Soil phosphorus as a factor controlling vegetation composition in moist hayfields: a statistical evaluation of some methods for the determination of available phosphate

Summary In vegetation research selection of a technique for the analysis of phosphate in soil frequently presents problems with respect to selection criteria. In plant communities of hayfields in Drenthe (The Netherlands) such a criterium has been found by means of drawing axes with different ordination techniques (9 techniques, 43 axes). Each numerical approach obtained one or two axes, which were evidently related to variation in phosphate. Five different extraction techniques were used (Ammonium lactate, EDTA, KF, NaHCO₃ and water). The outcome of these analyses could be classified in two groups: 1. P-KF and P-W (correlated with the ground water level) and 2. P-AL, P-EDTA and P-NaHCO₃ (correlated with pH). The first group showed clearer relations with the relevant ordination axes. Phosphate indicator values according to Kruijne (P-citric acid) were related to the first group only. P-KF had higher correlations with the ordination axes.