Annals of Botany 80 : 223230, 1997

BOTANICAL BRIEFING

The Brazilian Cerrado Vegetation and Threats to its Biodiversity

J. A. R A T T E R*, J. F. R I B E I RO and S. B R I D G E W A T ER* * Royal Botanic Garden Edinburgh, 20A In erleith Row, Edinburgh EH3 5LR, UK and EMBRAPA, Centro de Pesquisa AgropecuaT ria dos Cerrados (CPAC), BR 020-km 18, CEP 73301970, Planaltina, DF, Brazil
Received : 6 March 1997 Accepted : 22 May 1997

The Brazilian cerrado (savanna) biome covers 2 million km# representing 23 % of the area of the country. It is an ancient biome with rich biodiversity, estimated at 160 000 species of plants, fungi and animals. There are about 800 species of trees and large shrubs in the savanna vegetation and several times that number of ground species (herbs and subshrubs). When the ora of gallery forests, mesophytic forests and other habitats occurring in the biome are included, the total number of vascular plant species is estimated to reach about 10 000. During the last 25 years modern agriculture has been developed in the cerrado to produce soya, maize, rice, etc and enormous numbers of cattle are raised in planted pastures. Charcoal production for the Brazilian steel industry also causes great destruction of the cerrado. By 1994 an estimated 695 000 km# of cerrado (representing 35 % of its area) had been converted to anthropic landscape . This compares to the destruction of about 400 000 km# of Brazilian Amazonian forest representing 12 or 13 % of the area of this biome. Conservation initiatives are now desperately needed. Only 1n5 % of the cerrado biome is preserved as Federal Reserves and this area needs to be at least tripled. Surveys of the vascular ora aimed at discovering biogeographic patterns are now in progress with the objective of choosing representative areas and biodiversity hot spots for conservation. # 1997 Annals of Botany Company Key words : Brazil, cerrado, biodiversity, conservation.

INTRODUCTION Mention the vegetation of Brazil and the perils of its destruction and the picture conjured up in most minds is that of the tall Amazonian forest, the green hell of the more lurid travel writers, with lurking jaguars, anacondas and uncontacted Indian tribes. Few realize that the country also contains an immense species-rich savanna area, threatened to an even greater extent by destruction than the Amazonian forest. The present article is devoted to this unjustly neglected savanna with the intention of focusing attention on its importance and its present plight. Clearly, in such a brief communication, much of the eld has to be omitted, and for more detailed information, readers are referred to such works as Eiten (1972, 1982), Goodland and Ferri (1979), Novaes Pinto (1990), Dias (1992) and Alho and Martins (1995).

THE CERRADO The Brazilian savanna vegetation is called cerrado and covers some 2 million km# of Central Brazil (Fig. 1), representing about 23 % of the land surface of the country (the same size as Western Europe). In terms of area it is exceeded by only one vegetation formation in Brazil, the Amazonian forest covering approx. 3n5 million km#. The cerrado region extends from the margin of the Amazonian
For correspondence.

forest to outlying areas in the southern states of Sa 4 o Paulo and Parana! , occupying more than 20m of latitude and an altitudinal range from sea-level to 1800 m ; about 700 000 km# of the total area of cerrado vegetation is within the Amazon basin. In addition, there are areas of cerrado in Bolivia and Paraguay, while related, but species-poor, savanna vegetation such as the Roraima and Rupununi savannas and the Venezuelan llanos occur north of the Amazon. The cerrado climate is typical of the rather moister savanna regions of the world, with an average precipitation for over 90 % of the area of 8002000 mm and a very strong dry season during the southern winter (approx. April September), while average annual temperatures are 1828 mC (Dias, 1992). The soils of most of the area are dystrophic, with low pH and availability of calcium and magnesium, and high aluminium content (Lopes and Cox, 1977 a ; Furley and Ratter, 1988) ; most are oxisols (ferralitic soils). They are always well-drained and cerrado vegetation is intolerant of waterlogging. In this respect, the species-rich cerrado contrasts with the seasonally inundated, speciespoor hydrologic savannas common north of the Amazon. The typical vegetation landscape within the cerrado biome consists of cerrado on the well-drained interuves with gallery forests following the watercourses. The cerrado itself is very varied in form, ranging from dense grassland, usually with a sparse covering of shrubs and small trees, to an almost closed woodland with a canopy height of 1215 m. It is dicult to discover the causes of these dierences in vegetation structure. Some workers have related them to # 1997 Annals of Botany Company

