated is most likely available in literature now. It also models particle dispersion which
would enable a simulation for the volume of nutrients added to the pond through the fish
itself. It also contains a water quality model to simulate oxygen and nitrogenous substances,
which s important as nitrogen is a limiting factor in water quality.
1.5 Conclusions
Any of the visually dynamic software packages would be of use to create an environmental
model. However the one chosen for the SEAT project is Powersim. The Powersim software
Page 20

suite has been developed for use in larger businesses and has been used with confidence for
decision making. The other software suites do not have any major disadvantages in relation
to the Powersim suite and even spreadsheets may be used for linking model data to a GIS
program. Using an object-oriented dynamic software package has its advantages such as the
ability to visually display the steps undertaken to produce the final model as well as the
ability to display the results graphically. The majority of the visually dynamic software has a
user friendly interface, however it is notable that the Powersim suite has a large array of
tools to assist in model making.
Three models which have potential to be used or built upon for the production of the SEAT
environmental models include the APEM, IAAS and MOM. These have proven to be effective
in their application and have been developed in a general capacity allowing application to a
multitude of culture methods, specifically ponds. Although the MOM model was originally
developed for cage aquaculture, models are being adapted all of the time and MOM is no
exception.
Page 21

SECTION 2 – REVIEW OF GIS MODELS AND APPLICATIONS IN AQUACULTURE
DEVELOPMENT
2.1 Introduction
Work Package 4 in the SEAT project is assessing the environmental aspect of farming the
four case study species in the selected study areas. Environmental models will be developed
to predict the impact and fate of wastes associated with aquaculture. Within the case study
areas/regions there may be multiple farms releasing wastes into the environment and the
combined impact of these farms may be greater than the individual impact. Furthermore
there may also be other anthropogenic sources of wastes/nutrients/chemicals into the
environment such as agriculture that could also interact with aquaculture output. The
advantage of using GIS to develop models is that the impact from many farms can be
analysed simultaneously on a regional scale as well as taking into account inputs from other
sources, therefore the results are truly representative of the activities taking place in the
area and the subsequent environmental conditions. Farms that share a common water body
or source could benefit from co-ordinated management as the effects of multiple farms
could be cumulative and lead to problems at a larger scale than just farm level such as
eutrophication (Aguilar-Manjarrez, Kapetsky & Soto, 2010). Modelling the catchment area
using GIS allows the potential additive effects of clusters of farms to be assessed and
mitigation procedures put in place if necessary to try to ensure minimal environmental
impact and encourage sustainable aquaculture.
2.2 Geographical Information Systems (GIS)
Due to the variety of sectors and diversity of tasks that can be performed using GIS there are
many definitions of GIS available depending on the user’s background and purpose
(Bernhardsen, 1999). One commonly used definition of GIS is from Duecker & Kjerne (1989),
recently this definition has been slightly modified by Brimicombe (2010) (modifications are
included in brackets) to define GIS as “A system of hardware, software, data, people,
organisations and institutional arrangements for collecting, storing, analysing, [visualising]
and disseminating [spatial] information about areas of the earth”. It is important to
remember it is not just the end product or the software but that the GIS encompasses the
processes that are carried out, the information and resources that are used and the people
that are involved.
Longley et al (2005) describes the anatomy of a GIS as having six components; people,
hardware, data, procedures, network and the software. When describing a GIS the people
involved are often ignored yet they are a fundamental component as they plan and organise
the GIS in addition to making decisions on the operations and output (Heywood, Cornelius &
Carver, 2002). The hardware refers to the computer system that is used to run the GIS; this
can range from a small laptop to a large desk-based supercomputer. The data is the
information that is input into the GIS and the procedures are the operations carried out on
the data. The network (internet and intranet) is the method to communicate and share
information (Longley et al, 2005); this could involve downloading satellite images to input
into a GIS, communicating with stakeholders and sharing the final outcomes and models.
Finally, the sixth component described by Longley et al (2005) is the software; there are
many software packages available ranging from basic entry level to professional high-end
products.
Page 22

GIS technology has been available since the 1960’s, but it was not until the 1980’s when
computers powerful enough to use the applications became affordable that the commercial
GIS industry really began to grow (Heywood, Cornelius and Carver, 2002; Longley et al,
2005). Today, there are many GIS products available from very simple tools that can be
found for no cost on the internet to the expensive but highly specialised commercial
software suites with multiple extensions and tools. Some of the popular commercial GIS
software packages are as follows: -
ArcGIS ® [ESRI, California, USA] – ESRI (founded in 1969 as Environmental Systems
Research Institute Inc) is the market leader in terms of GIS software and they offer a
wide range of GIS products in their ArcGIS range including desktop, server and mobile
GIS in addition to specialised extensions. The latest version, ArcGIS 10 was released in
the summer of 2010 in addition to a newly developed ArcGIS.com online gateway which
users can access via ArcGIS desktop, browser based applications and mobile devices
(including a specially developed iPhone application). ArcGIS.com can be used to browse,
share, organise and use resources such as maps, applications and developer tools that
have been published by the ESRI community (ESRI, n.d.; Longley et al, 2005).
AutoCAD ® Map 3D [Autodesk, California, USA] - Autodesk are famous for their AutoCAD
products; primarily computer-aided design (CAD) software; however they also have a
range of geospatial software that integrates both CAD and GIS; AutoCAD Map 3D which
is mainly used by the engineering and utilities industries. As Map 3D is built using
AutoCAD technology, organisations with workforces already trained in CAD can easily
use that experience to work with geospatial data using tools that they are already
familiar with (Autodesk Inc, 2010).
ERDAS Imagine [ERDAS, Georgia, USA] – ERDAS is the world’s leading remote sensing
application developer. ERDAS Imagine incorporates remote sensing, photogammetry
and radar into one product and can be used to prepare imagery for further analysis
within ERDAS Imagine or more specialist GIS software. ERDAS have also worked with
ESRI to produce extensions for ArcGIS allowing easy integration of spatial imagery from
ERDAS to ArcGIS. In addition to the Imagine software group, ERDAS have also developed
a suite of photogrammetric production tools within LPS, a product that allows the
transformation of raw imagery into data layers required for digital mapping, GIS analysis
and 3D visualisation (ERDAS Inc. 2010).
IDRISI [Clark Labs, Massachusetts, USA] – IDRISI was first developed by Clark Labs (Clark
University) in the late 1980’s and is a popular GIS product among academic institutions,
both for teaching and for research as it is affordable and easy to use but can also
undertake highly complex spatial analysis if required. The most recent version, IDRISI
Taiga, was released in February 2009 and along with a wide range of geospatial tools it
includes two specially developed modules; an Earth Trends Modeler that allows the
analysis of environmental trends over time, in addition to the Land Change Modeler
from previous versions which can be used for land cover change analysis (Clark Labs,
2009).
Geomedia ® [Intergraph, Alabama, USA] – Intergraph have a range of GIS products under
their Geomedia range including specialised products for transportation, communication,
water/wastewater projects, public safety and security and also emergency response
management. Intergraph also have contracts to supply geospatial products and services
Page 23

to many US federal and government agencies, particularly in the area of defence and
intelligence (Intergraph Corporation, 2010).
Global Mapper [Global Mapper Software LLC, Colorado, USA] – Global Mapper is a
commercial GIS software based on software produced by the United States Geological
Survey (USGS). A major feature of Global Mapper is the active online community, where
users can communicate with each other and Global Mapper developers via a forum.
Users can get help and support, suggest useful features for newer versions and try Beta
versions of future software releases; version 12 of Global Mapper is due for release in
late summer 2010. The USGS also provides a free version of Global Mapper with limited
features as dlgv32 Pro (Global Mapper LLC, 2009; USGS, 2010).
Manifold ® [Manifold, Nevada, USA] – Manifold is a GIS software package that is
increasing in popularity due to its low cost compared to many other commercially
available packages. The most recent edition, Manifold 8.0, was released in summer 2007
and is available in either 32-bit or 64-bit versions. All Manifold editions (apart from the
most basic, Personal edition) include an Internet Map Server through which users can
publish GIS projects online allowing visitors to interact with the data. The Internet Map
Server enables users to communicate with stakeholders and share model outcomes and
projects without needing to be in the same room (Manifold, 2009).
MapInfo Professional ® [Pitney Bowes Business Insight, New York, USA] – MapInfo
Professional ® is used in a wide range of industries including insurance,
telecommunications and the public sector. The latest version, MapInfo Professional
10.5, was launched in June 2010 and is designed to integrate easily into existing IT
infrastructure making it a useful product for many businesses looking to expand their
scope to include geospatial analysis (Pitney Bowes Software Inc, 2010).
Maptitude [Caliper Corporation, Massachusetts, USA] – Caliper Corporation produces
both GIS (Maptitude) and transportation software (TransCAD). Maptitude is aimed more
at business customers rather than personal and academic users. The most recent version
is Maptitude 5 and it is available in numerous international versions that include
geographical datasets and maps specific to the selected country. Having data already
available is an attractive prospect for novice users looking to learn how to use GIS with
no hassle (Caliper, 2010).
Smallworld [GE Energy, Atlanta, USA] – Smallworld software focuses on utilities
(particularly gas and electric) and was developed in the late 1980’s by people with
previous experience in CAD. In 2000 Smallworld was sold to GE Energy (then GE Power
Systems) a division of General Electric. The most recent release, Smallworld 4 has been
designed for remote access so that it can be easily and reliably used in the field (GE
Energy, 2010; Longley et al, 2005).
It is important to state that the GIS software industry is evolving faster than ever and with
products updating and changing frequently it is difficult to keep up to date with all of the
current versions and features provided by individual companies and products. Furthermore,
free, open-source GIS software products available in the public domain are increasing in
popularity. These GIS products used to be very simple, lacked user support and were poorly
constructed, however these days there are some notable products available that have many
features and a strong online community to help with problems (Longley et al, 2005).
Geographic Resources Analysis Support System (known as GRASS GIS) [GRASS Development
Page 24

