Application of Multivariate Analysis to Site Quality Evaluation for Coniferous Plantations

Abstract

SYNOPSIS Multivariate analysis was performed on tree growth and environmental data from sites afforested with conifers in Ventania, southern Buenos Aires, Argentina. Pinus halepensis was investigated in 22 sites, Pinus radiata in 20 sites and Cedrus deodara and Cupressus sempervirens f. horizontalis in 15 sites each. Site index expressed as height of dominants at age 25 years was used as growth indicator. Four site quality classes were set up for each of the four species. Twenty three topographic and soil characteristics were used to group the sites into six groups by means of a non-hierarchical cluster analysis technique. The groups were then used to test the ability of the environmental variables to discriminate among site quality classes. Standardised mean site index of stands on cumulic soils were significantly higher than those on shallow soils. A hierarchical technique used to establish intergroup relationships produced similar results to cluster analysis. Seven variables were selected as potential ...