0305-7364\97\090223j08 $25.00\0

bo970469

224

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

F. 1. Distribution of cerrado vegetation in Brazil (letters are state abbreviations). Those referred to in connection with geographical patterns are : DF, Federal District ; GO, Goia! s ; MA, Maranha 4 o ; MG, Minas Gerais ; MS, Mato Grosso do Sul ; MT, Mato Grosso ; PA, Para! ; TO, Tocantins.

soil fertility gradients (Goodland and Pollard, 1973 ; Lopes and Cox 1977 b) ; the data of others, however, fail to demonstrate this correlation, and show well-developed woodland on no more fertile soils than sparser forms nearby (e.g. Ribeiro, 1983). In many cases, re is undoubtedly an important factor. The cerrado ora, like that of all savannas, is, of course, resistant to re and shows all the usual adaptations, such as thick corky bark, xylopodia (lignotubers), tunicate leaf-bases in grasses etc. Nevertheless, too

frequent burning causes destruction, particularly of trees and shrubs, and thus favours the herbaceous element of the ora at the expense of the woody. A brief description of the most widespread form of cerrado will give the reader a better idea of its appearance. It consists of a community of trees and large shrubs, usually about 28 m in height, belonging to many species and producing 1060 % cover, with a well-developed grassy ground layer between. The trees are of characteristic savanna

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

225

F. 2. A, Cerrado vegetation in the dry season ; B, charcoal burning following complete clearance of cerrado vegetation ; C, soya cultivation in a featureless cleared landscape ; D, area cleared and prepared for cultivation in the Federal District.

form, typically with contorted trunk, thick corky bark and hard sclerophyllous leaves ; some are evergreen while others are deciduous for variable periods of the dry season. The ground layer is usually up to about 60 cm tall and consists of many species of grasses and sedges mixed with a great diversity of forbs, amongst which Leguminosae, Compositae, Myrtaceae and Rubiaceae are particularly important families. Sometimes the appearance of this low tree cerrado is not unlike an orchard. Many of the trees and smaller plants have strikingly large and colourful owers, so virgin cerrado often resembles a beautiful wild garden. Where areas of richer (mesotrophic) soils occur in the cerrado landscape, they are clothed in mesophytic deciduous

and semi-deciduous forests. Such areas range from tiny patches where valleys have cut into base-rich rocks, to such extensive enclaves as the Mato Grosso de Goia! s, estimated to cover 40 000 km# (approx. twice the size of Wales) before its destruction for agriculture. Another conspicuous vegetation type found throughout the cerrado landscape is the wet campo, frequently occurring as a belt between cerrado and gallery forest. Here the water table rises so high during the rainy season that the area is often waterlogged, but falls drastically during the dry season ; such uctuation in moisture availability does not allow tree growth. However, such campos often contain long lines of tall, graceful, Mauritia exuosa L.f. fan palms

