Advances in farming systems analysis and intervention

In this paper, we recognize two key components of farming systems, namely the bio-physical ‘Production System’ of crops, pastures, animals, soil and climate, together with certain physical inputs and outputs, and the ‘Management System’, made up of people, values, goals, knowledge, resources, monitoring opportunities, and decision making. Utilising upon these constructs, we review six types of farming systems analysis and intervention that have evolved over the last 40 years, namely: (1) economic decision analysis based on production functions, (2) dynamic simulation of production processes, (3) economic decision analysis linked to biophysical simulation, (4) decision support systems, (5) expert systems, and (6) simulation-aided discussions about management in an action research paradigm. Biophysical simulation modelling features prominently in this list of approaches and considerable progress has been made in both the scope and predictive power of the modelling tools. We illustrate some more recent advances in increasing model comprehensiveness in simulating farm production systems via reference to our own group's work with the Agricultural Production Systems Simulator (APSIM). Two case studies are discussed, one with broad-scale commercial agriculture in north-eastern Australia and the other with resource poor smallholder farmers in Africa. We conclude by considering future directions for systems analysis efforts directed at farming systems. We see the major challenges and opportunities lying at the interface of ‘hard’, scientific approaches to the analysis of biophysical systems and ‘soft’, approaches to intervention in social management systems.     

Validation of the species composition index

This paper is a validation of the Species Composition Index (SCI). The SCI is a data-handling technique which expresses the botanical composition of a sward as a vector. Inclusion of these vectors allows the incorporation, in response models, of the dynamic changes in species composition of sods which occur over time. The SCI was developed originally with a data set which was obtained in a beef (Bos taurus L.) steer 3-year experiment where several dependent variables (forage growth, consumption and animal performance) were measured concurrently with several environmental, plant and animal characteristics. The SCI was then validated with a different data set, obtained over another 3-year period, with the same species combined in different ways and more complex treatments.

The SCI was related to the effects, if significant, of year, season, grazing pressure, forage quality, precipitation and air temperature on the dependent variables in the two data sets independently. The SCI was superior to the traditional classification variable ‘treatments’ in accounting for variation in the dependent variables in both data sets. When ‘treatments’ was entered into the model, coefficients of determination of 0·26, 0·43, 0·58 and 0·44 were obtained for forage growth and consumption, average daily gain and beef production with the first data set, and of 0·13, 0·40, 0·36 and 0·37 in the second data set, respectively; when SCI was used, coefficients increased to 0·40, 0·52, 0·68, 0·55, 0·42, 0·61, 0·58 and 0·58, respectively. Thus, the effectiveness of the SCI was similar in the two data sets and the SCI appears to be a useful tool for quantitatively describing dynamically changing sward compositions.     

Water productivity in smallholder irrigation schemes in South Africa

This study investigates the productivity and value of water in two smallholder irrigation schemes (Zanyokwe and Thabina) in South Africa. We apply the residual valuation method (RVM), willingness to pay (WTP) and cost-based approaches (CBA) (i.e. accounting costs of operation and maintenance) to evaluate water productivity and values per crop, per farm, and by scheme. In both schemes, water value estimated by the RVM for vegetables (cabbage, tomatoes and butternuts) is greater than water value for dry maize. At the farm and scheme levels, a comparison was made between gross margin per m³ of water, WTP per m³ and accounting cost per m³ to estimate the relative value of water productivity. The active farmers in the Zanyokwe scheme have lower WTP per m³ (R0.03) than the gross margin of output (R0.69). Also, the accounting cost (R0.084) per m³ of water is less than the gross margin. In the Thabina scheme, active farmers are willing to pay (R0.19) per m³ of water. Hence farmers in Thabina are ready to pay as much as three times the proposed costs of O&M (R0.062) per m³ of water used. Both the accounting cost and willingness to pay are less than the gross margin per m³ of water in the Zanyokwe scheme. Our findings indicate that extension and training may be required to improve the productive use of water for those farmers whose returns are insufficient to cover the cost of supply.     

