Equations for predicting the biomass of Acacia cyclops and Acacia saligna in the western and eastern Cape regions of South Africa: Part 2: stand-level models: scientific paper

Abstract

The study deals with stand-level models for estimating wood biomass for different tree sizes and different biomass categories. It was carried out on farms and nature reserves in three regions. The stratification was identical with that in part 1 of the study, which dealt with tree-level models. The three regions (Western Cape Plains, Agulhas Plains, and Eastern Cape Plains) were sampled independently. Within each region the two species involved, A. Cyclops and A. saligna, were also sampled independently. No sample plots were measured in A. saligna in the ECP region. Within each of the five strata, the number of sample plots was approximately proportional to the size of the locality. In consequence the sampling method can be described as stratified random sampling with Probability of Selection Proportional to Size. In addition, the leaf biomass per unit area was estimated for each stratum. Equations were fitted for each stratum, which produced estimates for the parameters of the decreasing diameter distributions. An analysis of variance based on a nested model was carried out to estimate the between-locality and between-plots within localities variances.


Key Words: Acacia cyclops, Acacia saligna, Inventory, Stand level regression models, Wood and leaf biomass


Southern African Forestry Journal Vol.201, 2004: 35-42