See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/223468790

Land use optimization in watershed scale

Article in Land Use Policy · April 2009

DOI: 10.1016/j.landusepol.2008.02.007

CITATIONS READS

66 475

3 authors, including:

Seyed Hamidreza Sadeghi Khalil Jalili

Tarbiat Modares University Academic Center for Education, Culture and …
281 PUBLICATIONS 1,323 CITATIONS 5 PUBLICATIONS 67 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

PhD: Effective Sediment Particle Size Distribution during Erosion Process at Plot Scale View project

Soil and Water Loss Control through Inoculation of Some Soil Microorganisms View project

All content following this page was uploaded by Seyed Hamidreza Sadeghi on 12 January 2018.

The user has requested enhancement of the downloaded file.

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Land Use Policy 26 (2009) 186–193

Contents lists available at ScienceDirect

Land Use Policy

journal homepage: www.elsevier.com/locate/landusepol

Land use optimization in watershed scale

S.H.R. Sadeghi a,∗ , Kh. Jalili a,1 , D. Nikkami b
a
Department of Watershed Management Engineering, College of Natural Resources and Marine Sciences,
Tarbiat Modares University, Noor 46417-76489, Mazadaran Province, Iran
b
Soil Conservation and Watershed Management Research Institute, Tehran, P.O. Box 13445-1136, Iran

a r t i c l e i n f o a b s t r a c t

Article history: Managing a watershed for satisfying the inhabitant’s demand is a difficult task if one has to maintain a
Received 16 July 2007 reasonable balance between usually conflicting environmental flows and demands. The solution to these
Received in revised form 24 January 2008 complex issues requires the use of mathematical techniques to take into account conflicting objectives.
Accepted 19 February 2008
Many optimization models exist for general management systems but there is a knowledge gap in linking
practical problems with the optimum use of all land resources under conflicting demands in a watershed.
Keywords:
In the present study, an optimization problem has been formulated for the Brimvand watershed, Iran, com-
Watershed management
prising ca. 9572 ha to find out the most suitable land allocation to different land uses, viz. orchard, irrigated
Optimization
Land use allocation
farming, dry farming and rangeland targeting soil erosion minimization and benefit maximization.
Soil erosion Soil erosion, net benefit and land capability maps were provided as inputs to formulate the objective
Linear programming functions and governing constraints in a multiobjectives linear optimization problem. The problem was
Brimvand watershed then solved using the simplex method with the help of ADBASE software package and the optimal solu-
Kermanshah Province tion was ultimately determined. Additionally, sensitivity analysis was also conducted to recognize more
Iran effective land use in reducing soil erosion and increasing benefit. The results of the study revealed that
the amount of soil erosion and benefit could, respectively reduce and increase to the tune of 7.9 and 18.6%,
in case of implementing optimal allocation of the study land uses. The results of sensitivity analyses also
showed that the objective functions were strongly susceptible to the constraint of maximum summation
of irrigated farming and orchard areas.
© 2008 Elsevier Ltd. All rights reserved.

Introduction human needs continually and effectively. The watershed optimiza-

Soil erosion is a widespread environmental problem threatens
human being in the developing countries. Each year, 75 billion t
of soil is removed in the globe due to erosion with most coming
from agricultural land. An average rate of soil loss for Asia is also
138 t/(ha year) (Ananda and Herath, 2003). It has therefore eco-
nomic, political, social and environmental consequences due to
both on-site and off-site damages. The conflict between environ-
mental protection and the economic development by different land
uses within a watershed are challenges facing land use planners
in many developing countries (Chang et al., 1995; Peel and Lloyd,
2007; Gezelius and Refsgaard, 2007). Proper environmental plan-
ning at different levels has been called for by watershed managers
during last years to integrate natural resources limitations and
∗ Corresponding author. Tel.: +98 122 6253101 3; fax: +98 122 6253499.
E-mail addresses: sadeghi@modares.ac.ir, shrsadeghi@yahoo.com
(S.H.R. Sadeghi), khaliljalili@yahoo.com (Kh. Jalili), nikkami@scwmri.ac.ir
(D. Nikkami).
1
Present address: Hydraulic and Water Resources Department, Jihad-e-
Daneshgahi, Kermanshah Province, Iran.
tion for each land use, especially agriculture as one of the significant
contributors to the environmental degradation, is therefore neces-
sary to achieve sustainable development (Keristofer, 2001; Seppelt
and Voinov, 2002; Heilman et al., 2003; Wang et al., 2004).
Given the strong concern about escalating cost of high rates
of soil erosion along with the necessity of cooperative solutions
among various users in a watershed, there is a need for analyti-
cal tools that can be easily used by researchers and environmental
managers at several levels. In order to achieve this there is a need for
modeling that emphasizes links between economic policy changes
and environmental outcomes at a landscape scale (Onal et al., 1998;
Shively and Coxhead, 2004). The essence of management science is
manifested in modeling approach; moreover planning methodol-
ogy to specify optimal use of scare resources is the most important
practical approach (Sadeghi et al., 2005a) in watershed manage-
ment to gain the optimal benefit.
Over the past few decades, the vital resources of almost entire
watersheds in Iran have been subject to rapid deterioration result-
ing from expanding anthropogenic activities. The Iran Forest and
Rangeland Nationalization, Act of 56 (fixing and controlling of
governmental national land) was effective in reducing land use

