Ecological Economics 106 (2014) 111

Contents lists available at ScienceDirect

Ecological Economics
journal homepage: www.elsevier.com/locate/ecolecon

Analysis

Sustainable harvest of a native species and control of an invasive species:

A bioeconomic model of a commercial shery invaded by a
space competitor
Marjolaine Frsard a,, Carole Ropars-Collet b,c
a
b
c

Universit de Brest, UEB, UMR AMURE, 12 rue de Kergoat, CS 93837, 29238 Brest Cedex3, France
Agrocampus Ouest, UMR1302, F-35000 Rennes, France
INRA, UMR1302, F-35000 Rennes, France

a r t i c l e

i n f o

Article history:
Received 2 December 2009
Received in revised form 28 March 2014
Accepted 29 June 2014
Available online 28 July 2014
Keywords:
Biological invasion
Space competition
Bioeconomics
Optimal control
Scallop shery

a b s t r a c t
Biological invasions are nowadays an important challenge to biodiversity and human welfare. This paper deals
with the control of an invasive species, void of market value, and acting as a space competitor for a valuable native
harvested species. It presents a theoretical bioeconomic model describing the interacting dynamics of the two
species and accounting for the undesirable consequences of native stock harvesters' behaviour on the spread
of invasion. Dynamic optimisation of the model displays the existence of a time-path leading to an optimal
stationary steady-state solution where the native species is sustainably harvested and the invasive species is
kept under control, provided unit costs of native species harvesting and of invaded areas cleaning are quite
low, natural and anthropogenic dispersal coefcients of invasion, and time-discount rate are moderate.
Moreover, the problem should be addressed early enough. The model is applied to the Bay of Saint-Brieuc scallop
shery invaded by slipper-limpet. We show that it is nearly always optimal to control the invasion in that case
study.
2014 Elsevier B.V. All rights reserved.

1. Introduction
Biological invasions are nowadays an important challenge to biodiversity and human welfare (Convention on Biological Diversity, article
8h, 1992). The worldwide spread of invasive alien species has increased
in the second part of the 20th century (Mack et al., 2000; Mooney and
Cleland, 2001; Williamson, 1996) and is now considered as one of the
most serious threats to biodiversity (Didham et al., 2005; Lvei, 1997;
Wilcove et al., 1998). This evolution is mainly due to human activities
(Carlton, 1987; Carlton and Geller, 1993; Vitousek et al., 1997), particularly trade (Carlton, 1989; Krcmar-Nozic et al., 2000; Ruiz and Carlton,
2003), and human-induced disturbance of ecosystems (Dalmazzone,
2000; Williamson, 1996, 1999). By altering biodiversity and the services
it supports, biological invasions may impose signicant cost in terms of
forgone output and ecosystem services (Perrings et al., 2000, 2002).
Thus, in the Black Sea the cost of Mnemiopsis leidyi invasion has been estimated at about $16.8 million per year (Knowler, 2005; Knowler and
Barbier, 2000), and biological invasion damage and control costs in
the USA have been evaluated to $128 billion per year (Pimentel et al.,
2005).

Corresponding author. Tel.: +33 298 641 934; fax: +33 298 016 935.
E-mail address: marjolaine.fresard@univ-brest.fr (M. Frsard).

http://dx.doi.org/10.1016/j.ecolecon.2014.06.020
0921-8009/ 2014 Elsevier B.V. All rights reserved.

Mostly, biological invasions management, which encompasses both

prevention and control, is considered as a public good (Perrings et al.,
2002) and thus may requires public policies. In this paper, we focus on
the control aspect of such policies in the case of a particular interspecic
competition, exerted by an invasive alien species on a native one, explicitly modelled. This paper deals with the control of an invasive species,
void of market value, and acting as a space competitor for a native
valuable harvested species. It presents a theoretical bioeconomic
model describing the interacting dynamics of the two species and accounting for the undesirable consequences of native stock harvesters'
behaviour on the spread of invasion. The aim of the paper is to describe
the impact of invasion on the native species harvest, through space
competition, and to display the possibilities of sustainable harvest of
the native species by controlling the invaded areas. The analysis is
derived from a real case study: the invasion of the Bay of Saint-Brieuc
scallop shery by slipper limpets.
Most optimal control models of biological invasions usually
relate the damage caused by the invasion to the invasive stock size
(Eiswerth and Johnson, 2002; Junqueira-Lopes et al., 1996; and others).
In those models, the state variable is the invasive stock. The control variable is the effort of control applied to the invasive species. The objective
of the social planner is to minimise the discounted ow of damage and
control costs. Wilman (1996) studies explicitly the interaction between
native and invasive species dynamics by combining the dynamics of a

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

native untargeted valuable species and an invasive species, acting as a

predator. In her model, state variables are the native and invasive stocks,
and control variables and target-functions are the same as previous. Our
model adopts the same approach with two state variables (the native
and invasive species), but (i) we assume a space competition relationship between species, (ii) the control variables are the levels of harvesting effort of each species, and (iii) the target-function maximises
the resource rent provided by harvesting the native stock, minus
the cost of controlling the invasive species. Perrings (2002) analyses
biological invasion in aquatic system, and shows that the dynamical
characteristics of the invasion depends on the costs and benets both
of native and invasive species as well as population dynamics. His
model used measures of the space occupied by invasive and native
species for the state variables rather than population or biomass, assuming the total space to be constant. Like Perrings, we use a measure of
space, but only for the invasive species state variable. Our model derives
from Flaaten's (1991) competing species bioeconomic model. While
competing species terrestrial models could also be relevant, Flaaten's
model of a shery is closed to our interspecic competition problem.
It is based on a simple analytical model (Gordon-Schaefer), classically
used in sheries economics, and it has already been applied to the Bay
of Saint-Brieuc scallop shery to deal with management issues
(Frsard and Boncoeur, 2006; Guyader et al., 2004; Mah and Ropars,
2001). Unlike Flaaten's model, we consider (i) an asymmetric competition between species, (ii) a competition inuencing the ecosystem's
carrying capacity for the native species, (iii) an invasive species void of
market value, and (iv) an invasive species dispersal coefcient depending on a natural component and on an anthropogenic one. Following
Macpherson et al. (2006), our model considers the unintentional effects
of native stock shers' behaviour on the spread of invasion. The differences between our model and Macpherson et al.'s (2006) model are
the following: in our model, (i) the invasion dispersal depends on natural and anthropogenic components, (ii) the shers' behaviour is exogenous as regards invasion and its control, and (iii) a logistic growth
plurispecies bioeconomic model is used (a two-species case of the standard Gordon-Schaefer model; Gordon, 1954; Schaefer, 1954, 1957).
The paper is organized as follows. Section 2 presents the
bioeconomic model. Section 3 exposes the dynamic optimisation
model. In Section 4, we illustrate our model numerically with reference
to the Bay of Saint-Brieuc scallop shery invaded by slipper-limpet.
Section 5 concludes.
2. Bioeconomic Model of the Invaded Fishery
2.1. Assumptions
We study a bay where a shing eet harvests a benthic native stock.
The native stock is negatively affected by the spread of an invasive
alien species, void of market value. This invasion is the only environmental disturbance in the shery and only the consequences of the
invasion on the shery are studied.
The invasive species is a benthic species and acts as a space competitor
for the native species: the invasion of the bay decreases the size of the
suitable area for native species recruitment. The whole square
meterage of the bay area can therefore be divided into two parts:
the unharmed ones and the invaded ones.
The space competition between species is asymmetric: the presence
of any number of native individuals on a square metre of benthos
does not hinder the establishment and the development of invasive
individuals.
The dispersal of the invasion has two components: a natural one and
an anthropogenic one, linked to shers' behaviour.
The natural dynamics of native stock biomass is represented by a
logistic model, where the carrying capacity of the ecosystem for
the native stock is proportional to the unharmed areas of the bay.
Catch per unit of effort (CPUE) is assumed to be proportional to