Pasture and forage crop systems for non-irrigated dairy farms in southern Australia: 3. Estimated economic value of additional home-grown feed

Dairy systems in southern Australia rely on grazed feed from pasture to supply between 50% and 70% of total herd feed requirements on an annual basis. However, the dominant pasture type in the region, which is based on perennial ryegrass (Lolium perenne), commonly results in feed deficits in summer which must be filled with supplements purchased off-farm, and feed surpluses in spring which must be conserved. Both of these strictures impose costs on farm businesses. It is likely, therefore, that additional grazeable feed available to dairy herds in southern Australia may have different economic value when interactions between season, stocking rate, calving date, and locality are taken into account. The analysis reported in this paper aimed to estimate, using the farm systems simulation model UDDER, the effect of these interactions on the efficiency with which extra feed can be converted to extra milk production, and therefore the possible gross economic value of the additional feed.‘Base’ farm simulations for ‘average’ and ‘top 10%’ farms (ranked according to farm profitability) in two localities (Terang: average annual rainfall 796 mm, 8 month growing season; and Ellinbank: average annual rainfall 1085 mm, 9-10 month growing season) were created to mimic the physical production and profitability of these farms as seen in regional farm benchmark datasets. These simulations were then altered to add the equivalent of 10% of the total annual herbage accumulation used in the Base simulation either on a pro-rata basis all year round, or in autumn only, in winter only, in spring only, or in summer only. The additional feed amounted to 620 and 780 kg DM/ha for Terang average and top 10% farms respectively, and 735 and 905 kg DM/ha for Ellinbank average and top 10% farms respectively. The management policies used in the Base simulations were then adjusted to harvest as much of the extra feed as possible, either by direct grazing or through silage conservation, while keeping the key system state indicators of cow condition score and average farm pasture cover within the limits known to result in long-term sustainable production.The efficiency with which extra feed was utilised was greatest in summer in all scenarios (80-100% of the extra feed supplied was harvested, all by direct grazing). This translated into consistently high gross economic returns of between $0.26 and $0.34 per kg DM of extra feed added to the model. Utilisation efficiency was lower in all other seasons and/or required marked increases in silage conservation, both of which resulted in lower gross economic returns per kg DM of additional feed. The impact of interactions between locality, season, stocking rate (higher in top 10% farm simulations than average farm simulations) and calving date (earlier at Terang than at Ellinbank) were clearly captured in the model. These interactions have very large effects on the profitability of growing extra feed at different times of the year. Agronomic research for the southern Australia dairy industry should focus on low-cost ways for supplying additional grazeable feed in summer, since current forage species options for this time of year are limited.     

Sustainable agriculture: A case study of a small Lopez Island farm

Overreliance on fossil fuel based inputs, and transport of inputs and products is seen by many as a threat to long-term agricultural and food system sustainability. Many organic, biodynamic, and low-input farmers limit off-farm inputs, attempting instead to farm within the carrying capacity of their land or local environment. These farmers often accept lower farm productivity because they see reduced reliance on non-renewable inputs as more sustainable. Documentation of low-input agricultural systems through both replicated research trials and case studies is needed in order to better understand perceived and real advantages and tradeoffs. The goal of our study was twofold: (1) to compare liming and biodynamic (BD) preparations in improving pasture on a moderately acidic pasture soil through stimulation of soil microbial activity; (2) to place these findings within the context of a whole farm analysis of economic, plant, and animal health. Treatments included lime, the Pfeiffer Field Spray plus BD compost preparations, and untreated controls. Soil pH, total C and N, microbial activity, forage biomass, and forage quality were evaluated over two growing seasons. Both lime and the Pfeiffer Field Spray and BD preparations were only moderately effective in raising soil pH, with no effect on soil microbial activity or forage yield. Lime significantly reduced forage crude protein but the practical implications of this are questionable given the overall low quality of the forage. While the farm is profitable and economically stable and the animals healthy, the need for future targeted nutrient inputs cannot be ruled out for sustainable long-term production.     

FSSIM, a bio-economic farm model for simulating the response of EU farming systems to agricultural and environmental policies

The disciplinary nature of most existing farm models as well as the issue specific orientation of most of the studies in agricultural systems research are main reasons for the limited use and re-use of bio-economic modelling for the ex-ante integrated assessment of policy decisions. The objective of this article is to present a bio-economic farm model that is generic and re-usable for different bio-physical and socio-economic contexts, facilitating the linking of micro and macro analysis or to provide detailed analysis of farming systems in a specific region. Model use is illustrated in this paper with an analysis of the impacts of the CAP reform of 2003 for arable and livestock farms in a context of market liberalization. Results from the application of the model to representative farms in Flevoland (the Netherlands) and Midi-Pyrenees (France) shows that CAP reform 2003 under market liberalization will cause substantial substitution of root crops and durum wheat by vegetables and oilseed crops. Much of the set-aside area will be put into production intensifying the existing farming systems. Abolishment of the milk quota system will cause an increase of the average herd size. The average total gross margin of farm types in Flevoland decreases while the average total gross margin of farms in Midi-Pyrenees increases. The results show that the model can simulate arable and livestock farm types of two regions different from a bio-physical and socio-economic point of view and it can deal with a variety of policy instruments. The examples show that the model can be (re-)used as a basis for future research and as a comprehensive tool for future policy analysis.     