0264-8377/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.landusepol.2008.02.007
 Author's personal copy
S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193
conversion and restoring many of resources. Despite its progress,
overexploitation and mismanagement of watershed resources still
remain as major threats to the watersheds in Iran. This is a major
challenge because of their complex nature and the existence of
diverse and diffuse contributing land uses within the watersheds
(Farahpour et al., 2004). Protecting the health of watershed ecosys-
tem is therefore critical issues facing watershed communities and
managers. A critical factor in watershed-based approaches in Iran
is the ability to prioritize land and to target programs and policies
to areas with maximum benefits. Since various environmental and
socio-economic conditions may lead to conflicts among stakehold-
ers, strategies and policies that are fragmented in scope may not
be effective for sound watershed management. Frequently, com-
promises are necessary in order to obtain overall optimal land and
water use for the entire watershed. Such effort involves government
and other stakeholders in a structured and focused process. Con-
sequently, an integrated modeling approach such as optimization
that incorporates individual system components within a general
framework, instead of examining or presenting them in isolation,
is useful for providing holistic and comprehensive analysis. This
allows gain in economic efficiency by allocating limited resources
to those areas that contribute or have potential to impair watershed
health.
Application of optimization approaches has been started since
the human faced low efficiency production of the system. It was
further extended to applied sciences like agriculture since long ago.
However, optimization techniques have been applied to watershed
resources recently. In the recent literature, Onal et al. (1998) incor-
porated environmental impacts and income distribution goals in
economic analysis of watershed management policies on a small
watershed in Illinois, USA, via a conventional programming con-
straint and a chance-constrained programming formulation. They
indicated that farm costs were increased notably by restricting
agricultural pollution and soil erosion. The income distribution con-
straint also reduced economic efficiency, but the efficiency loss
due to implementing this constraint was less than 10% of the costs
resulting from environmental regulations. Singh and Singh (1999)
in a case study in Mahi Comand watershed, India, maximized the
production and benefit using linear optimization. The results of
their study revealed that the application of optimal plan could
increase the agricultural productions and net return almost 36 and
3%, respectively. Amir and Fisher (1999) introduced an optimizing
linear model for analyzing agricultural production under various
water quantities, qualities, timing, prices and pricing policies. The
model was successfully designed to serve as a decision-making
tool for planners of agricultural production of eight watersheds in
Israel. Recatala et al. (2000) carried out a land-use-planning exer-
cise in a representative area of the Valencian Mediterranean Region
using LUPIS system which facilitated the generation of alterna-
tive land use plans by adjusting the relative importance attributed
by multiple stakeholders to preference and avoidance guidelines.
The results of their study suggested that comprehensive land use
planning could play a vital role in solving land use conflicts in the
region. Salman et al. (2001) presented a linear programming opti-
mization model for analyzing inter-seasonal allocation of irrigation
water in quantities and qualities and their impact on agricultural
production and income. The model generated an optimal mix of
water-demanding activities that maximized the net agricultural
income of the districts and gave the water demands under various
prices in a watershed in Jordan. Randhir et al. (2001) developed
a watershed land prioritization model for water supply optimiza-
tion by integrating geographic information, relationship between
land criteria and effects, and travel-time of runoff water in Ware
River watershed in Massachusetts, USA. They observed that focus-
ing on fewer areas in the watershed with high priority maximized
187
benefits to water quality and resulted in lower expenditures. They
also declared that the given approach could be adapted to prior-
itize a wide variety of land-protection and land use decisions by
adjusting criteria and weights. Nikkami et al. (2002) used multiob-
jective linear programming to minimize soil erosion and maximize
agricultural benefit for Siahrood area in Damavand watershed, Iran.
They reported the respective reduction and increment rate of 5 and
13.4% for sediment yield and income through implementing the
optimal decision. A constrained optimization model was built by
Heilman et al. (2003) that simulated the effect of imposing a con-
straint to reduce watershed sediment yield. The model calculated
a rancher’s profit, erosion and sediment yield for the Walnut Gulch
Experimental Watershed, USA, as a single ranch. The results indi-
cated little scope to reduce erosion in the short run by cutting herd
numbers without reducing the income of an already economically
stressed enterprise. Mohseni Saravi et al. (2003) used goal pro-
gramming in Garmabdasht watershed in Golestan province, Iran,
to determine the optimal land use pattern in the study watershed
based on economic, environmental and social criteria. Four land
uses, viz. industrial forest, pasture, park and protected areas were
optimized to satisfy a reasonable degree multiobjectives of max-
imization of benefit, production, employment opportunities, and
minimization of total investment and sediment yield. They then
notified that the scenario based on economic aspect could fulfill
other requirements in the study area. Kralisch et al. (2003) and
Riedel (2003) have successfully combined artificial neural network
and geographical information system with linear programming to
maximize benefits gained from land utilization in a watershed in
Germany and mountainous area of North Thailand, respectively.
Benli and Kodal (2003) developed a linear and non-linear opti-
mization model in South-east Anatolian watershed in Turkey for
the determination of optimum cropping pattern, water amount
and farm income under adequate and limited water supply con-
ditions. The objective function of the model was based on crop
water-benefit functions and model was solved using Microsoft
Excel Solver package. They reported that, the non-linear optimiza-
tion model could give higher farm income values than the linear
optimization model under deficit irrigation conditions. The effec-
tiveness of farm household land and forest allocation was also
evaluated with the help of linear programming by Tra and Egashira
(2004) at Tran Yen district in Japan. The results of their study
verified the significant rise in crop yield and forest production
after proper land and forest allocation. A planning support system
has also been applied by Farahpour et al. (2004) to the Chadegan
rangeland sub-region in Iran to illustrate its potential as a decision-
making tool using different scenarios. An optimization model was
introduced in a form of planning sub-module that examined the
degree of realization of various objectives of stakeholders and gen-
erated alternative solutions. The results suggested that maximum
land cover was the preferred scenario with the highest income and
carrying capacity, and the lowest level of subsidization. However,
since it eliminated wheat production, this scenario was culturally
less attractive. A watershed optimization model was also developed
on linear programming basis by Wang et al. (2004) that specified
the amount of land for each land use at a sub-area level in Lake
Erhai basin, China. It was incorporated into a GIS-based spatial allo-
cation model to provide specific location recommendations based
on existing land use, slope, distance to surface water and conver-
sion preference. Xevi and Khan (2005) developed a multi-criteria
decision-making framework to analyze production targets under
physical, biological, economic and environmental constraints. This
approach was successfully applied to the hypothetical irrigation
area using real data at Berembed weir on the Murrumbidgee River,
Australia. Ducourtieux et al. (2005) also criticized the present
land policy and farming practices in Laos and mentioned that the
 Author's personal copy

