Environmental Modelling & Software 22 (2007) 149e157 www.elsevier.com/locate/envsoft Combining farm and regional level modelling for Integrated Resource Management in East and South-east Asia Reimund P Roetter a,*, Marrit van den Berg b,c, Alice G Laborte d, Huib Hengsdijk e, Joost Wolf a, Martin van Ittersum b, Herman van Keulen b,e, Epifania O Agustin f, Tran Thuc Son g, Nguyen Xuan Lai h, Wang Guanghuo i a Alterra, Soil Science Centre, Wageningen UR, P.O Box 47, 6700 AA Wageningen, The Netherlands Plant Production Systems, Wageningen University, P.O Box 430, 6700 AK Wageningen, The Netherlands c Development Economics, Wageningen University, P.O Box 8130, 6700 EW Wageningen, The Netherlands d International Rice Research Institute (IRRI), DAPO, P.O Box 7777, Metro Manila, Philippines e Plant Research International, Wageningen UR, P.O Box 16, 6700 AA Wageningen, The Netherlands f Mariano Marcos State University, Batac, Ilocos Norte, Philippines g National Institute for Soils and Fertilizer, Hanoi, Vietnam h Cuu Long Delta Rice Research Institute, Omon, Cantho, Vietnam i Zhejiang University, Hangzhou, PR China b Received 29 July 2004; received in revised form February 2005; accepted 14 July 2005 Available online 18 November 2005 Abstract Currently, in many of the highly productive lowland areas of East and South-east Asia a trend to further intensification and diversification of agricultural land use can be observed Growing economies and urbanization also increase the claims on land and water by non-agricultural uses As a result, decisions related to the management and planning of scarce resources become increasingly complex Technological innovations at the field/farm level are necessary but not sufficient e changes in resource use at regional scale will also be essential To support decision-making in such situations, we advocate a multi-scale modelling approach embedded in a sound participatory process To this end, the Integrated Resource Management and Land use Analysis (IRMLA) Project is developing an analytical framework and methods for resource use analysis and planning, for four sites in Asia In the envisaged multi-scale approach, integration of results from field, farm, district and provincial level analysis is based on interactive multiple goal linear programming (IMGLP), farm household modelling (FHM), production ecological concepts and participatory techniques The approach comprises the following steps: (i) inventory/quantification of current land use systems, resource availability, management practices and policy views, (ii) analysis of alternative, innovative land use systems/technologies, (iii) exploration of the opportunities and limitations to change resource use at regional scale under alternative future scenarios, (iv) modelling decision behaviour of farmers and identification of feasible policy interventions, and (v) synthesis of results from farm to regional level for negotiation of the most promising options by a stakeholder platform In the current paper, the operationalisation of dual-scale analysis is illustrated by the outputs (development scenarios, promising policy measures and innovative production systems) from various component models for the case study Ilocos Norte, Philippines An approach is discussed for the integration of results from the different model components at two different decision making levels (farm and province) Ó 2005 Elsevier Ltd All rights reserved Keywords: Land use conflicts; Scenario analysis; Linear programming; Bio-economic models; Rice-based farming; Philippines * Corresponding author Tel.: C31 317 474229; fax: C31 317 419000 E-mail address: reimund.roetter@wur.nl (R.P Roetter) 1364-8152/$ - see front matter Ó 2005 Elsevier Ltd All rights reserved doi:10.1016/j.envsoft.2005.07.015 150 R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 Software availability Name of the software: IRMLA land use analysis tools/LP models for Dingras/Ilocos Norte IRMLA land use analysis tools contain farm household and regional LP models and a generic technical coefficient generator LP models: linear programming models for case study Dingras/Ilocos Norte Developer and contact address: (i) IRMLA tools e developed by the IRMLA Project; Coordinator: Soil Science Centre, Wageningen UR, Droevendaalsesteeg 3; 6700 AA Wageningen, The Netherlands, www.irmla alterra.nl; (ii) Developer LP models for specific case study: Marrit van den Berg, Social Sciences Group, Wageningen University Tel.: +31 317 47 4327 or 4229; fax.: +31 317 419 000 E-mail: reimund.roetter@wur.nl; (Alterra) marrit.vandenberg@ wur.nl (Development Economics Group) Year first availability: 2005 Hardware required: A platform to run MS WindowsÔ 32-bit applications Software required: GAMS 2.50 Professional Version, GAMS Development Cooperation; and Microsoft Office (Excel) Program language: GAMS (LP models) Program size: 10 MB (whole set) Availability and cost: at request Introduction Agricultural systems in East and South-east Asia are being challenged by the simultaneous requirements for increased productivity, more diversified products and reduced environmental impact, creating potential conflict situations among various stakeholder groups regarding land use objectives and resource use Current land use