stock abundance.
The dynamics of invaded areas is represented by a logistic model,
where the intrinsic growth rate of dispersal has two components: an
exogenous natural one and an anthropogenic one, which is proportional to the native stock harvesting effort. The dispersal of invaded
areas can be controlled by cleaning operations. We assume that the
average productivity of cleaning operations (the number of square
metres cleaned per unit of cleaning effort) is proportional to the
whole number of square metres of invaded areas.
Ex-vessel unit price of native species catch, unit cost of native species
harvesting effort, and unit cost of invaded areas cleaning effort are
exogenous.
There is no technical progress.
2.2. Equations
We consider the combined dynamics of two harvested species, a
native valuable one (i = 1) and an invasive one (i = 2), void of commercial value and acting as a space competitor. Then, there are two state
variables: the native stock biomass X1 and the invaded share of
the whole area of the bay X2, and two control variables: the harvesting
effort of the native stock E1 and the cleaning effort of the invaded
areas E2. All these variables are subject to a non-negativity constraint
and X2 1.
The two equations of motion describing the dynamics of X1 and X2
respectively are written as:



dX 1
X1
q1 E1 X 1
rX 1 1
dt
K 1X 2

dX 2
s gE1 X 2 1X 2 q2 E2 X 2
dt

where r is the intrinsic growth rate of native stock, K is the carrying

capacity of the non-invaded ecosystem for the native stock, q1 is the
catchability coefcient, i.e. the relationship between CPUE and native
stock biomass, s is the natural dispersal coefcient of invasion, g is the
anthropogenic dispersal coefcient of invasion, and q2 is the productivity of cleaning operations, i.e. the ratio between the number of square
metres cleaned per unit of effort and the whole invaded areas. All theses
parameters are constant and positive.
The immediate global surplus GS which is the sum of the prot of
harvesting the native stock minus the cost of cleaning invaded areas is
given by:
GS C 2 E2 Pq1 E1 X 1 C 1 E1 C 2 E2

where P, C1 and C2 are respectively the ex-vessel unit price of native

species catch, the unit cost of effort devoted to harvesting the native
stock, and the unit cost of cleaning effort, assumed constant.1
3. Sustainable Harvest of the Native Species and Control of the
Invasive Species
The damage imposed by the proliferation of the invasive species has
public goods characteristics since the community pay for the cleaning
cost of the invaded areas. Then the problem for the regulator of the invaded shery is to determine2 the effort levels E1 and E2 maximising

1
The total cost of cleaning effort depends implicitly of the invaded areas size. Indeed, at
equilibrium the cleaning effort, dened in Eq. (12), is increasing as this invaded areas size
decreases. The marginal cost of cleaning effort is constant, but the marginal cost of
cleaning the invaded areas is increasing as the invaded areas size decreases. Indeed, the total cost function can be expressed as C2Y2(q2X2)1, with the invaded areas cleaned being
Y2 = q2E2X2.
2
We drop the subscript t in the following equations to simplify notations.

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

the total discounted ow of global surplus generated by the exploitation

of the commercial
species and the invasive one, over an innite time

horizon, GS eat dt, subject to equations of motion (1) and (2) and
0
the following
Eqs. (4) and (5). The argument a is the time discount
rate, assumed constant and positive.
X i 0 X i0

i 1; 2

0 Ei t Ei max

i 1; 2 :

4
5

Eqs. (4) and (5) are respectively the given initial states of native
stock and invaded areas, and the domains of admissible values that
constrain the control variables. Eimax denotes the maximum available
effort i.
Following the standard maximum principle (Pontryaguine et al.,
1961), the current Hamiltonian for the problem is:
H 1

dX 1
dX
2 2
dt
dt

where 1 and 2 are respectively the current costate variables associated with the state variables X1 and X2. We suppose 1 0 and 2 0 as
species 2 is only considered as a pest. 1 represents the marginal value
of stock X1. Similarly, 2 is the marginal value of the invaded areas X2 (or
2 is the marginal value of the unharmed areas (1 X2)).
According to the Maximum Principle, effort Ei must be such that the
H
Hamiltonian is maximised. Singular controls arise when E
0 (for i =
i
1,2) leading to Eqs. (7) and (8).
Pq1 X 1 C 1 1 q1 X 1 2 gX 2 1X 2

space due to this effort unit. According to Eq. (8), for all t, the value at
sea of one unit of the invaded areas cleaning effort must be equal to
its unit cost one. Then, we must have:

C2
q2 X 2

1 P


1
C
C 1 2 g 1X 2 :
q1 X 1
q2

8

< Ei max when X i t N X i

Ei t
0
when X i t b X i
:


Ei
when X i t X i

i 1; 2 :

11

It is rather difcult to determine the optimal path which consists of

some combination of bangbang solution and singular solution. For all t,
we have a total of nine combinations of control variables as described in
Table 1.
Using the growth functions, when E1(t) and E2(t) are singular controls, which need to be the equilibrium solution, we have:

7

E1

Condition (7) means that the net private value of one unit of effort
for harvesting the commercial shery must be equal to the sum of its
value at sea plus the gain at sea in terms of reduction of the invaded

10

As the Hamiltonian is linear in the two control variables, our problem is a singular optimal control one. Thereby, the Most Rapid Approach
Path (MRAP) is optimal. Assuming a steady state exists (Xi for i = 1,2)
and given an initial population Xi0, the optimal policy is to use the effort
level that drives the population towards Xi as rapidly as possible (Clark,
1990). The optimal effort scheme is given by:

E1 t
C 2 2 q2 X 2

q1

E2 t E2

X
1  1  
K 1X 2

1X 2
r
sg
q2
q1

12

!!
X
1  1  
:
K 1X 2

Table 1
Possible congurations of the optimal control at time t.

Case 1:

Case 2:

Case 3:

Case 1:

Program 1-1

Program 1-2

Program 1-3

Case 2:

Program 2-1

Program 2-2

Program 2-3

Case 3:

Program 3-1

Program 3-2

Program 3-3

13

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

Moreover, according to the Maximum Principle, the following conditions must be satised.
di
H
ai
dt
X i
dX i H

dt
i

14

i 1; 2

15

i 1; 2

The transversality conditions for this problem are:3

at

limt X i  i  e

i 1; 2 :

16

From Eq. (14), we have the two following equations of motion for
each costate variable.