Bio-economic evaluation of farmers’ perceptions of viable farms in western Kenya

Arable land in western Kenya is under considerable pressure from increasing human population. Rural households depend on farming for at least part of their livelihood, and poverty rates are among the highest in Kenya. Land is often depleted of nutrients, and for most farmers, access to inputs and markets is poor. There is a need to identify options that are manageable within the context of the farmer’s resource base and the household’s objectives that could improve farm household well-being. In this study we integrated qualitative informal participatory approaches with quantitative mathematical programming and biophysical simulation modelling. Households in four sub-locations in Vihiga District were clustered and pilot cases identified. Meetings were held with farmers to elicit their perceptions of what their ideal farm would look like, and how its performance might compare with their own farm’s performance. With farmers’ help, a range of scenarios was analysed, relating to changes in current enterprise mixes, changes in current farm sizes, and changes in prices of staples foods and cash crops. A considerable mismatch was found between farmers’ estimates of their own farm’s performance, and what was actually produced. There seems to be a threshold in farm size of 0.4 ha, below which it is very difficult for households to satisfy their income and food security objectives. Even for larger farms whose households are largely dependent on agriculture, the importance of a cash crop in the system is critical. There is a crucial role for extension services in making farmers aware of the potential impacts on farm revenue of modest changes in their farm management systems. We are monitoring nine households in the district, whose farmers have made some changes to their system in an attempt to increase household income and enhance food security.     

The potential contribution of forage shrubs to economic returns and environmental management in Australian dryland agricultural systems

In face of climate change and other environmental challenges, one strategy for incremental improvement within existing farming systems is the inclusion of perennial forage shrubs. In Australian agricultural systems, this has the potential to deliver multiple benefits: increased whole-farm profitability and improved natural resource management. The profitability of shrubs was investigated using Model of an Integrated Dryland Agricultural System (MIDAS), a bio-economic model of a mixed crop/livestock farming system. The modelling indicated that including forage shrubs had the potential to increase farm profitability by an average of 24% for an optimal 10% of farm area used for shrubs under standard assumptions. The impact of shrubs on whole-farm profit accrues primarily through the provision of a predictable supply of ‘out-of-season’ feed, thereby reducing supplementary feed costs, and through deferment of use of other feed sources on the farm, allowing a higher stocking rate and improved animal production. The benefits for natural resource management and the environment include improved water use through summer-active, deep-rooted plants, and carbon storage. Forage shrubs also allow for the productive use of marginal soils. Finally, we discuss other, less obvious, benefits of shrubs such as potential benefits on livestock health. The principles revealed by the MIDAS modelling have wide application beyond the region, although these need to be adapted on farm and widely disseminated before potential contribution to Australian agriculture can be realized.     

MOPECO: an economic optimization model for irrigation water management

Water is a natural, sometimes scarce, and fundamental resource for life, essential both for agriculture in many regions of the world and also to achieve sustainability in production systems. Maximizing net returns with the available resources is of the utmost importance, but doing so is a complex problem, owing to the many factors that affect this process (e.g. climatic variability, irrigation system configuration, production costs, subsidy policies). The MOPECO model is a tool for identifying optimal production plans, and water irrigation management strategies. The model estimates crop yield, production and gross margin as a function of the irrigation depth. Finally, these gross margin functions are used to determine an optimum cropping pattern and irrigation strategy to maximize the gross margin on a farm in a specific scenario. Since the relationships between the variables are generally non-linear and the number of alternative strategies is quite large, the optimum process is complex and computationally intensive. Genetic algorithms are therefore used to identify optimal strategies. This paper describes the MOPECO model, which comprises three computing modules: (1) estimation of net water requirements; (2) derivation of the relationship between gross margin and irrigation depth; and (3) identification of the crop planning and the water volumes to be applied. The results obtained by applying the MOPECO model to a specific irrigable area in a semi-arid area of Spain, with great deficits and high water costs, are also included and discussed. These results usually show that the irrigation depth for maximum benefits is lower than that necessary to obtain maximum production. In some areas of Spain, horticultural crops are nearly always part of the optimum alternative. The crops that become part of the optimum alternative are mainly horticultural crops with a high gross margin and low water requirements. The irrigation depths selected for the ideal crop rotation are included among the irrigation depth of maximum economic efficiency and the maximum gross margin irrigation depth. Both are lower than that necessary for the maximum yield. This model helps farmers, extension services, and other agents to analyse, make decisions and optimize water management.Communicated by A. Kassam     

Intensive versus low-input cropping systems: What is the optimal partitioning of agricultural area in order to reduce pesticide use while maintaining productivity?

