Modelling of nutrient concentrations in roundwood based on diameter and tissue proportion: Evidence for an additional site-age effect in the case of Fagus sylvatica

摘要

Climate-mitigation strategies encourage the use of forest products and wood energy. This will likely increase the risk of soil depletion through nutrient exportation. Given the high variability in nutrient concentrations within and between trees, it is crucial to better understand the biochemical patterns of nutrient distribution and the related driving factors. Thus, predictions of nutrient concentrations can be improved, and used for assessing nutrient exportation through roundwood exploitation and for developing forest-management rules to maintain soil fertility. Our aims were to better determine (1) the shapes of the relationships between the total concentrations of different nutrients (N, S, P, K, Ca, Mg, Mn) in roundwood and roundwood diameter; (2) the variation of these relationships between different tissues (wood, bark); and (3) potential effects of site conditions and/or stand age on nutrient concentrations. We therefore developed a new modelling approach that relied on empirical and mechanistic relationships, and allowed for sound predictions of both concentrations in individual tissues and in total. We applied the approach to Fagus sylvatica using 134 trees from 11 plots that covered large ranges of ecological conditions in terms of geographic region (from the northwest to the northeast of France), site index (from 25 m to >35 m), stand age (form 10 years to 159 years), and roundwood diameter with nutrient concentrations being measured separately for wood and bark tissues partly up to a minimum diameter of 1 cm. Relationships between total nutrient concentrations and roundwood diameter showed several different shapes: a reverse-J-shape for N, S and P; a U-shape for K; and no clear, more complex shapes for Ca, Mg and Mn. These shapes were related to concentration changes with roundwood diameter in both wood and bark for N, S and P (but with low goodness of fit for the models of N and S in the bark), while they were only related to concentration change in the wood for K, and in the bark for Ca and Mn (but with low goodness of fit for the model of Mn in the bark). Moreover, shifts of concentration-diameter curves (additive effect) were likely related to site conditions (chemical fertility), as assessed based on foliar nutrient concentrations. An additional effect of stand age could not be excluded, but so far it appeared difficult to clearly distinguish between influences of site and age. We argue that our modelling approach is particularly suitable to further tackle the issue of site and age effects, by combing data from several studies on the same species and/or on other species (genericness). Comprehensive forest-management rules based on silvicultural scenario simulation and assessment may finally be developed by coupling nutrient, biomass and site-sensitive tree-growth models