See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/215689861

A Simple Example of How System Dynamics Modeling Can Clarify, and

Improve Discussion and Modification, of Model Structure

Conference Paper · January 1999

CITATIONS READS

13 321

All content following this page was uploaded by Richard George Dudley on 01 June 2014.

The user has requested enhancement of the downloaded file.

 A Simple Example of How System Dynamics Modeling Can Clarify, and
Improve Discussion and Modification, of Model Structure1

Richard G. Dudley2 and Chris S. Soderquist3

The field of system dynamics grew out of engineering feedback control systems
and electronics and, over the past 40 years, use of this approach has spread to a
number of other fields. Within the past 10 years further use of this approach has
been aided by the development of graphical modeling environments such as
Stella, VenSim, and PowerSim. While mathematical and computer modeling
have long been an integral part of fisheries management, use of system dynamics
modeling in fisheries is relatively rare. The biomass dynamic model of Schaefer
is a simple model familiar to fisheries professionals, and forms the basis of
several similar models. It provides a simple example of how system dynamics
modeling encourages an explicit representation of model structure, and mediates
discussion and modification of a model.

Introductory Comments
Ecological and natural resource systems have been subjects of interest for fifty years or more.
Over this same period techniques for the analysis of system structure and dynamics have been
refined. Approaches for the study of systems emerged as a distinct field within engineering:
system dynamics. Subsequently system dynamics was applied to management science and
other fields. Most system dynamics workers recognize Jay W. Forrester as the father of
system dynamics and his classic, Industrial Dynamics (Forrester 1961), was probably the first
highly detailed application of system dynamics techniques to non-engineering problems. This
was later followed by Urban Dynamics (Forrester 1969) and by World Dynamics (Forrester
1971) which was a precursor to the well known Limits to Growth models (Meadows et al
1972)4. These and other works helped to establish the idea that system dynamics modeling
not only helps us describe and understand systems, but can be useful in exploring possible
scenarios to solve complex real world problems, including those involving human behavior
and soft variables.

Various systems approaches have been used in ecology and natural resource management for
many years (e.g. Watt 1968, Patten 1971, Holling 1978, Walters 1986). The need for, and
utility of, these approaches have recently been summarized by Grant (1998). Academic
interest in systems approaches to natural resource management seems inconsistent with the
relative scarcity of actual application of these approaches for solving fishery management
problems. Increased application of the relatively standardized, simple, yet rigorous approach
of Forrester may help alleviate this scarcity. Some recent examples of application of the
system dynamics modeling approach to fisheries management are Ruth and Lindholm (1996)
and Holland and Brazee (1996).

The importance of improving our ability to manage dynamic natural resource systems has
been recently pointed out by Moxnes (1998a, b). In simulation settings he found that would-
be managers, including those well trained in population dynamics, consistently overharvested
a model stock even when given cash rewards for proper management.

In some senses system dynamics can be viewed as a quasi- standardized framework (i.e. the
stock flow modeling paradigm and associated rules) within which systems, and the human

1
For presentation at the 129th Annual Meeting of the American Fisheries Society, Charlotte,
North Carolina. August 1999.
2
Jalan Pangrango 6, Bogor, Indonesia. [email protected]
3
One Smith Road, Hanover, NH 03755 Chris.S.Soderquist@@pontifexconsulting.com
4
Which, more recently, was followed up with Beyond the Limits (Meadows et al 1992).
policies that drive many of them, can be rigorously examined. By stretching this point a bit
we could say that, just as statistics provides a framework for examining data, system
dynamics could provide a framework for examining systems.

Forrester's system dynamics approach utilizes stock and flow modeling which follows
variables over time. Calculations are accomplished via numerical integration over short
solution intervals (e.g. ∆t) which requires computing power and appropriate software. During
the 60's and 70's the software used was usually DYNAMO. During the 1980's system
dynamics software developers took advantage of the new computing capabilities. Current
software allows onscreen creation of the model structure of stocks, flows and auxiliary
variables, and has convenient user interfaces allowing detailed examination of very complex
model structures and output.5 These improvements have drastically improved the
accessibility of this modeling approach.

An Example Using the Schaefer Model

The Schaefer biomass dynamic model provides a simple example, with which we are all
familiar, of how a system dynamics modeling approach can help to:

1) clarify the structure of a model and

2) lead to a clearer discussion of possible model modifications.

