Indigenous forests of South Africa are few, small in size and highly fragmented, yet they face intense exploitation particularly in the rural areas. Management of these forests is challenging. High rural dependency on forests and the need to ensure the maintenance of the forest ecological processes that maintain biodiversity and ecosystem integrity are at odds with one another. Rural needs from forests are mainly short-term and interfere with the longer cycle of ecosystem maintenance. In Umzimvubu District of Eastern Cape Province forest management through sustainable use is hampered by a lack of information about the forests' ecology, resource availability and exploitation patterns. Thus it is difficult to set quotas or sustainable harvesting levels. This study addresses these challenges by tackling crosscutting themes of (1) forest policy and use, (2) forest resource availability and exploitation and, (3) the ecological processes of forest regeneration - all vital components for sustainable forest management. Using questionnaires it was found that all forest stakeholders (i.e. forest resource users and managers) were opposed to a ban on forest resource use but agreed to regulated harvesting. Although the new forest policy advocates the devolution of forest management from the state to communities, resource users preferred a greater role for the state in forest management than expected. Given the choice stakeholders selected state forest management over community forest management. However, the combination of roles of the state and communities in forest management, as exemplified by the new policy of participatory forest management (PFM) is probably the most applicable management practice, although it is not without its problems. Tree species are the focus of this study. Trees were largely used for fuelwood, medicinal purposes, craftwork, fencing posts and building poles and involved twenty species. Poles and posts were indiscriminately harvested from the medium (10-20cm dbh) tree size-class. Fuelwood harvesting was selective and only certain species were used. Fuelwood harvesting is unsustainable because the average amount of deadwood produced by the forest marginally balanced that removed from forest as fuelwood. Similarly medicinal tree harvesting (largely through stem debarking) was unsustainable and at least 28% of the debarked trees died. Only one species (Macaranga capensis) could withstand the current stem bark harvesting pressure. Species suitable for pole and post harvesting were determined by a graphic method, based on linear-programming approach that examined the spatial scale or grain of regeneration of a species. The grain of a species is established by comparing the density of stems from a species at the forest canopy and sub-canopy levels and sanctions harvesting only if a species was adequately represented at both levels i.e., fine-grained. Only one species (Englerophytum natalense) met these requirements in all forests and could sustain high levels of pole and post exploitation. Harpephyllum caffrum and Heywoodia lucens are among the most coarse-grained species and their use is discouraged. A relatively high percentage of the forest is under gaps (7.8%) created via natural disturbances of windthrow (50%), breaking tree branches (20%) and snags (13%). Another 17% result from selective tree cutting activities. The gap-phase dynamics paradigm appears to play a minor role in forest tree regeneration, as gap-size niche-differentiation is weak and there is no gap-filling guild of pioneer species. A lottery paradigm best explains tree regeneration in gaps in the forests of Umzimvubu District. Although selective tree harvesting creates gaps, in moderation gap creation is unlikely to change the forests' species composition since there is no gap-filling guild that is favoured by gaps and recruitment is a chance event. Resource use in the forests of Umzimvubu District is unsustainable and PFM offers a viable option for managing these forests. Sustainable use of forest cannot be achieved without an integration of the multifaceted social and ecological issues of forest management and more importantly without prioritising and understanding the ecology of forests. / Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 2002.