Team] is one of the more popular open source GIS. It was originally developed in the early
1980’s by a branch of the US Army Corp of Engineers for environmental planning and land
management analysis but it is now used by many people across the World as it is free to use
and easy to access (GRASS Development Team, 2010).
2.3 GIS Aquaculture projects
GIS has become increasingly more important to aquaculture since its introduction in the late
1980’s and projects using GIS and remote sensing have become more diverse in the species
and areas studied in addition to the overall purpose and impact of the research. Site
selection studies using GIS were among the first applications of GIS in the aquaculture sector
with Meaden (1987) looking at potential sites for trout farms in Britain and Kapetsky, Hill &
Worthy (1988) using GIS to identify suitable locations for catfish (Ictalurus punctatus) farms
in an area of Louisiana, USA. GIS allows the identification of many sites depending on
selected criteria and as a result it is a highly suitable tool in aquaculture site selection and
planning projects (Valavanis, 2002).
The amount of geospatial data available has increased in recent years and along with
improvements in technological capabilities the potential for advanced and complex GIS
projects has also grown. As the use of GIS in aquaculture has increased so has the amount of
research published and some key studies have been reported previously by other authors
(Aguilar-Manjarrez, 1996; Nath et al, 2000; Kapetsky & Aguilar-Manjarrez, 2007; Aguilar-
Manjarrez, Kapetsky & Soto, 2010). Most of the studies using GIS in aquaculture research
have been using GIS as a tool for site selection or land suitability, however there have been a
few studies using GIS as an environmental management tool assessing waste dispersion and
environmental impact.
There are many benefits of using GIS for environmental management, including: the ability
to handle large amounts of data, better efficiency than analysing the data manually, use of
data in different formats (graphic, textual, numerical etc), and in addition to simulating
current and future conditions GIS can also model, test and compare alternative scenarios
including worst-case (Corbin & Young, 1997). Shih, Chou & Chiau (2009) used GIS to assess
benthic environmental impact through two environmental indicators (sediment redox
potential and sediment sulphide) at a cobia (Rachycentron canadum) cage site in the Bay of
Penghu, Taiwan. By modelling the two indicators using GIS the environmental impact can be
visualised in the geographic setting. The advantage of using GIS to measure benthic
environmental impact in such a way is that it can help both producers and authorities to
visualise, plan and manage environmental capacity, stock rotation and cage location (Shih,
Chou & Chiau, 2009). Output models can be simplified versions of reality and when shown in
a spatial setting it can be easier for stakeholders with little or no scientific background to
understand the scenario that the model represents rather than a series of equations or
graphs of the same results.
In addition to using environmental indicators to assess environmental impact, GIS and
remote sensing tools can be used to analyse impact through land use classification.
Guimarães et al (2010) used remote sensing and GIS to evaluate the impact of aquaculture
on mangrove areas in northeast Brazil. The impact of aquaculture on mangroves has been
documented in the past and it is estimated that in the past few decades, development of
culture ponds for shrimp and fish accounts for the destruction of around 20-50% of
mangroves worldwide (Primavera, 1997). Guimarães et al (2010) produced thematic maps
Page 25

for selected years spanning a period of over 30 years (1973 – 2005) quantifying the areas
that mangrove vegetation and aquaculture nurseries occupied. Often aquaculture is blamed
solely for the destruction of mangroves, however although aquaculture does contribute to
the problem, other activities such as agriculture, logging, urban expansion, industrial
development and tourism are also responsible for the problem (Guimarães et al, 2010;
Neiland et al, 2001; Pillay 2004). The advantage of using GIS for such an activity is that it can
take into account other land use factors and Guimarães et al (2010) found that shrimp
farming was responsible for just 9.6% reduction of the total area of mangrove vegetation,
therefore there must have been other anthropogenic activities that were also responsible
for the destruction of the mangrove vegetation and aquaculture was not the only activity
responsible. GIS can use data from maps, satellite images and photographs to quickly and
efficiently analyse land use and also identify potential land use conflicts; saving time and
money that a manual land survey of the whole area would take.
Although GIS is an extremely useful tool it is also important to verify information. Ferriera et
al (2008) used GIS as part of the SPEAR (Sustainable Options for People, Catchment and
Aquatic Resources) project looking at two coastal systems in China; Sanggou Bay in a rural
area in the north and Huangdun Bay in an industrialised area in the south. Satellite imagery
was used to identify areas of aquaculture (pond structures, fish cages etc) in both Sanggou
Bay and Huangdun Bay and then local partners were used to ground truth, verify and update
the original identification of aquaculture systems (Ferriera et al, 2008). This verification was
essential as it identified some coastal ponds that were not used for aquaculture and this was
taken into account when using the information in further analysis. This highlights the
importance of verifying information as if these ponds had been identified as being active
rather than inactive then the results of the project would not have been representative.
There are many potential areas in which GIS can be applied to aquaculture both at present
and in the future. Table 2.1 contains selected examples of GIS studies published in the last 5
years (2005-2010).
The following section illustrates five selected examples of aquaculture projects that have
used GIS and remote sensing tools. Both Table 2.1 and examples 1 to 5 illustrate the
diversity of GIS projects in aquaculture. As technology increases and digital spatial data
becomes more easily accessible then the use of GIS and remote sensing tools in aquaculture
will continue to grow and the diversity of projects will be even greater than it is now.
Page 26

Table 2.1: Selected articles regarding aquaculture and GIS between 2005 and 2010
Year Author Title Species Type of Study Location
2010 Guimaraes et
al
Impact of aquaculture on mangrove areas in the northern
Pernambuco Coast (Brazil) using remote sensing and
geographic information system
Shrimp Environmental
Impact
2009 Hossain et al Integration of GIS and multicriteria decision analysis for
urban aquaculture development in Bangladesh
Carp Site selection Bangladesh
2009 Shih, Chou & Geographic information system applied to measuring Cobia (Rachycentron Environmental Taiwan
Chiau benthic environmental impact with chemical measures on
mariculture at Penghu Islet in Taiwan
canadum)
Impact
2008 Longdill, An integrated GIS approach for sustainable aquaculture Greenshell mussell (Perna Site Selection New
Healy & Black management area site selection
canaliculus)
Zealand
2008 Radiarta, GIS-based multi-criteria evaluation models for identifying Japanese scallop
Site Selection Japan
Saitoh & suitable sites for Japanese scallop (Mizuhopecten
(Mizuhopecten yessoensis)
Miyazono yessoensis) aquaculture in Funka Bay, southwestern
Hokkaido, Japan
2007 Hossain et al Multi-criteria evaluation approach to GIS-based land- Nile tilapia (Oreochromis Land
Bangladesh
suitability classification for tilapia farming in Bangladesh niloticus)
suitability
2006 Corner et al A fully integrated GIS-based model of particulate waste Atlantic salmon (Salmo Waste Scotland
distribution from marine fish-cage sites
salar)
dispersion
2005 Buitrago et al A single-use site selection technique, using GIS, for Mangrove Oyster
Site Selection Venezuela
aquaculture planning: choosing locations for mangrove
oyster raft culture in Margarita Island, Venezuela
(Crassostrea rhizophorae)
2005 Giap, Yi & GIS for land evaluation for shrimp farming in Haiphong of Shrimp Land
Vietnam
Yakupitiyage Vietnam
suitability
2005 Perez, Telfer Geographical information systems-based models for Seabream (Sparus aurata) Site Selection Tenerife
& Ross offshore floating marine fish cage aquaculture site Sea bass (Dicentrarchus
selection in Tenerife, Canary Islands
labrax)
2005 Salam, Khatun Carp farming potential in Barhatta Upazilla, Bangladesh: a Carp Land
Bangladesh
& Ali
GIS methodological perspective
suitability
Page 27
Brazil

2.3.1 Example 1: Shellfish scenarios, Brazil
Models based on environment and infrastructure data were used to determine
opportunities for culture of molluscs, principally mussel and oyster, in Baia de Sepetiba,
Brazil (Scott, Cansado, and Ross, 1998). In this study, extensive use was made of specialist
macros, particularly for generation of potential wave heights from wind and fetch data. This
approach is essential in all such siting studies where system structures will be placed in the
water. The outcome images (Fig. 2.1.) revealed that much of the bay could be exploited for
mollusc culture, but that different areas are better for different culture systems. These
results were partly verifiable from data on artesanal collection of molluscs in recent times.
This is, of course, no real substitute for field verification of any model, but careful use of
inventory data from fisheries departments and similar agencies enables at least partial
verification of models during the creation process.
Figure. 2.1: Shellfish culture scenarios in Baía de Sepetiba, Brazil showing opportunities for
mussel culture in the outer bay (A) and onshore shrimp culture (B). Green = Most suitable,
Yellow = Suitable, Blue = Marginal, Red = Unsuitable, Mauve = Urban areas. (after Scott,
2004).
Page 28