226

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

families in terms of species numbers, using the gures of Heringer et al. (1977), are : Leguminosae (153 spp., all three subfamilies), Malpighiaceae (46 spp.), Myrtaceae (43 spp.), Melastomataceae (32 spp.) and Rubiaceae (30 spp.). But in many areas the vegetation is dominated by Vochysiaceae (an almost entirely South American family, with 23 spp in the cerrado), largely because of the abundance of three species of Qualea, one of the familys largest genera. Heringer et al. (1977) analysed the geographic anities of the 261 genera they recorded and found that 205 had species in common with the Brazilian Atlantic Forest, 200 with the Amazonian forest, 30 with the mesophytic forests, 51 with species of the cerrado ground layer, while seven (three of which were monotypic) only occurred as cerrado trees. The strongest anities of the woody vegetation of the cerrado lie then almost equally with the Atlantic and Amazonian rainforests, and in our opinion the cerrado may represent the progenitor of both of these formations. Recent work by Ratter and Dargie (1992), Ratter et al. (1996) and Castro (1994 a, b) have been directed towards discovering patterns of distribution of cerrado vegetation by comparisons using multivariate techniques of oristic surveys covering the entire cerrado region (Ratter and colleagues compared 98 areas, while Castro has worked on a similar number). The results of the two studies are much in accord. Our research demonstrated a strong geographic pattern in the distribution of the ora and allowed the recognition of southern (Sa 4 o Paulo and S. Minas Gerais), southeastern (largely Minas Gerais), central (Federal District, Goia! s and parts of Minas Gerais), central-western (largely Mato Grosso, Goia! s and Mato Grosso do Sul) and northern regions (principally Maranha 4 o, Tocantins and Para! ), as well as a disjunct group of Amazonian savannas (see Fig. 1). Diversity of trees and large shrubs occurring at a single site (alpha diversity) may reach as many as 120 species per hectare, but is generally considerably lower than this, while at the other extreme it can be less than ten species in the isolated Amazonian savannas. On the other hand, a remarkable intersite heterogeneity (beta diversity) occurs, as demonstrated by our recent work comparing 98 sites throughout the cerrado region (Ratter et al., 1996). In total, 534 species were recorded, of which 158 (30 %) occurred at a single site only, while none occurred at all sites and only 28 species (5 %) were present at 50 % or more sites. The extreme oristic heterogeneity (beta diversity) of cerrado vegetation has important consequences for conservation planning, since it will necessitate the establishment of many protected areas to preserve the biodiversity adequately. The sites we have recorded with the highest species numbers ( biodiversity hot-spots ) are in Mato Grosso, Goia! s (Alto Araguaia region), Tocantins and in the Federal District (see Fig. 1). The number of congeneric species growing together in a small area of cerrado can be surprisingly high : for instance, at the ecological reserve of the University of Bras! lia in the Federal District, one can nd six species of Miconia and Byrsonima, and ve species of Erythroxylum and Kielmeyera. Recent research has thrown light on the mechanisms underlying the maintenance of so much diversity in the

and are then called veredas, one of the most beautiful features of the Central Brazilian landscape. HISTORY AND BIODIVERSITY The cerrado is a vegetation formation of great antiquity and there are even suggestions that it existed in prototypic form in the Cretaceous, before the nal separation of the South American and African continents (Ratter and Ribeiro, 1996). Unfortunately, however, there is little fossil evidence of the geological history, although deposits which can be interpreted as typical grassy savanna are known from the Tertiary in South America (Van der Hammen, 1983). As is now well known, a dynamic relationship has existed between forest and savanna type vegetation during the Pleistocene with expansion of savannas [probably including mesophytic (deciduous) forest as a dominant element (Prado and Gibbs, 1993 ; Ratter et al., 1988)], and contraction of the Amazonian forest during glacial periods, and ice ersa during interglacial periods (see, for example, Prance, 1973, 1982). Such processes have produced complicated patterns in both ora and fauna, and fragmentation of populations has probably led to exuberant speciation. A great deal of study has been devoted to this subject in the Amazonian forest but, as yet, such aspects have received only limited attention in the cerrado. The combination of the great age of the cerrados and the relatively recent (Pleistocene) dynamic phase in distribution patterns has probably led to their rich overall biodiversity, estimated by Dias (1992) as totalling 160 000 species of plants, animals and fungi. The long history of the biome has allowed the evolution of complex interactions between organisms, e.g. the exceptionally diverse rust ora showing high specicity to the native angiosperms (J. C. Dianese, University of Bras! lia, pers. comm.) and the great numbers of insect and other plant galls. More recent disruption of populations probably led to the evolution of many microendemic species in the isolated gene pools. The tree and large shrubby species* of the cerrado are comparatively well known. The base list for these was provided by Rizzini (1963) and added to by Heringer et al. (1977). In all, they recorded 774 species belonging to 261 genera, of which 336 species (43 %) were regarded as endemic to the vegetation. They considered only the vegetation of the cerrado (tree savanna proper), excluding the associated gallery and mesophytic forests of the biome. Since 1977, much research has been carried out on the oristics and phytosociology of the cerrados and the number of species recorded has increased considerably. Some workers give an estimate of about 1000 species ; however, one has to be careful not to include extraneous species (e.g. weedy trees such as cecropias, gallery forest species, etc.) and it is possible that the number of true cerrado tree and large shrub species does not exceed 800. The most important
* In cerrado, and other savanna vegetations, it is dicult to distinguish between trees and shrubs, and the same species can often be found in ower and fruit between 12 m and 15 m in height. We regard the denition of a large shrub as attaining 2 m in height and having perennial aerial woody shoots (i.e. exclusive hemixyles which have woody shoots of short duration).