The typical formulation of the Schaefer model (Schaefer 1954) equates the rate of change of
population biomass to inflows of biomass minus biomass outflows. It is typically presented
as follows:

dB  B
= rB1 −  − qEB
dt  k

Where B is population biomass, r is intrinsic rate of increase, k is biomass of the un-fished

stock. The expression qEB (often presented as C) is the catch where q is the fraction of the
stock caught by one unit of effort and E is the number of units of effort.

But even if presented as follows: dB rB 2

= rB − − qEB
dt k

this simplest of models is not something we could present profitably in a public forum for
discussion.

This illustrates one way in which system dynamics modeling differs from many other forms
of modeling: the system dynamics modeler generally tries to formulate each component of a
model separately, defines the structure linking components and then "lets the software do the
talking". Mathematical modelers, on the other hand strive to develop one summary "elegant"
equation that will calculate the answer for a particular set of inputs. Other forms of modeling
are, perhaps, somewhere in between. While other forms of computer modeling appear to

5
Some examples of such software are:
Stella - High Performance Systems, Inc., 45 Lyme Road, Suite 200, Hanover, NH 03755,
USA. http://www.hps-inc.com
VenSim - Ventana Systems, Inc., 60 Jacob Gates Road, Harvard MA 01451, USA.
http://www.vensim.com
PowerSim - Powersim Corporation, 1175 Herndon Parkway, Suite 600, Herndon, VA 20170
http://www.powersim.com/

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 2

parallel the above approach, they emphasize step by step programming which often takes
precedence over a careful examination of system structure.

We have all seen the above equation, and probably have examined the underlying logic. We
recall that the rate of increase with respect to time is equal to biomass times the constant r.
That -qEB represents the removal of biomass as catch is also obvious. But why is that extra r
in the outflow side, and what exactly is the meaning of B2?

Presented in a typical system dynamics stock and flow diagram (Figure 1) the biomass
dynamic model seems much clearer and the mathematical formulation of each component is
explicitly stated.6

In answer to the questions posed above, firstly we can clearly see that r, the rate of increase,
also equals the rate of decrease in the absence of fishing. That is, the inflow to biomass must
equal normal outflow when there is no catch.
UNITS OF
FISHING GEAR
GEAR E EFFICIENCY q
catch fraction

Current Fish catch C

additions Biomass B

deaths

RATE OF INCREASE r

ratio of current
biomass to
unfished NORMAL
biomass DEATH
FRACTION

BIOM ASS OF
UNFISHED
STOCK k <RATE OF INCREASE r>

Figure 1. A system dynamics model formulation (here the Schaefer biomass dynamic model) facilitates
examination of model structure and assumptions. The rectangle labeled Current Fish Biomass B is a
stock, or accumulation. Metaphorically, it is similar to a bathtub containing all the biomass. The thick
arrows are flows. Whatever accumulates in the bathtub flows in and out via these pipes. For example, the
"additions" flow adds biomass to the stock by flowing into the Current Fish Biomass B stock at a rate of
some amount of biomass per unit time (where time is the chosen time interval, ∆t).

To someone not familiar with the model it becomes clearer that in this model we have
assumed a proportional effect of fish biomass on deaths. That is , we have assumed that the
effect of current biomass on death fraction is equal to the biomass ratio B/k. In other words
we assume that natural death of biomass from the population is equal to the normal death
fraction r multiplied by B further multiplied by B/k.
rB 2
So the origin of becomes apparent as
k
B
"normal deaths" rB times the "effect of biomass on death" .
k
6
See page 13 for model equations.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 3

Of course this model was originally developed to calculate equilibrium yields under certain
conditions. System dynamics models, on the other hand, are typically used to gain an
understanding of system behavior over time. Looking at this familiar model with a system
dynamics approach helps us understand its structure better. A selection of fishing pressures
for the above model result in the following graph (Figure 2). Note that the traditional
parabolic curve of equilibrium yield vs. biomass can also be produced from the various
equilibrium yields and corresponding biomass levels (Figure 3).

Catch Over Time at Six Fishing Levels

20,000

15,000

5
10,000

4 56
3
4 5
3 4 6
3 45 45
5,000 3 4 4
2 6 3 3 5 3 5 34 3
4
34 34 34 34 3
2 6 5 5 5
2 2 5 5 5
2 2 6 2 2 2 2 2 2 2 2
6
6 6 6 6
1 1 1 1 1 1 1 1 1 1 1 1 61 61

0
0 5 10 15 20 25 30 35 40 45 50
Time (Year)

catch C : 20 Units 1 1 1 1 1 1 1 1 1 1 1 kg/Year

catch C : 50 Units 2 2 2 2 2 2 2 2 2 2 2 kg/Year
catch C : 80 Units 3 3 3 3 3 3 3 3 3 3 3 kg/Year
catch C : 100 Units 4 4 4 4 4 4 4 4 4 4 4 kg/Year
catch C : 150 Units 5 5 5 5 5 5 5 5 5 5 5 kg/Year
catch C : 200 Units 6 6 6 6 6 6 6 6 6 6 6 kg/Year

Figure 2. Example output from the Schaefer model for 6 fishing levels emphasizes the fact that system
dynamics models typically follow model variables over time.