2.3.2. Example 2: Sea urchin fishery management, Chile
Del Campo Barquin (2002) used GIS coupled with external simulation models that were
produced using Powersim dynamic modelling software to depict spatially the current state
of natural stocks of the red sea urchin Loxechinus albus and other important commercial
benthic resources in a small area of the central coast of Chile. Coupled with data on the
major physical characteristics of the coastal environment, such as bathymetry and seabed
type, models were developed which optimised sites for restocking of hatchery- reared seed
of red sea urchin and which accorded with the aims of the local area management plan (Fig.
2.2). By integrating the biological and socio economic outcomes in an external simulation
model, the ultimate objective of the GIS- based model for restocking and exploitation of the
red sea urchin fishery was achieved.
Figure. 2.2: Suitability of areas for fishery restocking in Quintay bay, Chile. The image shows
environmentally suitable areas for restocking cross-tabulated with existing populations of
keyhole limpet Fisurella sp, the red sea urchin Loxechinus alba and the Chilean abalone
Concholepas concholepas. Dark green = Highly suitable areas for L. Alba restocking. (after del
Campo Barquin, 2002).
Page 29

2.3.3. Example 3 : Aquaculture and Tourism
GIS has been useful for guiding decisions where aquaculture may be developed alongside an
important tourism industry (Salam et al, 2000). Pérez, et al, (2003, 2005) used GIS and
related technologies to build a spatial database using those criteria which were considered
to have any influence in integrating marine fish-cage culture within the tourism industry in
Tenerife. Criteria were grouped into three sub-models (distance to beaches, nautical sports
and viewsheds), which were combined to generate a final output showing the most suitable
areas for cage culture development in coexistence with this industry (Fig. 2.3). Most areas of
the coastline of Tenerife were identified as being suitable (56%) and very suitable (46%),
suggesting that marine cage aquaculture could be developed in the island in coexistence
with the well established tourism industry.
Figure. 2.3: Suitability of areas for aquaculture development in Tenerife, Canary Islands,
Spain. The model combines environmental aspects of sites, the physical requirements of the
ongrowing systems and is optimised using Viewshed analysis to minimise impacts on
tourism. The highest scores, in blue, have the highest potential. (after Pérez et al, 2005).
Page 30

2.3.4. Example 4 : Mangroves and aquaculture, Bangladesh
Salam, (2000), Salam and Ross (1999,2000) and Salam et al (2003) investigated management
scenarios for coastal development in the Khulna area of Southwestern Bangladesh which
includes the Sunderbans mangrove forest, a coastal belt, part of the Bay of Bengal and
several major rivers and their tributaries. They developed a series of GIS models in order to
identify and prioritise the most suitable areas for brackish water shrimp and crab farming.
Using qualitative and quantitative output from the models, Multi-Objective Land Allocation
was used, based on gross production, economic output and employment potential, to
compare and trade-off relative benefits from such developments and to consider their
economic and social impact. Comparisons were made of brackish water shrimp and crab
culture with moderately saline tolerant tilapia and prawn culture, fresh water carp culture
and traditional rice production systems. Shrimp was identified as the most capital intensive
and risky production system. The trade-offs between tilapia/carp and shrimp/crab cultures
are shown in Fig.2.4.
Figure. 2.4: Resolving the trade-off between the uses of land in south west Bangladesh for
agriculture, Penaeus culture and crab culture using Multi-Objective Land Allocation tools.
(after Salam, 2000).
Page 31

2.3.5. Example 5: Tilapia cage location, El Infiernillo, Mexico
Ross et al (in press) used GIS to develop models indicating suitable locations for tilapia cages
in a large reservoir in Mexico. The Presa Adolfo Mateos Lopez (El Infiernillo) was once the
main Mexican tilapia fishery, however, in recent years both the quality and catch of stocks in
the reservoir has declined. Research is currently underway to improve the quality and
quantity of the stocks and the preferred approach is to improve stocks through rigorous
hatchery control and management with on-growing in cage systems within the reservoir,
Good site selection is essential as the improved cages should be located in areas with
optimal conditions. Ross et al (in press) incorporated aspects of topography, climate,
hydrography, water quality and quantity, land use, infrastructure and socioeconomics to
develop spatial models indicating suitable locations for cage farms within the reservoir. The
reservoir has a total area of 312km 2 when full during the wet season, however, a reduction
in water level of 13m was recorded between the wet and dry season which significantly
reduces the surface area available for cages to 265km 2 . Three cage sizes were modelled, 5m,
10m and 15m diameter, for both high and low water levels. The outcome images indicated a
significant difference in availability of sites between the seasons for each of the cage sizes
(Fig 2.5). Modelling this change in water level and the subsequent loss of available sites
provides qualitative and quantitative guidelines for development of aquaculture sites, their
seasonal management and future monitoring, outcomes which can only be achieved using
GIS and associated technologies.
Figure 2.5 : Potential locations for 5m cages at high and low water level. Expanded area
indicates further detail of an area in the southern sector of the lake (after Ross et al, in press)
Page 32

2.4 Integrating environmental models within GIS
A further step in using GIS as an environmental management tool in aquaculture is to
integrate environmental models within GIS. Environmental models simulate environmental
scenarios and can be used to explain the processes involved in an environmental scenario or
extrapolated through time as a predictor of potential environmental conditions (Skidmore,
2002). The advantage of using environmental models within GIS is that the environmental
simulation can be examined and analysed within a geographical context (Argent, 2004;
Pullar & Springer, 2000). It is important when integrating models into GIS that the data is
compatible and there are no issues with scale when merging data (Argent, 2004; Bregt,
Skidmore & Nieuwenhuis, 2002). Coupling refers to the interoperability and
interdependence of software working together (Brandmeyer & Karimi, 2000; Tsui & Karam,
2007). A model can be loosely coupled to a GIS when the model is run in software that is
external to the GIS and information is exchanged between the model and GIS via files
(sometimes files need to be converted into compatible formats), a model can be tightly or
closely coupled to a GIS when the model and the GIS share the same files without the need
for translation and a model is considered to be embedded when it is operating as a GIS script
or via a graphic GIS interface (Karimi & Houston, 1996; Longley et al 2005).
Mathematical models are commonly used in environmental impact modelling as they can be
used to predict and quantify impact rather than conceptual models which are used as a
qualitative descriptor of a system or process (White, Mottershead & Harrison, 1992; Bruns &
Wiersma, 2004). Mathematical models can be integrated into GIS effectively using a number
of methods. The term mathematical model normally refers to an equation or sets of
equations that quantify and explain an observation or system and these models can be
integrated entirely within a GIS or performed in a separate model linked with a GIS (Logan &
Wolesensky, 2009; Johnston, 1998). There are many specialised programs available that can
be used in modelling. Some of the more popular programs used for environmental
modelling are Surfer (Golden software Inc, Colorado, USA), Powersim (Powersim Software
AS, Bergen, Norway), STELLA (isee systems inc, New Hampshire, USA), Soil Water
Assessment Tool (SWAT) (Grassland, Soil & Water Research Laboratory, Texas, USA) and
Water Quality Analysis Simulation Program (WASP) (EPA, Washington DC, USA). Many GIS
software packages have the ability to import data from statistical and modelling programs
either directly through data conversion (e.g. IDRISI Andes can import SurferGRID raster files
into IDRISI image format and it can also export IDRISI point files to SURFER DAT files).
Additionally, models can be integrated into GIS via independently developed programs and
interfaces; e.g. AVSWAT (Blackland Research Centre, Texas, USA) is an ArcGIS extension that
can be used to integrate SWAT models into ArcGIS. The use of extensions such as AVSWAT
is a very effective and efficient way to incorporate software into a GIS and increases the
functionality of both the modelling program and the GIS.
The tools within GIS software can be used to integrate models into a GIS; this is particularly
useful when modelling simple mathematical models. The soil science sector has focused a
lot on integrating soil erosion mathematical models within GIS. Many of these models use
equations based on the Universal Soil Loss Equation (USLE) that was developed by
Wischmeier & Smith (1978). Fistikoglu & Harmancioglu (2002) integrated USLE with GIS to
assess soil erosion and the transport of nonpoint source solution loads in the Gediz River
basin along Aegean coast of Turkey. USLE was integrated into GIS by converting all of the
variables in the equation into raster-based formats and then evaluating these layers using
the USLE calculation by overlaying the layers in raster format. Yue-qing, Jian & Xiao-mei
(2009) used RUSLE (Revised Universal Soil Loss Equation) to assess soil erosion of the
Maotiao River watershed, southwestern China. As had been done by Fistikoglu &
Page 33

