Evaluating competitive interactions between trees in mixed forests in the Western Carpathians: Comparison between long-term experiments and SIBYLA simulations

Abstract

The competition interactions between trees and species have already been thoroughly studied, but most often for mono-species and even-aged forests. Several recent studies have been undertaken, prompted by changes in forest management objectives, which now favour mixed-species forests to increase biodiversity. However, there have been no studies focused on the competition interactions for the mixed-species forests in the Western Carpathians. Recently, the single tree-based growth simulator SIBYLA has been developed for use in the forests of this area. The development of this model was based upon the growth simulator SILVA. Since there were no data available from long-term experiments for the Western Carpathian forests, SIBYLA was only calibrated using large-scale data from the Slovakian forest monitoring network. The aim of this study was (i) to define the potential maximum radial growth of silver fir (Abies alba L.), European beech (Fagus sylvatica L.) and Norway spruce (Picea abies L. Karst.) in mixed forests, (ii) to test the above-ground competition effects on the radial increment of these species using a competition index based on competition light cone, and (iii) to evaluate the simulator SIBYLA using seven long-term experimental plots established in the 1960s, in the mixed-species uneven-aged forests of the Western Carpathians. These plots were repeatedly measured between five to eight times at intervals of 5–16years, from 1967 to 2010. Long-term climatic records were obtained to allow simulation of tree growth in real climatic conditions. The results showed the differences in potential maximum radial growth between the species. The results also suggested that the competition effect on radial increment predicted by the SIBYLA is an overestimate compared to the real observed effect on silver fir, Norway spruce, and European beech. This, together with other factors, is suggested to be the cause of the observed bias between measurements and SIBYLA simulations, which ranged from −2.2% to −11% for silver fir, −0.2 to −3.7% for Norway spruce, and 0.9% to 13.9% for European beech (average bias for the measurement periods).