1 1 a q1 E1 r 1


2X 1
Pq1 E1
K 1X 2

2 2 as gE1 12X 2 q2 E2 1 r

17
2

X1
K 1X 2 2

18

Note that the dynamic of the costate variable for the invasive species
depends on both species shadow values.
Finally, condition (15) leads to equations of motion for native
species biomass and invaded areas.
i
i
0 and dX
0 for (i =
Assuming a steady state exists, it occurs if d
dt
dt
1, 2). From these four equations and from conditions (9) and (10), the
control variables vanished and we may derive two implicit functions
 
 
e  X  and X
^  X  (see Appendices A and
of the optimal steady state X

B for mathematical details on these functions). The analytical solution

is rather complicated but it can be studied graphically. We determine
d

the locus of points in the (X1, X2) plane, along which dt1 0 and dX
dt
 
d2
dX 2
^
e
0, corresponding to X 1 X 2 and dt 0 and dt 0, corresponding to X
1
 
X 2 . The curves are drawn in Figs. A.1 and B.1. The steady state solution
exists if the two curves have a common point (we have a maximum of
two common points). It is difcult to determine the behaviour of trajectories away from equilibrium, we can only make qualitative description
of the solution with the use of a phase diagram.
In Fig. 1, we have two steady-state solutions corresponding to the
two common points of the two curves. At these points, the optimal
level of the native stock X1 for the given level X2 of invaded areas is
high enough to justify economically the cost of controlling the invaded
areas.
A solution is stable if it represents the issue of a convergence process.
In order to study the stability of a solution, we assume that the equality
between the two curves is not satised for a given level of the invaded
areas. This implies that at least one of the two equalities between the
unit bank value and value at sea of the two species is not met,
which corresponds to a level of optimal effort set to zero or to its maximum. For example:

if

e X NX
^ X q X NC E E
X
2
1
2
1
2
2
2 max dX 2 =dtb0;
2 2 2

if

e X bX
^ X q X bC
X
1
2
1
2
2
2 2 2

E2 0

Michel (1982).

until X2B is reached. If the initial level X20 of the invaded areas is higher
than X2A, then the optimal effort applied to X2 is zero and a steady-state
solution could not be reached.
Thus, only the steady-state solution B is a stationary one. Moreover,
this stationary steady-state solution will be reached only if the problem
is addressed early enough (X20 b X2A).
In our model, the optimal effort of controlling X2 includes two steps:
the rst one is either a laissez-faire stage (no control) or a rollback
stage (high level of invaded areas cleaned), depending on the initial
level X20 of invaded areas, where 0 b X20 b X2A; the second one is a
containment stage aimed at stabilizing the level of invaded areas to
its optimal value by a constant level of effort. According to X20, the
optimal control effort is either zero as the low initial level of invaded
areas grows and reaches the stationary steady-state where it is stabilized by a positive and constant level of effort, or maximum as the
high initial level of invaded areas decreases and reaches the stationary
steady-state where it is stabilized.
In Fig. 3, we have no common points of the two curves. This corresponds to a situation where values of unit cost of native species harvesting, unit cost of invaded areas cleaning, natural and anthropogenic
dispersal coefcients of invasion, and time-discount rate are too high
to control the invasion. When unit cost of native species harvesting is

dX 2 =dtN0:

If the initial level X20 of the invaded areas is somewhere between

zero and X2B (Fig. 2), then the optimal effort of cleaning the invaded
areas is zero during the convergence process until X2B is reached. If
this initial level is somewhere between X2B and X2A, then the optimal
effort applied to X2 is at maximum during the convergence process
3

Fig. 1. Economically sustainable harvest of the native species and control of the invasive
e 1 X 2 represents the optimal level
species. A and B are the two steady-state solutions. X
^ 1 X 2 represents the level of native
of native stock X1 for a given level X2 of invaded areas. X
stock X1 that economically justies the cost of controlling the invaded areas at a given level
X2. X2sup is the breakeven point for harvesting the native stock.

Fig. 2. Stability of a stationary steady-state solution. A and B are the two steady-state
e 1 X 2
solutions. X2A and X2B are the levels of invaded areas corresponding to A and B. X
represents the optimal level of native stock X 1 for a given level X 2 of invaded
^ 1 X 2 represents the level of native stock X 1 that economically justies
areas. X
the cost of controlling the invaded areas at a given level X 2 .

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

areas (that affects the recruitment probability of the scallop) was 0.1065
(Hamon and Blanchard, 1994, 2006). Void of commercial value, this invasive species threatens the long term sustainability of the shery by reducing the size of suitable areas for scallop beds (Chauvaud, 1998).
Facing this invasion and the space competition it implies, a control programme, nanced by public aids (in 90%), based on yearly dredging
campaigns, was established by the local shery committee in 2002
(Anon, 2005).
We discuss the main hypothesis of the model with regard to this
case.

e 1 X 2 represents the optimal level of

Fig. 3. Quasi-eradication of the native species. X
^ 1 X 2 represents the level of native
native stock X1 for a given level X2 of invaded areas. X
stock X1 that economically justies the cost of controlling the invaded areas at a given level
X2. X2sup is the breakeven point for harvesting the native stock.

too high enough, there is no exploitation of the native species. Then the
native stock should be relatively high. But there is no incentive to
control the invasion which continues to grow, leading to quasieradication4 of the native species. We have the same result if unit cost
of invaded areas cleaning or natural and anthropogenic dispersal coefcients of invasion are too high enough to control the proliferation of the
invasive species. A high time discount rate leads to a situation that is
quite similar to the dissipation of the resource rent. In these cases, the
optimal time-path leads to an asymptotic eradication of the native
stock (X1 tends to zero and X2 tends to one). The shery will close
once the invaded areas have reached a level corresponding to the breakeven point X2sup for harvesting the native stock (cf. Fig. A.1).
4. Numerical Illustration
We now illustrate numerically our optimal control bioeconomic
model through the case of the scallop shery in the Bay of SaintBrieuc (France).5 Firstly, we present the case study, and discuss the
assumptions of the model, we then expose the data. Secondly, we
present our numerical results and sensitivity tests.
4.1. The Bay of Saint-Brieuc Scallop Fishery Invaded by Slipper-limpet
Our illustration is based on a regulated common pool resource based
on a control of shing effort: the common scallop shery (Pecten
maximus) of the Bay of Saint-Brieuc, located on the northern coast of
Brittany (France). It is the second largest scallop shery in France,
representing about 29% of the French scallop landings during the
2007/2008 harvesting season (MAP/FRANCEAGRIMER, 2008). During
the 2007/2008 scalloping campaign, shers landed 7099 tons of
common scallop, representing an ex-vessel value of 13.5 million
(Anon, 2008). The shery is seasonally operated by some 250 artisanal
boats under strict regulation scheme including numerus clausus licence
system, global allowed quotas, shing calendar, limitations of engine
power of vessels, mesh size and short allowed shing time (45 min by
vessel) by trip (Fifas et al., 2003). The bay has been invaded by a
slipper-limpet (Crepidula fornicata) that was rst noticed in 1974
(Dupouy and Latrouite, 1979). Slipper-limpet stock was estimated
once to be 250000 tons in 1994 and the corresponding level of invaded
4
In our model, derived from the GordonSchaefer one, the eradication of a species is asymptotic, called here quasi-eradication.
5
As much as possible, numerical simulations are run with an estimated model, based on
parameters previously estimated using shery data (see Frsard and Boncoeur, 2006, and
Frsard, 2008, for details on parameters estimations). Nevertheless, mean of some variables is considered when available data are to scarce. Sensitivity tests are carried out to
face with this limit.