This is a predictable result: the model is not dynamic, the rates do not change (except when
the user changes them). As expected, for each selected "units of fishing gear" an equilibrium
yield is reached, and as expected there is a maximum equilibrium yield obtained when the
stock biomass is 50% of the un-fished biomass.

While this result is not particularly surprising, it nevertheless, leads us to three important
points about this modeling approach:

1) Current system dynamics modeling tools present a format that encourages

discussion and suggestions for model construction both from specialists and non-
specialists. This is because:
(a) The visual layout encourages an overall understanding of the system
being modeled.
(b) Each component is clearly identified.
(c) The relationships among components are clearly specified.
(d) Programming is handled by the software, leaving the developer to work
on model, rather than programming, logic.

2) System dynamics modeling techniques, which use computer-based numerical

integration, allow us to easily examine complex systems that do not have explicit
mathematical solutions.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 4

 3) Consequently, such models are easily modified to examine both changes in model
parameters as well as changes to model structure.
20,000
Time lines of
Biomass vs. Catch for
different numbers of
fishing units
15,000 Time

Catch

10,000
150

200 units 100

80
5,000
50

Equilibrium catch
endpoints at each fishing level 10
0
0 25000 50000 75000 100000
Current Fish Biomass

Figure 3. Of course, data can also be presented in other formats. Here, time lines for each fishing
level terminate at each equilibrium catch. The traditional format of the model illustrates that
equilibrium catch is maximized at an intermediate biomass.

A model is comprised of cause-and-effect assumptions. Pure mathematical models are

difficult to interpret regarding what particular assumptions have been made in their
formulation. System dynamics models make those assumptions visible. Looking at the model
again we can see several areas for possible improvement that might not be obvious in the
mathematical formulation. In a classroom situation, for example, it might be possible to ask
students for suggestions for improving the model, which will also encourage thought about
how fishery management works.

One might ask, for example:

1) Might biomass ratio also have an effect on inflow to the stock? Perhaps as the
stock drops further from k then the intrinsic rate of increase would increase? Or
is the density dependency of the deaths a sufficient description of the system?

2) In the model the number of fishing units is fixed by the model operator. What
happens in the real world? Perhaps number of fishing units could be modeled as
being dependent on catch per unit?

3) How might one model agency decisions about numbers of fishing permits to be
issued in a limited entry fishery?

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 5

Model Modification 1: Catch per Unit Effort Affects Vessel Entry and Exit from the
Fishery

AVERAGE
VESSEL LIFE
SPAN IN FLEET

Units of
Fishing
entering or leaving fleet Gear E retiring from fleet
change efficiency
NORMAL RATE
OF VESSEL
ENTRY effect of cpue GEAR
EFFICIENCY
on vessel q
entry
catch fraction
cpue
look up table
cpue ratio

NORMAL
CPUE

Current Fish catch C

additions Biomass B

deaths

RATE OF INCREASE r

effect of
BIOMASS OF relative NORMAL
UNFISHED biomass on DEATH
STOCK k death fraction FRACTION

<RATE OF INCREASE r>

Figure 4. A modification of the original model to reformulate "Units of fishing Gear E" as
a stock which grows more rapidly if cpue (catch per unit effort) is high.

This second suggestion (above) could be formulated as presented in Figure 4.7 Here we have
substituted a small sub-model in place of "units of fishing gear" which appeared in the
original model. In this sub-model "units of fishing gear" has been made dependent on catch
per unit effort (cpue). While we could use the actual ratio (of current cpue to that at
maximum equilibrium) or some multiplier, we chose instead to use a "lookup table or graph
input" which is presented in Figure 5.

Such graphs can be adjusted by the model user to indicate how the user thinks cpue (for
example) might affect the entry of new units of fishing gear into the fishery. This can make
the model clearer and facilitates increased participation from non-specialists. The lookup
table also reminds us that system dynamics modeling frees us from reliance on
mathematically tractable formulations.