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

cerrado vegetation. Various workers have demonstrated that the vast majority of cerrado species so far investigated are obligate outbreeders and that self-incompatibility is very widespread (see Gibbs, 1990), indicating that the breeding systems are such as to promote a high level of outcrossing and thus much diversity and exibility. In addition, strong barriers to interspecic hybridization in the form of seasonal isolation, dierences in pollination systems, and internal genetic barriers, seem to be common. This type of information, detailing how the cerrado really functions at the genetic level, is vital for the long-term planning of conservation strategies. For example, knowledge of breeding systems is important in determining the size of populations which should be conserved to maintain the genetic viability of a species. The diversity of the ground layer (herbs, subshrubs and small shrubs) is much richer than for taller plants, and species numbers are so high that detailed oristic lists are only available for comparatively few localities. Rizzini (1963) gives gures of more than 500 genera of smaller plants against less than 200 (increased to 261 by Heringer et al., 1977) for taller woody vegetation, while the results of detailed studies in the Brazilian Federal District (Ratter, 1986 ; Pereira, Silva and Mendona, 1993) show between four and seven times the number of ground layer species compared to trees and large shrubs. A wild guesstimate of the ground layer might suggest about 4000 species, but clearly more research is needed to produce accurate gures. A large number of families are important in the ground layer, particularly Gramineae, Compositae, Leguminosae (all three subfamilies) and Rubiaceae. So far we have only dealt with species diversity of vascular plants of the true cerrado vegetation (i.e. forms of savanna), but it must be remembered that the cerrado biome also contains an abundance of gallery forests and, where richer soils occur, great extensions of mesophytic forests. A number of recent surveys have shown a much greater diversity of tree\large shrub species in the gallery forests than in the cerrado itself ; for instance Cesar (1994) found 260 species in the galleries of the Bras! lia National Park, but only 109 species in the cerrado (a ratio of 2n4 : 1), while Pereira et al. (1993) list 183 species from the galleries and 84 from the cerrado of the Reserva do Roncador, also in the Federal District (ratio 2n2 : 1). Furthermore, there are great dierences in oristic composition of galleries across the vast cerrado region : those of the west and north show strong links to the Amazonian forests, while those of the centre and south show stronger anity with the montane semi-deciduous forests of south eastern Brazil (OliveiraFilho and Ratter, 1995). The great majority of the species of the gallery forests occur in other vegetation formations, but a signicant number are gallery endemics . The mesophytic forests have a very characteristic ora, but with a greater tendency for species dominance and a much lower oristic diversity than occurs in the cerrado. We do not intend to discuss the diversity of groups other than vascular plants here, but some mention is necessary. Dias (1992) has attempted to give estimates for all groups and, as already mentioned, his overall total for the cerrado biome is 160 000 species of plants, fungi and animals, of