policies in general inadequately take into consideration multiple objectives and the increasingly complex nature of resource management decisions (Walker, 2002; Lu et al., 2004) In such situations, effective systems analysis tools at different scales may be useful to identify conflicts and design sustainable land use systems and supportive policy options (Van Ittersum et al., 1998) Since the early 1980s, a range of complementary analytical frameworks and operational tools have been developed (Kuyvenhoven et al., 1998; Stoorvogel and Antle, 2001; Van Ittersum et al., 1998, 2004) On the basis of their objectives we can distinguish explorative and predictive tools Explorative tools analyse the potential for different natural resource use strategies, often at regional or farm scale To this purpose, a frequently used procedure is interactive multiple goal linear programming (IMGLP) (Spronk and Veeneklaas, 1983; De Wit et al., 1988) IMGLP models generate optimal land use options under different sets of objectives and constraints Regional IMGLP models as operationalised in the SysNet project (Roetter et al., 2005) form one of the major building blocks of the Integrated Resource Management and Land use Analysis (IRMLA) approach to multi-scale analysis So-called predictive tools are required to analyse the likely land use changes in the short term as a result of introducing alternative agricultural policies and technologies (Bouman et al., 2000) For example, the technique of farm household modelling (FHM) is applied for simulating the impact of feasible changes in policy and technology choice for different (model) farm groups in a study area (Kruseman and Bade, 1998) FHM is the second major tool in the IRMLA project In most cases, the various modelling approaches, whether exploratory or predictive have been applied separately at a single scale This can only shed partial light on solutions to agricultural and environmental policy problems which are essentially of a multi-scale nature Policy makers at the provincial level, for instance, have only a limited number of variables that they can control Variables such as choice of crop, area cultivated and fertilizer and pesticide rates are decided by a huge number of other decision makers, i.e farmers, which apply different criteria Candler et al (1981) addressed this problem and examined the potential contribution of multilevel programming to solve two-level (publiceprivate interest) conflicts They detected a range of algorithmic problems in multilevel programming Solutions were only found for special cases To make things even more complicated, public interest at one level (e.g province) may be in conflict with the public interest at another level (e.g municipality) Integration of results from different scales, hence, remains a research challenge (Bouman et al., 2000) In this paper we not intend to resolve the problems inherent to multilevel programming Rather we want to demonstrate that, as a first step, combination of farm household modelling and regional multiple goal linear programming embedded in participatory processes can overcome shortcomings of single-level modelling This will be illustrated by confronting results from regional level explorations with farm household level analysis of the best land use strategy in Ilocos Norte province, Philippines The result from this dual-scale analysis will help to identify the options for promoting more resource use-efficient production technologies than presently practiced Case study characterization 2.1 Site description Ilocos Norte province, in north-western Luzon, Philippines, has a population of nearly 0.5 million people and a total land resource of 0.34 million ha, of which 46% is covered by forests Mean annual rainfall ranges between 1650 mm in the southwest to more than 2400 mm in the eastern mountain ranges On average, 6e7 typhoons per year cross the province (mostly between August and November) Soils have developed from very diverse parent materials In the lowlands, sandy loams developed from alluvial deposits are predominant (Fig 1) About 38% of the total area is classified as agricultural R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 151 Fig Ilocos Norte province: location in the Philippines (left) and map showing soil texture groups and location of Dingras municipality indicated (right) land (Roetter et al., 2000) Agricultural activities are most intensive in the central lowlands Rice-based production systems prevail Rice is grown in the wet season (JuneeOctober), whereas diversified cropping (tobacco, garlic, onion, maize, sweet pepper and tomato) is practiced in the dry season, using irrigation (mainly) from groundwater A fairly-developed marketing system facilitates this diverse production system of rice and cash crops (Lucas et al., 1999) Yet agricultural productivity and income are low compared to other areas in East and South-east Asia, mainly due to a relatively low cropping intensity, underdeveloped irrigation systems and a small average farm size The major environmental problems are soil erosion on hilly land in the eastern parts and groundwater pollution in the lowlands (Lansigan et al., 1998; Tripathi et al., 1997) For analysis at the household level, we focus on only one of the 