188 S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193

proper decision-making has to be made according to the governed

conditions emphasizing on local social relationships. A multiob-
jective optimization model of Smart Growth to land development
was presented by Gabriel et al. (2006) based on a formulation
and employing linear and convex quadratic objective functions
subject to polyhedral and binary constraints in a GIS environ-
ment for the stakeholders in Montgomery County, Maryland. The
results of the optimization problems for the study area were con-
vex, quadratic mixed integer programs. And recently, a modeling
approach was presented by Luo and You (2007) to investigate water
quality trading in soil erosion control, based on watershed sim-
ulation and optimization models in which various uncertainties
were reflected within the Swift Current Creek watershed, Canada.
The simulation–optimization approach was successfully bridged
through an assumed linear relationship between simulation out-
comes and decision variables in optimization models.
Despite the plethora of literature that exists on the optimization
field, frequent application of this applied mathematical approach in
the management of soil erosion as the single most important envi-
ronmental degradation problem in the developing world is lacking.
There are many efficient techniques for optimization of watershed
management out of which linear programming as a basic method
for many other optimization programs, impossibility of weighing
relative importance to different objectives required for goal pro-
gramming (Jianbo et al., 2002; Gabriel et al., 2006) in watershed
scale owing to inaccessibility of reliable data, and its simplicity
and applicability in different aspects (Chang et al., 1995; Amir and Fig. 1. General schematic view of Brimvand watershed, Iran.
Fisher, 1999; Benli and Kodal, 2003) as well as more availability of
associated softwares is usually preferred for application. Consid-
Table 1
ering scarcely documented researches in watershed management Some of important geographical characteristics of the Brimvand watershed in Iran
aspects applying optimization approaches, the present study has (Kermanshah Watershed Management Office, 2000)
been conducted to optimize land resources allocation to orchard,
Characteristics Value
range, irrigated and dry farming land uses within the Birmvand
watershed in Kermanshah province, Iran, using linear program- Total area (ha) 9572
Orchard 38.32
ming to minimize soil erosion and to maximize economic return. Rangeland 4001.27
The area represents the western Zagross part of Iran which is sup- Irrigated farming 4049.27
posed to be one of the most dynamic regions of Iran in terms of Dry farming 952.97
water and energy supply, industrial–urban development and agrar- Residential and outcrops 530.17
Maximum elevation (m amsl) 1310
ian activities. Consequently, land use conflicts and environmental
Minimum elevation (m amsl) 510
issues are also emerging increasingly in this region. The aforesaid Average elevation (m amsl) 651
watershed was specifically selected by virtue of its importance in Average slope (%) 10.6
the viewpoint of agricultural products in the region, availability Annual precipitation (mm) 485.4
of prerequisite studies, easy accessibility and threatening of the
existing water convey system in the watershed as a results of soil
ity/supply, soil characteristics, slope steepness and aspect, present
erosion.
land use, soil erosion and sediment yield, socio-economical con-
ditions were extracted from the available comprehensive studies
Materials and methods
(Kermanshah Watershed Management Office, 2000) conducted
for the area. In addition to the above-mentioned information,
Area description
some other field studies and land surveys were also conducted for
checking, and further details and information. The general benefit
The Brimvand watershed is located in upstream of agricultural
maximization problem was formulated as below:
canal of Brimvand dam and 4 km from north of Sarpole Zahab City in
Kermanshah Province, Iran. It comprises 9572 ha distributed within 
n

15 subwatersheds and extends between 34◦ 28 24 to 34◦ 36 08 N Max(Z1 ) = CBi Xi (1)
latitude and 45◦ 47 45 to 45◦ 54 46 E longitude (Fig. 1). Some of i=1
important geographical characteristics of the Brimvand watershed where Z1 is the total annual income in million Iranian Rails (mIR),
have been summarized in Table 1. The slope, land use and soil depth CBi is annual income for each land use (mIR/ha), Xi is the area of
distribution maps in the study watershed have been shown in Fig. 2. each land use in ha and n stands for numbers of land uses. If the
annual gross benefit, production cost and soil erosion destruction
Data acquisition and problem formulation cost per each hectare of different land uses is given by Ai1 , Ai2 and
Ai3 , the above equation can be rewritten as
The problem was structured in the study area to maximize
economic return and minimize soil loss. The information and 
n

data required for defining constants and coefficients of objective Max(Z1 ) = [Ai1 − (Ai2 + Ai3 )] Xi (2)
functions and constraints, viz. land availability, water availabil- i=1
 Author's personal copy
S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193
Fig. 2. Slope (top), land use (center) and soil depth (bottom) distribution map for
determination of allocable land resources within Brimvand watershed, Iran.
The general soil erosion minimization problem can also be
expressed as following form in which Z2 is the total annual soil
erosion (t) and CEi is the annual rate of soil erosion (t/ha) resulting
from different land use.

