Jean-Marc Barbier

Prospective and participatory integrated assessment of agricultural systems from farm to regional scales: Comparison of three modeling approaches

Evaluating the impacts of the development of alternative agricultural systems, such as organic or low-input cropping systems, in the context of an agricultural region requires the use of specific tools and methodologies. They should allow a prospective (using scenarios), multi-scale (taking into account the field, farm and regional level), integrated (notably multicriteria) and participatory assessment, abbreviated PIAAS (for Participatory Integrated Assessment of Agricultural System). In this paper, we compare the possible contribution to PIAAS of three modeling approaches i.e. Bio-Economic Modeling (BEM), Agent-Based Modeling (ABM) and statistical Land-Use/Land Cover Change (LUCC) models. After a presentation of each approach, we analyze their advantages and drawbacks, and identify their possible complementarities for PIAAS. Statistical LUCC modeling is a suitable approach for multi-scale analysis of past changes and can be used to start discussion about the futures with stakeholders. BEM and ABM approaches have complementary features for scenarios assessment at different scales. While ABM has been widely used for participatory assessment, BEM has been rarely used satisfactorily in a participatory manner. On the basis of these results, we propose to combine these three approaches in a framework targeted to PIAAS.

Growth and yield formation of irrigated, direct seeded rice as affected by nitrogen fertilizer

In two experiments performed in 1989 and 1990 in the South of France, the growth and yield formation of direct seeded rice as affected by fertilizer nitrogen (N) management were studied to gain insight in the possibilities of improving fertilizer N efficiency. Initial plant density markedly influenced the time course of N uptake rate. An early decrease in N uptake rates resulted in low shoot N concentrations during the tillering and reproductive growth phases and consequently, in a decrease of the length of the active tillering period. However, appearance rates of leaves and tillers were not clearly affected by N management. The relative mortality rate of tillers during the reproductive phase was linearly related to the decrease in shoot N concentration during that period. Low grain numbers per panicle were related to a low crop N status. The leaf area of the crops was determined by : the size of leaf laminae, the duration of the active tillering phase, the relative death rate of stems and the total number of leaves produced on main stems. Despite these different processes, pooled data from both experiments could be used to establish simple linear relationships between leaf biomass and cumulative shoot N until flowering, and between leaf area index and cumulative shoot N. This implied that until flowering the N use efficiency for leaf production and specific leaf weight were constant. Hence, crop N status and shoot growth rate were directly related to the formation of crop components. However, N uptake rate per se was not a reliable variable to relate to production. Despite the latter conclusion, we argued that the risk of a sink-limited yield might be minimized by proper N management. To subsequently reduce the risk of source-limited grain production, a moderate leaf area with high N content during the post-flowering period is recommended.