7
See page 15 for model equations.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 6

Of course, in addition to the insights gained from building and modifying models, there is the
usual benefit to be gained from examining the consequences of changing model parameters.
As an example we have presented some scenarios using the modified model.
p
10 1

Lookup table providing relationship

of cpue ratio to effect of cpue ratio
1
on vessel entry
5 1
effect of cpue on
vessel entry

1
1
1
0 1

cpue ratio below about 0.5 will

cause additional vessels to leave
1
-5
the fishery

-10
0 1 2 3 4
cpue ratio
Figure 5. A lookup "table" showing the proposed relation between "cpue ratio" (x) and "effect of
cpue on vessel entry " (y). As with most lookup functions the point (1,1) defines the normal value; in
this case where there is no additional effect of cpue. Explicit statements regarding how each model
component is calculated allow improved discussion of the model by all stakeholders. Values of points
in the lookup function are provided in the model equation for "look up table " on page 15.

A) Baseline Scenario B) Start with One Fishing Unit and Virgin Stock
200 units
10,000 kg/Year
100,000 kg 6
4

150 units 4
7,500 kg/Year
75,000 kg 6
2 2 4
2
6 2
100 units 4 4
2
5,000 kg/Year 1 3 5 12 3 5 1 3 5 1 3 5 61 3 5 1 34 5 1 3 5 1 23 5 12 3 5 61 23 5 6 12 34 56 123 45 61 2
4 6 4 6 4 4
6 6
50,000 kg

50 units
2,500 kg/Year
25,000 kg

0 units 2
0 kg/Year
0 kg
0 5 10 15 20 25 30 35 40 45 50
Time (Year)

Units of Fishing Gear E : S2 baseline 1 1 1 1 1 1 1 1 1 units

Units of Fishing Gear E : S2 starting with one unit of gear2 2 2 2 2 2 2 2 units
catch C : S2 baseline 3 3 3 3 3 3 3 3 3 3 3 kg/Year
catch C : S2 starting with one unit of gear 4 4 4 4 4 4 4 4 4 kg/Year
Current Fish Biomass B : S2 baseline 5 5 5 5 5 5 5 5 5 5 kg
Current Fish Biomass B : S2 starting with one unit of gear 6 6 6 6 6 6 6 6 kg

Figure 6. Two scenarios of the modified model showing the "baseline" (the straight lines), and a
scenario starting with one boat and a virgin stock of 100,000 kg. In the latter case the model will (as
currently structured) ultimately stabilize at the maximum equilibrium yield but only after overshooting
this goal by a considerable amount in terms of both catch and numbers of gear units.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 7

 Scenario Where Gear Effectiveness Increases by 50% Between Years 10 and 15
100 kg/(units*Year)
8,000 kg/Year
200 units 2
100,000 kg

2
2

3 3 2
1 1 3
50 kg/(units*Year)
3 3
5,000 kg/Year 3
1 2 34 1 2 34 1 2 34 1 2 34 1
100 units 4 2 3 1 1 1
1 3 1 3
50,000 kg 1 1
3 1 3 3 3
1 1
4
2 2
2
4 2 2
2 4
4 2 2 2 4 4 4
4 4 4 4 4

0 kg/(units*Year)
2,000 kg/Year
0 units
0 kg
0 5 10 15 20 25 30 35 40 45 50
Time (Year)

cpue 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 kg/(units*Year)
Catch C 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 kg/Year
Units of Fishing Gear 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 units
Current Fish Biomass 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 kg

Figure 7. This figure illustrates a scenario where the stock in in equilibrium with 50 vessels. Between
years 10 and 15 the effectiveness of the gear gradually increases by 50%. Here catch and number of
vessels initially increase, but of course eventually drop below the original value and eventually
approach a new equilibrium.

The baseline scenario uses equilibrium baseline values where the model is started with 100
vessels, and a stock size of 50,000 kg. This results in the straight lines shown in Figure 6.
Also presented in Figure 6 is a scenario of a new fishery where the model starts with one
vessel and a virgin stock. In this case the fishery grows beyond its maximum capacity and
later stabilizes.

In Figure 7 is another scenario, where the fish stock begins in equilibrium, but then, between
years 10 and 15, the gear effectiveness is increased by a total of 50%. As cpue rises more
vessels enter the fishery which, after a short peak, drive down the cpue, the overall catch, and
the fish population.

Even this simple model can be useful in gaining an understanding of how a fishery works. A
number of different scenarios could be investigated, and model structure could be modified.
How would product price effect the entry of vessels into the fishery? What about competing
products? What about other factors causing fluctuations in stock size?

Model Modification 2: Decisions by a Fishery Management Entity

Rather than having the number of vessels entering and leaving the fishery determined by the
cpue, we could assume that a management agency compares the actual stock to optimal stock
density and adjusts the number of units in the fishery accordingly. Such a model might be
formulated as in Figure 8.