n
Min(Z2 ) = CEi Xi
i=1
Since these two objective functions (Eqs. (2) and (3)) are conflicting
demands for land use, and often the use and degradation of land in
189
one sector has serious implications for other uses of the resource,
resulting in a variety of economic and social conflicts and costs (e.g.
soil conservation in uplands costs money for upland farmers, but
benefits downstream), there is a need to allocate land resources
to different land use to aim at satisfying the existing natural, leg-
islation, cultural and social limitations. The above two objective
functions are therefore subject to the following constraints:
• Land capability constraints:
X 1 ≤ B1 (4)
X3 ≤ B2 (5)
X4 ≤ B3 (6)
X1 + X3 ≤ B4 (7)
• Land availability constraint:
X1 + X2 + X3 + X4 ≤ B5 (8)
• Social and legislation constraints:
X1 ≥ B6 (9)
X2 ≥ B7 (10)
• Non-negativity constraint:
X1 , X2 , X3 and X4 ≥ 0 (11)
where B1 to B4 are the maximum allowable area to orchard (X1 ), irri-
gated farming (X3 ), dry farming (X4 ), and summation of orchard and
irrigated farming. B5 denotes the maximum arable land resources.
B6 and B7 also, respectively represent the minimum area of orchard
and rangeland (X2 ) in ha. Since there are sufficient and accessible
water supply systems in Brimvand watershed, no constraint was
defined for water availability. There are 10 springs with discharges
from 2 to 453 l/s (16.9 Mm3 /year) and 128 wells with the total dis-
charge capacity of 11.2 Mm3 /year in the study watershed. The main
irrigation canal of Brimvand Dam with the average discharge of
5 m3 /s also passes along the entire watershed.
The above objective functions (Eqs. (2) and (3)) and constraints
(Eqs. (4)–(10)) were quantified in coordination with site-specific
information which directly or indirectly obtained for the study area.
Since the economic losses resulting from soil erosion have also to
be considered in the first objective function, the soil erosion esti-
mation in different land uses was conducted on the first onset. The
estimation of soil erosion was made using Pacific South-West Inter-
agency Committee method (PSIAC, 1968; Johnson and Gebhardt,
1982; Tangestani, 2005; Haregeweyn et al., 2005) for 15 hydro-
logic units after applying the concept of sediment delivery ratio
(Kermanshah Watershed Management Office, 2000). The model
consisted of nine factors, viz. surface geology, soil erodability, cli-
mate, runoff, topography, vegetation cover, land use, upland erosion
and gully erosion which were totally determined for the entire sub-
watersheds. The erosion estimates were then incorporated to the
land uses existed either thoroughly or partially in each hydrologic
unit (subwatershed) and the soil erosion rate associated with each
land use was ultimately found out. The possible land use modi-
fication was then incorporated in order to minimize soil erosion
rate based on land capability criteria (Rastegar, 1992; Farshi et al.,
1997a,b) and cultural, social and legal constraints.
The annual gross benefit, production cost and soil erosion
(3) destruction cost per each hectare of different land uses were esti-
mated according to Kermanshah Watershed Management Office
(2000) studies and interviews made with the inhabitants. Based
on the collected data and information, grape is the main orchard
product planted in terraced lands. There is no interest among the
 Author's personal copy
190 S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193
watershed inhabitants to invest on gardening for long term period
return. Meanwhile, the farmers are not up till convinced to have
commercial gardening through which they can be benefited much
more. The low level economic conditions of the watershed inhab-
itants make them to scare large investments with high risk. The
individual contracts between almost two thirds of the farmers and
the landowners to guarantee the low but reliable income are one of
the evident of such unreliable benefit to the farmers. The irrigated
areas are mainly used for wheat, corn, melon, alfalfa, cotton, and
bean plantation. Wheat, barely and peas are cultivated in dry farm-
ing areas as well under low tillage precaution. They usually prefer
dry farming land use, since according to them, it needs low atten-
tion and tillage activities through which they also ascertain their
ownership. The rangeland areas are also mostly being utilized for
sheep and goat grazing purposes and the rates of benefits were then
calculated based on the forage productions and the total digestible
nutrients (TDN) which feeds a particular number of animal units.
The dry forage production amounts of less than 50, 50–120 and
more than 120 kg/ha (Sadeghi et al., 2005b) were respectively con-
sidered for rangeland classification in three categories of light,
moderate and heavy grazing. Erosion destruction cost was esti-
mated by calculating the area lost to erosion in each land use
considering rooting depth and soil bulk density (Nikkami et al.,
2002). The coefficients of maximization objective function (Eq. (2))
were ultimately calculated using net benefit obtained through sub-
tracting total cost from gross benefit. The right hand side values of
Eqs. (4)–(10) were then determined based on land capability stan-
dards (Rastegar, 1992; Farshi et al., 1997a,b; Berengel, 2000) defined
according to slope steepness, soil depth and water availability as
well as cultural and legal constraints with the help of geographic
information system (Wang et al., 2004).
The benefit maximization and soil erosion minimization in the
Brimvand watershed were solved with the help of ADBASE model
which is capable to solve multiobjective problems (Steuer, 1994,
1995a,b) using the simplex method. In order to obtain the most
effective constraint as well as land use on changing objective
functions, which facilitates decision makers/managers to address
various alternatives (Chang et al., 1995), the sensitivity analysis was
also performed through subjecting the objective functions to a par-
ticular change of input resources (Nikkami et al., 2002) within the
permissible range of variation. The permissible ranges (the extend
up to which a individual variable can practically be increased or
decreased) were approximately assigned with respect to the poten-
tial of change of the variables under consideration. The percentiles
of changes were then depicted against each other’s and the most
sensitive land use was ultimately distinguished in both objective
functions.
Application of the model to Brimvand watershed
As was already explained, only two broad planning objectives
of economic development and soil erosion reduction were consid-
ered to be optimized in the Brimvand watershed. The soil erosion
rates were estimated to be 7.39, 8.14, 7.39 and 21.11 t/(ha year)
for orchard, rangeland, irrigated farming and dry farming land
uses, respectively. Since, the rooting depth and soil bulk density
in orchard, rangeland, irrigated farming and dry farming land uses
based on field studies and lab experiments were measured to be
1.00 ± 0.2, 0.15 ± 0.05, 0.50 ± 0.1 and 0.15 ± 0.06 m, and 1.08 ± 0.04,
1.11 ± 0.06, 1.08 ± 0.07 and 1.09 ± 0.05 t/m3 , respectively, the area
depleted owing to soil erosion found to be 6.84 ± 1.7, 48.91 ± 7.0,
13.68 ± 3.2 and 129.13 ± 16.6 m2 /year per each unit area (ha) of land
uses at sequence. The mean net benefit of orchard, rangeland, irri-
gated farming and dry farming land uses were therefore calculated
to be respectively some 8.50, 0.16, 4.88 and 0.32 mIR/ha. So that,
the objective functions of the benefit maximization and the soil
erosion minimization problems in the Brimvand watershed were
formulated as follows:
Max(Z1 ) = 8.5042X1 + 0.1562X2 + 4.8758X3 + 0.3215X4 (12)
Min(Z2 ) = 7.389X1 + 8.144X2 + 7.389X3 + 21.112X4 (13)
The above two objective functions were then subject to the fol-
lowing constraints. Considering no limitation for water availability
for all land uses, the maximum allocable area of 518.81 with slope
below 12% and soil depth beyond 0.65 m (Nikkami et al., 2002) was
contemplated for orchard as,
X1 ≤ 518.81 (14)
Almost 59% of the area lies between altitude ranges of
500–600 m amsl. Most of the watershed appears as hilly, plateaus
and alluvial fan land types. The slope of some 38% of the area is
below 2%. The maximum area of 4044.64 could therefore be desig-
nated for irrigated agriculture with slope below 5% and very deep
soil (>100 cm) (Rastegar, 1992 and Farshi et al., 1997a,b). That was,
X3 ≤ 4044.64 (15)
The upper slope limit of 12% based on the existing standards
(Rastegar, 1992; Farshi et al., 1997a,b) and government regulation
was applied for determining the maximum allocable land for dry
farming agriculture as given below:
X4 ≤ 1464.34 (16)
Based on the similarities between recommended standards for irri-
gated and orchard land uses and easiness of having access to water
resources, the following constraint was also formulated for the
study area.
X1 + X3 ≤ 4563.37 (17)
It was not also possible to change the utility of inhabitant, roads
and outcrops areas and these areas had to be therefore subtracted
from the entire watershed area and the rest area could be used
for optimization. In the other words the total land available for
development in the watershed was 9041.83 ha. That was
X1 + X2 + X3 + X4 ≤ 9041.83 (18)
According to the current cultural tendency of the people in this
region toward household gardening mainly for self-sufficiency and
amusement, the area under orchards could be limited at least at the
level of existing area of 38.32 ha.
X1 ≥ 38.32 (19)
Government regulation (Iran Forest and Rangeland Nationalization,
Act of 56) required that the rangeland area should be legitimately
no less than 4001.27 ha in this watershed for the purpose of natural
resources conservation. Therefore,
X2 ≥ 4001.27 (20)
The allocable areas to different land uses were also ultimately deter-
mined based on land capability, land availability, and social and
legislation constraints as explained before, under data acquisition
and problem formulation section, and it has been shown in Fig. 3.
The corresponding simplex method table (Steuer, 1995b) was
therefore extracted according to the formulated problem (Eqs.
(11)–(20)) for the study watershed as given in Table 2.
 Author's personal copy