Slipper-limpet is a benthic species and acts as a space competitor for

the common scallop (Blanchard and Hamon, 2006; Chauvaud et al.,
2003). All scallop areas of the bay can be invaded by slipper-limpet
(Blanchard et al., 2001). When the density of slipper-limpet is high
enough in a given area, the scallop juveniles cannot settle down. We
assume that, on a square metre of benthos, the recruitment probability of the scallop is not affected by the presence of slipper-limpet
individuals as long as their density is below a rate of 40% (Hamon
and Blanchard, 2006), and becomes zero when this threshold is
reached. Thus, the whole square meterage of the bay area can be divided into two parts: the unharmed ones and the invaded ones.
The space competition between slipper-limpet and the common
scallop is asymmetric: when the threshold of slipper-limpet is
reached, scallop juveniles cannot settle down but the presence of
scallop on a square metre of benthos does not hinder the development of slipper-limpet. On the contrary, the invasive species can
develop colonies on scallop shells (Chauvaud et al., 2003; Hamon
et al., 2002).
The dispersal of the invaded areas has two components: a natural one
and an anthropogenic one. Slipper-limpet is a bycatch which is
discarded at sea. Thus, during the scalloping campaign, the discards
of slipper-limpet and the use of scallop dredges contribute to the
dispersal of the invasive species6 (Hamon et al., 2002).
The benthos of the bay is assumed to be spatially homogenous. This
assumption is an oversimplication as regards larvae dispersal,
which depends mainly on hydroclimatic conditions. These conditions
are not predictable at the moment in this bay. However, the invasion
takes the form of new spots appearing in various places of the bay
(Hamon et al., 2002). Then, the control of the invasion cannot be considered as defending a frontline between invaded and unharmed
areas (Frsard and Boncoeur, 2006), which makes this simplied
assumption more acceptable.
The natural stock dynamics of the common scallop is represented by a
logistic model, where the carrying capacity of the ecosystem for the
native stock is proportional to the unharmed areas of the bay. The
logistic model (without the space competition effect) has already
been applied to the scallop stock of the Bay of Saint-Brieuc (Frsard
and Boncoeur, 2006; Mah and Ropars, 2001). Catch per unit of effort
(CPUE) is assumed to be proportional to stock abundance (Frsard
and Boncoeur, 2006).
A logistic model7 is used to represent the dynamics of invasion by
slipper-limpet. With limited quantitative data, this model can be considered as an acceptable simplication for a rst approach. This model
is consistent with the standard logistic growth model applied frequently to biological invasions (Williamson, 1996). The dispersal of
slipper-limpet can be controlled by cleaning operations (cleaning
one square metre of invaded areas consists of decreasing the density

6
In this case study, ban on sea discards wouldn't be a relevant invasion control tool because these inshore shing boats are quite small (9 m on average) and allowed shing
time by trip is very short. Thus, shermen have few possibilities of storage on board and
no time to unload slipper-limpets on a large barge under the current management regime.
7
In a logistic model, the natural dynamics of invasion is driven, in a stable environment,
by the intrinsic growth rate of dispersal and the potential invaded areas of the ecosystem
(implicitly set to one here).

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

Table 2
The Bay of Saint-Brieuc scallop shery invaded by slipper-limpet: parameters value.
Parameter

Description

Value

r
K
q1
P
C1
s
g
q2
C2

Intrinsic growth rate of native stock

Carrying capacity of the non-invaded ecosystem for the native stock
Catchability coefcient of native stock
Ex-vessel unit price of native species catch
Unit cost of shing effort
Natural spatial dispersal coefcient of invasion
Anthropogenic spatial dispersal coefcient of invasion
Productivity of cleaning operations
Unit cost of cleaning effort

0.649
54252 (tons)
7.05 105
2000 ( per ton)
113 ( per shing hour and per boat)
[0; 0.045]
[0; 6 106]
3.675 104
1423 ( per hour of cleaning)

Main economic and technical data concerning harvesting activities and biological data concerning scallop dynamics were econometrically estimated by Frsard and Boncoeur (2006). The
productivity of cleaning operations and the intrinsic growth rate of dispersal of invasion are derived from Frsard (2008). We assume the values of its two components, s and g. Ex-vessel
unit price of native species, unit cost of shing effort and unit cost of cleaning effort are calculated on the basis of empirical observations by Frsard and Boncoeur (2006).

of the invasive species to below the threshold of 40%). We assume that

the number of cleaned square metres per unit of effort is proportional
to the whole number of metres squared of invaded areas (this
assumption, which is similar to the one of proportionality of CPUE to
native stock biomass, is justied by the random nature of the distribution of the invaded square metres (Frsard, 2008)).
We assume a 5% time discount rate.
The parameters value is summarized in Table 2. We use economic
and technical data concerning harvesting activities, and biological data
concerning common scallop dynamics econometrically estimated by
Frsard and Boncoeur (2006). The parameter q2 (the productivity of
cleaning operations) is the one which has been estimated by Frsard
(2008). Ex-vessel unit price of native species, unit cost of shing effort
and unit cost of cleaning effort are calculated on the basis of empirical
observations by Frsard and Boncoeur (2006). Biological data
concerning invaded areas dynamics are derived from Blanchard and
Hamon (2006), and Hamon and Blanchard (1994, 2006). The level of invaded areas of the bay was estimated to be 0.1065 in 1994 (Hamon and
Blanchard, 1994) and 0.1350 in 2000 (Blanchard and Hamon, 2006).
From these data, Frsard (2008) has estimated the corresponding intrinsic growth rate of dispersal of invasion (derived from the logistic
model) to be 0.045. This rate equals to s + gE1 in our model, but it
seems quite difcult to estimate these two components separately. Nevertheless, by taking the average value of the level of the harvesting effort
of the native stock over the studied time period (19902002), we have
30000

Native stock biomass (tons)

25000

20000

15000

10000

5000

0
0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Invaded share of the whole area

Fig. 4. Steady state solutions of the invaded shery. Optimal level of native stock
X1 (common scallop) for a given level X2 of invaded areas by slipper-limpet. Level
of native stock X1 that economically justies the cost of controlling the invaded areas at
a given level X2.