In this formulation we assume that data are collected about the stock size which is compared
to the optimum stock size (or goal) which, in the basic model, is assumed to be 0.5 of the
maximum stock size. We assume that the agency considers changes as they occur and forms a
perception of the stock's status. This perception is the basis of new decisions about numbers

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 8

 TIME TO
IMPLEMENT
UNITS OF
Units of FISHING
GEAR E GEAR
Fishing Gear catch fraction EFFICIENCY q
changing
fishing gear
numbers

Current Fish catch C

proposed Biomass B
change in gear additions
numbers
deaths

effect of perception
on fishing gear RATE OF INCREASE r
numbers
ratio of current
biomass to
perception unfished NORMAL
vs decision biomass DEATH
lookup FRACTION

BIOMASS OF
UNFISHED
STOCK k
<RATE OF INCREASE r>
difference
in status
perception

latest perception
of fishery status
Current change in
Perception of perception
Fishery by change in perception lookup
Management
TIME NEEDED
TO CHANGE
PERCEPTION

Figure 8. Additions (shown in bold) to the original model simulate adjustment to fishing
gear numbers by a management entity. In this case the management entity reviews population
status and then makes appropriate adjustments in fishing gear numbers.

of fishing units, and these decisions are eventually implemented. Two stocks are added to
the model: one specifying the "current perception by fishery management" and the other
specifying the current "units of fishing gear". The times over which perception and the units
of fishing gear gradually change are also specified.

In this model there are two look-up functions which are presented in the equations on page 16
and graphically on page 18.

If this model is started with one fishing gear unit and a virgin stock the scenario evolves as
presented in Figure 9.

A second scenario, where fishing gear effectiveness is gradually increased 50% between years
10 and 15, is presented in Figure 10. (This is the same scenario that was used in the first
model modification presented in Figure 7). As larger amounts of fish are removed from the
population by the improved gear the population ratio decreases and management adjusts the
fishing gear numbers accordingly. Unlike the first model modification, this structure
eventually returns the stock to the point of maximum equilibrium yield. Even with such
constant adjustment, however, the population takes many years to return to equilibrium.

One might feel that a faster response time by the management entity in changing its
perception and in implementing the new fishing gear numbers would improve the response of

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 9

the system. As indicated in Figure 11 this is only partially true. Shortening the two
management response times increases the amplitude of the damped oscillations that are
inherent in the system. Although these keep the stock nearer to the target level, an
unexpected consequence is that larger fluctuations in fishing gear numbers also result.

Conclusions

Our point here is not that these models are particularly appropriate or special. In fact they are
merely simple examples. Our point is that the system dynamics approach is simple and
convenient for looking at a variety of complex issues facing fishery scientists. When used for
this purpose unexpected consequences of our decisions become clear, and reasons for these
consequences are revealed. The system dynamics approach provides a somewhat
standardized framework for developing models of many types. The approach is not limited to
biological systems and is ideal for interdisciplinary teamwork. Several authors have provided
guidelines for developing such models in group environments (e.g. Andersen and Richardson
1994, Richardson and Andersen 1995).

In today's world fishery management is often carried out by agencies with a mandate to obtain
and use input from a variety of stakeholders. Interested elements of the public and special
interest groups insist on having input into the fishery management decision making.
Enmeshed in this reality we have a real need to develop mechanisms which allow these
interest groups or stakeholders to participate in the overall management scheme, even when
this scheme is complex. System dynamics modeling provides one such mechanism where
modeling can be made more transparent to a variety of stakeholders, and will permit the
incorporation of their ideas and desires into management planning and execution.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 10

 Response of the Managed Fishery When it is Started with a Virgin Stock and One Unit of Gear
4
200 kg/(units*Year)
4
200 units
8,000 kg/Year
100,000 kg
4

3
3

3
4 3
3 3 3 3 3 3 3 3 3
3

100 kg/(units*Year) 3 4
2 2 2
100 units 2 4
2 2 2 4 4 4 4
1 1 4 4 2 4 2 4 2 2 2 2 2
4,000 kg/Year 1
4 4 4
50,000 kg
2
1

1
1
1 1 1 1 1 1 1 1 1 1 1 1
2
3
0 kg/(units*Year)
0 units
0 kg/Year 2
0 kg 2
3

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Time (Year)

cpue : start with one unit of gear 1 1 1 1 1 1 1 1 1 1 1 1 1 1 kg/(units*Year)

UNITS OF FISHING GEAR E : start with one unit of gear 2 2 2 2 2 2 2 2 2 2 2 units
catch C : start with one unit of gear 3 3 3 3 3 3 3 3 3 3 3 3 3 3 kg/Year
Current Fish Biomass B : start with one unit of gear 4 4 4 4 4 4 4 4 4 4 4 4 4 kg

Figure 9. When started with a single unit of gear and a virgin stock (100,000 kg) the
Managed Fishery formulation of the model approaches the expected equilibrium with only
minimal overshoot of the ultimate number of fishing gear (line 2).