S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193 191

explained before, the conversion of the present land uses to optimal

solution may not be assuredly implemented in the field and it may
be culturally less attractive because of the governing conditions on
farmers’ life in the study watershed. The socio-economic studies in
the area show that there are 32 villages with the total population of
7382 in the watershed out of which four villages were completely
evacuated due to migration to the cities vicinity to the area and
five more villages have also tribal/nomadic life system. Providing
the educational and extensional services by the local authorities
are therefore seriously required to encourage the farmers to really
obey the laws and keep in line with the optimal solution. How-
ever, an optimal solution as same as many other programs is not
always successful since uncertainties pertaining to watershed man-
agement are often not well acknowledged as mentioned by Luo
and You (2007) in case of soil erosion control. Such uncertainties
can be caused by both physical and socio-economical characteris-
tics of a watershed. Additionally, according to the interviews with
the farmers, the governmental support through subsidization and
providing agricultural assurance may positively affect changing
their manner of agricultural system from less investment – low
Fig. 3. Maximum allocable areas to different land uses in Brimvand watershed, Iran, income to more time and money investment – more efficient ben-
based on existing land use, slope, soil depth, and social and legislation limitations. efit. Meanwhile, stakeholders whose activities are on water or land
within the watershed have to involve in defining problems, set-
ting priorities and implementing solutions as noted by Wang et al.
Table 2
Simplex table of land use optimization in Brimvand watershed, Iran
(2004).
The results of the study verified that the linear optimization
Functions X1 X2 X3 X4 Modality Right hand side problem was successfully solved using the multipurpose ADBASE
Objectives software program (Steuer, 1995b) and the results (Table 3) led to
1 8.5042 0.1562 4.8758 0.3215 Max 0.00 maximize benefit as well as minimize soil erosion in the Brim-
2 −7.389 −8.144 −7.389 −21.112 Max 0.00
vand study watershed. The results also showed the successful
Constraints linkage between economic aspects and environmental outcomes
1 1 0 0 0 ≤ 518.81 at a watershed scale as emphasized by Recatala et al. (2000) and
2 0 0 1 0 ≤ 4044.64
3 0 0 0 0 ≤ 1464.37
Shively and Coxhead (2004), Peel and Lloyd (2007) and Gezelius
4 1 0 1 0 ≤ 4663.37 and Refsgaard (2007). Besides that, the applicability of linear pro-
5 1 1 1 1 ≤ 9041.83 gramming in solving optimization problem was proved as verified
6 1 0 0 0 ≥ 38.32 by Singh and Singh (1999) in maximizing production and benefit
7 0 1 0 0 ≥ 4001.27
in part of India, Amir and Fisher (1999) in introducing an optimal
linear model for analyzing agricultural production under various
scenarios in Israel, Salman et al. (2001) through presenting a lin-
Optimization results and analysis ear programming optimization model for analyzing inter-seasonal
allocation of irrigation water in Jordan, Nikkami et al. (2002) to
Optimal solution minimize soil erosion and maximize benefit for Damavand water-
shed in Iran, Benli and Kodal (2003) in optimizing cropping pattern,
From Tables 1 and 2, it can be seen that there is no serious change water amount and farm income in Turkey, Tra and Egashira (2004)
in irrigated farming and rangeland areas whereas the orchard area in allocating farm household and forest in Japan and Xevi and Khan
with a very small quantity of land occupancy has been increased (2005) in analyzing agricultural production targets under sundry
by 13.5 times and the dry farming area has been declined by 50%. constraints in Australia. Because of changes considered in land use
All these possible changes can be made within the areas qualified areas through optimization, the annual benefit suppose to grow
for each land use as depicted in Fig. 3. It is environmentally as well up from 21,001 to 24,911 mIR, i.e. some 18.62% growth, whereas
as economically preferred that some uses, mostly dry farming land the annual soil erosion would decrease from 82,910 to 76,380 t, i.e.
use, to be changed to orchards, since the steep slope areas with high about 7.87% reduction. Such type of land use allocation not only
erosion rate and less production are traditionally converted to level satisfied all governing constraints exist in the study watershed but
terraces by the farmers when they develop orchards. It has to also also ascertained socio-economic improvement, legitimate fulfill-
be clarified here that prior to change, the land needs to be precisely ment and environmental sustainability (Shababi Tabari, 1993; Onal
surveyed to avoid any hazardous unexpected problem. As it was et al., 1998; Chang et al., 1995; Chapi, 1998; Singh and Singh, 1999;