22 municipalities of the province: Dingras Farm household analysis requires accurate classification of farm households on the basis of their resources, activities and market access Given the large heterogeneity of farms in the province, specification of a limited number of relevant farm types is not feasible on the provincial scale The IRMLA project has selected two municipalities (Fig 2) for analysis that represent both ends of the land use range in Ilocos Norte: Batac, with an extremely diversified cropping pattern, and Dingras, with a relatively strong focus on rice production As indicated, this paper presents results for the latter municipality, which is located in the inner lowlands of Ilocos Norte about 10e15 km to the East from the main road connecting the provincial capital Laoag with Ilocos Sur The municipality has a population of 33,300 persons and a total land resource of 17,310 ha, of which 55% is agricultural land 2.2 Land use issues and agricultural development perspectives for Ilocos Norte According to current local government views, agriculture will maintain its central role in the economic development of Ilocos Norte province However, agriculture will have to become more productive and increasingly compete for land with industrial and tourism areas Competition for scarce natural resources, particularly land and water, is evident in the most recent provincial development plan, which includes conversion of some agricultural areas into other uses Such conversion will not spare the strategic zones for agriculture and fisheries development identified in earlier plans, such as Dingras municipality 152 R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 Fig Location of municipalities within Ilocos Norte province Views on development objectives and resource use are, at least, partly conflicting, as previous dialogues between scientists and Ilocano stakeholders had revealed (Roetter and Laborte, 2000) In three stakeholder workshops held between 2002 and 2004, different stakeholder groups were involved in discussing future land use objectives and development goals: (1) Provincial planners and policy makers (2) Municipal, agricultural, planning and development officers (Dingras) (3) Farmers’ and people’s organizations (operating within Dingras municipality) (4) Local scientists (MMSU) The presentations from and discussions among the different stakeholder groups e which were all documented (e.g in: Roetter, 2002; Roetter and Wolf, 2003) e are summarized below Provincial planners and policy makers expressed the need to maintain rice production at current levels to feed the population, while promoting intensification and diversification of production to raise farmer’s income could help achieve a boost in local income The long-term objective is transformation into an industrialized province and tourist destination with a wellmanaged environment Municipal officers largely agreed with the provincial objectives However, they did not consider their municipality as suitable for developing tourist attractions A major concern of the farmers was the high input costs (especially for fertilizer, biocides) vis-a`-vis low prices for agricultural produce Furthermore, for shifting to other crops such as fruits and vegetables, lack of capital/access to credit was seen as a major constraint Scientists as well as stakeholders at provincial level stressed that environmental issues, such as nitrate pollution and excessive pesticide residues needed to be addressed Public awareness of current and possible future negative environmental effects resulting from high input use of diversified agricultural systems, especially in the lowlands, was created through research on groundwater pollution by the RLRRC, the Rainfed Lowland Rice Research Consortium, coordinated by IRRI (Lucas et al., 1999; Tripathi et al., 1997) One question that came up in subsequent discussion was, if there was a policy regulating the use of chemicals (fertilizers and biocides), how would this affect farmers’ income and production level (targets for different crops)? This subsequently led to the proposal to examine the consequences of different production technologies for agricultural production and land use at provincial level, and, analyse the impact of a number of policy interventions on the chance that improved technologies would be adopted at the farm level These considerations led to formulation of a limited number of scenarios (land use scenarios at provincial and policy interventions at farm level) as presented in Section 3.1 Development and implementation of models and databases 3.1 Scenario construction At the regional level, four major agricultural development goals were considered for scenario analysis: maximizing farmer income and rice production, and minimizing nitrogen fertilizer and biocide use (while maintaining a minimum level of income and/or crop production) This was to address the specific ‘what-if question’ put forward during stakeholder discussions: how does goal attainment (rice production, income, etc.) and land use allocation change, if under given resource availability and a set of available production activities, the production technologies change Three basic model runs were performed for analysing effects of changes