The purpose of this study was to build a mathematic predictive model for the formation of injuries and to try its usability in predicting. The empirical data consisted of 24 thinning sites, which were situated mainly in forests owned by forestry companies. The timber was made by strip roads. The terrain transportation of timber was carried out by forwarders or by agricultural tractor-trailer combinations.
The model has been compiled by multiple regression analysis. The predictive model has been interpreted so that the formation of injuries is, for one part, dependent of the number of trees exposed to damages, and, on the other hand, on the factors influencing the probability of injuries, the possibility for this existing. The number of trees exposed to injuries is dependent on the density of strip roads and on the amount of standing timber left after logging. The probability of injuries is influenced by the quality of the standing timber, the type of soil and the intensity of the timber harvesting process.
The predictive model has been interpreted in similar way as the number of injured trees. The possibility of formation of injuries is mainly affected by the length of strip roads and the amount of the standing timber. The probability of injuries is influenced, for the part of the number of root injuries by the strength of the soil type, and probably also by transportation equipment. For the amount of stem injuries no valid predictive equation was found in this study.