Pesticide use should be reduced for sustainable agriculture. Low-input cropping systems, centered on hardy varieties that maintain their yield in the presence of pests, allow pesticide use to be reduced. Since yield potential is generally lower for hardy varieties than for high-yielding varieties, a balance must be found between production and pesticide reduction. In order to compute the optimal partitioning of agricultural area between intensive and low-input cropping systems, we present a model that allows yield and gross margins to be computed at the landscape scale, as a function of the proportion of the area under intensive and low-input systems. The model shows that two cases must be distinguished, depending on inoculum production by each of the coexisting systems. If the low-input system produces less inoculum (e.g. because resistant varieties are used), coexistence can be optimal, whereas if the low-input system produces more inoculum (e.g. because tolerant varieties are used), it is best to devote the whole area to a single system. The model gives the gross margin for each cropping system as a function of the proportion of low-input systems - and so predicts the proportion to which the farmers’ choices will lead - and illustrates the use of different (simplified) policies that would ensure that the optimum proportion is reached.     

Comparison of two pasture growth models of differing complexity

Two pasture growth models that shared many common features but differed in model complexity were refined for incorporation into the Integrated Farm System Model (IFSM), a whole-farm model that predicts effects of weather and management on hydrology, soil nutrient dynamics, forage and crop yields, milk or beef production, and farm economics. Major differences between models included the explicit representation of roots in the more complex model and their effects on carbon partitioning and growth. The simple model only simulated aboveground processes. The overall goal was to develop a model capable of representing forage growth and ecosystem carbon fluxes among multiple plant species in pastures while maintaining a relatively simple model structure that minimized the number of required user inputs. Models were compared to observed yield data for 12 site-years from three experiments in central Pennsylvania, USA. Both models underestimated observed yield by 6% when averaged across site-years. However, the simple model provided a better fit to the one-to-one line between observed and simulated yield than did the complex model. The models also showed similar relationships between yield and gross primary productivity (GPP), despite the fact that the complex model was specifically developed to optimize simulation of GPP. The simple model predicted much greater shoot respiration and carbon partitioning to above ground plant tissues, but less shoot senescence than the complex model. Published data on the proportion of GPP consumed in aboveground or total plant respiration exhibit a wide range of values, making it impossible to determine which model provided the best representation of respiration rates and, thus, of the entire carbon budget.     

A stochastic simulation model for evaluating forage machinery performance

The selection of machinery on a dairy farm is the focus of the study. A simulation model is constructed that evaluates alternative forage machinery complements on New York State dairy farms. Attention is focussed on the machinery use, forage crop production and concentrate purchase. The key measure of performance is the total cost of acquiring feed for the dairy herd. The total cost includes the machinery investment and operating costs and the cost of feeds required to supplement that produced on the farm.The simulation model, written in Fortran, utilises a daily timestep. Each day the feasibility of machinery operation is determined, machinery operations are scheduled and the end of day status is determined.     

GAMEDE: A global activity model for evaluating the sustainability of dairy enterprises Part I – Whole-farm dynamic model

 Crop-livestock farms are complex systems. The interactions operating in such systems involve decisional, biophysical, structural, and environmental factors. Moreover, as farmers face a large range of management options, tools are needed to support their decision-making to enable them to reach production levels meeting their objectives and compatible with their human and physical resources, while controlling their effects on the environment. Gamede, a whole-dairy-farm model, has been developed to explore this complexity and to represent dynamically the effect of management decisions on biomass and nitrogen flows and on numerous sustainability indicators, such as milk and forage crop productivity, labour requirements, nitrogen balance, and nitrogen efficiency.This article describes the integration of six modules accounting for biophysical processes in a dairy farm (forage production; forage conditioning; herd demography; milk, excreta and animal biomass productions; grazing, quality of fertilisers; and nitrogen gaseous emissions) together with a decision system accounting for the farmer’s strategy and technical operations. Most of the six biophysical modules incorporate mathematical models from the literature, but the decision system stems from our own original work.Six commercial farms with different structures, agro-climatic conditions and management strategies were used for validation. The model can explain the differences found in their sustainability indicators at the year scale. The intra-year variability of the main biomass stocks and flows is also well explained. This quantitative validation was completed by a qualitative validation from researcher, adviser and farmer points of view, including simulations of prospective scenarios.     