in production technologies on the different land use objectives at provincial level This would R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 Table Characteristics of the three different production technologies Characteristic Production technology (1) Current (2) High input (3) Improved (SSNM C IPM)a Target yields Amount of fertilizers Low Current Recovery fraction of applied fertilizer nutrients Labour requirements for crop management Labour requirements for other tasks Machine and fuel use Biocide use Not used High High Calculated Calculated (Zrelatively high) (Zrelatively low) Low High Current Slightly increased Strongly increasedb Current About current About current Current Current Increased Increased Increased Reduced a SSNM is site-specific nutrient management; IPM is integrated pest and disease management b Definition of technology differs in terms of labour requirements between farm household and regional model (see, text) illustrate gains and losses in terms of income, food production and environmental goals At the farm level, the effects of changes in fertilizer and biocide prices, access to credit and off-farm employment on technology choice, farmer income, rice production and fertilizer and biocide use were examined for representative farm household types in Dingras 3.2 Three alternative production technologies Three production technology levels were evaluated in both the regional and farm models: (technology 1) ‘average farmer practice’, (technology 2) ‘high yield/high input’ and (technology 3) ‘high yield/improved practice’ The relevant inputeoutput coefficients for technologies and were derived from farm surveys in Ilocos Norte province and Dingras municipality (Roetter et al., 2000; Roetter and Wolf, 2003) Technology involves average values for all farms (after data cleaning) Technology depicts the land use of a group of survey farmers obtaining higher than average output through intensive use of inputs For yields, the mean of the values with a yield level between the 90th and 95th Table Main land use types (LUT) in the regional scale analyses for Ilocos Norte province and the farm household analyses for Dingras municipality Rice Corn Garlic RiceeCorn RiceeTomato RiceeEggplant RiceeSweet potato RiceeRiceeRice RiceeRiceeSweet pepper RiceeRiceeTobacco CorneCorn RiceeCorneRice RiceeCotton RiceeRice RiceeGarlic RiceeSweet pepper RiceeMung bean RiceeTobacco RiceeRiceeCorn RiceeRiceeMung bean RiceeBitter gourd RiceeCorneCorn RiceeGroundnut RiceeOnion 153 percentile of the survey data was applied Fertilizer and pesticide use were assumed 100% higher and labour 70% higher, other inputs remaining identical to those in the average practice For the ‘improved practice’ (technology 3), the same, high, yields as in technology were assumed, but biocide inputs were reduced by 20% compared to those in ‘average farmers’ practice We, moreover, assumed higher fertilizer use efficiency than in the first two technologies For defining realistic improvements in fertilizer efficiency, we screened data from fertilizer experiments at Mariano Marcos State University (MMSU), Batac, Ilocos Norte (Pascua et al., 1998), at IRRI’s long-term continuous cropping experiment (LTCCE), and other pluri-annual and multi-site experiments on rice in the framework of the RTDP and RTOP Projects (Dobermann et al., 2004) The resulting efficiencies were used to compute fertilizer use in TechnoGIN, a program for generating inputeoutput coefficients using QUEFTS procedures to assess fertilizer requirements (Ponsioen et al., 2003) In comparison to technology 1, average applications of N, P, and K were reduced by 20% for non-rice crops For rice, a more balanced NPK application was assumed: N was reduced by 40%, P by 15% and K increased by 20% In the first set of technical coefficients, also labour was reduced by 20% as compared ‘average farmers’ practice Discussion of preliminary results with stakeholders during in-country workshops, and consultation with local experts from Dingras, however, revealed that our assumption on labour requirements for technology would be too optimistic Based on revised estimates for technology (representing a combination of site-specific nutrient management (SSNM) and integrated pest management (IPM)), labour requirements were set to be 30% higher than average farmers’ practice This revised set of technical coefficients for technology was only applied for farm household modelling in Dingras Characteristics of the three different technologies and main land use types (LUT) taken into account in the analyses are shown in Tables and 2, respectively 3.3 Regional level A model developed for Ilocos Norte province (Roetter et al., 2005) was applied The model maximizes a selection of development goals subject to constraints on land, water and labour resources, agricultural technology, and local demand for agricultural products Data on resource availability and local demand have been assembled and adapted from a number of secondary sources The total area available for agriculture for the year 2010 was estimated at 119,850 (assuming an overall land use conversion rate of 7% from agriculture to non-agricultural uses) (Roetter et al., 2000) This area was divided into a total of 200 relatively uniform