Harmancioglu (2002) the study created raster layers for each variable within the equation
and then used the GIS to calculate the equation and produce the model output layer
mapping soil erosion in the study area. Zhang et al (2009a) went a step further than
Fistikoglu & Harmancioglu (2002) and Yue-qing, Jian & Xiao-mei (2009) and produced an
ArcGIS® [ESRI] tool, named ArcMUSLE using C# (programming language) and ArcObjects®
(the development platform for ArcGIS (Chang, 2008)) to model soil erosion risk using MUSLE
(Modified Universal Soil Loss Equation). ArcMUSLE integrates the MUSLE equation within GIS
and it works similar to the technique used by Yue-qing, Jian & Xiao-mei (2009) but instead of
the user being responsible for the whole process including the actual equation modelling
ArcMUSLE does the equation and the user is only responsible for inputting the required
layers and setting the parameters.
Models produced and developed within GIS are not solely used by scientists or those with
technological backgrounds. They can have an important practical influence on day to day
business operations such as aquaculture and agriculture where the majority of the
stakeholders such as the farmers and producers do not have mathematical or scientific
backgrounds to understand the complex theories behind models. GIS can be used to simplify
the decision making process, allowing the producer to input the required data and then the
GIS software will analyse the data and carry out the required operations to produce a final
output that is of use to the producer. By developing a module such as ArcMUSLE users with
little experience of GIS can perform the analysis as the module details the required input
layers and the user sets the parameters then the module carries out the operation itself,
whereas with the method used by Fistikoglu & Harmancioglu (2002) and Yue-qing, Jian &
Xiao-mei (2009) the user must do the analysis using the overlay function within the GIS
software and there is a greater opportunity for human error and consequently results could
be unrepresentative. Some GIS software developers have included commonly used models
as modules in newer versions of their software (e.g. the RUSLE module within recent
versions of IDRISI).
Another example of a practical use of an integrated model within a GIS setting is the
software developed by Georgoussis et al (2009) to help plan irrigation schedules for
agricultural areas. Georgoussis et al (2009) integrated the mathematical model
S.W.BA.CRO.S (Babajimoppulos, 1995) and GIS by developing a new piece of software that
interacts with ArcGIS using code created in Visual Basic (Microsoft, Redmond, Washington,
USA) and Visual Basic for Applications (Microsoft, Redmond, Washington, USA). The aim of
this software is to help irrigation scheduling of agricultural areas and by combining the
model with ArcGIS the software can produce maps of irrigation water needs and also
investigate potential future scenarios under different climatic conditions (Georgoussis et al,
2009). Georgoussiss et al (2009) developed this new open source software so that it is user
friendly and simple to use, they also created a new toolbar within ArcGIS allowing quick and
easy access. This is a useful tool for the agriculture industry and although improvements are
needed to the software in order to optimise the performance it is a good foundation to
develop a new management system with regards to irrigation scheduling.
The aquaculture industry could also benefit from integrating models with GIS in order to
develop new management strategies and monitoring programs particularly in assessing
environmental impact and dispersion of wastes. There has already been some research into
integrating environmental models into GIS. This research has focused on the modelling of
waste from fish cages; however the techniques used could be adapted and applied to other
systems and species. Pérez et al (2002) used both spreadsheet models and GIS to model the
distribution of particulate waste from Atlantic salmon (Salmo salar) cages in Scotland at farm
level. The model consisted of three main stages; the first sub-model was run in a
Page 34

spreadsheet and used mass balance techniques to quantify waste material (uneaten
feed/faeces), the second sub-model was also run in a spreadsheet and used Gowen’s
formula (Gowen et al, 1989) to calculate the distribution of waste and finally the third submodel
was carried out using GIS software involved the calculation and generation of the
final contour distribution diagrams (Pérez et al 2002). As noted by Brooker (2002) the main
limitation of the model developed by Pérez et al (2002) is that for a user experienced in GIS,
spreadsheets and image manipulation it takes some time to become familiar with the model
in order to operate it successfully and without previous experience it would be very difficult
to operate the model. Furthermore, the model is limited in its use as it is unable to model
large cage sites or handle large hydrographic datasets due to the size limitations of the
spreadsheet models (Brooker, 2002).
Although the model developed by Pérez et al (2002) had made some advances in integrating
a waste dispersion model (via loose coupling) into a GIS there were some limitations, mainly
due to the complex interaction between the spreadsheet sub-models and the GIS model.
Brooker (2002) and Corner et al (2006) further expanded on the work by Pérez et al (2002)
by fully integrating a particulate waste distribution model into GIS software using a
specifically developed program module and thus removing the use of separate spreadsheet
models. A typical output from the model (Figure 2.6) shows a double row of individual fish
cages and the quantified spread of waste material around them. The module was produced
using the development tool DELPHI 3 (Borland Software, California, USA) and was then fully
integrated into IDRISI32 (Clark Labs, Massachusetts, USA) using the IDRISI Application
Programming Interface (API) (Brooker, 2002; Corner et al, 2006). These studies were
analysing the impact of aquaculture systems on the local environment and are useful to
predict potential impacts in the future which is advantageous to regulators when it comes to
licensing new aquaculture systems and granting consent to discharge wastes (Pérez et al,
2002). The use of GIS and waste dispersion models together is an extremely useful
environmental management tool and could be applied in integrated coastal zone
management (ICZM).
In addition to being a useful management tool for regulators, the models are useful for the
farmers themselves providing they are in a simple, user-friendly format. Tironi, Marin &
Campuzano (2010) developed a management tool that could be used to assess the
particulate waste dispersal of salmon farming activities by combining a hydrodynamic
model, a particle tracking module and a GIS application. The hydrodynamic model was
developed using the free opensource MOHID Water Modeling System v.4.6 (Braunschweig
et al 2004; Neves, 1985), which was also used to simulate the dispersion of the particulate
wastes using the Lagrangian module. The results of the hydrodynamic model and the
particle tracking module were then combined within a GIS environment to create a
management tool using a modified ArcView 3.3 (ESRI) interface which was programmed
using AVENUE which is the scripting language for ArcView 3.3. Unlike the model developed
by Corner et al (2006) the model produced by Tironi, Marin & Campuzano (2010) has not yet
been fully validated due to a lack of available data, however as the management tool is
already used by local decision makers and the hydrodynamic model is able to simulate
dynamics within the Patagonian fjordic environment the authors note that it can be
considered partially validated. As the tool was being developed for salmon farmers in Chile
the text on buttons and in the help/information was translated into Spanish which is the
local language and throughout the development the developers checked the local manager’s
capabilities, needs and requirements to ensure efficient usability of the tool (Tironi, Marin &
Campuzano, 2010). By incorporating the target user(s) during the development process it
ensures that the tool can be understood and operated more efficiently and effectively by the
user(s) and that if they have any requirements then it is easier to adjust/ add/modify the
Page 35

system during model development rather than retrospectively once the model has been
completed.
Figure. 2.6: Modelling solid waste dispersal as Kg carbon.m 2 .yr from an intensive salmon
cage farm in Scotland, using GIS. Circles show the mean cage positions. The peak wastes are
directly underneath the cages, reducing to background levels within a short distance from
the cage group. (after Corner et al, 2006).
Other models have been developed to predict waste dispersion from fish cages without the
spatial aspect. These models include the popular DEPOMOD model (Scottish Association for
Marine Science, Oban, UK) which predicts the solids accumulation on the benthos through
particle tracking and resuspension models (Cromey, Nickell & Black, 2002). While DEPOMOD
is a useful tool in predicting environmental impact the advantage of a waste dispersion
model integrated with GIS is that the model can be georeferenced enabling features such as
coastlines, beaches and other structures to be added, this would allow the prediction of
where the waste outputs could impact local communities (Brooker, 2002; Corner et al,
2006). The model developed by Tironi, Marin & Campuzano (2010) was produced for a
specific location (Chacabuco bay in southern Chile), therefore was able to include some
relevant local features such as docks, food processing industries, sewage discharges and
towns allowing farmers and decision makers to be aware of activities that may impact
farming activities and industries/areas that may be impacted by the farms. Furthermore,
areas of biodiversity and special interest could be added to ensure that these areas are
protected from potential impacts from the aquaculture industry.
Page 36

2.5 GIS models and the SEAT project
There have been a few studies using GIS as a method of site selection for some of the SEAT
case study species (Hossain et al, 2007; Giap, Yi & Yakupitiyage, 2005); however there has
been a lack of research into using GIS as a tool to measure environmental impact and waste
dispersion with the selected species; the studies that were described previously (Pérez et al
2002; Brooker, 2002; Corner et al, 2006; Tironi, Marin & Campuzano, 2010) all focussed on
Atlantic salmon (Salmo salar). Although aspects of these models can be used with other
species, the environmental conditions and growing requirements vary greatly between
warm/cold water and freshwater/marine/brackish species, consequently new models would
have to be developed. Furthermore, there are differences between each of the case study
species and each species will have different requirements and will vary in the type and
degree of impact therefore models will have to be species specific rather than one model as
that would be too general and would not be representative. New models will be developed
within IDRISI GIS software that will integrate the system dynamics modelling (described in
section 1 of this report) into a spatial setting and also analyse wider field inputs.
Page 37

3.1 Study areas
Due to the large amount of work involved in creating spatial databases for an area it was
decided to have one study area per country which covered both of the case study SEAT
species. Study areas were identified using the information provided in Table 3.1 along with
field knowledge gained during Work Package 4’s scoping trips to each country. Using this
information the study areas have been selected and defined within a geographical context
using drainage basins to set boundaries for the study area in each country. Drainage basin
vector files were downloaded from the Hydro 1K dataset and the boundary for each study
area was then identified. Hydro 1K is a geographic dataset that was developed by the US
Geological Survey (USGS) to provide a set of comprehensive and consistent data at a global
scale using topographically derived data sets derived from the GTOPO 30 (a global raster
Digital Elevation Model (DEM)).
Table 3.1: Geographical coverage of farmer survey
(Source: SEAT extranet file : PSGFarmerSurveyPresentation2Dec10)
Page 38