Impact of resources and technology on farm production in northwestern China

Analyses were performed to evaluate the impact of resources and technological inputs on farm production in northwestern China, including five provincial districts of Shaanxi, Gansu, Ningxia, Qinghai, and Xinjiang. The gross value of farm production using 1978 constant prices was chosen to be the dependent variable in data analysis. Input variables included the irrigation ratio, farm labour, fertilizer application, and farm machinery. Multivariate statistical analysis (correlation and regression) was used to characterize the relationships between the gross value of farm production and various input variables. The gross value of farm production in the region increased fivefold during the study period, and was strongly associated with resource use and technological inputs. The gross value of farm production was significantly correlated with the irrigation ratio in Shaanxi, Gansu, and Ningxia, where the irrigation ratio was relatively low compared with the national averages. The increase in irrigation ratio in these provincial districts explained 9–19% of the growth in farm output from 1978 to 1998. Technological inputs of chemical fertilizer and farm machinery were significantly associated with farming output in all provincial districts. During the study period, the application of technological inputs increased two to eightfold, and contributed about 45% in the growth of farming output in the region. Farm production was also found to be significantly correlated with labour input. Results of this study indicated that water use efficiency and irrigation management need to be improved for future agricultural development, and further advances in farm mechanization and technology application are key for increasing farm production with limited water resources in this semi-arid area.     

Policy incentives for reducing nitrate leaching from intensive agriculture in desert oases of Alxa, Inner Mongolia, China

Excessive irrigation and nitrogen applications result in substantial nitrate leaching into groundwater in intensively cropped oases in desert areas of Alxa, Inner Mongolia. An integrated modelling approach was developed and applied to compare policy incentives to reduce nitrate leaching. The integrated model consists of a process-based biophysical model, a meta-model, a farm economic model and an assessment of policy incentives. The modelling results show that there are “win-win” opportunities for improving farm profitability and reducing nitrate leaching. We found that 4471 Yuan ha⁻¹ of farm gross margin could be obtained with a reduction in nitrate leaching of 373 kg ha⁻¹. Farmers’ lack of knowledge about water and nitrogen in soil, and on crop requirements for water and nitrogen could explain the differences, so that agricultural extension is an appropriate policy incentive for this area. When the economic optimum is obtained reductions in nitrate leaching are not achievable without profit penalties and there is a “trade-off” relationship between farm profitability and groundwater quality protection. The combination of low elasticity of nitrate leaching and large elasticity of farm gross margin against water price increases results in very high costs for reducing nitrate leaching (105.6 Yuan kg⁻¹). It is suggested that if the water price increases were coupled with subsidies for adopting nitrate leaching mitigation practices, environmental gains could come at a lower cost.     

节约型农机化生产体系的研究

农业机械化是农业现代化和农村经济发展的重要标志,是实现农业产业化、加快农业科技进步的重要措施.为此,通过对国内外农机化生产体系的阐述,分析了我国农业机械化发展现状和存在的主要问题.机械化农业生产是一种高投入、高产出的资源型、集约型生产,它能极大地提高劳动生产率和土地产出率,为改善农民的生活做出了重大贡献.     

Redirecting technology to support sustainable farm management practices

Agricultural technology has increased farm production to unprecedented levels. However, return on investment is diminishing and environmental concerns conflict with current input intensive farm practices. Conventional technologies and their application such as crop breeding and management practices have focused on monocultural systems that are dependent on chemical inputs to produce optimum yields. Current profit margins are low or non-existent with these conventional non-sustainable practices and must be changed if the family farm is to survive. We propose an ecologically based approach to farm management that strives to reduce reliance on chemically intensive inputs through better use of multiple attributes inherent within agroecosystems. This approach requires a redirection in the development and application of current and emerging technologies. Examples of redirections in research and development programs for pest management practices, genetic engineering, and precision agriculture necessary to provide a more ecologically-based and sustainable farming approach are illustrated.     