land units by over-laying biophysical characteristics (irrigated areas, annual rainfall and distribution, slope and soil texture) and administrative units, comprising 22 municipalities and one township Provincial demand for agricultural products was assessed on the basis of information on per capita demand and projected population from the 154 R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 Provincial Planning Office The demand for rice was estimated at 112,610 t Labour-force and irrigation water were quantified per month and per land unit and month, respectively, based on census data and hydrological data (rainfall, ground and surface water) from the province and trend projections Details on the procedures applied to assess resource availability and constraints have been described in previous studies (Roetter et al., 2000; Laborte et al., 2002) These so-called land use types (LUTs) are selected on the basis of farm surveys and comprise (i) single cropping of root crops, sugarcane, and rice followed by fallow; (ii) double cropping: two rice crops, rice in rotation with (yellow or white) corn, garlic, mung bean, peanuts, tomato, tobacco, cotton, potato, onion, sweet pepper, eggplant, and vegetables; (iii) triple cropping: three rice crops, and rice in rotation with garlic and mung bean, with (white or yellow) corn and mung bean, and with water melon and mung bean For each LUT, three different technologies were specified, as elaborated in Section 3.2 3.4 Farm level The farm household models have essentially the same structure as the provincial model, except that there is only a single objective One model is developed for each of four representative households, which are defined on the basis of a comprehensive survey among 150 farm households All models have the same structure but part of the coefficients is different The models try to mimic household production decisions They maximize income above subsistence, given the household specific endowment of resources, minimum consumption requirements, limits on off-farm employment and credit, and generic crop technology and prices Future extensions of the models will include the incorporation of risk However, the model in its present form is sufficient for illustrating the combined regional and farm level methodology for land use analysis Somewhat different land units and land use types are defined for Dingras than for Ilocos Norte province, as different scales of analysis require different classifications Six land units were distinguished based on drainage conditions and the presence and duration of surface irrigation Twenty-two major cropping systems were identified on the basis of the farm survey These include many triple cropping systems, which have a relatively high incidence in Dingras and are less important at the provincial level The 150 households covered in the survey were classified into four homogeneous groups based on their land, labour and capital resources (including land rented in from absentee landlords) The average resource endowments of each group were used to define the representative households used in the models The major characteristics of these households are as follows: Medium farm, well drained: 0.92 of cultivated land, 64% groundwater irrigation, 74% sharecropped Medium farm, poorly drained: 1.07 ha, 76% surface irrigation, 80% sharecropped Large farm: 1.63 ha, well drained, 85% surface irrigation, 86% sharecropped Small irrigated farm: 0.83 ha, well drained, 100% surface irrigation, 94% sharecropped Results 4.1 Regional level We consider two scenarios for presentation: (A) ‘maximize farmer income’, and (B) ‘maximize rice production’ (Table 3) For both scenarios, the satisfaction of provincial demand for major agricultural products and available land, labour and water were introduced as constraints Results for scenario A show, among others, that if all farmers in Ilocos Norte would apply high input/high output technology 2, their income would be considerably higher than with average-farmer technology However, this would be achieved at the cost of high use of agrochemicals If all farmers would apply the improved, more resource-efficient practice (technology 3), even higher income levels than with technology could be achieved at about 30% lower inputs of fertilizers and pesticides For all technologies, in scenario A, total rice production would exceed the current production levels Site-specific and more balanced nutrient and pest management practices could lead to considerably higher incomes at reduced environmental costs, while still satisfying local demand for the main food crops: a clear winewin situation In scenario B, rice production under technology was 295 thousand tons If all farmers would adopt technology or 3, rice production would increase by 67 and 74%, respectively This is more than four times the local demand for rice However, income reduction would range between 85 and 93% for the different all technologies, as compared to scenario A Results for scenario ‘‘Minimize N Fertilizer Use’’(not shown) indicate that application of technology could reduce nitrogen fertilizer use by almost 70% as compared to technology 2, while still meeting the local demand for agricultural products Income from farming would be