estimated an interval for the dispersal coefcient of invasion by shers:

g = [0; 6 106], and for the natural dispersal coefcient of invasion:
s = [0; 0.045]. For the sake of numerical illustration, we assume
different values of s and g over these intervals.
4.2. Numerical Results and Sensitivity Tests
The numerical illustration of the model based on the Saint-Brieuc
case shows that there are two steady state solutions for sustainable
harvest of the common scallop and control of the slipper-limpet, corresponding to the two common points of the two curves (Fig. 4). The rst
one corresponds to a level of the invaded areas (X2 = 0.9851) less than
the breakeven point X2sup (equal to 0.9852) but close to it. This solution
corresponds to a high level of invaded areas and is unstable. The second
one corresponds to a lower level of invaded areas (X2 = 0.0129) and is
stable. The value of the variables corresponding to this optimal stable
solution is given in Table 3. The numerical illustration of the model
shows that it is nearly always optimal to control the invasion: if the
initial level of the invaded areas is less than X20 = 0.98518 in the
Saint-Brieuc Bay, then the optimal solution can be reached (under current economic and biological conditions that guarantee its existence).
Sensitivity tests were carried out on ve parameters: the ex-vessel
price of the native species catch, the unit cost of the shing effort of native species, the unit cost of the invaded areas cleaning effort, the total
dispersal rate of invasion, and the time discount rate. Using realistic assumptions concerning the parameters value of the model, sensitivity
tests carried out show that the optimal stable solution exists, except
for a strict zero time discount rate case. This optimal solution seems
rather insensitive to variations in model parameters, particularly for
the optimal stock biomass of native species, except for extreme values
of some of these parameters.
Fig. 5 shows the sensitivity of the optimal level of invaded areas and
of the native stock biomass to the ex-vessel price of this native species.
For variation of +/ 100%, the optimum size of the scallop stock remains at about 25000 tons. The optimal level of invaded areas decreases
with the ex-vessel unit price. It doubles if the price is halved, and decreases by about 50% if the ex-vessel price doubles. However, the optimal level of invaded areas remains very low. With an extremely low
price (200 per ton, ten times lower than the main simulation price,
which is unlikely to happen), this level reaches 14% of the whole of
the bay area. For extremely high price values, the optimum size of the
invasion tends towards zero, while the size of the scallop stock
biomass does not fall below 25000 tons.
Changes in the shing effort unit cost of the native species have almost no effect on the optimal size of the invaded areas (Fig. 6). It always
remains close to 1.3%. The harvesting effort of the native species will
change. Ultimately if the cost is very low, this effort will never be extremely high and the size of the native species stock biomass remains
8
X2 = 0.9851is the unstable solution. Its level is rather insensitive to realistic variations
of parameters values (it remains up to 0.90).

Variable

Description

Optimal value

X1
E1
Y1
X2
E2
Y2

Native stock biomass

Harvesting effort of the native stock
Native stock catch
Invaded share of the whole area of the bay
Cleaning effort of the invaded areas
Invaded share of the whole area cleaned

25225 (tons)
4869 (hours)
8659 (tons)
0.0129
100 (hours)
0.00047

greater than 24750 tons. Indeed, it is conceivable that harvesting costs

could increase with the rising cost of fuel. Nevertheless, even if these
costs were doubled, the optimum size of the native species would
increase by no more than 2%.
Changes in the unit cost of cleaning operations of the invaded areas
mostly affect the optimal size of the invasion as shown in Fig. 7. If this
unit cost doubles, then the optimal invaded share of the whole area
doubles as well (but remains below 3%). However, this implies very
few changes of the native species optimal biomass. It decreases
by only 1%. Indeed, the protability of the scallop shery is still very
high compared to the costs of cleaning the invaded areas. Even if
these costs increase, it would have very little impact on the overall
protability.
The variation of the total dispersal coefcient of invasive areas
(Fig. 8) has very little effect on the optimal size of the invasion, as it
remains at around 1.3%. The optimal stock of the native species also
varies very little. But it will signicantly change the invaded areas
cleaning operations effort. If the dispersal coefcient doubles, then the
effort to keep the invasion at a very low level varies to the same proportions. However, even if the cost of stabilizing the invaded areas
increases, it remains relatively low compared to the high protability
of the scallop shery. The stabilization of the invasion at a small level
of invaded areas still remains economically attractive. Aside from that,
this result minimises the potential impact of native stock shers' behaviour on the optimal size of the invasion in this case study.
Fig. 9 presents the impact of time discount rate variations on the optimal solution. For a null time discount rate, there is no stable solution.
But as this rate tends towards zero, quasi-eradication of the invasion
seems to be optimal to let the native stock biomass stabilize around
27000

0.0140

26500

0.0136

26000

0.0132

25500

0.0128

25000

0.0124

24500
0

100

200

300

400

500

0.0120
600

Unit harvesting cost (/hour/boat)

Fig. 6. Impact of the unit cost of shing effort on the optimal stable solution. Native
stock biomass. Invaded share of the whole area.

an optimal size of about 27000 tons, which corresponds to the

Maximum Sustainable Yield (MSY). As the discount rate increases to
within a reasonable range, the optimal invaded share rises and the native stock biomass decreases, both at an increasing rate.9
For reasonable variations of the model parameters, we note that the
optimal size of the invasion is still very low and below the one that prevails today. The high protability of the native scallop shery, compared
to the relatively low level of the operating cost of the control programme, means that it is always worth cleaning the invaded areas at
an intermediate level to control the invasion. The optimal stable solution can be reached as long as the present level of invaded areas is
lower than the one corresponding to the unstable steady-state solution.10 However, the results appear to be sensitive to particularly extreme values tested: a very low ex-vessel unit price of native species
catch (with a price ten times lower than the main simulation one) and
a very high time discount rate (set to one), which are unlikely to
happen.

0.18

5. Discussion and Conclusion

0.16

In this paper, we have developed a theoretical model of optimal control analysing the sustainable harvest of a native species combined with
control of an invasive species, acting as a space competitor. We have
shown that a time-path leading to an optimal stationary steady-state
solution, where the native species is harvested sustainably and the invasion is kept under control by cleaning operations, exists and can be
reached. The stationary solution exists, provided unit cost of native species harvesting, unit cost of invaded areas cleaning, natural and anthropogenic dispersal coefcients of invasion, and time-discount rate are
not too high. The optimal solution can be reached only if the problem
is addressed early enough: with a high initial level of invasion, the optimal time-path leads to the quasi-eradication of the native species.
The originality of the paper was to address an invasion problem of a
particular kind: an asymmetric space competition relationship between

0.14

26000

0.12
25500

0.1
0.08

25000

0.06

Invaded share of the whole area

Native stock biomass (tons)

27000

0.2

26500

0.04

24500

0.02
24000

Native stock biomass (tons)

Table 3
Sustainable harvest of the native species (common scallop) and control of the invasive
species (slipper-limpet) in the Bay of Saint-Brieuc (France): variables value of the optimal
stable solution.

Invaded share of the whole area

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

1000

2000

3000

4000

ex-vessel unit price ( per ton)

Fig. 5. Impact of the ex-vessel unit price of native species catch on the optimal stable
solution. Native stock biomass. Invaded share of the whole area.

9
As it tends towards one (which is unlikely to happen), the optimal level of the invaded
areas controlled is very high (X2 = 0.9585 for a = 1), and is close to the unstable solution
(X2 = 0.9790 for a = 1). This case leads to a very low harvestable level of native species.
10
For reasonable variations of the model parameters, sensitivity tests show that this level has to be less than 90% of the bay invaded by slipper-limpet for the worst cases of the
main tested values.