3.1.1. Bangladesh
In Bangladesh the study area covers the districts of Khulna, Satkhira and Bagerhat. Cox's
Bazar was not included due to its distance from other districts. Shrimp and prawn farms are
located in this study area covering the two SEAT case study species for Bangladesh.
Figure 3.1 : Drainage basins from HYDRO1K dataset covering the study area in Bangladesh.
Figure 3.2 : Study area in Bangladesh
Page 39

3.1.2 China
In China the study area covers Zhanjiang and Maoming in Guangdong province. Hainan is a
large island to the south of the study area and was not included due to the distance from
other surveyed areas. Zhanjiang and Maoming neighbour each other and they include tilapia
and shrimp farms which are the SEAT case study species for China.
Figure 3.3: Drainage basins from HYDRO1K dataset covering the study area in China
Figure 3.4: Study area in China
Page 40

3.1.3 Thailand
In Thailand the study area will include Samut Sakhan, Samut Prakhan, Chachoengsao,
Chanthaburi and Petchaburi as these areas cover both tilapia and shrimp which are the SEAT
case study species for Thailand. Surat Thani was not included due to the large geographical
distance between Surat Thani and the other areas.
Figure 3.5 : Drainage basins from HYDRO1K dataset covering the study area in Thailand
Figure 3.6: Study area in Thailand
Page 41

3.1.4 Vietnam
The study area in Vietnam is the Mekong Delta; this study area covers all of the locations
surveyed in the SEAT farmer survey. The SEAT case study species for Vietnam are catfish and
shrimp.
Figure 3.7 : Drainage basins from HYDRO1K dataset covering the study area in Vietnam
Figure 3.8 : Study area in Vietnam
Page 42

REFERENCES
Aguilar-Manjarrez, J. (1996) Development and evaluation of GIS-based models for planning
and management of coastal aquaculture: a case study in Sinola, Mexico. PhD. thesis,
University of Stirling.
Aguilar-Manjarrez, J., Kapetsky, J.M. & Soto, D. (2010) The potential of spatial planning
tools to support the ecosystem approach to aquaculture. FAO/ROME. Expert Workshop. 19-
21 November 2009, Rome, Italy. FAO Fisheries and Aquaculture Proceedings. No. 17. FAO,
Rome. 176ppm
Ahmed, M., Dey, M.M. & Garcia, Y.T. (2007) The role of species and systems in the
development and growth of aquaculture in Asia: Needs and prospects. In: Leung, P.S., Lee,
C.S. & O’Bryen, P.J. eds. Species and system selection for sustainable aquaculture. Blackwell
Publishing Ltd, Oxford. 506pp.
Araujo, J.N., Mackinson, S., Stanford, R. J., Hart, P.J.B. (2008) Exploring fisheries strategies
for the western English Channel using an ecosystem model. Ecological Modelling 210, 465-
477
Argent, R.M. (2004) An overview of model integration for environmental applications –
components, frameworks and semantics. Environmental Modelling & Software, 19: 219 –
234.
Autodesk Inc (2010) Autodesk – Geospatial Solutions. [Internet]. Autodesk Inc. Available
from [Accessed:
27/08/10].
Babajimopoulos, C. (1995) SWBACROS: A model for the estimation of the water balance of a
cropped soil. Environmental Modelling Software, 10; 211–220
Barlund, I., Kirkkala, T. (2008) Examining a model and assessing its performance in
describing nutrient and sediment transport dynamics in a catchment in southwestern
Finland. Boreal Environment Research 13, 195-207.
Barlund, I., Kirkkala, T., Malve, O., Kamari, J. (2007) Assessing SWAT model performance in
the evaluation of management actions for the implementation of the Water Framework
Directive in a Finnish catchment. Environmental Modelling & Software 22, 719-724
Beder, S. (2006) Environmental principles and policies: an interdisciplinary introduction.
Earthscan, London. 336pp.
Bengtsson, M., Carlson, R., Molander, S. & Steen, B. (1998) An approach for handling
geographical information in life cycle assessment using a regional database. Journal of
Hazardous Materials, 61: 67-75.
Bernhardsen, T. (1999) Geographic Information Systems: An introduction. 2nd edition.
372pp.
Beveridge, M.C.M. (2004) Cage Aquaculture. 3 rd edition. Blackwell Publishing Ltd, Oxford.
368pp.
Page 43

Black, K. (2001) Sustainability of aquaculture. In: Black,K. ed. Environmental Impacts of
Aquaculture. Sheffield Academic Press, Sheffield. 199 - 212pp.
Boyd, C.E., Tucker, C.S. (1999) Pond Aquaculture Water Quality Management. Kluwer
Academic Publishers, Massachusetts, USA
Brandmeyer, J.E. & Karimi, H.A. (2000) Coupling methodologies for environmental models.
Environmental Modelling & Software, 15: 479 -488.
Braunschweig, F., Leitao, P.C., Fernandes, L., Pina, P., Neves R.J.J. (2004) The objectoriented
design of the integrated water modelling system mohid. In: Proceedings of the XV
international conference on computational methods in water resources, vol 2, pp 1079–1090
Bregt, A.K., Skidmore, A.K. & Nieuwenhuis, G. (2002) Environmental modeling: issues and
discussion. In: Skidmore, A.K. ed. Environmental Modeling with GIS and remote sensing.
Taylor & Francis, London. 268pp.
Brigolin, D., Pastres, R., Tomassetti, P., Porrello, S. (2010) Modelling the biomass yield and
the impact of sea bream mariculture in the Adriatic and Tyrrhenian Seas (Italy). Aquaculture
International 18, 149-163
Brimicombe, A. (2010) GIS, Environmental Modeling and Engineering. 2 nd edition. CRC Press,
Florida. 378pp.
Brooker, A.J. (2002) Development and integration of waste dispersion models for cage
aquaculture within the GIS framework. MSc thesis, University of Stirling.
Bruns, D.A. & Wiersma, G.B. (2004) Conceptual basis of environmental monitoring systems:
a geospatial perspective. In: Wiersma, G.B. (ed) Environmental monitoring. CRC Press,
Florida. 792pp.
Buitrago, J., Rada, M., Hernández, H. & Buitrago, E. (2005) A single-use site selection
technique, using GIS, for aquaculture planning: choosing locations for Mangrove Oyster raft
culture in Margarita Island, Venezuela. Environmental Management. 35 (5) : 544-556.
Caliper (2010) Maptitude Suite of Mapping Software Products. [Internet]. Newton MA,
Caliper Corporation. Available from
[Accessed: 27/08/10].
Cellier, F. E. (2008) World3 in Modelica: Creating System Dynamics Models in the Modelica
Frame- work. In: Proc. 6th International Modelica Conference, Bielefeld, Germany (2008)
Vol.2 393- 400.
Chamberlain, J., Stucchi, D. (2007) Simulating the effects of parameter uncertainty on waste
model predictions of marine finfish aquaculture. Aquaculture 272, 296-311
Chang, K.T. (2008) Programming ArcObjects with VBA: A task-orientated approach. 2 nd
edition. CRC Press, Florida .360pp.
Christensen, V., Walters, C.J. (2004) Ecopath with Ecosim: methods, capabilities and
limitations. Ecological modelling, 172, 109-139
Page 44

Clark Labs (2009) Clark Labs - IDRISI GIS and Image Processing Sofware. [Internet],
Worcester, MA, Clark Labs. Available from [Accessed
: 27/08/10].
Corbin, J.S. & Young, L.G.L. (1997) Planning, Regulation and Administration of Sustainable
Aquaculture. In: Bardach, J.E. ed. Sustainable Aquaculture. John Wiley & sons, Canada.
251pp.
Corner, R.A., Brooker, A.J., Telfer, T.C. & Ross, L.G. (2006) A fully integrated GIS-based
model of particulate waste distribution from marine fish-cage sites. Aquaculture. 258: 299-
311.
Costanza, R., Gottlieb, S. (1998) Modelling ecological and economic systems with STELLA:
Part II. Ecological Modelling 112, 81-84
Costanza, R., Voinov, A. (2001) Modelling ecological and economic systems with STELLA:
Part III. Ecological Modelling 143, 1-7
Cripps, S. & Kumar, M. (2003) Environmental and other impacts of aquaculture. In: Lucas,
J.S. & Southgate, P.C.eds. Aquaculture: farming aquatic animals and plants. Blackwell
Publishing Ltd, Oxford. 502pp.
Cromey, C.J., Nickell, T.D. & Black, K.D. (2002) DEPOMOD – modelling the deposition and
biological effects of waste solids from marine cage farms. Aquaculture, 214: 211-239.
Cromey, C.J., Nickell, T.D., Treasurer, J., Black, K.D., Inall, M. (2009) Modelling the impact of
Cod (Gadus morhua L.) farming in the marine environment – CODMOD. Aquaculture 289, 42-
53
Culberson, S.T., Piedrahita, R.H. (1996) Aquaculture pod ecosystem modelling: temperature
and dissolved oxygen prediction- mechanism and application. Ecological modelling 89, 231-
258
Del Campo Barquin, M. (2002). Restocking A bio-socio-economic simulation model for
management of the red sea urchin fishery in Chile. PhD. Thesis. University of Stirling.
De Silva, S.S. & Davy, F.B. (2009) Aquaculture successes in Asia: contributing to sustained
development and poverty alleviation. In: Silva, S.S. & Davy, F.B. eds. Success Stories in Asian
Aquaculture. Springer, Netherlands. 214pp.
Dias De Oliveira, M.E., Vaughan, B.E., Rykiel Jr, E.J. (2005) Ethanol as fuel: Energy, Carbon
dioxide balances and ecological footprint. Bioscience 55, 593-613
Duarte, P., Ferriera, J.G. (1997) Dynamic modelling of photosynthesis in marine and
estuarine ecosystems. Environmental Modelling and Assessment 2, 83-93
Dudley, R.W., Panchang, V.G., Newell, C.R. (2000) Application of a comprehensive
modelling strategy for the management of net-pen aquaculture waste transport.
Aquaculture 187, 319-349
Dueker, K., and D. Kjerne, (1989), Multipurpose Cadastre: Terms and Definitions. Falls
Church, VA: American Society for Photogrammetry and Remote Sensing and American
Congress on Surveying and Mapping (ASPRS-ACSM). Quoted in: Brimicombe, A. (2010) GIS,
Environmental Modeling and Engineering. 2 nd edition. CRC Press, Florida. 378pp.
Page 45

