Systems Approach to Exploratory Analysis of Agricultural
Land Use Options in Ilocos Norte, Philippines
Felino P. Lansigan
University of the Philippines Los Banos, 4031 College, Laguna, Philippines
Abstract: Increasing population in urban and rural areas continues to exert pressure on land and water
resources through changing land use, intensive and extensive agricultural production systems, land
conversion, and rapid urbanization. These changes transform the spatial and temporal distribution of the
quantity and quality of natural resources which threatens sustainable agricultural development. An
objective and efficient optimal allocation and management of these natural resources is required to ensure
Major advances have been made in recent years on the use of models or decision support systems
for natural resources management. Systems research tools such as dynamic simulation models, geographic
information systems (GIS), optimization technique and databases, which are in the past being used
separately in various practical applications are now integrated as a management decision tool to explore
alternative natural resources use and management options. Multiple goal analysis provides the framework
in which land and water resources management is approached as a multi-objective optimization problem
where development and management priorities and agenda including the biophysical as well
socio-economic and institutional constraints and the requirements for environment conservation are
identified and quantified. The procedure also provides a framework for dialogue and negotiations among
various stakeholders such as policy and decision makers, development planners, and representatives of
different administrative hierarchies (e.g. region, province, district and village).
This paper presents a systems research framework for exploratory analysis of agricultural land use
options and its application to regional (provincial) agricultural land use planning in a province in
northwestern Philippines as a case study. The framework for modeling as an multiple goal linear
programming problem is presented, and the important components are discussed. Achievements of target
values defined for certain objective functions and goals associated with some possible resource allocation
options and corresponding tradeoffs are described.
Keywords: Systems analysis; linear programming; land use options.
Increasing population in both urban and rural areas continues to subject several regions to the
pressure of land use change, land degradation, water shortage, and loss of bio-diversity. The demand for
food has continued to increase which exert pressure on limited natural resources like land and water
resources for crop and livestock production. While crop production has improved tremendously during the
post Green Revolution period resulting to significant increase in productivity per unit area, crop produced
per capita has remained constant due to growing population. Consequently, food production threatens the
natural resources through conversion of agricultural lands due to rapid urbanization, exploitation and
competition for uses of limited resources, and unsustainable land uses.
Sustainable development of an region through a more rational natural resources management
requires an optimal agricultural land use plan based on sound scientific knowledge base. An innovative
systems analysis-based approach to explore and analyze possible land use options which address the
development goals and priority objectives of the region is needed. Moreover, a major challenging issue in
efficient and effective agricultural land use planning and analysis is incorporating the views, goals and
priorities of various stakeholders as well as in communicating the results for stakeholders’ negotiation
and tradeoff analysis. The systems approach to exploratory land use analysis ensures that the stakeholders
such as agricultural development planners, decision- and policy-makers use a rational procedure in the
analysis and evaluation of various land use options.
The systems approach to the exploratory analysis of land use options for sustainable natural
resources management and agricultural development can be implemented following the Land Use Planning
and Analysis System (LUPAS) framework (Roetter et al., 2000). Land use planning is approached as a
multiple goal optimization problem which considers a set of objectives and policy views of stakeholders
(e.g. planners, agricultural officers and extension workers, advisors and policy makers, etc.), the
consequences of which and tradeoffs involved must be evaluated. The optimal land use options for a
specified development scenario is obtained using the Multiple Goal Linear Programming (MGLP)
framework (van Ittersum, 1997) by optimizing the set of priority objectives defined by stakeholders
considering the biophysical potentials of the resources available, and the socio-economic constraints in the
This paper presents the application of the LUPAS methodology to the exploratory analysis of
agricultural land use options for the province of Ilocos Norte in northwestern Philippines. The case study
area is briefly described, and the implementation of the study through multidisciplinary and
multi-institutional collaboration is presented. The case study also identifies some opportunities for
agricultural development of the region by exploring the optimal agricultural land use options under specific
development scenarios. These scenarios are translations of policy views in terms of objective functions
optimized. The analyses also highlights potential policy recommendations (e.g. sharing of available
resources like water and labor, use of high or improved agricultural technology level) to meet regional
Case Study Area: Ilocos Norte Province, Philippines
Ilocos Norte Province is located in the northern west tip of the Philippines and lies between 190 N
and 210 N (Figure 1). Total land area is 0.340 M has. of which 129, 500 has. are classified as
agricultural lands. It has two distinct climate types: wet and dry seasons, i.e. predominantly dry season
from November to April, and wet season from May to October. Mean annual rainfall is about 2000 mm.
