That tropical trees are extremely diverse is well established, and recognition of this fact has led to better exploitation of both their inter- and intra-specific genetic diversity. Taxa used in agroforestry, however, have generally been less studied than their industrial counterparts (Eucalyptus, Pinus). Leakey and Newton (1995) referred collectively to these neglected taxa as the Cinderella species. It is some of these largely wild and predominantly indigenous species that this project seeks to focus upon. The aspects of domestication of agroforestry trees to be studied are: (a) efficiency of phenotypic selection; (b) effective conservation of genetic diversity of high utility value; and (c) delivery pathways for superior germplasm.

ICRAF in collaboration with ISNAR and NARS have developed more objective procedures for priority-setting of agroforestry tree species for genetic improvement research (Jaenicke et al., 1995). These methodologies have now been used in a number of eco-regions, including the Humid Lowlands of West Africa (HULWA), the lowland forest of the Peruvian Amazon and the Yucatan Peninsula in Mexico. In these regions, and in the Miombo eco-zone of southern Africa, the top five species are predominantly fruit and timber trees. This project will focus on these agro-forestry trees of high economic value that primarily provide fruit or timber products.

High-value timber and fruit trees with improved performance identified with promotion of their use in agroforestry systems and determination of priority germplasm for conservation (in situ, ex situ and circa situ).

• Efficiency of targeted and random collections determined
• Identification for conservation of priority individuals or populations with high utility value.
• Understanding of delivery pathways for superior germplasm of agroforestry trees of high utility value.