Table 3
Result of land use optimization in Brimvand watershed, Iran

Land use Area (ha) Erosion rate (t/ha year) Total erosion (t/year) Net income (mIR/ha year) Total income (mIR/year)

Orchard 518.81 7.389 3833.487 8.5042 4412.064

Range 4001.27 8.144 32586.343 0.1562 624.998
Irrigated farming 4044.56 7.389 29885.254 4.8758 19720.466
Dry farming 477.19 21.112 10074.835 0.3215 153.417
Total 9041.83 – 76379.919 – 24910.945
 Author's personal copy

192 S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193

Keristofer, 2001; Ebrahimi, 2001; Hasanzadeh, 2001; Rabet, 2001;

Schroder et al., 2002; Gezelius and Refsgaard, 2007) and sustain-
able watershed development (Chang et al., 1995; Kralisch et al.,
2003; Wang et al., 2004; Sadeghi et al., 2005a; Ducourtieux et
al., 2005; Gabriel et al., 2006). According to Boserup’s theory of
population pressure, poverty and soil erosion (Ananda and Herath,
2003) which hypothesizes that the rural poor are dependent on
natural resources for survival and hence poverty is a major source
of soil erosion as well as Onal et al. (1998) who indicated that
farm costs has been increased notably by restricting soil erosion,
it is emphasized that the success in implementation of optimal
solution in the study watershed would not be ascertained until
unless the basic demands of the habitants are taken into account
by the government. Population growth has intensified the agrar-
ian activities in the area forcing many farmers to cultivate easily
eroded hillslopes, deforestation and shifting cultivation to fulfill
their instant and present essential demands. Additionally, based
on Lopez’s theory on environmental and institutional dynamics of
soil erosion (Ananda and Herath, 2003) which posits that if envi- Fig. 4. Sensitivity analysis of benefit maximization function in Brimvand watershed,
Iran (B1 , B2 , B3 , B4 , B6 and B7 are maximum allowable area to orchard, irrigated
ronmental dynamics dominate institutional dynamics, then the soil
farming, dry farming, summation of orchard and irrigated farming, minimum area
erosion problem will be exacerbated, the environmental dynamics of orchard and rangeland, respectively).
can however be controlled in the Brimvand watershed as same as
many other watersheds in less developed countries. The soil erosion
exacerbation can be controlled through proper institutional estab-
lishment, and defining enforceable property rights and appropriate
polices. In consequent new institutions that protect the land will
emerge, while improving the economic status of the farmers. Tran-
sition to private property rights also may create incentives to adopt
soil conservation and optimal solution. As it has been emphatically
mentioned by Ananda and Herath (2003), it should be noted how-
ever, that private property rights are not the only property rights
that help arrest soil erosion. Common property resources are used
sustainably by farmers through collective action and social capi-
tal. The implication is that these systems need to be supported by
governments by providing adequate legal recognition and facilitat-
ing the definition of legal communal rights over common property
rights. The policies should emphasize support and improvement of
community institutions that permit the most efficient utilization
of the environmental resources. This would encourage public par-
ticipation where the role of the government is limited to observer,
facilitator or regulator. Fig. 5. Sensitivity analysis of soil erosion minimization function in Brimvand water-
shed, Iran (B1 , B2 , B3 , B4 , B6 and B7 are maximum allowable area to orchard, irrigated
farming, dry farming, summation of orchard and irrigated farming, minimum area
Sensitivity analysis of orchard and rangeland, respectively).