Ecopath. (2010).[Internet]. Available from [Accessed April 2010]
EcoWin2000 Website.[Internet]. Available from [accessed
15 th May 2010]
Elshorbagy, A., Jutla, A., Barbour, L., Kells, J. (2005) System dynamics approach to assess
the sustainability of reclamation of disturbed watersheds. Canada Journal of Civil
Engineering 32, 144-158
ERDAS Inc (2010) ERDAS, Inc – The Earth to Business Company. [Internet]. Available from <
http://www.erdas.com/default.aspx> [Accessed 27/08/10]
ESRI (n.d.) ESRI – THE GIS software leader. [Internet]. Available from [Accessed : 27/08/10].
FAO. (2010). Penaeus indicus species Factsheet. An Annotated Catalogue of Species of
Interest to Fisheries. FAO Fisheries Synopsis No.125, Volume 1
FARM website. [Internet]. Available from [accessed 29 th April
2010]
Fernandes, T.F., Eleftherious, A., Ackefors, H., Eleftheriou, M., Ervik, A., Sanchez-Mata, A.,
Scanlon, T., White, P., Cochrane, S., Pearson, T.H. & Read, P.A. (2001) The scientific
principles underlying the monitoring of the environmental impacts of aquaculture. Journal of
Applied Icthyology, 17: 181-193.
Ferreira, J.G, Andersson, H.C., Corner, R.A., Desmit, X., Fang, Q., De Goede, E.D., Groom,
S.B., Gu, H., Gustafsson, B.G., Hawkins, A.J.S., Hutson, R., Jiao, H., Lan, D., Lencart-Silva, J.,
Li, R., Liu, X., Luo, Q., Musango, J.K., Nobre, A.M., Nunes, J.P., Pascoe, P.L., Smits, J.G.C.,
Stigebrandt, A., Telfer, T.C., De Wit, M.P., Yan, X., Zhang, X.L., Zhang, Z., Zhu, M.Y., Zhu,
C.B., Bricker, S.B., Xiao, Y., Xu, S., Nauen, C.E. & Scalet, M. (2008) SPEAR –Sustainable
Options for People, Catchment and Aquatic Resources. IMAR-Institute of Marine Research,
ISBN 978-972-99923-2-2, 184 pp.
Ferreira, J.G. (1995) ECOWIN- an Object-oriented ecological model for aquatic ecosystems.
Ecological Modelling 79, 21-34
Ferreira, J.G., Hawkins, A.J.S., Bricker, S.B. (2007) Management of productivity,
environmental effects and profitability of shellfish aquaculture- the Farm Aquaculture
Resource Management (FARM) model. Aquaculture 264, 160-174
Ferreira, J.G., Sequeira, A., Hawkins, A.J.S., Newton, A., Nickell, T.D., Pastres, R., Forte, J.,
Bodoy, A., Bricker, S.B. (2009) Analysis of coastal and offshore aquaculture: Application of
the FARM model to multiple systems and shellfish species. Aquaculture 289, 32-41
Fistikoglu, O. & Harmancioglu, N.B. (2002) Integration of GIS with USLE in assessment of soil
erosion. Water Resources Management. 16: 447-467.
Flynn, K.J., Fasham, M.J.R., Hipkin, C.R. (1997) Modelling the interactions between
ammonium and nitrate uptake in marine phytoplankton. Royal Society of Philosophical
Transactions: Biological Sciences 352, 1625- 1645
Ford, A. (1999) Modelling the Environment. Island Press, Washington DC, p319-352
Ford, A. (2009) Modelling the Environment 2 nd ed. Island Press, Washington DC, p319-352
Page 46

Forrester, J. (1961) Industrial Dynamics. Pegasus Communications, Walthan Publishing, MA.
Franco, A.R., Ferriera, J.G., Nobre, A.M. (2006) Development of a growth model for penaeid
shrimp. Aquaculture 259, 268-277
Frankic, A. & Hershner, C. (2003) Sustainable aquaculture: developing the promise of
aquaculture. Aquaculture International, 11: 517-530.
Frid, C. & Dobson, M. (2002) Ecology of Aquatic Management. Pearson Education Ltd, Essex.
274pp.
Gallego, E. (2006) MUD: a model to investigate the migration of 137 Cs in the urban
environment and drainage and sewage treatment systems. Journal of Environmental
Radioactivity 85, 247-264
Gallego, E., Brittain, J.E., Hakanson, L., Heling, R., Hofman, D., Monte, L. (2000) MOIRA: a
computerized decision support system for the restoration of radionuclide contaminated
freshwater ecosystems. IRPA-10 Congress, Hiroshima, 15-19 May 2000.
GE Energy (2010) Geospatial Enabled Solutions (Smallworld/MapFrame). [Internet] GE
Energy. Available from
[Accessed :
27/08/10].
Georgoussis, H., Babajimopoulos, C., Panoras, A., Arampatzis, G., Hatzigiannakis, E., Ilias,
A. & Papamichail, D. (2009) Regional scale irrigation scheduling using a mathematical model
and GIS. Desalination, 237: 108-116.
GESAMP (1996) Monitoring the ecological effects of coastal aquaculture wastes. Reports
and studies No. 57,FAO, Rome, 38 pp.
Geyer, R., Stoms, D.M., Lindner, J.P., Davis, F.W. & Wittstock, B. (2010) Coupling GIS and
LCA for biodiversity assessments of land use, Part 1: inventory modeling. International
Journal of Life Cycle Assessment, 15: 454-467.
Giap, D.H., Yi, Y. & Yakupitiyage, A. (2005) GIS for land evaluation for shrimp farming in
Haiphong of Vietnam. Ocean & Coastal Management, 48: 51 – 63.
Global Mapper LLC (2009) GIS Mapping Software -GPS Mapping software – Satellite maps –
Aerial photos (Global Mapper). [Internet]. Global Mapper LLC. Available from
[Accessed : 27/08/10].
Gowen, R.J., Bradbury, N.B. & Brown, J.R. (1989) The use of simple models in assessing two
of the interactions between fish farming and the marine environment. In: De Pauw, N.,
Jaspers, E., Wilkins, N. (Eds). Aquaculture – a biotechnology in progress. European
Aquaculture Society, Belgium. 1071-1080pp.
Grant, J., Curran, K.J., Guyondet, T.L., Tita, G., Bacher, C., Koutitonsky, V., Dowd, M. (2007)
A box model of carrying capacity for suspended mussel aquaculture in Lagune de la Grande-
Entrée, Iles-de-la-Madeleine, Quebec. Ecological Modelling 200, 193-206.
Grant, W. E., Pedersen, E.K., Marin,S.L. (1997). Ecology and natural resource management:
systems analysis and simulation. John Wiley and Sons Inc. USA
Page 47

GRASS Development Team (2010) GRASS GIS – The World Leading Free Software GIS.
[Internet]. GRASS Development Team. Available from
[Accessed : 27/08/10].
Guimarães, A.S., Travassos, P., Filho, P.W.M.E.S., Goncalves, F.D. & Costa, F. (2010) Impact
of aquaculture on mangrove areas in the northern Pernambuco Coast (Brazil) using remote
sensing and geographic information system. Aquaculture Research, 41: 828-838.
Guinée, J.B. (2002) Handbook on Life Cycle Assessment : Operational guide to the ISO
standards. Kluwer Academic Publishers, The Netherlands. 708pp.
Hall, C.A., Day, J.W. (1977). Ecosystem modelling in theory and practice. John Wiley &Sons,
Inc, USA.
Hansen, P.K., Ervik, A., Schaanning, M., Johannessen, P., Aure, J., Jahnsen, T. &
Stigebrandt, A. (2001) Regulating the local environmental impact of intensive, marine fish
farming II. The monitoring programme of the MOM system (Modelling-Ongrowing fish farms
– Monitoring). Aquaculture, 194: 75-92.
Heywood, I., Cornelius, S. & Carver, S. (2002) An introduction to Geographical Information
Systems. 2 nd Edition. Pearson Education Limited, Essex. 295pp
Hossain, M.S., Chowdhury, S.R., Das, N.G. & Rahaman, M.M. (2007) Multi-criteria
evaluation approach to GIS-based land-suitability classification for tilapia farming in
Bangladesh. Aquaculture International, 15: 425 – 443
Hossain, M.S., Chowdhury, S.R., Das, N.G., Sharifuzzaman, S.M. & Sultana, A. (2009)
Integration of GIS and multicriteria decision analysis for urban aquaculture development in
Bangladesh. Landscape and Urban Planning, 90: 119-130.
Huang, Z., Xue, B., Pang, Y. (2009) Simulation on stream flow and nutrient loadings in
Gucheng Lake, Low Yangtze River Basin, based on SWAT model. Quaternary International
208, 109-115
Huijbregts, M.A.J. (1998)Application of uncertainty and variability in LCA, Part 1: A general
framework for the analysis of uncertainty and variability in life cycle assessment.
International Journal of life cycle assessment, 3 (5): 273-280pp.
Huijbregts, M.A.J., Hellweg, S., Frischknecht, R., Hungerbϋhler, K. & Hendricks, A.J. (2008)
Ecological footprint accounting in the life cycle assessment of products. Ecological
Economics, 64: 798-807.
Intergraph Corporation (2010) Intergraph – GeoMedia. [Internet]. Intergraph Corporation.
Available from
[Accessed: 27/08/10].
Isee Systems Website (STELLA) (2010) [Internet]. Available from
[Accessed April
2010]
Jamu, D.M., Piedrahita, R.H. (2002a). An organic matter and nitrogen dynamics model for
the ecological analysis of integrated aquaculture/agriculture systems: I. model development
and calibration. Environmental Modelling and Software 17, 517-582
Page 48

