DECISION SUPPORT IN HORTICULTURAL PRODUCTION - IMPACT ON ACCURACY FROM A FULLY MONITORED TO A MINIMAL DATA FRAMEWORK

Current uncertainty propagation in models around agricultural production is heavily focused on climate change and its effect on production. More important for on farm use of model based decision support systems is the handling of the daily weather unpredictability. Growers learn from experience how to overcome this hindrance. As such, instead of programming climate change on the long run, practical use can come from methodologies which tackle day-to-day weather variations, giving the farmer a helpful tool, besides experience, for managerial decision. Open field vegetable production is an agricultural sector where quality and yield of the produce is essential to be competitive. Since these crops are very demanding in terms of nutrient supply and weather conditions, farmers spread the risk by generously applying inorganic fertilizers, often beyond the plants need. Yet, with present European legislation these practices are targeted because of their environmental impact. Fertilizers are already a major cost factor and, in the mindset of cleaner production, governments will reprimand offenders of environmental laws, especially for exceeding soil water nitrate thresholds caused by agricultural activities. To address this situation a cauliflower growth model was developed and calibrated on experimental fields, intensively monitored from 2009 to 2011, allowing simulation of daily growth, yield and N demand. Since intensive sampling would not be feasible outside an experimental context, farmers are not served by these detailed measurements. Thus, work has been done to interpret how our model responds on field level, going from a fully monitored to a minimal data framework. More so, this paper will demonstrate to which extent the models accuracy will be influenced when working with minimal datasets of information that can be gathered regularly by the farmer.