227

which the numerically most important groups are insects (90 000 spp), fungi (40 000 spp) and angiosperms (10 000 spp). His estimate for angiosperms seems reasonable but, of course, includes all the communities of the cerrado biome : cerrado (savanna) proper, gallery and mesophytic forests, wet campos, etc. The numbers for such well known groups as mammals and birds are 150 and 550 species, respectively, few of which are completely endemic to the biome. As a nal thought on biodiversity, an interesting aspect of the cerrado biome is that, unlike the African savannas, it has lost the fauna of large mammals with which it must have co-evolved throughout the Tertiary. The large herbivores (grazers and browsers) must have been eliminated as a result of competition with North American fauna which migrated across the Panama Land-Bridge in the Great American Interchange 3 million years ago in the late Pliocene, or later in Mans Pleistocene and Holocene Overkill. The only remnants of the ancient neotropical mammalian fauna now occurring in the cerrado biome are Edentates (such as the tamandua! ant-eaters and armadillos), marsupials (such as opossums), platyrrhine monkeys (such as marmosets, howlers and capuchins) and various rodents (such as agoutis, pacas, capybaras and many mouse-sized species). Many larger-fruited plant species probably lost their natural mode of dispersal as a result of the extinction of their native mammalian vectors (see Janzen and Martin, 1982). The reintroduction of grazers in the form of cattle and horses into the natural cerrado vegetation in the last few hundred years probably partially restored the balance of the vegetation to the situation prior to the Great American Interchange. MODERN AGRICULTURAL DEVELOPMENT In the past the cerrado domain was sparsely populated by Brazilian country-people (backwoodsmen and Indians). Much of it was so remote that it only became incorporated into modern Brazilian life relatively recently, with the construction of railways and roads. The population practised little more than subsistence agriculture based largely on low-density cattle-grazing in the cerrado vegetation, raising small crops in clearings in the gallery or deciduous forests, charcoal-burning (if there was an accessible market), and some hunting and shing. The native vegetation provided materials for housing (timber, palm thatch, etc.), seasonal fruits, bre, rewood and many other products for the rural economy. All this has now changed, however, and during the last 25 or so years the cerrados have been extensively developed for agriculture with the active encouragement of the Brazilian government. Such development is an important part of the policy to develop the empty centre of Brazil and incorporate it into the national economy : the best known elements of the same policy were the building of the new capital city, Bras! lia, right in the core of the cerrado area, and the construction of a vast system of national highways. Encouragement for the agricultural development of the cerrados consisted of various forms of subsidy, extremely generous tax incentives, low interest loans with no indexing (practically a donation in an economy suering from

228

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

Amazonian rainforest : to date approx. 450 000 km# of rainforest, representing some 13 % of the original area, have been destroyedan area far smaller in both absolute and relative terms than that of the cerrado. So much emphasis has been put on the emotive issue of the destruction of the rainforests that the world has largely forgotten the fate of their oristic cousins, the savanna woodlands ! CHARCOAL PRODUCTION Charcoal production for Brazils steel industry exerts pressure on the cerrado secondary only to agriculture. The two are often closely interlinked, as the trunks and roots of trees from the areas cleared for cultivation are usually used for charcoal production, partially o-setting the costs of preparation of the land. The process is carried out in situ and one sees huge extensions of cleared land with rows of domed kilns emitting smoke and adding to the holocaustlike appearance of the landscape ! The Brazilian steel industry has always depended on charcoal and this has traditionally been derived from native forests and woodlands, particularly those of the cerrado. Today the steel mills in the state of Minas Gerais, lying in the cerrado region, are the largest charcoal users in the world (Alho and Martins, 1995). However, because of increasing diculties in supply from native sources, charcoal from eucalypt plantations is now also being used and its contribution is growing. Figures for charcoal consumption from 1980 to 1991 (Alho and Martins, op. cit.) show a peak in 1989 of about 45 000 m$, of which 71 % was derived from native vegetation. By 1991 the gure had dropped to 30 500 m$, (no doubt as a result of a downturn in the economy) of which 57 % came from native vegetation. Probably more recent gures will show a strong upturn since steel production is closely linked to the performance of the economy. C O N S E R V A T I O N T HE W A Y A H E A D The value of the species-rich cerrado biome has always been undervalued as an important centre of biodiversity. It stands alone among the major biomes of Brazil in not being recognized in the Brazilian Constitution as a National Heritage (Alho and Martins, op.cit.) : a status accorded to the Amazon, Atlantic Rainforest, Pantanal and coastal areas. The cerrado has always been regarded as the poor relative of the so-called rainforests, and this is shown very strongly in the allocation of international aid, e.g. the Pilot Programme for the Protection of Tropical forests in Brazil to be nanced by G-7. However, this attitude is now changing and the Brazilian Ministry of the Environments National Biodiversity Programme (PRONABIO) and bodies interested in conservation and rational use of the environment are now acting in concert. Specialists of Brazilian Government-nanced institutes such as IBAMA (the Brazilian Environmental Institute), EMBRAPA CPAC (Centre for Agricultural Research in the Cerrado) and CENARGEN (National Centre for Genetic Resources and Biotechnology) and other scientists are well aware that urgent action must be taken to safeguard the existence and