The sensitivity analysis was also performed for the benefit max-
imization as well as soil erosion minimization objective functions
and the corresponding results have been depicted in Figs. 4 and 5,
respectively.
Scrutinizing the results of sensitivity analyses (Figs. 4 and 5) ver-
ify that the changes in objective functions in both cases are linear
and they are mostly controlled by reduction in rather than increas-
ing the resources. It can also be verified here that the change of
some specific allocations would create much more impact on the
final optimal solutions generated by the optimization programming
as mentioned by Chang et al. (1995) in connection with variations
of parameter values versus the relative changes of decision vari-
ables in Taiwan. It could also be implied from Fig. 4 that reduction
in benefit has the highest sensitivity to the reduction of orchard and
irrigated farming areas whereas benefit increment is only sensitive
to increase in orchard area. It is seen in Fig. 5 that the reduction
of irrigated farming and orchard areas increased soil erosion dras-
tically. On the other hand, reduction in rangeland area leads to
increase soil erosion. In over all, the changes in benefit and soil
erosion in Brimvand watershed is mainly controlled by variation in
orchard and irrigated land uses.
Concluding remarks
A benefit and soil erosion problem was formulated and then
solved to minimize soil erosion and maximize benefits using opti-
mization of allocable land resources to orchard, range, irrigated and
dry farming land uses within the Birmvand watershed in Kerman-
shah province, Iran. The ADBASE optimization software program
was successfully applied and led to determine appropriate areas
allotted to different land uses. The results obtained during the
study approved the applicability of optimization model in solving
problems which sometimes conflicting each other. It can also be
concluded that, contrary to single objective classical land use plan-
ning models, the multiobjective linear programming can be used
to tractably search for optimum land use scenarios with respect
to different governing constraints existing within a watershed.
On the study watershed there appears a significant reduction in
soil erosion and augmentation in profit from allocating the opti-
mal land uses. This approach needs a number of improvements
in institutional approaches and public beliefs, nevertheless the
approach provides a framework for assessing the benefit to water-
 Author's personal copy
S.H.R. Sadeghi et al. / Land Use Policy 26 (2009) 186–193
shed inhabitants of reducing erosion and changing land uses. The
approach could provide better information on where changes are
required, how large the changes need to be, and how much the
changes will benefit the people when improving. Further studies
in the line of optimization approach using different methodologies
such as goal programming, numerical optimization, transportation
programming, non-linear optimization, dynamic programming,
parametric optimization, integer programming and mixed integer
programming are strongly recommended to assess their ability in
formulating watershed problems particularly when many dynamic
decision variables are involved. The conjunction of optimization
techniques with other tools like geographical information sys-
tem, genetic algorithm, fuzzy logic, artificial neuron networks and
applying different soft wares and simulation techniques are also
suggested to be taken into account in further studies to draw ulti-
mate necessary conclusions.
Acknowledgments
The authors profoundly are grateful to Kermanshah Water-
shed Management Office and Mr. Janfeshan for supplying valuable
information and assistance. They also appreciate the reviews of
respected anonymous referees of the manuscript.
References
Amir, A., Fisher, F.M., 1999. Analyzing agricultural demand for water with an opti-
mizing model. Agric. Syst. 61, 45–56.
Ananda, J., Herath, G., 2003. Soil erosion in developing countries: a socio-economic
appraisal. J. Environ. Manage. 68, 343–353.
Benli, B., Kodal, S., 2003. A non-linear model for farm optimization with adequate and
limited water supplies: application to the South-east Anatolian Project (GAP)
Region. Agric. Water Manage. 62, 187–203.
Berengel, K.J., 2000. Dry farming operation and principles. Translated by Rashed
Mohasel, M.H., Koochaki, A., Jahad Daneshgahi, Mashhad University.
Chapi, K., 1998. Study on type and amount of soil erosion in relation to land use
management and determination of the sediment ratio for land use optimization.
M.Sc. Thesis. Tarbiat Modares University, College of Natural Resources, Iran (in
Farsi, with English Abstr.).
Chang, N.B., Wen, C.G., Wu, S.L., 1995. Optimal management of environmental and
land resources in a reservoir watershed by multiobjective programming. J. Env-
iron. Manage. 44, 145–161.
Ducourtieux, O., Laffort, J.R., Sacklokham, S., 2005. Land policy and farming practices
in Laos. Develop. Change 36 (3), 499–526.
Ebrahimi, M., 2001. The evaluation of four empirical models for sediment estima-
tion in Dareh Ghanbarloo watershed of Parsabad-Moghan. In: Proceedings of the
National Congress on Land Management, Soil Erosion and Sustainable Develop-
ment, vol. 20, Arak, Iran, 21–23 January (in Farsi).
Farahpour, M., Van Keulen, H., Sharifi, M.A., Bassiri, M.A., 2004. Planning support
system for rangeland allocation in Iran with case study of Chadegan sub-region.
Rangeland J. 26 (2), 225–236.
Farshi, A.A., Shariati, M.R., Jarollahi, R., Ghaemi, M.R., Shahabifar, M., Tavallaee, M.,
1997a. An estimation of water requirement of main field crops and orchards in
Iran. Field Crops: Agricultural Education, vol. 1. Karaj, Iran (in Farsi).
Farshi, A.A., Shariati, M.R., Jarollahi, R., Ghaemi, M.R., Shahabifar, M., Tavallaee, M.,
1997b. An estimation of water requirement of main field crops and orchards in
Iran. Orchards, Agricultural Education, vol. 2. Karaj, Iran (in Farsi).
Gabriel, S.A., Faria, J.A., Moglen, G.E., 2006. A multiobjective optimization approach
to smart growth in land development. Socio-Eco. Plan. Sci. 40, 212–248.
Gezelius, S.S., Refsgaard, K., 2007. Barriers to rational decision-making in environ-
mental planning. Land Use Policy 24 (2), 338–348.
Haregeweyn, N., Poesen, J., Nyssen, J., Verstraeten, G., 2005. Specific sediment yield