slightly higher Table Results of the regional explorations (year 2010): A Maximize farmers’ income and B Maximize rice production (constraints: land C water C labour and provincial demand for important food crops satisfied) Variable Unit (A) Maximize farmers’ income (B) Maximize rice production Tech Tech Tech Tech Tech Tech Income Rice Employment Biocide N fertilizer Land used 10 pesos 103 t 106 labdays 103 kg a.i 103 t % 15.3 119 9.5 75 13.5 100 30.4 226 17.8 161.6 33.8 91 36.6 241 12.1 79.5 15.9 96 1.9 295 6.2 9.7 11.0 30 2.3 494 15.0 30.2 26.6 44 3.6 514 9.2 19.0 8.5 46 155 R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 than for technologies and In this scenario, however, for all technologies only about one fifth of the available land would be used and income from farming would be marginal The regional model thus indicates that the use of improved technologies will lead to a winewin situation of increased income and productivity and decreased use of agrochemicals However, this does not mean that farmers will adopt these technologies A regional IMGLP model explores the window of opportunities in terms of agricultural potential and technical feasibility under different resource constraints Such explorations look at the consequences of optimally allocating land to different uses for a given set of objectives at provincial scale, using a selection of agricultural activities and production technologies that (by consensus) seem to be promising Objectives of farmers and other decision makers at lower scales are assumed subject to the provincial objectives In reality, there are many resource managers with different objectives and resource endowments, and groups using different sets of criteria for guiding their decisions If several of these stakeholder groups not behave conform the regional objectives, the government can use policies and programmes to stimulate a change in behaviour towards the desired direction While regional IMGLP models address the question ‘‘what would be possible, and what would have to be changed’’, farm household models (FHM) are used to provide answers to how options for policy and technical change can be best implemented In the following section, FHM is used as a tool for assessing the potential impact of various alternative policy interventions 4.2 Farm level The base-run simulations show which crops and technologies the representative households select at current prices and constraints assuming perfect knowledge about all technologies (Table 4) Farmers use the relatively sustainable improved technology only on dry land On irrigated land, they prefer high-input technology Besides, the small and large farmers use (average) farmer technology (technology 1) on part of their land due to credit constraints Income is clearly highest for the large farms, that also use most biocides and nitrogen fertilizer Income of the small farmer is 16% higher than for the medium farmer, whereas the small farmer uses more than four times as much nitrogen fertilizer and almost three times as much biocides This difference is explained by the use of the high-input technology on irrigated land and the (more resource use-efficient) improved technology on dry land Other than for the province as a whole, outputs of the household model show that the Dingras farmers use improved technology only on dry land and prefer high-input technology on irrigated land This is a result of using different sets of technical coefficients for the provincial and the farm household model Contrary to the provincial model, where technology is characterized by 20% lower labour requirements than technology 1, in the FHM for Dingras adjustments were made after stakeholder consultation: labour requirements in technology were set 30% higher than in technology (Section 3.4) e showing the importance of stakeholder dialogue for proper model development We ran four scenarios to assess the impact of different policies and developments on the model outcomes (Table 4) The first scenario simulates the removal of all credit constraints, which potentially leads not only to an increase in income but also in the use of agrochemicals Only the large and the small farmers were credit constrained in the base run The ultimate effect of increased credit availability is relatively low for the large farmer, but high for the small farmer The latter uses the additional credit to substitute high-input technology Table Results of the farm household simulations for Dingras municipalitya Income (10# pesos) Rice (t) N fertilizer (kg) Biocides (kg a.i.) Area (ha) Medium farm-well drained Base run Unlimited credit (%) Unlimited off-farm employment (%) 10% Increase in fertilizer prices (%) 10% Increase in biocide prices (%) 301.1 0b 16 0 5.0 0 0 47 0 0 187 40 0 0.92 0 0 Large farm Base run Unlimited credit (%) Unlimited off-farm employment (%) 10% Increase in fertilizer prices (%) 10% Increase in biocide prices (%) 656.0 12 ÿ1 ÿ2 11.8 ÿ8 ÿ4 ÿ7 376 ÿ5 ÿ7 1013 ÿ1 ÿ5 ÿ7 Small irrigated farm Base run Unlimited credit (%) Unlimited off-farm employment (%) 10% Increase in fertilizer prices (%) 10% Increase in biocide prices (%) 348.4 15 ÿ1 ÿ2 5.1 41 ÿ1 ÿ1 ÿ3 203 23 ÿ1 ÿ2 540 19 ÿ1 ÿ2 a b Tech (area share) Tech (area share) Tech (area share) 0 0 0.12 200 0 0.88 