Jamu, D.M., Piedrahita, R.H. (2002b). An organic matter and nitrogen dynamics model for
the ecological analysis of integrated aquaculture/agriculture systems: II. Model evaluation
and application. Environmental Modelling and Software 17, 583-592
Johnston, C.A. (1998) Geographic Information Systems in Ecology. Blackwell Science Ltd,
Oxford. 256pp.
Jusup, M., Gecek, S., Legovic, T. (2007) Impact of aquacultures on the marine ecosystem:
Modelling benthic carbon loading over variable depth. Ecological Modelling 200, 459-466.
Jusup, M., Klanjscek, J., Petricioli, D., Legovic, T. (2009) Predicting aquaculture-derived
benthic organic enrichment: Model Validation. Ecological Modelling 220, 2407- 2414
Kalantzi, I. & Karakassis, I. (2006) Benthic impacts of fish farming: Meta-analysis of
community and geochemical data. Marine Pollution Bulletin, 52: 484 – 493.
Kapetsky, J.M., Hill, J.M. & Worthy, L.D. (1988) A geographical information system for
catfish farming development. Aquaculture, 68: 311 -320.
Kapetsky, J.M.; Aguilar-Manjarrez, J. (2007) Geographic information systems, remote
sensing and mapping for the development and management of marine aquaculture. FAO
Fisheries Technical Paper No. 458. Rome, FAO. 2007. 125 pp.
Karimi, H.A. & Houston, B.H. (1996) Evaluating strategies for integrating environmental
models with GIS: current trends and future needs. Computers, Environment and Urban
Systems, 20(6): 413-425.
Leung, P. & El-Gayar, O.F. (1997) The role of modeling in the managing and planning of
sustainable aquaculture. In: Bardach, J.E. ed. Sustainable Aquaculture. John Wiley & sons,
Canada. 251pp.
Logan, J.D. & Wolesensky, W.R. (2009) Mathematical Methods in Biology. John Wiley &
Sons, New Jersey. 417pp.
Longdill, P.C., Healy, T.R. & Black, K.P. (2008) An integrated GIS approach for sustainable
aquaculture management area site selection. Ocean & Coastal Management, 51: 612 – 624
Longley, P.A., Goodchild, M.F., Maguire, D.J. & Rhind, D.W. (2005) Geographic information
systems and science. 2 nd edition. John Wiley & Sons Ltd, West Sussex. 517pp
Losordo, T.M., Piedrahita, R.H. (1991) Modelling temperature variation and thermal
stratification in shallow aquaculture ponds. Ecological Modelling 54, 189-226.
Lymer, D., Funge-Smith, S., Clausen, J. & Miao, W. (2008) Status & potential of fisheries and
aquaculture in Asia and the Pacific 2008. FAO document, Bangkok. 104pp.
Macedo, M.F. Duarte, P. (2006) Phytoplankton production modelling in three marine
ecosystems- static versus dynamic approach. Ecological modelling 190, 299- 316
Manifold (2009) Manifold® System – The Ultimate GIS and Mapping Package. [Internet],
Available from: [Accessed : 27/08/10].
Maxwell, T. (1999) A Parsi-Model approach to Modular Simulation. Environmental Modelling
and Software 14, 511-517
Page 49

Mazzonleni, S., Giannino, F., Colandrea, M., Nicolazzo, M., di Agraria, F. (2003) Integration f
system dynamics models and Geographic information Systems. Modelling and Simulation,
304-306
Meaden, G.J. (1987) Where Should Trout Farms be in Britain? Fish Farmer 10 (2), 33-35.
Midlen, A. & Redding, T.A. (1998) Environmental Management for Aquaculture. Chapman &
Hall, London. 223pp.
Muetzelfeldt, R., Massheder, J. (2003) The Simile visual modeling environment. European
Journal of Agronomy 18, 345-358.
Mulligan, M. & Wainwright, J. (2004) Modelling and model building. In: Wainwright, J. &
Mulligan, M. eds. Environmental Modelling: Finding simplicity in complexity. John Wiley &
Sons, West Sussex. 430pp;
Nath, S.S., Bolte, J.P., Ross, L.G. & Aguilar-Manjarrez, J. (2000) Applications of geographical
information systems (GIS) for spatial decision support in aquaculture. Aquacultural
Engineering. 23: 233-278.
Neiland, A.E., Soley, N., Varley, J.B., Whitmarsh, D.J. (2001) Shrimp aquaculture: economic
perspectives for policy development. Marine Policy, 25: 265 – 279.
Neves, R.J.J. (1985) Étude experimentale et modélisation des circulations trasitoire et
résiduelle dans l’estuaire du sado. Ph. D. Thesis, Univ. Liége
Nirmalakhandan, N. (2002) Modeling tools for environmental engineers and scientists. CRC
Press LLC, Florida. 328pp.
Normark, K. (1997) DYNAMO – A set of tools for dynamic modelling. Report to the Danish
Natural Science Research Council. Grant No. 9400911. 19 pp. Available from:
http://www.cs.aau.dk/~normark/dyn-models/tools-2/abstract.html.
Nunes, J.P., Ferreira, J.G., Gazeau, F., Lencart-Silva, J., Zhang, X.L., Zhu, M.Y., Fang,
J.G.(2003). A model for sustainable management of shellfish polyculture in coastal bays.
Aquaculture 219, 257-277
Panchang, V., Cheng, G., Newell, C. (1997) Modelling hydrodynamics and aquaculture waste
transport in coastal Maine. Estuaries 20, 14-41
Pauly, D., Christensen, V., and Walters, C. (2000) Ecopath, Ecosim, and Ecospace as tools for
evaluating ecosystem impact of fisheries. – ICES Journal of Marine Science, 57: 697–706.
Pérez, O.M., Telfer, T.C., Beveridge, M.C.M. & Ross, L.G. (2002) Geographical information
systems (GIS) as a simple tool to aid modeling of particulate waste distribution at marine fish
cage sites. Estuarine, Coastal and Shelf Science. 54: 761 – 768.
Pérez, O.M., Telfer, T.C. & Ross, L.G. (2005) Geographical information systems-based
models for offshore floating marine fish cage aquaculture site selection in Tenerife, Canary
Islands. Aquaculture Research, 36: 946 – 961.
Pérez, O. M. Telfer, T. C. and Ross, L. G. (2003). Use of GIS-based models for integrating and
developing marine fish cages within the tourism industry in Tenerife (Canary Islands).
Coastal Management. 34: (4), 355-366.
Page 50

Pillay, T.V.R. & Kutty, M.N. (2005) Aquaculture: principles and practices. Blackwell
Publishing Ltd, Oxford. 624pp.
Pillay, T.V.R. (2004) Aquaculture and the Environment. 2 nd edition. Blackwell Publishing Ltd,
Oxford. 196pp.
Pitney Bowes Software Inc (2010) Products: MapInfo Professional | Pitney Bowes Business
Insight. [Internet]. Available from
[Accessed : 27/08/10].
Powersim (2010). [Internet] Available from [Accessed April
2010]
Primavera, J.H. (1997) Socio-economic impacts of shrimp culture. Aquaculture Research, 28:
815-827.
Pullar, D. & Springer, D. (2000) Towards integrating GIS and catchment models.
Environmental Modelling & Software, 15: 451-459.
Pullin, R.S.V., Rosenthal, H., Maclean, J.L., Editors. (1993). Environmental and Aquaculture
in developing countries. ICLARM Conf. Proc. 31.
Radiarta, I.N., Saitoh, S.I & Miyazono, A. (2008) GIS-based multi-criteria evaluation models
for identifying suitable sites for Japanese scallop (Mizuhopecten yessoensis) aquaculture in
Funka Bay, southwestern Hokkaido, Japan. Aquaculture, 284: 127 – 135.
Randolph, J. (2004) Environmental Land Use Planning and Management. Island Press,
Washington DC. 704pp.
Ross, L.G., Falconer, L., Campos-Mendoza, A. & Martinez-Palacios, C. (in press) Spatial
modelling for freshwater cage location in the Presa Adolfo Mateos Lopez (El Infiernillo),
Michoacán, México. Aquaculture Research.
Salam, A. Md. (2000). GIS modelling of coastal aquaculture development in Khulna district,
Sunderbans, Bangladesh. Ph.D. University of Stirling. 2000. p400.
Salam, M.A., Khatun, N.A. & Ali, M.M. (2005) Carp farming potential in Barhatta Upazilla,
Bangladesh: a GIS methodological perspective. Aquaculture, 245: 75-87.
Salam, M.A. and Ross, L.G. (1999). GIS modelling for aquaculture in South-western
Bangladesh: Comparative production scenarios for brackish and freshwater shrimp and fish.
Proceedings of GIS’99. 13th Annual Conference on Geographic Information Systems.
Vancouver. p141-145.
Salam, M.A. and Ross, L.G. (2000). Optimizing site selection for development of shrimp
(Penaeus monodon) and mud crab (Scylla serrata)) culture in Southwestern Bangladesh.
Proceedings of GIS’2000, 14 th Annual Conference on Geographic Information Systems.
Toronto, Canada, 13-16 March’2000. p1-17.
Salam, M. A., Ross, L.G., and Beveridge, M.C.M. (2003). A comparison of development
opportunities for crab and shrimp aquaculture in South-western Bangladesh, using GIS
modelling. Aquaculture. 220: 477-494.
Sammet, J.E. (1969) Programming Languages: History and Fundamentals. Prentice Hall, USA
Page 51

