DYSPALLOC, a model to simulate farmers’ cropping plan decisions in their spatial and temporal dimensions Schaller N1, Aubry C1, Boussard H2, Joannon A2, Martin P1 1 AgroParisTech INRA UMR 1048 SAD-APT, BP 01, 78850 Thiverval-Grignon, France; [email protected] 2 INRA UR 980 SAD Paysage, 65 rue de Saint-Brieuc, 35042 Rennes Cedex, France Keywords: model for action, farm scale, decision rules, crop succession, landscape organization Introduction Farmers’ cropping plan decisions have an impact on the spatial and temporal crop organization at farm and landscape levels, which in turn strongly impact many environmental issues (soil erosion, biodiversity, coexistence between GM and non GM crops, crop pest management etc.). It is generally accepted that managing landscape organization, considering both landscape composition and configuration, is a way to reconcile agricultural production and preservation of ecosystem services at the landscape scale (Foley et al., 2005). We thus consider it necessary to understand farmers’ cropping plan decisions in their spatial and temporal dimensions, to anticipate their potential consequences on landscape organization. These decisions include choosing crops, allocation of crops to plots, and splitting agricultural plots (Schaller et al., 2010). The spatial dimension of cropping plan decisions is particularly important for understanding the spatial arrangement of crops in agricultural plots. Likewise, the temporal dimension of cropping plan decisions particularly matters for understanding when decisions are made during a year and when it could be opportune to coordinate farmers’ decisions to favorably orientate the landscape organization (Dury et al., 2011). In this paper, we present DYSPALLOC, a conceptual model of decisions for DYnamic-andSPatially-explicit-ALLOcation-of-Crops-to-land, built on the basis of a French case study. The aim of this study is (i) to represent and simulate farmers’ cropping plan decisions in their spatial and temporal dimensions at the farm scale, and (ii) to evaluate the model. Materials and Methods Farmers’ decisions were represented through a generic framework derived from the “model for action” (Sebillotte and Soler, 1990) and including decisional variables, determinants, and decision rules (Schaller et al., 2010). The necessary data for using such a framework requires specific on-farm surveys. We carried out the surveys in the Niort plain region (France). We chose a reduced sample of 9 farms, in order to carry out 3 successive and detailed surveys about farmers’ cropping plan decisions over time. We selected the farms so as to account for the regional diversity in farming systems and farm territory spatial structure. We indeed hypothesized that these two criteria highly influence farmers’ decisions. All surveys (in May, November 2009 and May 2010) were semi-structured and aimed at encouraging the farmer to explain how he spatially allocated crops to land and split his farming territory into plots and to specify the reasons of his choices over different time scales. The surveys thus gave the possibility to identify farmers’ plans regarding the choice of the cropping plan for 2010 and to check possible adjustments over time. After having made an inventory of the elements constituting cropping plan decisions in space and time (based on the 9 farms), we focused on 4 farms (saturating the diversity in farming systems and farm territory spatial structure) in order to build the conceptual model DYSPALLOC, before carrying out an evaluation procedure on the 5 other farms. Results and Discussion Figure 1 represents the global structure of the DYSPALLOC model. It simulates a planned cropping plan in year n for year n+1 at the farm scale. The model represents farmers’ decisions when the farm is in a “coherence phase” of its life cycle (Chantre et al., 2010), which is a time period during which the strategic decisions remain stable (labor, equipment, crop combination, etc.). We chose to represent the cropping plan decision process through the same model for all farmers, but the diversity in each farmer’s decision rules can be accounted for through the input data. DYSPALLOC requires input data regarding the farm territory (e.g. area, soil types), the possible crops in the farming system, and crop succession decision rules. The model proceeds in 3 time steps: (i) The first time step represents the strategic decisions which are made once for the entire “coherence phase”. The strategic decisions include the crop functional hierarchy, which determines if a crop or a crop category will have priority access to land; the definition of permanent plot boundaries and of strategic suitable crop areas, which determines in which plots crops will be possibly allocated; and the definition of crop blocks, which defines groups of plots having the same possible crop succession over time. (ii) The second time step represents annual decisions made by farmers during year n for planning the cropping plan for year n+1. These annual decisions consist first in identifying the possible crop allocations to existing plots considering constraints due to past allocations. Then the model proceeds in planning the compulsory crop allocations to plots (when only one choice is possible) and in planning preferential crop allocations to plots, according to preferential criteria (when several choices are possible). The main preferential criteria are: water access, spatial