ÿ27 0 1.62 ÿ14 0 0.07 ÿ100 29 114 143 0.77 18 ÿ9 ÿ14 0.16 31 ÿ100 ÿ6 0.84 ÿ1 0 0.29 ÿ100 0.7 41 ÿ1 ÿ3 0.01 ÿ100 0 The results for the medium farm-poorly drained are not listed, as they are very similar to those of the medium farm-well drained Numbers for scenarios represent percentage changes to the base run 156 R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 for average-farmer technology on almost one third of his land This results in an increase of income by 15% and of nitrogen and biocide use by 23% and 19%, respectively At present, there is little off-farm employment, which makes sustainable, labour-intensive production technologies relatively attractive This could change in the future Simulations show that the unlimited availability of off-farm employment would lead to an increase in income of 12e16% for all farmers, but to limited reduction in the use of biocides and N fertilizer except for the medium farmer, who increases his biocide use by 40% The latter shifts from the labour intensive improved technology to the more biocideintensive high-input technology for the production of single rice on his dry land The large farmer makes a shift from improved technology to high-input technology and averagefarmer technology However, this hardly results in a higher farm-level use of biocides as it is compensated by a decrease in land rented in Finally, we evaluated two price-change scenarios to assess the potential of reducing input of agrochemicals through input price policies The large farmer is affected most by this policy and shifts between technologies Changing biocide prices is most effective in decreasing the use of agrochemicals: a 10% increase in biocide prices results in a 7% decrease in both the use of fertilizers and biocides, while the same increase in fertilizer prices results in a decrease of 5% for both types of inputs The other changes are minor Discussion and outlook There is a need for tremendous agricultural productivity increases in the countries with high population densities in E and SE Asia, such as the Philippines Such increase can only be achieved durably by judicious use of external inputs and natural resources, and supportive policies Model results for the province show the high potential of the new technologies to improve income and sustainability at the same time, implicitly suggesting that investment in agricultural research and extension is the answer However, there are some pitfalls when generalizing findings on possible efficiency increases from a few experiments and survey data to a whole province, as was, for instance, illustrated by the required revision of our assumptions based on more in-depth surveys and iterative consultations in one sub-region (i.e Dingras municipality) Moreover, there are many constraints to optimizing resource use efficiency (such as limited access to credit), which cannot be analysed using the regional model Here, FHM comes in for analysing the constraints and possibilities to adoption of sustainable technologies at the farm level Analysis of the effectiveness of different policy instruments (investment to improve access to credit, off-farm employment and influence on prices in contributing to regional development goals) was performed For this, a prototype model for representative farm types in Dingras municipality was applied The model results show that the new technologies (with coefficients adjusted using best expertise of local conditions) are only attractive on dry land and not on land with surface irrigation The analysis of different policy instruments shows that trade-offs between income, rice production and ecological sustainability of agricultural production vary depending on farm type An important result is that increased availability of off-farm employment is likely to hamper adoption of sustainable technologies This illustrates that results from FHM clearly add value to the ‘‘broad strokes’’ of regional modelling When developed and applied in close interaction with stakeholders, such dual-scale modelling approach can provide valuable information for policy development in relation to natural resource management (Van Ittersum et al., 2004) Such process is currently underway in the case study regions of the IRMLA project (www.irmla.alterra.nl) In future studies for Ilocos Norte, however, conflicting land use objectives at the decision level of the farm, municipality and province should be taken into account Furthermore, there is scope to improve the methodology by considering production and price risks in farmers’ land use decisions References Bouman, B.A.M., Jansen, H.G.P., Schipper, R.A., Hengsdijk, H., Nieuwenhuyse, A (Eds.), 2000 Tools for Land Use Analysis on Different Scales With Examples from Costa Rica Kluwer Academic Publishers, Dordrecht, The Netherlands Candler, W., Fortuny-Amat, J., MacCarl, B., 1981 The potential role of multilevel programming in agricultural economics American Journal of Agricultural Economics 63, 521e531 De Wit, C.T., Van Keulen, H., Seligman, N.G., Spharim, I., 1988 Application of interactive multiple goal programming techniques for analysis and planning of regional agricultural development Agricultural Systems 26 (2), 11e230 Dobermann, A., Witt, C., Dawe, D (Eds.), 2004 Increasing