The province experiences frequent storms of very high intensity and duration associated with typhoons
during the wet season.
The province consists of 23 administrative units (22 municipalities and 1 city, Laoag City) (Figure
1). The population of the province is about 0.5 M with an annual growth rate of 1.68%, average
household size of 5 persons, population density of 136 per km-2, and a high employment rate of 92.8% of
which 51% of the total labor force is engaged agriculture, fishery and forestry (NSO, 1995).
The province has high degree of agricultural crop diversification and intensification, and thus offers
an opportunity for exploring possible cultivation of crops other than the staple rice crop and the high value
crops such as tomato, garlic and onion after rice. Ilocos Norte has produced two important documents
relevant to land use planning, namely: (i) the Ilocos Norte Provincial Physical Framework Plan /
Comprehensive Provincial Land Use Plan (1993-2002), and (ii) the Sustainable Food Security Action Plan
and Framework for Agro-fishery Modernization (1999-2001). These plans serve as excellent references
on the development goals and priority objectives for the province.
Systems Approach to Land Use Planning
The operational structure of the systems analysis approach to land use planning referred to as
LUPAS (Land Use Planning Analysis System) used in the study in illustrated in Figure 2. The
systems-based methodology for agricultural land use planning requires the delineation of land units (LUs)
and the characterization of their bio-physical attributes. Determination of the LUs for Ilocos Norte is
facilitated by using the geographic information systems (GIS) technique which delineated the
homogeneous areas in terms of rainfall distribution, annual rainfall volume received, soil properties (slope,
soil texture), and irrigation potentials. The capability of each LU is evaluated in terms of its potentials for
crop production under the given agro-environments, and under two levels of production technology.
Input-output tables were generated using relevant socio-economic data and information from a recent farm
survey as well as by rationalization based on scientific and expert judgment. Optimization analysis is
facilitated using the MGLP software such as XPRESS (Dash Associates, 1999).
The systems approach in the case study requires the participation of a multidisciplinary team of
researchers from academic and research institutions with the different stakeholders such as the provincial
planners, provincial agriculturist, municipal agricultural officers and extension workers, and farmer leaders
in the province. The participatory aspect of the case study was implemented through a series of
consultative meetings and workshops with stakeholders where they defined and articulated the policy views
based on available development plans in the different hierarchy of decision-making in the region
(provincial and municipal levels). The multi-disciplinary team was trained in the systems research tools like
MGLP, crop modeling, and GIS.
Resource balance and land evaluation. The total available area for agricultural development of
about 129,550 has. was determined by excluding the areas not suitable for agriculture. Areas identified as
unsuitable for agricultural development are those areas which are (1) severely eroded areas; (2) steep
(30-50% slope) and very steep (slope > 50%) areas; (3) mountainous soils, river wash, dune land, sand
and coral bed, and rock land; and (4) current land use such as forest, water bodies (rivers, lake), built-up
areas. For the future scenario (2010), the proposed land use map of the province was used to determine
additional land to be allocated for residential, commercial, and industrial uses. An additional 9,700 has. of
land for built-up areas is indicated. Hence, the total available land for agriculture in 2010 is reduced to
119,850 has. with the corresponding area for each LU determined by also using a GIS software.
The agro-ecological units (AEUs) for the province were obtained by overlaying maps of the
province describing its biophysical characteristics. Five (5) biophysical characteristic maps were digitized,
aggregated/disaggregated, and reclassified into corresponding classes, namely: (a) irrigated areas (2 classes;
irrigated and not irrigated); (b) monthly rainfall pattern (3 classes; 2-4 dry months, 5-6 dry months, and 7
dry months); (c) annual rainfall (2 classes; 2,000 mm. and > 2,000 mm.); (d) slope (2 classes; 18% and
> 18%); and (e) soil texture (3 classes; fine, medium, and coarse). Overlaying the 5 maps of biophysical
characteristics resulted to 37 AEUs. The resulting thematic map of AEUs was overlaid to the
socioeconomic characteristics taken as the boundaries of the 23 administrative (political) units (22
municipalities and 1 city) resulting to 200 land units (LUs) having homogeneous biophysical and
Development and crop production scenarios. Two crop production technology levels are
distinguished, namely: (1) the current technology (using average crop yields); and (2) the high technology
(based on high yield level). Input-output data for both levels were determined from the latest farm crop
production survey conducted during the case study as well as from the farm socio-economic survey and
monitoring of selected farms in several villages.