regrouping of crops, and minimization of the distance between the farmstead and forage crops. At this stage, the model provides a planned cropping plan for year n+1, with a crop or crop category allocated to each plot. (iii) The third time step represents infra annual decisions made by farmers during the year n+1. These decisions make it possible to adjust the previous plan to new events occurring along the year and giving farmers new information. These events can be related to climate, market prices or commercial opportunities, technical operations, water access etc. These infra annual decisions give the possibility to explain the differences between the planned cropping plan and the final one. In addition to these 3 time steps, another original feature of DYSPALLOC relies on the fact that it accounts for agricultural plot splitting: it describes 3 types of plot splitting. The first type represents the splitting of administrative CAP islets1 depending on their individual characteristics; they are split into homogeneous pieces of land regarding soil type and water access (the split plots are called “elementary islets”). The second type of plot splitting is related to the global farm structure and to crop successions. The “elementary islets” are indeed split into “permanent plots” when their area is too large to ensure both crop succession and the stability of crop areas over time. The third type of plot splitting is the definition of “temporary plots” inside the “permanent plots” in order to temporarily adjust annual crop areas. 1 CAP islet = spatially continuous land area used by agricultural authorities to calculate the European CAP subsidies The evaluation procedure consisted of a comparison of the simulated planned cropping plans (before the third time step) via DYSPALLOC with the real planned cropping plans, identified through the surveys. We applied this procedure to the 5 surveyed farms that were not used for building the model. The difference between simulated and real planned crop areas on farms was less than 6% in all cases, and less than 3% in 4 of the 5 farms. DYSPALLOC spatially allocated the correct crops in the correct plots in 69 to 100% of the cases (84% on average), which represented 71 to 100% of the total farm area (86% on average) (Figure 2). Inputs Inputs for infra annual decisions Strategic decisions Crops Elementary islets definition Crop blocks definition Annual Identification of temporary plots and of possible crop decisions allocations to plots Farm territory Crop successions rules Strategic suitable crop area definition Functional crop hierarchy Permanent plots definition Compulsory crop allocation to plots Plan of crop allocation to the other temporary plots Output = planned and spatialized cropping plan for year n+1 at farm scale Infra annual decisions Possible successive adjustments of the planned crop allocation to plots Explanation of the final spatialized cropping plan in year n+1 Figure 1. Global structure of the DYSPALLOC model: cropping plan decisions are made over 3 time steps (strategic, annual and infra annual) % of well allocated plots (number of plots) % of well allocated farm area 100% 80% 60% 40% 20% 0% Farm 1 Farm 2 Farm 3 Farm 4 Farm 5 Figure 2. Validation of the DYSPALLOC model (proportion of correctly allocated plots and farm area) Conclusions DYSPALLOC is a useful model to simulate farmers’ cropping plan decisions at the farm scale. It represents (i) its spatial dimension by accounting for agricultural plot splitting and crop spatial organization in those plots and (ii) its temporal dimension by accounting for the three type steps at which farmers’ decisions are made (strategic, annual, infra annual). This conceptual model combining decision rules and calculations could be implemented in computer tools in order to simulate the impacts of farmer decisions on spatial and temporal landscape organization. It could also be used to examine farmer leeway and flexibility in cropping plan choices for improving landscape organization. References Chantre E., Cerf M., Le Bail M. (2010) Farmers’ learning processes in implementing lowinput field crop agriculture, in: I. S.-T. J. Wery, A. Perrin (Eds), Agropolis International Editions (Ed.), Proceedings of 'Agro2010 the XIth ESA Congress', Montpellier, France. pp. 993-994. Dury J., Schaller N., Garcia F., Reynaud A., Bergez J.E. (2011) Models to support cropping plan and crop rotation decisions. A review. Agronomy for Sustainable Development. DOI: 10.1007/s13593-011-0037-x. Foley J.A., DeFries R., Asner G.P., Barford C., Bonan G., Carpenter S.R., Chapin S.C., Coe M.T., Daily G.C., Gibbs H.K., Helkowski J.H., Holloway T.H., Howard E.A., Kucharik C.J., Monfreda C., Patz J.A., Prentice I.C., Ramankutty N., Snyder P.K. (2005) Global Consequences of Land Use. Science 309:570-574. DOI: 10.1126/science.1111772. Schaller N., Aubry C., Martin P. (2010) Modelling farmers' decisions of splitting agricultural plots at different time scales: a contribution for modelling landscape spatial configuration, in: I. S.-T. J. Wery, A. Perrin (Eds), Agropolis International Editions (Ed.), Proceedings of 'Agro2010 the XIth ESA Congress', Montpellier, France. pp. 879-880. Sebillotte M., Soler L.G. (1990) Les processus de décision des agriculteurs : acquis et questions vives. , in: J. Brossier, Vissac, B., Lemoigne, J.L. (Ed.), Modélisation systémique et systèmes agraires, INRA Paris. pp. 93-102.
© Copyright 2026 Paperzz