Soil analyses
Directly after recording the plant species, on each subplot three soil cores of 1 cm diameter were taken in a systematic grid and mixed together. Soil samples were sieved to a minimum of 2 mm grain size. To determine water content, 5 g of soil was dried for 48 hours at 105°C. Ammonium and nitrate were extracted from 10 g of wet soil with 25 ml 1 mol KCl solution, shaking for 30 minutes. In the case of shallow soil, analyses were made with less soil, keeping the soil:solution ratio constant to avoid differences in extraction strength. If there was too little soil to keep the soil:solution ratio constant, a correction factor was empirically determined to account for the higher extraction strength. After filtration with the KCl extract, we measured the pH-value with a pH-electrode (719 S Titrino, Metrohm, Switzerland). With the remaining solution, we measured ammonium and nitrate with a microplate reader (Synergy mx, Biotek Instruments, Germany). We determined ammonium cholorometrically according to Baethgen and Alley (1989) and nitrate according to Miranda et al. (2001). Both concentrations were expressed as ppm of dry soil.
We analyzed concentrations of 20 cations and of phosphorous with an inductively coupled plasma spectrometer (ICP-OES, Spectroblue Ti, Spectro Analytical Instruments GmbH, Germany). As standards, we used a 20-cation solution (ICP multielement standard solution IV Merk KgaA, Germany) and a P solution (Single-element ICP standard solution, Phosphorus, Carl Roth GmbH + Co. KG, Germany). Only those cations known to be macro- or micro-nutrients for plants were further considered. To achieve this, we extracted 5 g wet soil with Mehlich 1 solution (Mehlich, A 1953). Samples exceeding determination threshold were diluted and re-measured.
We dried an aliquot of soil at 60 °C and milled with a pebble mill for 1.5 min. 50 µg were weighed in tin capsules to determine C and N content with a combustion gas detector (Vario EL cube, Elementar, Germany).