Studying temporal aspects of stand development is challenging because of the inherent duration of rotational cycles. Even in the case of CS in our example, it is theoretically 30 years but we included a 50-year-old outgrown stand. In HF, the rotational cycle is twice as long. Pickett [45] recommended a false chronosequence approach where he substitutes time for space. It is therefore crucial to find stands with similar management, species composition and other environmental conditions (microclimate, soil, topography), only differing in stand age. It must be assumed that the stands follow convergent succession trajectories [46], which was ensured in our case by use of inventory data from the past. The chronosequence approach is generally contested since it comes with a set of limitations (e. g. the problem of regional averaging, ignoring major disturbances or site-specific parameters as well as variation between hypothetical stands at the same age). Moreover, it assumes that there are no major disturbances (e. g. windthrows, insect attacks) during the rotational cycle. However, the method allows a researcher to successfully study temporal changes through the judicious use of chronosequences [46] and is often the only possible method to study long-term dynamics. Five plots were chosen for each chronosequence in HF and CS, ranging from 1-50 years in CS and 11-91 years in HF, respectively (Table 1).

A full biomass inventory was performed above — and belowground using allometric functions from Hochbichler [42]. In addition, belowground fine root biomass was determined to a depth of 50 cm by using soil cores. Additionally, soil macronutrient analysis was performed using these samples. Details for plot selection and setup as well as investigation of compartments and subsequent laboratory methods can be derived from Bruckman et al.[13]. The HF forest was chosen for comprehensive soil analysis, including exchangeable cations in soil and nutrient pools of different aboveground compartments, such as foliage, bark, wood and branches as well as regeneration and stems (see Figure 3). Exchangeable cations were determined at different soil horizons by using a BaCl2 extraction and subsequent Inductively Coupled Plasma — Optical Emission Spectroscopy (ICP-OES) analysis. Exchangeable phosphorus pools were estimated using data from the forest soil inventory of adjacent forest sites [47]. Macronutrients N, P and K were determined in biomass compartments (foliage, bark of stems > 8 cm diameter, wood of stems > 8 cm diameter, composite sample of stems <8 cm, branches > 2 cm diameter, branches < 2 cm diameter and regeneration < 1.3 m height) according to inventory data for the most abundant species at each forest site. Nutrient analysis is based on three full tree samples (aboveground compartments) for Quercus and three foliage and branch samples from different crown layers per plot and species. Foliage was collected in August 2011 to ensure sampling fully developed leaves. HNO3/HQO4 extraction according to ON L 1085 [48] followed by ICP-OES analysis was used to determine nutrient contents.