부산대학교 도서관

The paper deals with meta-analysis of 259 dendrochronological series of mountain trees – Bosnian pine, Norway spruce and beech from 14 locations in Bulgaria for eustress investigation. The chronologies were analyzed by the original SPPAM 2.0 application. The tree ring width sequences were approximated with polynomials with determination coefficients, R2 ≥ 0.45. The growth index (It – the ratio between measured and approximated value of tree ring width) was computed. Years with It less than the threshold values were considered as eustress ones. The four grade scales for assessment of frequency, duration and depth of eustress was proposed. For the studied tree species and locations, the obtained eustress periods are as follows, respectively for the Bosnian pine, Norway spruce and beech: 42 (K=9.1), 30 (K=3.4) and 37 (K=3) on average. The obtained eustress characteristics are respectively: Average Depth, Aav = 0.16, 0.26 and 0.28; Average Duration, Dav = 2, 2.5 and 2.3 years; and Average Frequency, Fav = 10, 33 and 34 years over 100 years. The risk for The stands of two studied locations of spruce and three locations of beech were established to be in some risk. They require further monitoring. The predominance of cold-dry and hot-wet years in total number of adverse climatic years (with eustress) for the Bosnian pine, as well as hot-dry and cold-wet years for spruce and beech stands was established. The influence of climatic types of years on the eustress appearance for tree years periods have also been investigated. Proposed holistic approach for meta-analyses of series (eustress identification in radial growth of stems) is convenient for the fast monitoring of forest communities state and for recognizing the stands under risk. The capability of approach allows expression of reactive functional types of trees (PFTs) as groups of trees with the similar behavior to the influence of climatic types of years, and with similar characteristics of eustress periods. These PFTs can support the development of different local, regional and global models. The originality of approach applied includes: meta-analysis, new software, characteristics, indixes and coefficients.