Response of the Managed Fishery to a 50% Increase in Gear Efficiency During Years 10 to 15
100 kg/(units*Year)
8,000 kg/Year
200 units
100,000 kg

1 1 1 1 1 1
1
1
2 1
1
2
1
1
1
50 kg/(units*Year) 1 1
1
5,000 kg/Year 4 4 4 4
2
4
2 4 2 4 2
12341234 3 4 2
100 units 3 2
4 2
2
50,000 kg 4 3 2
4 2
23 4 2
4 4 23 2
4
3
3 3 3 3 3 3 3 3 3

0 kg/(units*Year)
2,000 kg/Year
0 units
0 kg
0 10 20 30 40 50 60 70 80 90 100
Time (Year)

cpue : S3 50% increase in efficiency years 10 to 15 1 1 1 1 1 1 1 1 1 1 1 kg/(units*Year)

catch C : S3 50% increase in efficiency years 10 to 15 2 2 2 2 2 2 2 2 2 2 2 2 kg/Year
Units of Fishing Gear : S3 50% increase in efficiency years 10 to 15
3 3 3 3 3 3 3 3 3 3 3 units
Current Fish Biomass B : S3 50% increase in efficiency years 10 to 15 4 4 4 4 4 4 4 4 4 4 kg

Figure 10. The represents the response of the managed fishery when fishing gear
effectiveness increases by 50% over a five year period. Even though management regularly
monitors the population and adjusts fishing gear numbers, biomass and catch return to their
original maximum equilibrium values only after a considerable delay.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 11

 The Effect of Altering Management Response Time
12312312312 1
100 2 1
3
1 Units of Fishing Gear
2
1

85 2 1
3

1
2
1
3 3

70 1 3
2
3 3 3 3
1 3 3 2 23 2 2 2
3 3 2 2 3 3 3
2 1 2 1 1 1
1 3 3 2 1 1 1
1 3 2 1 1 1
2 12 1 1
3 2
2 2 2 2

55 3
3
3

40
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Time (Year)
60,000
Current Fish Biomass
3 3
3 2 2 2 2 2 1 1 1 1
3 2 1 12 2 1 1 1
2
3 2 31 3 3 3 2 2
50,000 1231231231 31 3 3 3 3 23 23 23
2 3 31
2 3
3
3 1
2
1 1
3 2
23 1
2 1
40,000 12
2
2 1
1 1
1 1

Line 1: baseline
30,000 Line 2: times to act are 0.5 years
Line 3: times to act are 0.1 years

20,000
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Time (Year)

Figure 11. The effect of changing the management response times: "time needed to change
perception", and "time to implement." In the baseline scenario these are 2 years and 1 year
respectively. In the test scenarios these are both set to 0.5 years (line 2) and 0.1 years (line
3). A more rapid response time maintains the current fish biomass nearer to the target level,
but induces unexpected larger oscillations in units of fishing gear.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 12