Scenarios indicating possibilities for sharing of limited resources in the province like labor and
water resources are also considered. Labor sharing scenario assumes that the labor force can move from a
municipality to other adjacent municipalities in the province which can easily be realized due to the ease in
mobility in most part of the region. On the other hand, water resources sharing is assumed possible
among adjacent land units served by the same existing irrigation systems. Expansion of irrigable service
area for a particular irrigation system can be realized through improvement in water use efficiency by
reducing water delivery losses.
Achieving the development goals such as maximizing rice production for the province, maximizing
farmers’ income as well as the optimum allocation of land resources to the different LUTs are compared
with different policy constraints for the current year (1999/2000) and for the future scenario (2010). For
the latter, values of the variables (land resource, demand for each crop, etc.) were correspondingly adjusted
based on provincial targets, and estimated demand considering projected population increase and per capita
consumption for each crop.
Labor. Considering its seasonal demand the available labor (in 1000 man-days/ha ) for crop
production activities or enterprises were calculated per land unit and for each month in the study area. It
is assumed that 45% of the rural population contributes to the total labor supply for each month with each
individual providing 23 working days per month. It is further assumed that mobility in the province
allows the sharing of labor force from the rural communities.
Water resources. Water resources available for crop production is assumed to come mainly from
rainfallreceived, from groundwater, and from existing irrigation systems in the area. Expected available
rainfall is calculated using existing rainfall stations in the study area. The mean monthly rainfall among
common years and the coordinates of rainfall gauging stations were linked into the digital map using GIS.
Then, using the inverse weighted-linear interpolation technique, isohyets were derived in the study area.
The available rainfall volume per month was calculated to be 85% of the GIS-estimated rainfall per month.
The amount of groundwater in 1000 m3 ha-1month-1 was estimated for the current year and for 2010 as
the product of the inflow adjusted (by a correction factor) for soil type and eco-region.
There are 24 existing irrigation systems in the province: 13 National irrigation systems (NIS) and 11
communal irrigation systems (CIS). Irrigation service areas were delineated based on the rice-based
cropping systems map reclassified using GIS. Two scenarios were explored on this resource, namely: (a)
no water-sharing; and (b) with water-sharing. The available irrigation water for each scenario was
Land use systems. Since the province is a highly crop-diversified area, there are 17 agricultural
crops considered, namely: rice, white and yellow corn, garlic, mungbean, peanut, tomato, tobacco, cotton,
potato, onion, sweet pepper, eggplant, vegetables, rootcrops, sugarcane, and melon. The combination of
these products in cropping systems resulted to 23 land use types (LUTs). The existing LUTs considered
were based on current practices, socio-economic surveys in irrigated, rainfed lowland and rainfed upland
ecoregions. Thus a detailed farmer’s level survey was conducted on the current input-output crop
productions systems in the three eco-regions in each municipality in cooperation with the municipal
agricultural officers. The high and current (average) levels of crop production technologies were assumed
in the study area based on the yield levels obtained from the farm survey.
The various input-output parameters for each crop production systems (by municipality by crop by
eco-region) and estimates of available resources (land, water and labor) were stored as a Microsoft Excel
file with sub-worksheets. Database Lookup functions of Excel was used for relational and referencing
items among main and sub-worksheets. In order to link the data in the input-output tables of the crop
production activities and output of optimization results, the MapLink tool (linking data in Excel files to a
GIS) is used.
Results and Discussion
Development Goals, Objectives and Priorities
Consultative meetings with stakeholders from provincial and municipal levels, and reviews of
available development plans revealed the following goals and priorities for agricultural development in the
(1) Maximizing crop production. Intensive and diversified cropping systems are already being
practiced in the area with an average farm size of 1 ha. per household.
(2) Maximizing employment generation. Labor supply principally from the rural population is not
fully utilized year-round since the cropping season is dependent on weather and climate as well as water
conditions for irrigation.
(3) Maximizing labor use efficiency. This aims to optimize use of labor for agricultural activities
while achieving targets for crop production so that excess labor can be used for other income generating,
non-agricultural enterprises. While demand for labor is seasonal, it is often quite limited during the
(4) Minimizing water use. Water for irrigation comes mainly from rainfall and surface water
during the wet season. Supplemental irrigation from subsurface water using shallow tube wells and water
pumps are also used particularly during the dry season.
