Modeling potential site productivity for Austrocedrus chilensis trees in northern Patagonia (Argentina)

Abstract

Sustainable management of native species is essential in regions where forest is continually decreasing, such as South America. A first step for sustainable management is to develop models of productivity and site quality, which are usually related to the height of dominant trees. The aim of this study was to model the height (h) of dominant trees of southern South American conifer Austrocedrus chilensis based on climate, topography and soil predictors, and tree age using a mixed-effect modeling approach under a multi-model inference framework. Tree data (h and age) were collected in 43 plots placed throughout the natural distribution range of A. chilensis in northern Patagonia (Argentina). Soil characterization was carried out in 32 out of 43 plots. Our results indicate that dominant trees are taller in cooler and wetter sites with more soil carbon and lower soil acidity. The model predicted h with ≈3 m (19 %) error and explained about 85 % of variability in h (conditional R2 = 0.84). When considering only climate variables, the explained variance was reduced by 7 % although the loss of predictive capability was not substantial (3.1 m prediction error). This study provides the first regional statistical model predicting productivity indicators in A. chilensis. With this model, site quality can be classified just using a few climatic variables available from satellite-based geospatial information and then improved by including edaphic information (soil carbon, pH). The model could have usefulness beyond forestry, for example to foresee climate change effects on ecosystem services associated to forest productivity.