Literature Cited

Andersen, David F., Richardson, George P. 1994. Scripts for Group Model Building.
International System Dynamics Conference Sterling, Scotland.
Forrester, Jay W. 1961. Industrial Dynamics. Portland, OR: Productivity Press. 464 pp.
Forrester, Jay W. 1969. Urban Dynamics. Portland, OR: Productivity Press. 285 pp.
Forrester, Jay W. 1971. World Dynamics. Portland, OR: Productivity Press. 144 pp.
Grant, W. E. 1998. Ecology and natural resource management: reflections from a systems
perspective. Ecological Modelling 108(1998) 67-76.
Holland, Daniel S. and Richard J. Brazee. 1996. Marine Reserves for Fisheries Management.
Marine Resource Economics. 11(3):157-171.
Holling, C.S. 1978. Adaptive environmental assessment and management. Wiley, New York,
377 pp.
Meadows, Donella H., Dennis L. Meadows, Jorgen Randers, and William W. Behrens III.
1972. The Limits to Growth. New York: Universe Books. 205 pp.
Meadows, Donella H., Dennis L. Meadows, Jorgen Randers, 1992. Beyond the Limits:
Confronting Global Collapse, Envisioning a Sustainable Future. Post Mills, VT: Chelsea
Green Publishing Co., 300 pp.
Moxnes, E. 1998a. Not only the tragedy of the commons, misperceptions
of bioeconomics. Management Science 44(9):1234-1248.
Moxnes, E. 1998b. Overexploitation of renewable resources: The role of
misperceptions. Journal of Economic Behavior and Organization 37(1):107-127.
Patten, B.C. (Ed) 1971. Systems analysis and simulation in ecology. Academic Press, New
York. 607 pp.
Richardson, George P., and Andersen, David F. 1995. Teamwork in Group Model Building.
System Dynamics Review 2(11) 113-137.
Ruth, Matthias, and James Lindholm. 1996. Dynamic modelling of multispecies fisheries for
consensus building and management. Environmental Conservation 23(4): 332-342.
Schaefer, M. B. 1954. Some aspects of the dynamics of populations important to the
management of commercial marine fisheries. Bull. Inter-Am. Trop. Tuna Comm. 1:27-56.
Walters, C.W. 1986. Adaptive management of renewable resources. Macmillan, New York,
374 pp.
Watt, K.E.F. 1968. Ecology and resource management. McGraw-Hill, New York, 450 p.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 13

Formulation of Schaefer Model (alphabetical order by component name):

(01) additions = Current Fish Biomass B*RATE OF INCREASE r

Units: kg/Year
Amount of biomass added to the population each year.

(02) BIOMASS OF UNFISHED STOCK k= 100000

Units: kg
Biomass that the stock has when no fishing takes place.

(03) catch C= catch fraction*Current Fish Biomass B

Units: kg/Year

(04) catch fraction = GEAR EFFICIENCY q*UNITS OF FISHING GEAR E

Units: 1/Year
Fraction of the current fish biomass caught.

(05) Current Fish Biomass B= INTEG (additions-catch C-deaths, 100000)

Units: kg
The current biomass of the fish population.

(06) deaths = Current Fish Biomass B*NORMAL DEATH FRACTION *

ratio of current biomass to unfished biomass
Units: kg/Year
Biomass of fish dying.

(07) FINAL TIME = 50

Units: Year
The final time for the simulation.

(08) GEAR EFFICIENCY q= 0.001

Units: 1/units/Year
Fraction of the current fish biomass caught by each fishing gear unit.

(09) INITIAL TIME = 0

Units: Year
The initial time for the simulation.

(10) NORMAL DEATH FRACTION = RATE OF INCREASE r

Units: 1/Year
This is the normal death fraction if the current fish biomass is equal to the biomass of
the un-fished stock. In that case it is the same as the rate of increase r.

(11) RATE OF INCREASE r = 0.2

Units: 1/Year
The intrinsic rate of growth in biomass.

(12) ratio of current biomass to unfished biomass =

Current Fish Biomass B/BIOMASS OF UNFISHED STOCK k
Units: dmnl
The effect of current fish biomass on the death fraction is the ratio of the current fish
biomass to the biomass of the stock of fish if no fishing too place.

(13) SAVEPER = TIME STEP

Units: Year

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 14

 The frequency with which output is stored.

(14) TIME STEP = 0.125

Units: Year
The time step for the simulation.

(15) UNITS OF FISHING GEAR E = 0

Units: units
Number of fishing gear units. Number can be specified by the user.

Components added for model where cpue affects fishing gear entry and exit

(01) AVERAGE VESSEL LIFE SPAN IN FLEET = 10

Units: Year
Life-time of vessels in the fishery.

(02) change efficiency = 0

Units: dmnl
An added convenience to turn on a change in gear efficiency. 0 is off, 1 is on.

(03) cpue = catch C/Units of Fishing Gear E

Units: kg/Year/units
This is the catch obtained by each unit of fishing gear.

(04) cpue ratio = cpue/NORMAL CPUE

Units: dmnl
The ratio of the current cpue to the normal cpue.

(05) effect of cpue on vessel entry = look up table(cpue ratio)

Units: dmnl
The effect that the cpue ratio has on the entry of gear units into the fishery.
When this ratio is 1 there is no effect.

(06) entering or leaving fleet = NORMAL RATE OF VESSEL ENTRY *

effect of cpue on vessel entry
Units: units/Year
number of vessels (or gear units) entering the fishery.

(07) GEAR EFFICIENCY q = 0.001+change efficiency*RAMP(2.5e-005, 10, 30)

Units: 1/units/Year
Fraction of the current fish biomass caught by each fishing gear unit. The ramp
function is used to examine different scenarios.