hyperination during much of the period), guaranteed prices, etc. As intended, such incentives have led to the establishment of a massive, highly mechanized, capital intensive system of agriculture. Cerrado vegetation, with its small trees, is much more easily cleared than tall forest and the soil is of a good structure for cultivation. However, before cultivation can take place a heavy application of lime and fertilizer is necessary to counteract soil acidity and to neutralize aluminium, which is present at toxic levels for virtually all cultivated crops. The cost of this application is about US $340 per hectare and represents nearly twice the purchase price of the virgin land. The arable crops planted are, in order of importance, soya, maize, rice and mandioca (manihot). Soya is grown largely for the export market and occupied 3n9 million hectares of the cerrado area in 1994, producing 8n8 million tonnes, while 4n9 million tonnes of maize were produced. However, far more of the cerrado is exploited as improved pasture, planted with such exotic grasses as Brachiaria, Hyparrhenia rufa and Panicum maximum, than as arable land. It is dicult to nd recent gures for the number of cattle raised on the cerrado, but in 1980 it was nearly 48 million and the present number is certainly much higher. The source of all the above gures is Alho and Martins (1995) except for cattle numbers from Wagner (1985). Clearly the intensive agribusiness practised in the cerrado is only possible with large nancial resources, so it is not surprising that a high percentage of the area consists of very large properties : 1985 gures showed that slightly more than 4 % of farms had more than 1000 ha but these covered 60 % of the total area, whereas more than 50 % had less than 50 ha but covered only 0n2 % of the area. The system of cultivation of the cerrado is far from environmentally friendly. The employment of intensive mechanization requires huge tracts of monoculture with great areas of bare soil and the concomitant problems of erosion by rain and wind, while the legally required reserve areas have to be kept in concentrated blocks so that the trees do not impede spraying aircraft. Clearly it would be better if such reserves were dispersed as a web to act as corridors for animals and a more widespread seed source for recolonization. Incidentally, the Brazilian laws for maintenance of reserve areas are extremely enlightened, requiring 50 % of land to be kept under natural vegetation in Legal Amazonia and 20 % in the rest of Brazil. If such laws were strictly observed, problems of conservation would be greatly reduced. Of the many other environmental problems, heavy, and often careless, use of pesticides and depletion of water reserves by giant rotating irrigators are among the most important. Obviously the cerrado biome has received a formidable agricultural onslaught and has been much altered during the last 25 or so years. It is estimated that by 1994 some 695 000 km#, representing 35 % of the total cerrado area, had been changed to arable land, planted pasture, or was in altered fallow. Projections for the year 2000 indicate that the gure will have risen to 817879 000 km#, representing 4144 % of the total area (gures modied from Alho and Martins, 1995). It is interesting to compare this with the well-publicized gures for destruction of the Brazilian

