Assessing phosphorus nutrition of the main European tree species by simple soil extraction methods

In the past few decades, high nitrogen emissions have reversed the nitrogen limitation of trees. Meanwhile, phosphorus (P) appears as a primary limiting nutrient of many forests in Europe. However, a simple and economic standard method for the determination of the plant-available phosphorus in forest soils, which drives the P supply of the trees, is not yet available. Therefore, we tested two soil extraction methods [citric-acid extract (P_cit), according to Hoffmann et al. (1991) and sodium bicarbonate extract (P_HCO3), referring to Olsen et al. (1954)] for analyzing all Bavarian soil samples of the second German National Forest Soil Inventory (NFSI), resulting in a big data set with 342 forest sites and 2232 samples. We combined these results with the standardly gathered soil parameters and tested them as predictors for P nutrition indicated by the foliar P concentrations of the four most important tree species: Norway spruce (Picea abies), European beech (Fagus sylvatica), Scots pine (Pinus sylvestris) and pedunculate/sessile oak (Quercus robur/petraea). During the statistical process, we stepwise reduced the complexity of the prediction models. In our analyses, we achieved the best prediction of foliar P using soil information from the organic layer and mineral soil down to depths of 10 cm and 20 cm. P_cit stocks provided by far the best predictions in nearly all cases. Based on this parameter, we derived parametric regression functions to estimate the P nutrition (spruce: R² = 0.52^∗∗∗; beech: R² = 0.49^∗∗∗; pine: R² = 0.39^∗∗∗; oak: R² = 0.35^∗∗∗). Based on logistic regressions, threshold values were deduced for P_cit stocks down to 10 cm soil depth (spruce: 33 kg ha⁻¹; beech: 43 kg ha⁻¹; pine: 32 kg ha⁻¹; oak: 41 kg ha⁻¹), above which P deficiency becomes unlikely. Since these models are based on a large number of sites with diverse geologies and soil conditions, they can be used to estimate the P availability of most forests in Bavaria. Furthermore, the applied methods can be extended to other soil inventories. Estimated P availability based on such functions and thresholds can support sustainable nutrient management of temperate forests.