Multi-factorial analysis of spruce forests decline in the Slovak part of Beskydy region

Paper deals with analysis of circumstances of so called "new type" of spruce forest decline in the Slovak part of Beskydy region (northern Slovakia). Stand damage level was visually assessed on more than 500 sample plots allocated along the sample transects across the Kysuce and the Orava region. Critical damage level was defined and probability of its occurrence was analysed by logistic regression in relation to 6 site-related, 5 stand-related and 2 anthropogenic factors. Cross validation of derived models showed correctness of classification about 80%, but only about 70% for critically damaged class 1. Some differences appeared between studied regions in circumstances of decline, but generally, it can be stated, that most affected by decline were older stands with higher proportion of spruce and located closer to the focus of biotic pests activity. Exploitability of the derived regression models for risk prediction in the forest management is discussed.