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

26000

0.045

25800

0.040

1.00

30000

0.90

0.035

25400
0.030
25200
0.025
25000
0.020
24800
0.015
24600
0.010

24400
24200

0.005

24000

0.000
5000

Native stock biomass (tons)

Native stock biomass (tons)

25600

Invaded share of the whole area

25000

0.80
0.70

20000

0.60
0.50

15000

0.40
10000

0.30
0.20

5000

0.10
0.00

0
0

1000

2000

3000

4000

Fig. 7. Impact of the unit cost of cleaning effort on the optimal stable solution. Native
stock biomass. Invaded share of the whole area.

450

0.019

400

0.018

350

0.017

300

0.016

250

0.015

200

0.014

150

0.013

100

0.012

50

0.011

0
0

0.05

0.1

0.15

0.2

Invaded share of the whole area

Cleaning effort (hours)

a native and an invasive species, based on the ecosystem carrying capacity. Moreover, the native species is a commercial one and thus the social
planner considers the combined harvest of both species (two state and
two control variables). Furthermore, we have considered the effect of
native species shers' behaviour on the invasion dynamics. This point
is important because human behaviour has to be considered when
management actions are addressed (Macpherson et al., 2006; Settle
and Shogren, 2002, 2006; Timar and Phaneuf, 2009). Unlike these
authors, in our model, accounting for human behaviour doesn't imply
important changes in the equilibrium level of invasion but can withdraw the existence of the steady-state solution. Besides, we have
shown that this optimal solution can be reached only if the problem is
addressed early enough. Thus, we have underlined the importance of
the initial invasion size, which is time dependent.
Finally, the optimal control model was applied to the Bay of SaintBrieuc scallop shery invaded by slipper-limpet and shows that, under
current economic and biological conditions that guarantee the existence
0.02

0.2

0.4

0.6

0.8

Time discount rate

Unit cost of cleaning operations (/hour of cleaning)

500

Invaded share of the whole area

0.01
0.25

Fig. 9. Impact of the time discount rate on the optimal stable solution. Native stock
biomass. Invaded share of the whole area.

of the optimal solution, it is nearly always optimal to control the invasion. Moreover, it could provide a useful background for the denition
of an efcient management. Results show that the control programme
implemented in the Bay of Saint-Brieuc since 2002 could be improved.
Frsard (2008) showed that this control programme allows the level
of invaded areas to be stabilized at 0.1130. Results of the simulation of
our model (Table 3) show that the optimal level of stabilized invaded
areas must be lower than 0.1130 (X2 = 0.0129). This is due to the
control programme of this bay: cleaning operations aim rstly to
reduce the invaded areas at a xed ratio, and then to stabilize them at
the level obtained after ve years. It would be more efcient to try to
reach the optimal steady-state as soon as possible, by drastically reducing the level of invaded areas in the rst stage of the control programme
(with a control effort set at its maximum until the optimal steady-state
is reached). Results of the numerical illustration show that the optimal
level of the common scallop stock biomass is lower than that of the
Maximum Sustainable Yield11 (Xmsy = K/2). This optimal level is higher
than the present stock biomass in the bay (23220 tons of common
scallops estimated by Fifas and Huet, 2007). However, this situation is
not so far from the optimal one (25225 tons of common scallops derived
from the model) and, as long as the present level of invaded areas is
lower than the one corresponding to the unstable steady-state solution
(X2 = 0.9851), the optimal solution could be reached.
Few theoretical analyses are followed by application on case studies.
However, they could provide some insights about invasions management. For example, Burnett et al. (2006) have shown that the current
population level of the invasive plant Miconia calvescens is higher than
the optimal one. This invasive population has to be reduced and then
maintained at the optimal level. Horan and Wolf's (2005) model of
control shows that a low level of disease (bovine tuberculosis) in the
Michigan white-tailed deer population is optimal, while current control
aims to eradicate it. Through a review of the economic literature on
biological invasion control, Frsard (2011) studies applications of
theoretical models to real case studies that allow discussions on policy
measures. Policy goals are in most case different from optimal levels
of control, and sometimes policy tools are not appropriate. Thus, current
control programmes could easily be improved by accounting for

Total spatial dispersal coefficient

Fig. 8. Impact of the spatial dispersal coefcient on the optimal stable solution. Cleaning
effort of the invaded areas. Invaded share of the whole area.

11
It reects the impact of the asymmetric competition exerted by slipper-limpet on scallop dynamics and the additional cost linked to cleaning operations in the social manager
problem.

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

outputs of the theoretical analyses. Furthermore, these analyses underline biological and economical key parameters that can inuence the
control: the level of invasion, the value of the damages of invasion, the
invasion dynamics, the potential value of the invasive species, the control cost of invasion, and the time discount rate. The application of our
optimal control model to the Bay of Saint-Brieuc case shows that the
results are sensitive to particularly extreme values tested: a very low
ex-vessel unit price of native species catch and a very high time discount
rate. But, for realistic variations of the model parameters, the optimal
size of the invasion is still very low and below its present one. Nevertheless, the application of our model to this case study remains limited by
the availability of data, and the deciency of biological knowledge, especially in the eld of slipper-limpet dynamics. Moreover, describing the
common scallop dynamics by means of a global deterministic model
could be open to criticism, due to the high variability of recruitment of
this species.
e 1 X 2 curve. X
e 1 X 2 represents the optimal level
Fig. A.2. Impact of time discount rate a on X
of native stock X1 for a given level X2 of invaded areas. X2sup is the breakeven point for
harvesting the native stock. represents an increase of a.

Acknowledgements
We acknowledge Jean Boncoeur (UMR AMURE, Universit de Brest,
France) for his invaluable help on a preliminary version of this work. We
thank Jennifer Ford for providing English language help. We also
acknowledge anonymous referees for their careful review and helpful
comments on a preliminary version of this paper.
e 1 X 2 Curve
Appendix A. The X
d

1
From dt1 0 and dX
dt 0, and using Eqs. (9) and (12), we nd the
following quadratic equation:








2X 1 2
a
1
C1
C g
a 1X 2
C1
C g
2
2
X1
0
1
K 1X 2
r Pq1 K 1X 2
q2
r Pq1 1X 2
q2

A:1

Solving this equation gives us two real roots. One of which is

e 1 X 2 .
positive, called here: X




 q
C1
C g
e
e X K 1X 2 1 a 1
X
2

1
2
4
r Pq1 K 1X 2
q2

where:







C1
C g 2
8 a
C1
C g
e 1 a 1

2 :
r Pq1 K 1X 2
q2
Pq1 K r 1X 2
q2

A:3

e 1 X 2 function is positive, decreasing and concave for any 0 b

The X
e 1 X 2 biomass is positive
X2 b 1. However, the shing effort linked to X
when X2 is below the threshold X2sup, where:
X 2 sup

C1
Pq1 KC 2 g=q2

A:4

e 1 X 2 is dened over 0 b X2 b X2sup (Fig. A.1). Over this interThen, X

e 1 X 2 biomass is below the carrying capacity K(1 X2) and
val, the X
equal to it when the threshold X2sup is reached.
e 1 X 2 function depends on the time discount rate a
The X
e 1 with respect to a is negative for
(Fig. A.2). The rst derivative of X

A:2

e 1 X 2 curve. It represents the optimal level of native stock X1 for a given level X2
Fig. A.1. X
of invaded areas. X2sup is the breakeven point for harvesting the native stock. K(1 X2sup)
is the carrying capacity for the native stock corresponding to a X2sup level of invaded areas.