(5) Minimizing environmental degradation. This is translated to reducing soil erosion through
proper choice of crops and use of appropriate tillage practices as well as minimizing use of chemicals from
fertilizer applications and for crop protection.
These goals and objectives are translated into objective functions to be optimized as an MGLP
problem imposing some relevant constraints such as limitations on land for agriculture, water for irrigation,
and labor supply, namely:
(1) Maximize rice production
(2) Maximize non-rice production
(3) Maximize employment in agriculture
(4) Maximize farmers’ income
(5) Maximize total provincial income
(6) Minimize soil erosion
(7) Minimize use of biocides
(8) Minimize fertilizer use
Optimization runs with the objective functions of maximizing rice production and maximizing
income considering the various development goals assuming no constraints (i.e. no limitations on land,
water, and labor) indicate that every million peso increase in income translates to about 9 ton decrease in
crop production (representing 60% decrease); to 16 man-days (10%) increase in employment; to 4 kg.
(1013%) increase in pesticide usage; and to 177 kg (27%) increase in fertilizer usage.
Under both scenarios of maximizing rice production and of maximizing income, all available land
for agriculture (129,650 has.) are fully allocated. However, the difference is that in the former, cropping
systems consist of double rice crops and triple rice crops under high (or improved) technology level.
This is also very labor intensive. On the other hand, the latter allocates the available land mainly for
rice-tomato cropping system under high technology level since tomato is a high value crop.
Maximizing employment in agriculture yields agricultural land uses which are more labor intensive
such as Rice-White Corn-Mungbean (RWM), Rice-Garlic-Mungbean (RGM), and Rice-Eggplant (REg)
cropping systems. Maximizing income results to land use option with high value crops like tomato, garlic
and onion after rice under high technology level.
Two scenarios for water availability are considered in the analysis: (1) without water-sharing, and
(2) with water-sharing among adjacent municipalities. The second scenario assumes that areas served or
adjacent to existing irrigation systems can access water for irrigation, and that current water use efficiency
can be improved. Analysis also shows that if water resources can be shared by neighboring municipalities,
rice production and income of the municipalities as well as the province can be increased. Rice
production will increase by 49%, and income will income by 26%.
With high technology level, increasing rice production, employment in agriculture and income can be
further enhanced. That is, the effect of technology on rice production an income is very significant. With
water sharing, switchi planted to rice decreases as more constraints (policy preferences) are imposed in the
Future Development Scenario (2010)
The province anticipates the conversion of about 9,700 has. from agricultural land for
non-agricultural purposes to meet the need of increasing population. Comparison of the values of the
objective functions and the land use allocations for the current and future (2010) scenarios indicate that
allowing land conversion due to increase in population translates to low income and low employment.
Although there is increased rice production for same scenarios, there is a tremendous pressure to the soil
resource base due to more intensive rice mono-culture. The area devoted to 3 rice cropping increases
from 21% to 35%. This has implications on the sustainability of rice production system as well as on the
overall goal of environmental stability.
The case study has demonstrated the applicability of the systems approach in exploring optimal ng
to a higher production technology increase rice production by 62%, employment by 26%, and income by
167%. Without water sharing, the effect is more pronounced in terms of percentage change. Rice
production increased by 104%, employment by 38%, and income by 181%. Comparison of rice
production under different scenarios imposing various constraints on resources such as land area for
agriculture, available labor, water resources, demand for crops and market ceilings. The area agricultural
land use options for Ilocos Norte. While policy measures can not be explicitly identified from the model,
the procedure can be used to evaluate and analyze the consequences of a specific policy on the multiple
goals for which opportunities for development may be suggested (e.g. water sharing scenario with
improved irrigation system; promoting of high technology in crop production systems in the province).
However, the methodology has specific limitations. It is not a prediction model nor dynamic but a
multiple goal optimization model based on a particular scenario. It should be emphasized that the
optimization is focused only on the agricultural sector only, and the analysis is at the provincial level only,
i.e. optimizing of objectives functions at the municipality levels is not included. The optimization
considers only few priority development objectives which are translated to mathematical objective
functions and solved as a MGLP problem.
Other development goals and objective functions may be incorporated in the analysis.
Optimization of land use options at both the provincial and municipal levels may also need to be considered.
This will provide useful information in tradeoff analysis and resolution of potential conflicts among
stakeholders on the optimal land use options considering the priorities and constraints at different levels.
It is also imperative that a decision support system (DSS) such as in Laborte et al. (2001) can be developed
using LUPAS framework to provide development planners and decision-makers a useful management tool
for analyzing land use options.