IN-Palm: an agri-environmental indicator to assess nitrogen losses in oil palm plantations

The Alliance for Sustainable Palm Oil supports the IN-Palm project developed by CIRAD, a model to help managers and scientists to estimate nitrogen losses in the environment and to identify best management practices.

Oil palm, is currently cultivated on 19 M ha and palm oil represents more than one-third of the global vegetable oil market. Addition of nitrogen via legume cover crop and fertilizers is a common practice in industrial oil palm plantations. Part of this added nitrogen is prone to loss from the field, contributing significantly to environmental impacts. To improve the sustainability of palm oil production, it is crucial to determine which management practices minimize nitrogen losses. Continuous field measurements would be prohibitively costly as a monitoring tool, and in the case of oil palm, available models do not account for all the potential nitrogen inputs and losses or management practices. In this context, CIRAD developed IN-Palm, a model to help managers and scientists estimate nitrogen losses to the environment and identify best management practices. The main challenge was to build the model in a context of knowledge scarcity. Given these objectives and constraints, they developed an agri-environmental indicator, using the INDIGO® method and fuzzy decision trees. We validated the nitrogen leaching module of IN-Palm against field data from Sumatra, Indonesia. IN-Palm is implemented in an Excel® file and uses 21 readily available input variables to compute 17 modules. It estimates annual emissions and scores for each nitrogen loss pathway and provides recommendations to reduce nitrogen losses. IN-Palm predictions of nitrogen leaching were acceptable according to several statistics, with a tendency to underestimate nitrogen leaching. However, they highlighted necessary improvements to increase IN-Palm precision before use in plantations.

Core ideas

– They used INDIGO® method and fuzzy decision trees to develop IN-Palm indicator

– IN-Palm is run in Excel® using 21 input variables to estimate 6 loss pathways of N

– N leaching predictions against field data were acceptable, although underestimated – IN-Palm outputs were sensitive to management changes and climate variability