THE IMPORTANCE OF PLANT DIVERSITY The diversity of plant life is an essential underpinning of most of our terrestrial ecosystems. Humans and most other animals are almost totally dependent on plants, directly or indirectly, as a source of energy through their ability to convert the sun's energy through photosynthesis. Worldwide tens of thousands of species of higher plants, and several hundred lower plants, are currently used by humans for a wide diversity of purposes as food, fuel, fibre, oil, herbs, spices, industrial crops and as forage and fodder for domesticated animals. In the tropics alone it has been estimated that 25,000-30,000 species are in use (Heywood 1992) and up to 25,000 species have been used in traditional medicines. In addition, many thousands of species are grown as ornamentals in parks, public and private gardens, as street trees and for shade and shelter. Another important role of plant life is the provision of ecosystem services the protection of watersheds, stabilization of slopes, improvement of soils, moderation of climate and the provision of a habitat for much of our wild fauna. While it is generally accepted today that the conservation of all biodiversity should be our goal, especially through the preservation and sustainable use of natural habitats, this is an ideal that is unlikely to be achieved and there are convincing scientific, economic and sociological reasons for giving priority to the conservation of the major centres of plant diversity throughout the world, especially as thiswill very often also lead to the conservation of much animal and micro-organism diversity as well. The concept of identifying centresof diversity and endemism The idea of seeking out high concentrations of diversity among plants, animals or both has a long history in biogeography in one form or another. Attention has frequently been paid to the floristic or faunistic richness of certain areas, such as the tropics of Asia, Africa and the Americas, the Mediterranean climatic regions, and the concentrations of species on islands, such as Madagascar and Indonesia. Particular emphasis has been given to the large numbers of species that are endemic to such areas. More recently, the concept of sites or centresof high diversity has attracted the attention of conservationists, both as a tool for helping determine which areas should receive priority attention, and also as a challenge as to how to undertake the conservation action necessary. Such efforts to seek out areas of high priority for conservation have acquired increased urgency in the light of the accelerating losses throughout the world of natural habitats and the biodiversity they contain, as a result of human action and the growth of the world's population. Determining priority areas for plants The problem of determining priority areas can be approached at different geographical scales global, regional, national or local. At a global level, Raven (1987) developed an approach based on analysis of the size of floras that are threatened, and highlighted the fact that about 170,000 of the world's estimated total of 250,000 species of angiosperms grow in tropical regions of the world, with an estimated 85,000 in Latin America, 35,000 in tropical and subtropical Africa (excluding the Cape), and at
least 50,000 in tropical and subtropical Asia. He drew attention to the remarkable fact that more than 40,000 plant species about a quarter of total tropical diversity occur in Colombia, Ecuador and Peru. Some of these regional figures have been modified subsequently: for example, the count for tropical Africa has been reduced from 35,000 to 21,000 in the light of more accurate assessments (A.L. Stork, pers. comm. to P. Raven 1991) and the figure for tropical Asia appears to have been under-estimated. The total number of single country endemics recorded by the World Conservation Monitoring Centre is a remarkable 175,976 species (see Table 1), an estimate which casts doubt on the generally accepted global total of about 250,000 species. The selection of sites The analyses of floristic richness and endemism described above, while providing useful general indications as to which areas might be considered for priority action, have severe limitations in that they are based essentially on species richness and endemism in selected areas, irrespective of the nature, relationships and values of the species concerned, the ecological diversity of the areas and socio-economic factors. Nonetheless they give useful pointers. Any top-down approach, no matter how sophisticated the science,is liable to fail unless full cognizance is taken of the detailed needs, perceptions, aspirations and political realities of the countries and regions concerned. We adopted from the beginning a principle of involving local experts and, wherever possible, national governmental and non-governmental conservation bodies. The objectives of the Centres of Plant Diversity (CPD) project are: to identify which areas around the world, if conserved, would safeguard the greatest number of plant species; todocument the many benefits, economic and scientific, that conservation of those areas would bring to society and to outline the potential value of each for sustainable development; to outline a strategy for the conservation of the areas selected. These objectives are fully consonant with the Convention on Biological Diversity. Although the original intention was to select between 150 and 200 sites of global priority, the total number finally chosen greatly exceeds these figures. In addition to the 234 priority sites selected for Data Sheet treatment, many more sites are treated in summary paragraphs in the Regional Overviews. By extending the Regional Overviews in this way, we overcame the problem of making an arbitrary selection of a single site where several potential sites occur in a particular area. The criteria and methodology used for selecting sites The criteria adopted for the selection of sites and vegetation types was based principally on a requirement that each must have one or both of the following two characteristics: (1) the area is evidently species-rich, even though the number of species present may not be accurately known; (2) the area is known to contain a large number of species endemic to it. The following characteristics were also considered in
the selection: the site contains an important genepool of plants of value to humans or that are potentially useful; the site contains a diverse range of habitat types; the site contains a significant proportion of species adapted to special edaphic conditions; and, the site is threatened or under imminent threat of large-scale devastation. To qualify for Data Sheet treatment, most mainland sites have (or are believed to have) in excess of 1000 vascular plant species, of which at least 100 (i.e. 10%) are endemic either to the site (strictly endemic) or to the phytogeographical region in which the site occurs. The criteria for the selection of islands treated as Data Sheets were somewhat different from those used for mainland sites. Many islands have depauperate floras compared with continental areas, but the level of endemism is often very high.
