Supervised Learning Method: Critical Analysis and Updating of Cubing Rate Formulas for Determining Bark Masses of <i>Prunus africana</i> (Hook. f.) Kalkman (Rosaceae) in Cameroon

Abstract

The stem barks of Prunus africana are used in the treatment of the benign prostate. Cameroon is one of the important exporters of the barks. Despite the important measures adopted in Cameroon for sustaining its harvesting, some many chalenges still remain. The objective of this work is to refine the forest management parameters in relation to P. africana in the regions of Adamaoua and the South-West by developing a volume rate which makes it possible to estimate the production for a new stem. The work took place in two phases: in the South-West in 2010 and in Adamaoua in 2011. Data collection used the semi-direct method, while the cubing equation was deduced by the multiple linear regression method. Two models for volume estimation and three models for mass prediction were developed. The predictive parameters retained are diameter, height of the bole and thickness of the bark. Results show that the average mass of the dry bark for a given P. africana tree species is 27.55 ± 14.44 kg and this varies according to the site. The strong adjusted coefficient of determination (adjusted R2) observed illustrates the reliability of the proposed models. These models provide a reliable tool that can be adopted as a standard in Cameroon for P. africana.