Evidence-based sustainable phosphorus use in agriculture in Flanders (Belgium)

Phosphorus (P) is an important nutrient for agriculture but excess P application to soils can contribute to eutrophication of surface waters. Phosphorus fertilisation recommendations rarely take environmental concerns into account (Jordan-Meille et al., 2012). Soil P tests used for soil P content measurements mostly build on chemical extractions for which empirical (not mechanistic) relations with crop responses are determined. Recently, the Flemish Land Agency (government of Flanders, Belgium) initiated a research project to stimulate sustainable phosphorus use in agriculture. Three research institutes (Institute for Agricultural and Fisheries Research (ILVO), Soil Service of Belgium and KU Leuven) started this four-year project at the beginning of 2015. In the first phase of the project, soil P tests are evaluated in order to select the (combination of) soil P test(s) that best reflect both (i) the P availability for plants and (ii) the risk of P losses towards surface waters at a relatively low cost. Several common and new tests (ammonium lactate extraction, 0.01 M CaCl2 extraction, oxalate extraction, Olsen extraction, etc) are compared in a pot experiment in a depletion scenario and on soil samples from 14 long term fertilisation field trials in NW Europe with reported yield differences due to soil P differences. Suitability of the test is assessed by the correlation with crop yield and the relative width of the 95% confidence interval of the critical soil P content, i.e. the soil P content corresponding with 95% relative yield. Not only single soil P tests but also combinations are evaluated. According to van Rotterdam-Los (2010), especially the combination of a soil P test reflecting P intensity (directly available P) and one reflecting P quantity (P available in the long term) shows promise for describing the behaviour and availability of P. The same tests are also evaluated for correlations with soil P losses by performing soil column leaching experiments under unsaturated conditions. The selected test(s) will be used in the second and third parts of the research project. In this second part, the soil P contents at which yields are optimal and losses still small, will be defined as the target zone. In the third part, we will derive soil P fertilisation advice in order to reach or remain in the target zone from sorption/desorption experiments and field trials. The outline of this project and its first results for sustainable P use will be presented and discussed.