Environmental factors affecting wood properties of Eucalyptus spp. grown on the Zululand coastal plain and along the Mpumalanga escarpment of South Africa.
The environmental factors affecting wood property formation of Eucalyptus spp. trees in two distinct geographic areas within South Africa were studied. Wood prop~rty data for trees from 43 sites (26 in Zululand and 17 in Mpumalanga) were collected from work conducted at the Forest and Forest Products Research Centre (FFPRC) at the CSIR (Council for Scientific and Industrial Research) in Durban. The wood properties considered included screened pulp yield, fibre length, wood density and active alkali chemical consumption during pulping. The effect of environmental factors on growth rate (expressed as Site Index at a base age of 5 years) was also measured. A detailed site description for each forest compartment was carried out. Detailed annual and monthly rainfall and minimum and maximum temperature estimates were calculated for each site by interpolating long term means of these variables by splining using the software package Anusplin. These estimates of climatic factors were validated by comparison to the data published in the South African Atlas of Agrohydrology and -Climatology. The outputs of the Anusplin model were used to derive surrogate bioclimatic parameters for each site using the computer program Bioclim. These parameters are considered as better descriptors of the energy-water balance experienced by the plant than normal measures of climate such as mean monthly or annual precipitation. Soil characteristics were measured on samples taken from the individual sites. The effects of these environmental and bioclimatic variables on wood properties were analysed using appropriate statistical techniques. Multiple regression models were used to predict wood properties and it is suggested that this approach could form part of a fibre management system. Wood property prediction models incorporating climate (and bioclimate) alone were preferred to those including soil data as no further site data are required. The effect of edaphic factors was considered to describe any further variation not accounted for by bioclimate alone. Particle size distribution of the soil, as an indication of the water holding capacity of that soil, was not found to effect wood properties or growth significantly. A weak influence of organic matter content in the topsoil on wood density was noted in Mpumalanga. In Zululand, a multiple linear regression using both rainfall of the wettest quarter and mean diurnal temperature range as inputs yielded the best predictive model for growth rate. In this region a combination of precipitation seasonality and mean diurnal temperature range gave the best linear regressi'o,n model describing variation in screened pUlp' yield and fibre length. In Mpumalanga effective rooting depth was found to have a pervasive effect on plant development. Solar radiation (as a measure of energy supply), calculated from a function of latitude, aspect, slope and time of year, was also found to significantly affect the growth rate and SPY of plant material in Mpumalanga. Measures of temperature in both geographic regions were found to significantly affect wood density.