Site index functions using site descriptors for Pinus patula plantations in South Africa

Abstract

The quantification of the complex interaction of water and nutrients on forest productivity has long been a challenge to forest scientists worldwide. South Africa is no exception, and an industry that is experiencing increased management intensity, requires reliable decision support systems to remain competitive. Empirical growth models as well as site quality prediction models have played a significant role in forest management to date. However, there is a definite need for models incorporating process-based information to permit better assessment of management options as well as to capture dynamics caused by climate changes. The Mpumalanga escarpment area, one of the major timber growing areas in South Africa, was used as a study area, with data on Pinus patula growth. Two site quality prediction models were developed, using multiple regression techniques. Both models use a range of site factors that express the moisture and nutrient balance of the specific site. These models were found to predict site index values with a high degree of accuracy to be used as management tools. A third model was produced by combining a site quality prediction model with a Chapman–Richards type model in order to provide a site-specific growth curve. The performance of these models was compared with a process-based model called 3-PG, using independent site and growth data from the study area. Although some encouraging results were found with the 3-PG model for the prediction of growth curves, significant deviations existed between observed and predicted values of MAI 20 values. Possible reasons for the difference in accuracy between the models are discussed. Future research is required to enable the proper parameterisation of the 3-PG model for South African conditions. The 3-PG model can potentially contribute substantially to a wide spectrum of applications to, and interpretations of, existing high resolution site classification systems in South Africa, especially in the fields of growth and yield prediction and economic analysis of forest site potential.