Abstract

Ricinodendron heudelotii kernels are an important source of income for rural women in the humid tropics of Africa. Manual kernel extraction is time consuming and efforts have been made to mechanize the process. However, mechanical damage to kernels remains an issue. Within the socio-economic and resource ecology context in the central region of Cameroon, a prototype machine for extracting njansang kernels was developed and has been tested since 2007, then evaluated using a learning selection model. Surveys with users and non-users suggested a shift towards male and younger users. The advantages of rapid kernel extraction are partly offset by the requirement to sort out broken kernels, leading to a total time requirement of 60.87 min for manual processing and 2.08 + 39.56 min for mechanical extraction + manual sorting. The technology needs further improvement to deliver the expected increase in labour efficiency and economic advantages for rural women.

Key words: Cameroon, adoption, agroforestry, NTFPs, smallholder producers.