A stratification of the South African forestry landscape based on climatic parameters

Abstract

The benefits of high-resolution forest site classification systems have been widely acknowledged in South Africa as a prerequisite for decision-making in forestry management. Although substantial research and development work have been done on categorising and mapping the soil and geological environment, surprisingly little has been done to improve our understanding of the role of climate in plantation ecology and to incorporate climatic variables into spatial databases. A stratification of the forest landscape of South Africa is proposed, using growth days and growth temperature as primary variables. The appropriateness of these variables is discussed in terms of its relation to plant physiology and landscape ecology. The provinces of KwaZulu-Natal and Mpumalanga are used as a case study for demonstration purposes. A total of nine growth day classes and 17 growth temperature classes were identified. An analysis of variance illustrated a significant difference between all classes in terms of related factors expressing moisture and temperature regimes, indicating the relevance of growth days and growth temperature for identifying areas of homogenous climatic conditions. The covariation between growth days and growth temperature and factors expressing moisture and temperature, respectively, also indicates the usefulness of these variables for landscape stratification in complex biophysical environments such as South Africa. The application potential of this stratification system is discussed in terms of its value as a resource inventory, as well as for species choice, site quality prediction, growth modelling, site ecology analysis and various aspects of risk management in plantations. It is proposed that this system be regarded as a first step towards a more holistic multifactor national forest site classification system.

Keywords:

• climate classification
• forest site classification
• growth days
• growth temperatures