Plants as Indicators of Soil Chemical Properties

Abstract

We investigated the question whether plant species can be used to indicate quantitative chemical properties of soil in physical units. The investigation is based on the data of the second National Forest Soil Inventory (NFSI II) in Germany. Generalized linear models were used to calibrate species response curves for soil parameters that significantly affect vegetation. These response curves are used to indicate the most probable value of the soil parameters based on Bayes’ formula. For the major soil habitat factors [soil acidity, nutrient stocks (e.g. potassium)], an indicator system was developed that can predict the values of these habitat factors in physical units based on the vegetation present. The indicator system was validated with an independent dataset. At present, the indicator system is valid for Germany only, but the concept could be extended to other regions and other habitat factors as well. The indicator system is built on intensive quantitative soil analyses and vegetation surveys. The values of habitat conditions can be predicted based only on the plant species occurring in a site. This way, a rapid assessment of major habitat factors is possible. This is helpful when soil analyses are lacking.