Productivity of Intensive Rice Systems Through Site-specific Nutrient Management Science Publishers, Inc., and International Rice Research Institute, Makati City, Philippines, p 410 Kruseman, G., Bade, J., 1998 Agrarian policies for sustainable land use: bio-economic modelling to assess the effectiveness of policy instruments Agricultural Systems 58, 465e481 Kuyvenhoven, A., Bouma, J., Van Keulen, H (Eds.), 1998 Policy analysis for sustainable land use and food security Agricultural Systems Special Issue 58, 281e481 Laborte, A.G., Roetter, R., Hoanh, C.T., 2002 The land use planning and analysis system of the systems research network in Asia, paper presented at MODSS’99, 1999 Brisbane, Australia, August 1e6 In: Lawrence, P.A., Robinson (Eds.), Proceedings Report QNRM02143 Queensland Department of Natural Resources and Mines, Australia Lansigan, F.P., Orno, J.L., Pascual, C.M., Navarrete Jr., N.R., Francisco, S.R., Laborte, A.G., Hoanh, C.T., Roetter, R., 1998 Systems methodology development for exploratory agricultural land use planning for Ilocos Norte Province, Philippines In: Roetter, R.P., Hoanh, C.T., Luat, N.V., Van Ittersum, M.K., Van Laar, H.H (Eds.), Exchange of Methodologies in Land Use Planning SysNet Research Paper Series No International Rice Research Institute, Los Ban˜os, Philippines, pp 73e80 Lu, C.H., Van Ittersum, M.K., Rabbinge, R., 2004 A scenario exploration of strategic land use options for the Loess Plateau in Northern China Agricultural Systems 79, 145e170 Lucas, M.P., Pandey, S., Vilano, R.A., Culanay, D.R., Obien, S.R., 1999 Characterization and economic analysis of intensive cropping systems in rainfed lowlands of Ilocos Norte Experimental Agriculture 35, 1e14 R.P Roetter et al / Environmental Modelling & Software 22 (2007) 149e157 Pascua, S.R., Ventura Jr., W., Augustin, E.O., Padre, A.T., Valencia, D.A., Marcos, T.F., Sta Cruz, P.C., Obien, S.R., Ladha, J.K., 1998 Sustainability of diverse lowland rice-based cropping systems with special reference to nutrient utilization RLRRC, Final Report 1996e97, Phase II Rainfed Lowland Rice Research Consortium, Batac Site Accomplishment Report 1995e97 Mariano Marcos State University, Batac, Ilocos Norte, Philippines, pp 25e39 Ponsioen, T.C., Laborte, A.G., Roetter, R.P., Hengsdijk, H., Wolf, J., 2003 TechnoGIN-3: a technical coefficient generator for cropping systems in East and South-east Asia Quantitative Approaches to Systems Analysis No 26 Wageningen University, The Netherlands, 69 pp Roetter, R.P., Van Keulen, H., Van Laar, H.H (Eds.), 2000 Synthesis of methodology development and case studies SysNet Research Paper Series No International Rice Research Institute, Los Ban˜os, Philippines, p 94 Roetter, R.P., Laborte, A.G., 2000 Stakeholder meetings in the Philippines, Malaysia and Vietnam In: Roetter, R.P., Van Keulen, H., Van Laar, H.H (Eds.), Synthesis of Methodology Development and Case Studies SysNet Research Paper Series No International Rice Research Institute, Los Ban˜os, Philippines, pp 69e76 Roetter, R.P (Ed.), 2002 First Annual Scientific Report IRMLA Project Report No Wageningen UR, The Netherlands, p 95 (CAnnexes) Roetter, R.P., Wolf, J., 2003 Second Annual Scientific Report IRMLA Project Report No Wageningen UR, The Netherlands, 102 pp (CAnnexes) 157 Roetter, R.P., Hoanh, C.T., Laborte, A.G., Van Keulen, H., Van Ittersum, M.K., Dreiser, C., Van Diepen, C.A., De Ridder, N., Van Laar, H.H., 2005 Integration of systems network (SysNet) tools for regional land use scenario analysis in Asia Environmental Modelling and Software 20, 291e307 Spronk, J., Veeneklaas, F., 1983 A feasibility study of economic and environmental scenarios by means of interactive multiple goal programming Regional Science and Urban Economics 13, 141e160 Stoorvogel, J.J., Antle, J.M., 2001 Regional land use analysis: the development of operational tools Agricultural Systems 70, 623e640 Tripathi, B.P., Ladha, J.K., Timsina, J., Pascua, S.R., 1997 Nitrogen dynamics and balance in intensified rainfed lowland rice-based cropping systems Soil Science Society of America Journal 61, 812e821 Van Ittersum, M.K., Rabbinge, R., Van Latensteijn, H.C., 1998 Exploratory land use studies and their role in strategic policy making Agricultural Systems 58, 309e330 Van Ittersum, M.K., Roetter, R.P., Van Keulen, H., De Ridder, N., Hoanh, C.T., Laborte, A.G., Aggarwal, P.K., Ismail, A.B., Tawang, A., 2004 A systems network (SysNet) approach for interactively evaluating strategic land use options at sub-national scale in South and South-east Asia Land Use Policy 21, 101e113 Walker, D.H., 2002 Decision support, learning and rural resource management Agricultural Systems 73, 113e127 ... productivity and income are low compared to other areas in East and South -east Asia, mainly due to a relatively low cropping intensity, underdeveloped irrigation systems and a small average farm size... scenario analysis: maximizing farmer income and rice production, and minimizing nitrogen fertilizer and biocide use (while maintaining a minimum level of income and/ or crop production) This was to... the models will include the incorporation of risk However, the model in its present form is sufficient for illustrating the combined regional and farm level methodology for land use analysis