in Tigray-Northern Ethiopia: assessment and semi-quantitative modeling. Geo-
morphology 69, 315–331.
Hasanzadeh, M., 2001. The role of effective information in soil erosion and sus-
tainable development. In: Proceedings of the National Congress on Land
Management, Soil Erosion and Sustainable Development, vol. 85–93, Arak, Iran,
21–23 January (in Farsi).
Heilman, P., Duan, Y., Miller, R., Guertin, D.P., 2003. Calculating the cost of reducing
erosion from a small rangeland watershed. In: Proceedings First Interagency
Conference on Research in the Watersheds, Benson, AZ, 27–30 October, pp.
398–404.
Jianbo, L., Zhaoqian, W., Penning de Vries, F.W.T., 2002. Application of interactive
multiple goal programming for red soil watershed development: a case study of
Qingshishan watershed. Agric. Syst. 73, 313–324.
Johnson, C.W., Gebhardt, K.A., 1982. Predicting sediment yields from Sagebrush
rangelands. In: Proceedings of the Workshop on Estimating Erosion and Sed-
View publication stats
193
iment Yield on Rangelands, Tucson, Arizona, 7–9 March 1981. USDA-ARS,
Agricultural Reviews and Manuals, ARM-W 6, 145–156.
Kermanshah Watershed Management Office, 2000. The studies of Brimvand Water-
shed, Jihad Engineering Co.
Keristofer, J.B., 2001. Principles and methods of environmental management. Trans-
lated by Andaroodi, M., Kongereh, Tehran, Iran.
Kralisch, S., Fink, M., Flugel, W.A., Beckstein, C., 2003. A neural network approach
for the optimisation of watershed management. Environ. Model. Software 18,
815–823.
Luo, B., You, J., 2007. A watershed-simulation and hybrid optimization modeling
approach for water-quality trading in soil erosion control. Adv. Water Resour. 30
(9), 1902–1913.
Mohseni Saravi, M., Farzanegan, M., Koopaee, M., Kholghi, M., 2003. The determina-
tion of optimal land use pattern in watershed resources using goal programming.
Iranian Nat. Resour. J. 56 (1/2), 3–16 (in Farsi, with English Abstr.).
Nikkami, D., Elektorowicz, M., Mehuys, G.R., 2002. Optimizing the management of
soil erosion. Water Qual. Res. J. Canada 37 (3), 577–586.
Onal, H., Algozin, K.A., Isık, M., Hornbaker, R.H., 1998. Economically efficient water-
shed management with environmental impact and income distribution goals. J.
Environ. Manage. 53, 241–253.
Peel, D., Lloyd, M.G., 2007. Neo-traditional planning. Towards a new ethos for land
use planning? Land Use Policy 24 (2), 396–403.
PSIAC, 1968. Report of the water management subcommittee on factors affecting
sediment yield in the Pacific Southwest Area and selection and evaluation of
measures for reduction of erosion and sediment yield. ASCE, 98, Report No. Y12,
p. 27.
Rabet, A., 2001. The role of watershed management to streamline the tourism attrac-
tions and recreational activities (A case study in the North and South of Zanjan
Watershed). In: Proceedings of the National Congress on Land Management, Soil
Erosion and Sustainable Development, Arak, Iran, January 21–23, pp. 258–266
(in Farsi).
Randhir, T.O., O’Connor, O., Penner, P.R., Goodwin, D.W., 2001. A watershed-based
land prioritization model for water supply protection. Forest Ecol. Manage. 143,
47–56.
Rastegar, M., 1992. Dry Farming, First ed. Berahmand Publication, Tehran, Iran (in
Farsi).
Recatala, L., Ive, J.R., Baird, I.A., Hamilton, N., Sanchez, J., 2000. Land-use planning
in the Valencian Mediterranean region: using LUPIS to generate issue relevant
plans. Journal of Environmental Management 59, 169–184.
Riedel, C., 2003. Optimizing land use planning for mountainous regions using LP and
GIS towards sustainability. J. Soil Conserv. 34 (1), 121–124.
Sadeghi, S.H.R., Jalili, Kh., Nikkami, D., 2005a. Land use planning: a practical approach
in soil erosion control. In: Proceedings of International Conference on Human
Impacts on Soil Quality Attributes, Isfahan, September 12–16, p. 5.
Sadeghi, S.H.R., Fattahi, B., Safaeian, N.A., Erfanzadeh, R., 2005b. Effects of grazing on
some of soil physico-chemical characteristics in Lar Region National Park, Iran.
Hydrol. J. Indian Assoc. Hydrol. 28 (1/2), 79–88.
Salman, A.Z., Al-Karablieh, E.K., Fisher, F.M., 2001. An inter-seasonal agricultural
water allocation system (SAWAS). Agric. Syst. 68 (3), 233–252.
Schroder, P., Huber, B., Olazabal, U., Kammerer, A., Munch, J.C., 2002. Land use
and sustainability: FAM research network on agroecosystems. Geoderma 105,
155–166.
Seppelt, R., Voinov, A., 2002. Optimization methodology for land use patterns using
spatially explicit landscape models. Ecol. Model. 151, 125–142.
Shababi Tabari, H., 1993. Optimal use of soil and water resources as a national
necessity for sustainable agriculture development. In: Proceedings of the First
Congress on Soil and Water Infrastructural Affairs Programs in Agricultural Sec-
tor, Tehran, Iran, September 16–17 (in Farsi).
Shively, G., Coxhead, I., 2004. Conducting economic policy analysis at a landscape
scale: examples from a Philippine watershed. Agric. Ecosyst. Environ. 104 (1),
159–170.
Singh, A.K., Singh, J.P., 1999. Production and benefit maximization through optimal
crop planning—a case study of Mahi Command. Indian J. Soil Conserv. 27 (2),
152–157.
Steuer, R.E., 1994. Random problem generation and the computation of efficient
extreme points in multiple objectives linear programming. Comput. Optim. J. 3,
333–347.
Steuer, R.E., 1995a. The ADBASE Multiple Objective Linear Programming
Package/Multiple Criteria Decision-Making. SCI-Tech, Windsor, England,
pp.1–6.
Steuer, R.E., 1995b. Manual for ADBASE, multi objective linear programming package,
Faculty of Management Science, 297 Brooks Hall. University of Georgia, Athens,
USP.
Tangestani, M.H., 2005. Comparison of EPM and PSIAC models in GIS for erosion and
sediment yield assessment in a semi-arid environment: Afzar Catchment, Fars
Province, Iran. J. Asian Earth Sci. 27, 585–597.
Tra, N.T., Egashira, K., 2004. Land use effectiveness by farm households after land
and forest allocation at Tran Yen district, Yen Bai province. J. Fac. Agric. Kyushu
Univ. 49 (2), 461–466.
Wang, X.H., Yu, S., Huang, G.H., 2004. Land allocation based on integrated GIS-
optimization modeling at a watershed level. Landscape Urban Plan. 66 (2),
61–74.
Xevi, E., Khan, S., 2005. A multi-objective optimisation approach to water manage-
ment. J. Environ. Manage. 77, 269–277.