Soil phosphorus testing for intensive vegetable cropping

Environmental concerns and rapidly decreasing phosphorus (P) resources caused a renewed interest in improving soil P tests for a more efficient P fertilization. This led to the development of better P fertilizer recommendation systems for major arable crops and grass. Nevertheless, these P fertilizer recommendation systems seem to fail for intensive vegetable crops, with often a very short growing season and limited rooting system. This leads to low P use efficiencies in the horticultural sector. In order to address this problem we set up a study to answer following questions: (1) which soil P test predicts the plant available P content for intensive vegetable crops the best and (2) can new insights, such as combining different soil P tests, improve P fertilizer recommendations for intensive vegetable crops? To this end, bulk samples of 41 soils with very different P status (based on ammonium lactate extractable P) were collected. The plant available P content of these soils was determined using six commonly used soil P tests (P‐CaCl₂, P‐water, P‐Olsen, P‐acetate, P‐lactate, and P‐oxalate) and a P fertilizer pot experiment with endive (a very P sensitive vegetable crop) was conducted. Six pots of each soil were planted with endive. Three of these pots received no P fertilization (0P) and three pots received ammonium polyphosphate equivalent to 24 kg P ha^?1 (24P). All other factors were kept constant. Relative crop yield of the 0P fertilized plants compared to the 24P fertilized plants was determined. Plotting these relative yields against the P status of the soil per soil P test allowed to fit a Mitscherlich curve through the data. Also the combination of two different soil P tests to predict the relative yield with a Mitscherlich equation was evaluated. The coefficients of variation of the soil P tests, the R² values and the relative standard errors of the parameter estimates revealed that P‐acetate and P‐water predicted the relative yield of the 0P plants the best and that combining two different soil P tests gave no extra predictive power. This finding may form the basis for the development of a new P fertilizer recommendation system for intensive vegetable crops, leading to an improved P use efficiency in horticulture. In order to develop this new system more data relating soil P test values with RY of intensive vegetable crops should be collected.