Thaumatococcus daniellii, the source of the super-sweet protein, thaumatin, is a perennial understory forest herb of West and Central Africa. This thesis investigates its natural variation and potential as an intercrop with rubber (Hevea brasiliensis) in West Africa. Field work undertaken in Ghana and Cameroon, resulted in two distinct locations per country being used as sites for the collection of morphological data, and the supply of planting material for experimental sites. Two experimental sites were established: the main
experimental site was based at the Missellele Rubber Estate, Cameroon, under plantation rubber; two secondary experimental field trials were established under rubber smallholdings near Akona Junction, Ghana. Natural variation in the plant was studied through the collection of morphological data from four locations in Cameroon and Ghana. Data collected included petiole length, lamina width and breadth, fruit size and mass, and aril number and mass. Arils were extracted and thaumatin was recovered from them to assess if
thaumatin content differed amongst populations. Significant differences in morphology were found amongst populations. Thaumatin content, varied significantly amongst populations but the range (1.08-2.00%) was within industry standards for all accessions. A molecular study of variance based on leaf material collected from the four populations detected 364 AFLP
markers using three primer sets which were used to investigate phenetic and phylogenetic variance within and among populations. The study revealed that populations geographically close to one another were more closely related genetically, than those with similar phenotypic morphology. There was no evidence to suggest that speciation within T. daniellii is occurring despite some reports of this in the literature. Genetic differentiation between
populations, 58%, was greater than within regions and within populations. φRT values of 0.777 (p = 0.01) indicate a strong genetic structure and suggests some isolation amongst populations. Three field trials were established in Cameroon. 1n the first trial sections of T. daniellii rhizomes from each of the four populations studied above were planted under nine year old rubber in a randomized block design to study whether the phenotypic expression of wild populations would be stable maintained across environments, to document phenology
and quantify fruit production. The second trial investigated the effect of early growth of T. daniellii on rubber production using a second randomised block design. The trial was planted with material from one population each from Ghana and Cameroon, plus material found at the experimental site. The final trial studied the effect of shade on the growth of T. daniellii.
Two blocks of six plots were established with trees surrounding plots lopped by different amounts. Different methods to measure the shade cast by rubber canopy were compared. Fruit were produced two and a half years after planting with locally adapted material producing significantly more and larger fruit (p<0.05) than germplasm obtained from further away from the site. Phenotypic differences measured across locations were not maintained
when the material was all grown at the same site, suggesting that the traits under study were more under environmental than genetic control. T. daniellii growth was unaffected by rubber tree pruning but intercropping reduced cup lump rubber production by up to 12%. Growth of T. daniellii planted from rhizomes and seedlings was investigated in trials established in Ghana under rubber smallholdings. The trials were useful for small holders to
gain experience with the plant and an indication of potential fruit production from small holder rubber was obtained. There was no difference in growth and fruit production amongst types of planting material. Overall, the study has demonstrated the practicality of growing T. daniellii under rubber and the significant genetic and phenotypic variation found in fruit
yield indicate that selection of appropriate material for use at potential cultivation sites may have potential to increase productivity.
Keywords: Thaumatococcus daniellii, Hevea, Rubber, AFLP, West Africa, thaumatin, shade response.