The current state of global biodiversity is one of sustained and increasing decline especially in developing countries such as South Africa, where, medicinal plants face a particular threat due the herbal medicine trade, and because in situ conservation measures have not stemmed the exploitation of these plants (Chapter 1). Furthermore, seed storage, which offers an efficient ex situ conservation technique, cannot presently be applied to many medicinal plants, either because these species produce short-lived, recalcitrant seeds, or the post-shedding behaviour of the seeds is altogether unknown. This study investigated three medicinal plant species indigenous to Kenya and South Africa: Trichilia dregeana and T. emetica, of which no population inventories exist and no wild populations were encountered locally during the course of this study; and Warburgia salutaris, one of the most highly-utilised medicinal plants in Africa, and which is currently endangered and virtually extinct in the wild in some countries such as South Africa. Aspects of post-shedding seed physiology (Chapter 2) and the responses of the germplasm of the three species to cryopreservation (Chapter 3) were studied using viability and ultrastructural assessment, with the aim of establishing methods for both short-term and the long-term preservation, via appropriate seed storage and cryopreservation, respectively. The effect of cryopreservation on genetic fidelity, a crucial aspect of germplasm conservation, was assessed by polymerase chain reaction (PCR) based random amplified polymorphic DNA (RAPDs), using W. salutaris as a case-study (Chapter 4). The seeds of all three species were found to exhibit non-orthodox behaviour. On relatively slow-drying, seeds of T. dregeana and T. emetica lost viability and ultrastructural integrity at axis water contents of 0.55 g g-l (achieved over 6 d) and 0.42 g g-l (after 3 d) respectively, while flash-drying of embryonic axes facilitated their tolerance of water contents as low as 0.16 g g-l (T. dregeana, flash-dried for 4 h) and 0.26 (T. emetica, flash-dried for 90 min). Seeds of W. salutaris were relatively more tolerant to desiccation, remaining viable at axis water contents below 0.1 g g-l when desiccated for 6 d in activated silica gel. However, excised embryonic axes flash-dried to similar water contents over 90 min lost viability and were ultrastructurally damaged beyond functionality. In terms of storability of the seeds, those of T. dregeana could be stored in the fully hydrated state for at least 5 months, provided that the quality was high and microbial contamination was curtailed at onset of storage, while those T. emetica remained in hydrated storage for about 60 d, before all seeds germinated in storage. Seeds of W salutaris, even though relatively tolerant to desiccation, were not practically storable at reduced water content, losing viability within 49 d when stored at an axis water content of 0.1 g g-l. The seeds of all three species were sensitive to chilling, suffering extensive subcellular derangement, accompanied by loss of viability, when stored at 6 °C. Thus, T. dregeana and T. emetica are typically recalcitrant, while those of W. salutaris are suggested to fit within the intermediate category of seed behaviour. For either recalcitrant or intermediate seeds, the only feasible method of long-term germpalsm preservation may be cryopreservation. Subsequent studies established that whole seeds of W. salutaris could be successfully cryopreserved following dehydration in activated silica gel. However, whole seeds of T. dregeana and T. emetica were unsuitable for cryopreservation, and excised embryonic axes were utilised. For these, in vitro germination methods, as well as cryoprotection, dehydration, freezing and thawing protocols were established. Post-thaw survival of the axes of both species was shown to depend on cryoprotection, rapid dehydration and cooling (freezing) in cryovials. Embryonic axes of T. dregeana regenerated only as callus after cryopreservation, while those of T. emetica generated into apparently normal plantlets. Thawing/rehydration in a 1:1 solution of 1 µM CaC12.2H2O and 1 mM MgC12.6H2O increased the percentage of axes surviving freezing, and that of T. emetica axes developing shoots. The effect of the extent of seed/axis development on onward growth after cryopreservation was apparent for seeds of W. salutaris and excised axes of T. emetica. The seeds of W. salutaris surviving after cryopreservation germinated into seedlings which appeared similar to those from non-cryopreserved seeds, both morphologically and in terms of growth rate. Analysis using PCR-RAPDs revealed that there were no differences in both nucleotide diversity or divergence, among populations of seedlings from seeds which had been sown fresh, or those which had either been dehydrated only, or dehydrated and cryopreserved. Thus, neither dehydration alone, nor dehydration followed by cryopreservation, was associated with any discernible genomic change. The above results are reported and discussed in detail in Chapters 2 to 4, and recommendations and future prospects outlined in Chapter 5. / Thesis (Ph.D.)-University of Natal, Durban, 2002.