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

viability of the biome. In addition, non-governmental organizations such as FUNATURA, ISPN and WWF Brazil are active in the eld and the last has recently produced a valuable assessment of the situation (Alho and Martins, 1995) and is doing pioneering work in establishing extractive reserves. One matter of urgency is to increase the area of Federal Conservation Units. At present they represent only 1n5 % of the total area and Alho and Martins (op.cit.) suggest that this should be at least tripled. However, research is required to choose the location of such reserves as the cerrado biome is oristically very heterogeneous and constitutes a biological mosaic (Ratter and Dargie, 1992 ; Castro 1994 a ; Ratter et al., 1996). Teams from the University of Bras! lia, CPAC and the Royal Botanic Garden Edinburgh have been collaborating on this project for a number of years funded by Brazilian, European Community and British funds. The project has recently been expanded into a major AngloBrazilian initiative, Conser ation and Management of the Biodi ersity of the Cerrado Biome, with UK Overseas Development Administration funding. Its aim is to survey the oristic patterns of cerrado vegetation and to discover representative areas and biodiversity hot-spots . The method used covers the whole of the cerrado area with a grid of 1m latitudei1m30h longitude and plots how much information exists for each grid rectangle. Priority is given to areas where there are (a) little or no data, or (b) indications of exceptional diversity. Experienced eld teams then carry out rapid surveys with techniques designed to provide the maximum data in the minimum time. The rst two priority areas identied were the states of Mato Grosso do Sul and Tocantins. The survey of the former was completed last year and work in Tocantins will commence this year. A second team of the project led by Professor A. J. F. Castro (Federal University of Piau! ) is studying the little-known cerrados of north-eastern Brazil (Castro et al., 1998). In addition, a separate initiative, the long-standing Biogeograa dos Cerrados project led by Prof. Jeanini Felli-Fagg (University of Bras! lia) concentrates on the Chapada Pratinha (in the central cerrado core area) and is producing complementary results. In conclusion, a growing body of scientic and lay opinion is realizing the urgency of protecting the rich biodiversity of the cerrado. Although vast areas of the biome have been exploited, and in some southern states, e.g. Sa 4 o Paulo, only tiny vestiges remain, one hopes that with increased scientic eort and political will a satisfactory conservation situation can be achieved. A C K N O W L E D G E M E N TS We thank the sta of the University of Bras! lia Departments of Engenharia Florestal, Biologia Vegetal, and Ecologia and the Centro de Pesquisas Agropecua! rias do Cerrado (EMBRAPA-CPAC) for their collaboration, help and logistic support over many years of research. WWF Brazil and UK have also taken a great deal of interest in our work and have encouraged us in various ways. Our recent research has been nanced by the Baring Foundation, European Community nancial contribution

