A research report submitted to the Faculty of Science, University of Witwatersrand, in partial fulfillment of the requirements for the degree of Master of Science of the School of Animal Plant and Environmental Sciences Johannesburg, 3rd November 2017. / Only one tetraploid (four sets of chromosomes) species of baobab (Adansonia digitata, L.) was recognized on mainland Africa. However, a study published in 2012 reported the existence of a new diploid baobab species known as Adansonia kilima which was said to co-exist on mainland Africa with A. digitata. This new species was identified on the basis of morphology (mainly floral and stomatal features), ploidy level and molecular phylogenetics. The two species of African baobabs were also differentiated according to their elevation preferences of between 650 m –1500 m above sea level (a.s.l.) and below 800 m a.s.l. for A. kilima and A. digitata respectively, which were said to rarely overlap. The report of this new species and the need to accurately determine the exact number of species of baobabs existing on mainland Africa have brought about a renewed interest in the study of the African baobabs and have necessitated this study. This research therefore compared A. digitata and A. kilima to assess the latter’s authenticity as described by work. The objectives were to (i) examine floral and stomatal traits from samples across mainland African baobab populations to establish whether there are distinct differences that can be correlated with altitudinal differences in order to distinguish the species present in Africa, and (ii) use stomatal size and density to infer ploidy levels of baobabs that occur at low and high altitudes. Herbarium and fresh specimens of A. digitata and A. kilima from across Africa with both leaves and flowers were borrowed from various herbaria to represent the widespread distribution of baobab in Africa. Cluster analysis (CA), Principal Component Analysis (PCA) and Non-metric Multi-dimensional Scaling (NMDS) of seven floral traits of 124 African baobab specimens were used to analyze baobabs that occur at low or high altitude. These specimens did not form distinct clusters or separate groupings correlated with either low or high altitude. The inference of ploidy using stomatal size categories as per previous study showed no differences in ploidy level among the baobabs studied and the difference in morphological features between baobabs found at low and high altitudes also could not be linked to inferred differences in ploidy levels. Specimens were assigned diploids or tetraploids based on their stomatal length as defined by an earlier work and the altitude of each specimen was used to ascertain whether the ploidy level correspond to the altitudinal category reported by a study published in 2012. Box and whisker plots were also used to compare the floral (pollen grain diameter, volume and density) and stomatal traits of baobabs that occurred at low and high altitudes. There was no statistical difference in the variation in floral features among African mainland baobabs found at low and high altitudes except in staminal tube length and style length. The stomatal traits results reveal that the mean stomatal length and density (per 1000 μm2) between baobabs that occur at low and high altitudes were not statistically different except when mean stomatal length and density were grouped into the two putative ploidy categories as per previous study. Morphological features results suggest that the differences in floral size, and stomatal size and density of mainland African baobabs could be related to differences in climate, water availability and geographical variation. Based on the data analyzed for this work, it is concluded that A. kilima is not distinct enough from A. digitata, to be recognized as a new species. However, further research should investigate potential variation in genome size among African baobab that is correlated with altitude using flow cytometry analysis and/or chromosome counts. Key words: altitudinal variation, mainland African baobab, multivariate analysis, ploidy level, pollen grains, species concept, stomatal density, sympatric speciation. / LG2018