^ 1 X 2 curve. It represents the level of native stock X1 that economically justies

Fig. B.1. X
the cost of controlling the invaded areas at a given level X2. C1/Pq1 is the ratio between the
unit cost of shing effort, and the ex-vessel unit price of native species catch multiply by
the catchability coefcient of native stock.

10

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

any 0 b X2 b X2sup. When a tends to zero or to innite, the extreme

e 1 X 2 are:
values of X




K 1X 2
1
C1
C g
1
2
2
Pq1 K
1X 2
q
2 
1
C g
C 1 1X 2 Pq1 K 2

2Pq1
q2

e X
lim X
1
2

a0


C1
C g
e X 1X 2
lim X
2
1
2
a
q2
Pq1 1X 2


1
C g
C 1 1X 2 2 :

Pq1
q2

A:5

^ X :
lim X
1
2

A:6

2
2
From d
0 and dX
0, and using Eqs. (9), (10), (12) and (13) we
dt
dt
nd the following quadratic equation:


C1
C
a
gr
2K
X 1 2 1X 2
s
0:
r X2
q1
Pq1
Pq2

B:1

Solving this equation gives us two real roots. One of which is posi^ 1 X 2 .
tive, called here: X


p
^
^ X 1 C 1
X
1
2
2 Pq1

B:2

where:
^

C1
Pq1

2


C2
a
gr
2K
1X 2
s
:
r X2
q1
Pq2

2
3

s

2



1
C
C
C
K
s
g
2
1
1
2
^ X 4
5
1X 2

4
lim X
1
2
2 Pq1
Pq2
r q1
a0
Pq1

Appendix B. The X 1 X 2 Curve

X1

any 0 b X2 b 1. When a tends to zero or to innite, the extreme values

^ 1 X 2 are:
of X

B:3

^ 1 X 2 function is positive, decreasing and convex for any 0 b

The X
X2 b 1 (Fig. B.1). This function has an innite asymptotic limit as X2
tends to zero (for any a strictly positive).
^ 1 X 2 function depends on the time discount rate a
The X
^ 1 with respect to a is positive for
(Fig. B.2). The rst derivative of X

^ 1 X 2 curve. X
^ 1 X 2 represents the level of
Fig. B.2. Impact of time discount rate a on X
native stock X1 that economically justies the cost of controlling the invaded areas at a
given level X2. C1/Pq1 is the ratio between the unit cost of shing effort, and the exvessel unit price of native species catch multiply by the catchability coefcient of native
stock. represents an increase of a.

B:4

B:5

References
Anon, 2005. Programme de valorisation des crpidules en Bretagne. Bilan d'activits 2004.
AREVAL/Ctes d'Armor Dveloppement, Saint-Brieuc, France.
Anon, 2008. Evolution de la production de coquille Saint-Jacques sur le gisement class de
Saint-Brieuc. Bilan de la campagne 2007/2008. Ctes-d'Armor Dveloppement, SaintBrieuc, France.
Blanchard, M., Hamon, D., 2006. Study results of the industrial exploitation project on
slipper-limpet, along northern coasts of Brittany, 20022005 (in French). Final Report
AREVAL/Ifremer DYNECO/EB/06-01. Ifremer, France.
Blanchard, M., Blanchet, A., Gaffet, J.D., Hamon, D., 2001. Population dynamics for the
slipper-limpet (Crepidula fornicata) in the Bay of Saint-Brieuc (Western Channel)
(in French). Scientic Report RST/DEL/00.08. Ifremer, France.
Burnett, K.M., Kaiser, B., Pita, B.A., Roumasset, J., 2006. Prevention, eradication, and
containment of invasive species: illustrations from Hawaii. Agric. Resour. Econ. Rev.
35 (1), 6377.
Carlton, J.T., 1987. Patterns of transoceanic marine biological invasions in the Pacic
Ocean. Bull. Mar. Sci. 41 (2), 452465.
Carlton, J.T., 1989. Man's role in changing the face of the ocean: biological invasions and
implications for conservation of near-shore environments. Conserv. Biol. 3 (3),
265273.
Carlton, J.T., Geller, J.B., 1993. Ecological roulette: the global transport of non-indigenous
marine organisms. Science 261, 7882.
Chauvaud, L., 1998. La coquille Saint-Jacques en rade de Brest: un modle biologique
d'tude des rponses de la faune benthique aux uctuations de l'environnement(PhD
Thesis) Universit de Brest, France.
Chauvaud, L., Thouzeau, G., Grall, J., Paulet, Y.M., 2003. La crpidule en rade de Brest:
un paradoxe pour le devenir de la coquille Saint-Jacques. In: Laubier, L. (Ed.),
Exploitation et Surexploitation des Ressources Marines Vivantes. Editions Lavoisier,
Acadmie des Sciences RST n17, Paris, pp. 307318.
Clark, C.W., 1990. Mathematical Bioeconomics: The Optimal Management of Renewable
Resources, 2nd ed. John Wiley & Sons, New York.
Dalmazzone, S., 2000. Economic factors affecting vulnerability to biological invasions. In:
Perrings, C., Williamson, M., Dalmazzone, S. (Eds.), The Economics of Biological
Invasions. Edward Elgar, Cheltenham, pp. 1730.
Didham, R.K., Tylianakis, J.M., Hutchison, M.A., Ewers, R.M., Gemmell, N.J., 2005. Are
invasive species the drivers of ecological change? Trends Ecol. Evol. 20 (9), 470474.
Dupouy, H., Latrouite, D., 1979. Le dveloppement de la crpidule sur le gisement
de coquilles Saint-Jacques de la baie de Saint-Brieuc. Sci. Pche Bull. Inst. Pches
Mar. 292, 1319.
Eiswerth, M.E., Johnson, W.S., 2002. Managing nonindigenous invasive species: insights
from dynamic analysis. Environ. Resour. Econ. 23 (3), 319342.
Fifas, S., Huet, J., 2007. Gisement de coquilles Saint-Jacques de la baie de Saint-Brieuc.
Campagne 2007 d'valuation directe. Rsultats et perspectives. Scientic report
STH/LBH. Ifremer, France.
Fifas, S., Guyader, O., Boucher, J., 2003. La pcherie de coquilles Saint-Jacques en baie de
Saint-Brieuc: productivit et gouvernance. In: Laubier, L. (Ed.), Exploitation et
Surexploitation des Ressources Marines Vivantes. Editions Lavoisier, Acadmie des
Sciences RST n17, Paris, pp. 221234.
Flaaten, O., 1991. Bioeconomics of sustainable harvest of competing species. J. Environ.
Econ. Manag. 20 (2), 163180.
Frsard, M., 2008. Economic Analysis of Controlling a Biological Invasion. Theoretical
Model and Application to the Common Scallop Fishery of the Saint-Brieuc Bay
Invaded by Slipper-limpet (in French)(PhD Thesis) Universit de Brest, France.
Frsard, M., 2011. The economic analysis of biological invasions control: a literature
review (in French). Rev. Econ. Polit. 121 (4), 489525.
Frsard, M., Boncoeur, J., 2006. Controlling the biological invasion of a commercial shery
by a space competitor: a bioeconomic model with reference to the Bay of St-Brieuc
scallop shery. Agric. Resour. Econ. Rev. 35 (1), 7897.
Gordon, H.S., 1954. The economic theory of a common property resource: the shery. J.
Polit. Econ. 62 (2), 124142.
Guyader, O., Daurs, F., Fifas, S., 2004. A bioeconomic analysis of the impact of
decommissioning programs: application to a limited-entry French scallop shery.
Mar. Resour. Econ. 19 (2), 225242.
Hamon, D., Blanchard, M., 1994. Etat de la prolifration de la crpidule (Crepidula
fornicata) en baie de Saint-Brieuc. Scientic report DEL 9414Ifremer, France.
Hamon, D., Blanchard, M., 2006. Expert estimates. Approximation of the density of
slipper-limpet that induces a zero recruitment probability of scallop. DYNECO/
EBIfremer, France.
Hamon, D., Blanchard, M., Houlgatte, E., Blanchet, A., Gaffet, J.D., Cugier, P., Mnesguen, A.,
Bassoulet, P., Cann, P., Domalain, D., Haubois, A.G., 2002. La crpidule: identier les
mcanismes de sa prolifration et caractriser ses effets sur le milieu pour envisager

