Simulating the effects of thinning and species mixing on stands of oak (Quercus petraea (Matt.) Liebl./Quercus robur L.) and pine (Pinus sylvestris L.) across Europe

Abstract

Tree species mixing of oak (Quercus petraea (Matt.) Liebl./Quercus robur L.) and pine (Pinus sylvestris L.) has been shown to have positive effects on ecosystem service provision. From a management perspective, however, it is still uncertain which thinning regime provides the highest possible productivity of mixed oak–pine forests in the long term. Because of a lack of empirical studies dealing with thinning and species mixing effects on oak–pine forests, we simulated forest growth in order to test which thinning type and intensity may provide the highest productivity in the long-term. To achieve this, we simulated the growth of pure and mixed stands of oak and pine for 100 years in 23 triplets located on an ecological gradient across Europe. For this purpose, we applied four different growth simulators and compared their results: the distance-independent single-tree simulator PROGNAUS, the distance-dependent single-tree simulator SILVA, the gap model ForCEEPS, and the process-based simulator 3D-CMCC-FEM. We investigated the effects of species mixing and thinning from the upper (thinning from above) and lower tail (thinning from below) of the diameter distribution by reducing the stand basal area to 50 and 80% of the maximum basal area. We compared simulated results of the relative volume productivity of mixed versus pure stands and of thinned versus unthinned stands to empirical results previously obtained on the same set of triplets. Simulated relative volume productivity ranged between 61 and 156%, although extremes of 10% and of 300% could be observed. We found the relative volume productivity to be influenced by stand age, but not by stand density, except for PROGNAUS. Relative volume productivity did not increase with the site water supply of the triplet location. Highest long-term productivity for oak, pine and oak–pine stands can be expected in consequence of thinning from above, but the effect of thinning intensity differed between simulators. Thinning effects were positively affected by stand density, but not by stand age, except for thinning from above predicted by PROGNAUS. Predicted thinning effects showed good approximation of results from thinning experiments for oak, but not for pine stands. We hypothesize the results might be caused by the insufficient simulator representation of climate and its interaction with other site variables and stand structure. Further work is needed to reduce the revealed limitations of the existing growth models, as we currently see no alternative to such kind of studies and simulators.