only be revealed through exhaustive floristic analysis, a task that has been accomplished for very few species groups in Africa. Even in South Africa, which has benefited from more than 200 years of biological survey, new centers of endemism of considerable importance are only now being detected.
During the past 30 years, the International Union for the Conservation of Nature and Natural Resources has developed a comprehensive data bank on the protected area systems of the world. It has recently completed a detailed listing of all protected areas of the Afrotropical realm, providing exhaustive information on the geographic, faunal, floral, and management attributes of each of some 620 protected areas greater than 50 square kilometers in the realm (IUCN/UNEP, 1987). This data base served as the foundation for the review of Afrotropical protected areas undertaken by MacKinnon and MacKinnon (in press), whose analysis indicates that 4.7% of the realm falls within protected areas totalling 949,500 square kilometers, considerably larger than the State of Texas. By comparing the total area of each mapped vegetation unit with that falling within protected areas, they found that only one of the seven major centers of endemism of the Afrotropical realm has more than 10% of its area protected. The other six centers have between 3.6 and 7.0% of their area within national parks and reserves. Even at the extremely coarse scale of resolution afforded by this analysis, the finding that as little as 3.6% of the biotic resource is protected gives cause for concern. On closer inspection, the situation is even worse—many of the so-called national parks in Africa are little more than yellowing documents in government archives. The Giant Sable Integral Nature Reserve in central Angola is occupied by more than 20,000 peasant farmers, several trading villages, and until guerilla activity prevented it, extensive diamond prospecting. The Reserve has not seen a game ranger in 10 years.
Factors such as the above necessitate a more objective evaluation of the effective protection afforded each biogeographic unit and protected area. In recent years, a variety of scoring systems have been proposed. These systems take account not only of the relative area protected but also of the effectiveness of government action to provide long-term security to the area (Clarke and Bell, 1986; Cumming, 1984). MacKinnon and MacKinnon (in press) developed a scoring system based on the size, protection objectives, and management effectiveness reported for each site. These data were then summarized by biogeographic unit and weighted for the number of distinct habitats and altitudinal ranges represented within the conservation network. Assessment of priority for action was based on the principle that action should be taken where it could have the best effect, not on lost causes—an all too common failing of conservation efforts based on sentiment rather than science.
At a finer scale of resolution, an analysis of the protected area cover of 189 vegetation units in 10 southern African states demonstrated the existence of major deficiencies in many of the 24 major vegetation divisions recognized (Huntley and Ellis, 1984). The most seriously threatened systems included the lowland forests