Economic significance of wool in meat production systems

The contribution of wool to ewe output declines in relative terms as the number and value of lambs reared per ewe increases. This is illustrated by the spectrum of sheep production systems in Great Britain, from extensive hill production where wool accounts for 18% of ewe output, to intensive production out of the main lambing season, where wool only accounts for 6% of ewe output. The ratio of wool production to sheep meat production and their respective prices vary greatly between countries.In Great Britain, as technical efficiency increases, the relative, but not necessarily the absolute, contribution of wool to output declines. Changes in production systems and management, aimed at improving slaughter lamb output per hectare, will also bring about a consequential increase in wool production per hectare.Because wool represents an appreciably lower part of the output in meat producing systems than lamb sales, increases in wool prices have a relatively small effect on gross margins. In lowland flocks a 20% increase in wool prices only increases the gross margin per ewe by 3·1%.Although wool output is considerably less important than lamb sales in Britain it is, nevertheless, worthwhile for the producer to pay close attention to the fleeces produced in order to ensure that he receives the highest returns possible. This is illustrated by reference to the variation in wool returns per ewe between flocks.Fleece weights and quality have a high heritability and are rapidly improved by selection. However, the relative economic value of the annual genetic improvement in increasing the number of lambs reared per ewe is worth five times as much as the annual genetic improvement in fleece weight in lowland flocks; in hill flocks this falls to only twice the value.     

Economic evaluation of current and alternative dual-purpose cattle systems for smallholder farms in the central Peruvian highlands

In four communities in the Peruvian Andes, 56 farmers were interviewed every three months over a period of one year. Information linked to milk and cattle production such as activities, inputs (labour, means of production, capital) and outputs (milk, cheese, animals) were recorded using a closed-ended questionnaire. The communities were divided into two groups with low (LC) and high (HC) level of dependence on income from milk and animal sales. The survey results showed that cattle production on the LC farms was based on less land and a smaller herd (3.32 ha/farm, 1.06 lactating cows) than on HC farms (10.28 ha/farm, 4.19 lactating cows). The data from the survey and the results of the nutritional analyses of 74 feed samples were introduced into a model that applied linear programming techniques in order to estimate the farm household income under the current production systems and evaluate the economic impact of improved forage varieties for hay production. Furthermore, the economic viability of other changes in fodder and herd management was tested. Both groups were characterised by a dual-purpose system generating a gross income from the sale of both, milk and live animals in the amount of -21 (LC) and +1057 US$/farm and year (HC). Due to higher production costs for forages and better access to markets, LC communities were characterised by an integrated crop–livestock system whereas in the HC group income was mainly based on livestock. Introduction of improved and fertilized barley for hay production, was estimated to increase the annual farm income to 127 and 1257 US$ for LC and HC, respectively. This increase was accompanied by an increment of the animal number. Maintaining the animal number but increasing the milk production/cow by feeding additional forage was a less profitable option generating 50 and 1221 US$ of income per farm and year for LC and HC, respectively. The production of hay was limited by high costs (external labour) in LC communities and the restricted availability of family labour in the HC group. A scenario based on the use of improved cow genotypes led to the highest estimated annual farm income for HC communities (1280 US$) but was less favourable for LC. The modelling results showed that the best development strategy depends on various factors such as production costs, access to the markets and to irrigation and availability of different feed resources.     

An entropy approach to spatial disaggregation of agricultural production

While agricultural production statistics are reported on a geopolitical – often national – basis we often need to know the status of production or productivity within specific sub-regions, watersheds, or agroecological zones. Such re-aggregations are typically made using expert judgments or simple area-weighting rules. We describe a new, entropy-based approach to making spatially disaggregated assessments of the distribution of crop production. Using this approach, tabular crop production statistics are blended judiciously with an array of other secondary data to assess the production of specific crops within individual ‘pixels’ – typically 25–100 square kilometers in size. The information utilized includes crop production statistics, farming system characteristics, satellite-derived land cover data, biophysical crop suitability assessments, and population density. An application is presented in which Brazilian state level production statistics are used to generate pixel level crop production data for eight crops. To validate the spatial allocation we aggregated the pixel estimates to obtain synthetic estimates of municipality level production in Brazil, and compared those estimates with actual municipality statistics. The approach produced extremely promising results. We then examined the robustness of these results compared to short-cut approaches to allocating crop production statistics and showed that, while computationally intensive, the cross-entropy method does provide more reliable estimates of crop production patterns.