(08) look up table([(0,-10)-(4,10)],(0,-10),(0.108761,-4.91228), (0.265861,-1.84211),

(0.44713,-0.0877193), (0.700906,0.614035), (1,1),(1.32931,1.57895),
(1.90937,2.7193), (2.67069,5),(3.28701,6.92982),(4,10))
Units: dmnl
This is a graphical function which takes as x the cpue ratio and
produces as y the effect on gear entry into the fishery.

(09) NORMAL CPUE = 50

Units: kg/(Year*units)
This the the catch obtained by each unit of fishing gear under normal conditions.
Which here have been defined as the catch per unit obtained at maximum

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 15

 equilibrium yield with 50 units fishing.

(10) NORMAL RATE OF VESSEL ENTRY = 10

Units: units/Year
This is the normal rate of entry at start of the model (at max equilibrium
yield in this case).

(11) retiring from fleet = Units of Fishing Gear E /

AVERAGE VESSEL LIFE SPAN IN FLEET
Units: units/Year
Number of vessels retiring from the fishery.

(12) Units of Fishing Gear E= INTEG (entering or leaving fleet-retiring from fleet, 1)
Units: units
Number of units of fishing gear in the fishery.

Components added for model where a management entity determines fishing gear
numbers.

(01) change in perception = difference in status perception /

TIME NEEDED TO CHANGE PERCEPTION
Units: dmnl/Year
Changing perception of the fishery

(02) change in perception lookup([(0,-10)-(1,10)],(0,-10), (0.0996979,-5),

(0.247734,-2.19298),(0.338369,-0.789474),(0.425982,-0.263158),(0.5,0),
(0.58006,0.263158),(0.655589,0.789474),(0.752266,1.92982),
(0.897281,4.91228),(1,10))
Units: dmnl
A graphical lookup function describing the relationship between biomass ratio and
perception of fishery status.

(03) changing fishing gear numbers = proposed change in gear numbers/TIME TO

IMPLEMENT
Units: units/Year
Change in fishing gear numbers.

(04) cpue = catch C/UNITS OF FISHING GEAR E

Units: kg/units/Year
Catch fishing gear unit (note: not shown on diagram)

(05) Current Perception of Fishery by Management = INTEG (change in perception, 0)

Units: dmnl
The currently held perception of the fishery by the management entity.

(06) difference in status perception = latest perception of fishery status-Current Perception

of Fishery by Management
Units: dmnl
The difference between the current perception of the fishery and the new perception.

(07) effect of perception on fishing gear numbers =

perception vs decision lookup(Current Perception of Fishery by Management)
Units: dmnl
The effect that the management entity's perception has on fishing gear numbers.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 16

(08) latest perception of fishery status =
change in perception lookup(ratio of current biomass to unfished biomass)
Units: dmnl
The management entity's perception of fishery status as based on biomass ratio.

(09) perception vs decision lookup([(-10,0)-(10,2)],(-10,0.1),(-8.54985,0.482456),

(-6.85801,0.745614),(-4.98489,0.868421), (0,1),(5,1.1),(7,1.15),(10,1.25))
Units: dmnl
A graphical function describing the relationship between the current perception of the
fishery and the effect on fishing gear numbers.

(10) proposed change in gear numbers= (Units of Fishing Gear * effect of perception on
fishing gear numbers)-Units of Fishing Gear
Units: units
The change in fishing gear numbers proposed by the management entity.

(11) TIME NEEDED TO CHANGE PERCEPTION = 2

Units: Year
The time needed for the management entity to change its perception of the fishery
status.

(12) TIME TO IMPLEMENT = 1

Units: Year
Time needed to implement the new fishing gear numbers.

(13) Units of Fishing Gear= INTEG (changing fishing gear numbers, INITIAL
NUMBERS OF GEAR)
Units: units
Number of units of fishing gear in the fishery.

(14) UNITS OF FISHING GEAR E = Units of Fishing Gear

Units: units
Number of fishing gear units.

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 17

 10 1

Lookup function describing the effect of biomass

ratio on latest perception of fishery status
latest perception of fishery
5 1

1
1
0 1
1
1

-5 1

-10
0 0.25 0.5 0.75 1

biomass ratio

2
Lookup function describing the effect of
current perception of fishery status
on
1.5
effect of perception on fishing gear numbers
effect of perception on
fishing gear numbers

1
1

1 1

0.5 1

0
-10 -5 0 5 10

current perception of fishery status

A SIMPLE EXAMPLE DUDLEY & SODERQUIST PAGE 18

View publication stats