M. Frsard, C. Ropars-Collet / Ecological Economics 106 (2014) 111

sa gestion. Chantier baie de Saint-Brieuc. Final report of LITEAU programme. DEL,
Ifremer, France.
Horan, R.D., Wolf, C.A., 2005. The economics of managing infectious wildlife disease. Am.
J. Agric. Econ. 87 (3), 537551.
Junqueira-Lopes, R., Michel, P., Rotillon, G., 1996. Bioeconomic management of red
swamp craysh (Procambarus clarkii) in the presence of environmental externalities.
Mar. Resour. Econ. 11 (1), 19.
Knowler, D., 2005. Reassessing the costs of biological invasion: Mnemiopsis leidyi in the
Black Sea. Ecol. Econ. 52 (3), 187199.
Knowler, D., Barbier, E.B., 2000. The economics of an invading species: a theoretical model
and case study application. In: Perrings, C., Williamson, M., Dalmazzone, S. (Eds.), The
Economics of Biological Invasions. Edward Elgar, Cheltenham, pp. 7093.
Krcmar-Nozic, E., van Kooten, G.C., Wilson, B., 2000. Threat to biodiversity: the invasion of
exotic species. In: van Kooten, G.C., Bulte, E.H., Sinclair, A.R.E. (Eds.), Conservating
Nature's Diversity: Insights from Biology. Ethics and Economics. Ashgate, Burlington,
pp. 6887.
Lvei, G.L., 1997. Global change through invasion. Nature 388 (6643), 627628.
Mack, R.N., Simberloff, D., Lonsdale, M., Evans, H., Clout, M., Bazza, F.A., 2000. Biotic invasions: causes, epidemiology, global consequences and control. Ecol. Appl. 10 (3),
689710.
Macpherson, A.J., Moore, R., Provencher, B., 2006. A dynamic principal-agent model of
human-mediated aquatic species invasions. Agric. Resour. Econ. Rev. 35 (1), 144154.
Mah, L.P., Ropars, C., 2001. L'exploitation rgule d'une ressource renouvelable:
inefcacit d'un traitement factoriel et efcacit des quotas individuels transfrables.
Econ. Prvis. 148 (2), 141156.
MAP/FRANCEAGRIMER, 2008. Bilan annuel de production 2008 des pches et de
l'aquaculture. Report MAP/FRANCEAGRIMER, France.
Michel, P., 1982. On the transversality condition in innite horizon optimal problems.
Econometrica 50 (4), 975985.
Mooney, H.A., Cleland, E.E., 2001. The evolutionary impact of invasive species. Proc. Natl.
Acad. Sci. U. S. A. 98 (10), 54465451.
Perrings, C., 2002. Biological invasions in aquatic systems: the economic problem. Bull.
Mar. Sci. 70 (2), 541552.

11

Perrings, C., Williamson, M., Dalmazzone, S., 2000. Introduction. In: Perrings, C.,
Williamson, M., Dalmazzone, S. (Eds.), The Economics of Biological Invasions. Edward
Elgar, Cheltenham, pp. 113.
Perrings, C., Williamson, M., Barbier, E.B., Delno, D., Dalmazzone, S., Shogren, J.,
Simmons, P., Watkinson, A., 2002. Biological invasion risks and the public good: an
economic perspective. Conserv. Ecol. 6 (1).
Pimentel, D., Zuniga, R., Morrison, D., 2005. Update on the environmental and economic
costs associated with alien invasive species in the U.S. Ecol. Econ. 52 (3), 273288.
Pontryaguine, L., Boltyanski, V., Gamkrlidz, R., Mitchenko, E., 1961. The Mathematical
Theory of Optimal Processes (in Russian). Fizmatguiz; Moscou (translation into English, 1962, J. Wiley & Sons; New-York).
Ruiz, G.M., Carlton, J.T., 2003. Invasive Species: Vectors and Management Strategies.
Island Press, Washington.
Schaefer, M.B., 1954. Some aspects of the dynamics of populations important to the
management of commercial marine sheries. Bull. Inter Am. Tropical Tuna Comm.
1, 2556.
Schaefer, M.B., 1957. Some considerations of population dynamics and economics in
relation to the management of marine sheries. J. Fish. Res. Board Can. 14, 669681.
Settle, C., Shogren, J.F., 2002. Modeling native-exotic species within Yellowstone Lake.
Am. J. Agric. Econ. 84 (5), 13231328.
Settle, C., Shogren, J.F., 2006. Does the integration of biology and economics matter for
policy? The case of Yellowstone Lake. Top. Econ. Anal. Policy 6 (1) (article 9).
Timar, L., Phaneuf, D.J., 2009. Modeling the human-induced spread of an aquatic invasive:
the case of the zebra mussel. Ecol. Econ. 68 (12), 30603071.
Vitousek, P.M., D'Antonio, C.M., Loope, L.L., Rejmanek, M., Westbrooks, R., 1997.
Introduced species: a signicant component of human-caused global change. N. Z.
J. Ecol. 21 (1), 116.
Wilcove, D.S., Rothenstein, D., Dubow, J., Philips, A., Losos, E., 1998. Quantifying threats to
imperilled species in the United States. Bioscience 48 (8), 607615.
Williamson, M., 1996. Biological Invasions. Chapman and Hall, London.
Williamson, M., 1999. Invasions. Ecography 22 (1), 512.
Wilman, E.A., 1996. Pests: sustained harvest versus eradication. J. Environ. Manag. 46 (2),
139147.