Species composition and specific aspects of the Green Tree Retention structure in unmanaged mixed forests in Asturias (northern Spain)

Abstract

Aims of the study: i) To characterize the species composition and stand variables in a Green Tree Retention (GTR) structure obtained as result of applying an uneven aged silviculture system based on the BDq method; ii) to conduct a biometric study of these components of the GTR.
 Area of study: Asturias (northern Spain).
 Material and methods: The GTR system included three groups of trees: non-commercial species, trees destined for production of edible fruits (Castanea sativa) and veteran trees (with DBH>52.5 cm, established by the BDq method) in 6 strata of the fourth Spanish National Forest Inventory (SNFI4). The method was applied to trees up to DBH at which are then classified as veteran. The other groups are excluded from management. A biometric study of these components was conducted by analysis of the extreme value distributions of maximum and minimum diameters by fitting the Gumbel Cumulative Distribution Function (CDF). The goodness of fit was evaluated with the Kolmogorov-Smirnov statistic (Dn), the Cramér von Mises criterion (W2), the bias and the mean squared error (MSE).
 Main results: The Gumbel CDF generally produced good results, and the best fits were obtained for maximum diameters. The Dn values were generally low or normal. The values of W2 and the MSE are consistent with the Dn. As the mean value of the biases is always positive, the Gumbel CDF tends to underestimate the real distributions.
 Research highlights: The values obtained for the parameters can help in determining several aspects of the structure of the GTR and thus help in forest management decision-making.