SAMS. (2004). Ecosystem Approach for Sustainable Aquaculture (ECASA). [Internet] Available
from [accessed 5 th May 2010]
Schaffner, M., Bader, H.P., Scheidegger, R. (2009) Modelling the contribution of point
sources and non-point sources to Thachin River water pollution. Science of the Total
Environment 407, 4902-4915
Schaffner, M., Hurni, H., Weingartner, R. (2007) Applying a Material Flow Analysis Model to
assess river water pollution and mitigation potentials: a Case-study in the Thachin River
Basin, Central Thailand.
Schneiderman, E. M., Pierson D. C., Lounsbury, D.G., Zion, M.S. (2002) Modelling the
hydrochemistry of the Cannonsville watershed with generalized watershed loading functions
(GWLF) Journal of American Water Resources Association. 38 (5). 1323-1348
Scott, P. (2004) Aquaculture development interactions in Sepetiba Bay, Rio de Janeiro,
Brazil. A GIS study. PhD. Thesis. University of Stirling.
Scott, P.C., Cansado, S. and Ross, L.G. (1998) A GIS-assisted mollusc culture potential
determination for Sepetiba Bay, Brazil. GIS Planet '98. Lisbon. Portugal.
SEPA. (2010) Regulation and monitoring of marine cage fish farming in Scotland- a manual of
procedures. Annex F. Seabed Monitoring and Assessment. Scottish Environment Protection
Agency
Shih, Y.C., Chou, C.L. & Chiau, W.Y. (2009) Geographic information system applied to
measuring benthic environmental impact with chemical measures on mariculture at Penghu
Islet in Taiwan. Science of the Total Environment, 407: 1824-1833.
Silvert, W. & Cromey, C.J. (2001) Modelling Impacts. In: Black, K.D. ed. Environmental
Impacts of Aquaculture. Sheffield Academic Press. 214pp.
Simachaya, W. (2003). Lessons Learned on Integrated Watershed and Water Quality
Management in the Thachin River Basin, Thailand. Proceedings First Southeast Asia Water
Forum. November 17-21, 2003.
Simulistics (Simile). (2010) [Internet] Available from
[Accessed June 2010]
Skidmore, A.K. (2002) Taxonomy of environmental models in the spatial sciences. In:
Skidmore, A.K. ed. Environmental Modeling with GIS and remote sensing. Taylor & Francis,
London. 268pp.
Smith, J., Smith, P. (2007) Environmental Modelling: An introduction. Oxford University
Press, New York. USA
Southall, T., Telfer, T.C. & Hambrey, J. (2004) Environmental Capacity modeling in
Aquaculture Development. TROPICA working paper No 1, DFID.
Spruill, C.A., Workman, S.R., Taraba, J.L. (2000) Simulation of daily and monthly stream
discharge from small watersheds using the SWAT model. American society of Agricultural
Engineers 43, 1431-1439.
Page 52

Stave, K. A. (2002) A system dynamics model to facilitate public understanding of water
management options in Las Vegas, Nevada. Journal of Environmental Management. 67. 303-
313
Stigebrandt, A., Aure, J., Ervik, A., Hansen, P.K. (2004) Regulating the local environmental
impact of intensive marine fish farming III. A model for estimating the holding capacity in the
Modelling- Ongrowing fish farm- Monitoring System. Aquaculture 234, 239- 261
Tironi, A., Marin, V.H. & Campuzano, F.J. (2010) A management tool for assessing
aquaculture environmental impacts in Chilean Patagonian fjords: integrating hydrodynamic
and pellets dispersion models. Environmental Management. 45: 953-962.
Tolson, B.A., Shoemaker, C.A.. (2007) Cannonsville Reservoir Watershed SWAT2000 model
development, calibration and validation. Journal of Hydrology 337, 68-86
Tsui, F.F. & Karam, O. (2007) Essentials of Software Engineering. Jones and Bartlett
Publishers, London. 387pp.
Ullrich, A., Volk, M. (2009) Application of the Soil and Water Assessment Tool (SWAT) to
predict the impact of alternative management practices on water quality and quantity.
Agriculture Water Management 96, 1207- 1217
US EPA. (2010) Water Quality Analysis Simulation Program (WASP) [Internet]. Available from
[Accessed 15th April 2010]
USGS (2010) USGS – dlgv32 Pro Viewing Software. [Internet]. Available from
http://mcmcweb.er.usgs.gov/drc/dlgv32pro/ [Accessed: 27/08/10].
Valavanis, V.D. (2002) Geographic information systems in oceanography and fisheries.
Taylor & Francis, London. 209pp.
Vanclay, J.K. (2006) Spatially-explicit competition indices and the analysis of mixed-species
plantings with the Simile modeling environment. Forest Ecology and Management 233, 295-
302.
Ventana Systems (Vensim)[Internet]. Available from [Accessed
May 2010]
Vlachos, D., Georgiadis, P., Lakovou, E. (2007) A system dynamics model for dynamic
capacity planning of remanufacturing in closed-loop supply chains. Computers and
Operations Research 34, 367- 394.
Vuksanovik, V., De Smedt, F., Van Meerbeeck, S. (1996) Transport of polychlorinated
biphenyls (PCB) in the Scheldt Estuary simulated with the water quality model WASP. Journal
of Hydrology 174, 1-18
Wackernagel, M. & Rees, W. (1996) Our ecological footprint: reducing human impact on the
earth. New Society Publishers, Canada. 160pp.
Wackernagel, M. & Yount, J.D. (2000) Footprints for sustainability: the next steps.
Environment, Development and sustainability, 2: 21-42.
Ward Jr, G.H., Benaman, J. (1999) Models for TMDL application in Texas Water sources:
Screening and model review. Online report: CRWR-99-7. Centre for Research in Water
Page 53

Resources. The University if Texas at Austin, Austin, Texas 78712;
http://www.crwr.utexas.edu/reports/pdf/1999/rpt99-7.pdf. [Accessed 27th May 2010]
Weise, A.M., Cromey, C.J., Callier, M.D., Archambault, P., Chamerlain, J., McKindsey, C.W.
(2009). Shellfish-DEPOMOD: modelling the biodeposition from suspended shellfish
aquaculture and assessing benthic effects. Aquaculture 288, 239-253.
White, I.D., Mottershead, D.N. & Harrison, S.J. (1992) Environmental Systems: an
introductory text. 2 nd edition. Chapman & Hall, London. 632pp.
Wischmeier, W.H. & Smith, D.D. (1978) Predicting rainfall erosion losses: a guide to
conservation planning. United States Department of Agriculture – handbook no. 537. United
States Government Printing Office, Washington DC.
Wool, T.A., Davie, S.R., Rodriguez, H.N. (2003) Development of a three-dimensional
hydrodynamic and Water Quality models to support Total Maximum Daily Load decision
process for the Neuse River Estuary, North Carolina. Journal of Water Resources Planning
and Management 129, 295-306
WWF (2010) WWF-Aquaculture – Aquaculture Dialogues. [Internet] WWF. Available from: <
http://www.worldwildlife.org/what/globalmarkets/aquaculture/aquaculturedialogues.html>
[Accessed: May 2010].
Yeh, S.C., Wang, C.A., Yu, H.C. (2006) Simulation of soil erosion and nutrient impact using an
integrated system dynamics model in a watershed in Taiwan. Environmental Modelling &
Software 21, 937-948
Yue-qing, X., Jian, P. & Xiao-mei, S. (2009) Assessment of soil erosion using RUSLE and GIS: a
case study of the Maotiao River watershed, Guizhou Province, China. Environmental
Geology, 56: 1643-1652.
Zhang, Y., Degroote, J., Wolter, C. & Sugumaran, R. (2009a) Integration of modified
universal soil loss equation (MUSLE) into a GIS framework to assess soil erosion risk. Land
degradation & development, 20: 84 -91.
Zhang, J., Hansen, P.K., Fang, J., Wang, W., Jiang, Z., (2009b). Assessment of the local
environmental impact of intensive marine shellfish and seaweed farming- application of the
MOM system in the Sungo Bay, China, Aquaculture 287, 304-310
Zhang, L., Mitsch, W.J. (2005) Modelling hydrological processes in created freshwater
wetlands: an integrated system approach. Environmental Modelling and Software 20, 935-
946
Page 54