229

B92\43040\9304 and the ODA-funded Conser ation and Management of the Biodi ersity of the Cerrado Biome project. LITERATURE CITED
Alho CJR, Martins E de S. 1995. Bit by bit the Cerrado loses space. Bras! lia, DF, Brazil : WWF. Castro AAJF. 1994 a. Comparaa g o orT stico-geograT ca (Brazil) e tossocioloT gica (PiauT -Sa g o Paulo) de amostras de cerrado. PhD Thesis, Universidade Estadual de Campinas, SP, Brazil. Castro AAJF. 1994 b. Comparaa 4 o Flor! stica de Espe! cies do Cerrado. Sil icultura 14 : 1618. Castro AAJF, Martins FR, Fernandes AG. 1998. The woody ora of cerrado vegetation in the state of Piau! , northeastern Brazil. Edinburgh Journal of Botany (in press).. Cesar PCM. 1994. Vegetation communities and soils in the National Park of BrasT lia. PhD Thesis, University of Edinburgh, UK. Dias BF de S. 1992. Cerrados : Uma Caracterizaa 4 o. In : Dias BF de S, ed. Alternati as de Desen ol imento dos Cerrados : Manejo e Conser aa g o dos Recursos Naturais Reno a eis. Bras! lia, DF, Brazil : FUNATURA. Eiten G. 1972. The cerrado vegetation of Brazil. Botanical Re iews 38 : 201341. Eiten G. 1982. Brazilian Savannas . In : Huntley, BJ, Walker BH, eds. Ecology of tropical sa annas. Ecological Studies 42. BerlinHeidelberg, Germany : Springer Verlag : 2547. Furley PA, Ratter JA. 1988. Soil resources and plant communities of the Central Brazilian cerrado and their development. Journal of Biogeography 15 : 97108. Gibbs PE. 1990. Self-incompatibility in owering plants : a neotropical V nica 13 : 125136. perspective. Re ista brasileira de Bota Goodland RJ, Ferri MG. 1979. Ecologia do Cerrado. Sa 4 o Paulo, Brazil : Editora Universidade de Sa 4 o Paulo. Goodland RJ, Pollard R. 1973. The Brazilian cerrado vegetation : a fertility gradient. Journal of Ecology 61 : 219224. Heringer EP, Barroso GM, Rizzo JA, Rizzini CT. 1977. A Flora do Cerrado. In : Ferri MG, ed. IV SimpoT sio sobre o Cerrado. Sa 4o Paulo, Brazil : Editora Universidade de Sa 4 o Paulo : 211232. Janzen DH, Martin P. 1982. Neotropical anachronisms : what the gomphotheres ate. Science 215 : 1927. Lopes AS, Cox FR. 1977 a. A survey of the fertility status of surface soils under cerrado vegetation in Brazil. Soil Science Society of America Journal 41 : 741747. Lopes AS, Cox FR. 1977 b. Cerrado vegetation in Brazil : an edaphic gradient. Agronomical Journal 69 : 828831. Novaes Pinto M. Organizer. 1990. Cerrado : Caracterizaa g o, Ocupaa go e Perspecti a. Bras! lia, DF, Brazil : Editora Universidade de Bras! lia\SEMATEC. Oliveira-Filho AT, Ratter JA. 1995. A study of the origin of central Brazilian forests by the analysis of plant species distribution patterns. Edinburgh Journal of Botany 52 : 141194. Pereira BAS, Silva MA, Mendona RC. 1993. Reser a EcoloT gica do IBGE, BrasT lia (DF) : Lista das Plantas Vasculares. Rio de Janeiro, Brazil : Fundaa 4 o Instituto Brasileiro de Geograa e Estat! stica. Prado DE, Gibbs PE. 1993. Patterns of species distribution in the dry seasonal forests of South America. Annals of the Missouri Botanic Garden 80 : 902927. Prance GT. 1973. Phytogeographic support for the theory of Pleistocene forest refuges in the Amazon basin, based on evidence from distribution patterns in Caryocaraceae, Dichapetalaceae and V nica 3 : 529. Lecythidaceae. Acta Amazo Prance GT. 1982. A review of the phyteogeographic evidences for Pleistocene climate changes in the neotropics. Annals of the Missouri Botanic Garden 69 : 594624. Ratter JA. 1986. Notas sobre a Vegetaao da Fazenda AT gua Limpa (Brasilia-DF). Editora Universidade de Brasilia, textos universitarios no. 003. 136 pp. Ratter JA, Bridgewater S, Atkinson R, Ribeiro JF. 1996. Analysis of the oristic composition of the Brazilian cerrado vegetation II : comparison of the woody vegetation of 98 areas. Edinburgh Journal of Botany 53 : 153180.

230

Ratter et al.Brazilian Cerrado and Threats to its Biodi ersity

egetaa g o arboT rea e nos solos de um cerrado e um cerrada g o no Distrito Federal, Brasil. MSc Thesis, University of Bras! lia, Brazil. Rizzini CT. 1963. A Flora do Cerrado. Ana! lise Flor! stica dos Savanas Centrais. In : Ferri MG. SimpoT sio sobre o Cerrado. Sa 4 o Paulo, Brazil : Editora Universidade de Sa 4 o Paulo : 127177. Van der Hammen T. 1983. The palaeoecology and palaeography of ' re F, ed. Tropical sa annas. Amsterdam : savannas. In : Bourlie Elsevier : 1935. Wagner E. 1985. Desenvolvimento da regia 4 o dos Cerrados. In : Goedert WJ, ed. Solos dos Cerrados : tecnologias e estrateT gias de manejo. Sa 4 o Paulo, Bras! lia : EMBRAPA (CPAC)\Nobel : 1931.

Ratter JA, Dargie TCD. 1992. An analysis of the oristic composition of 26 cerrado areas in Brazil. Edinburgh Journal of Botany 49 : 235250. Ratter JA, Pott A, Pott VJ, Cunha CN, Haridasan M. 1988. Observations on woody vegetation types in the Pantanal and at Corumba! , Brazil. Notes of the Royal Botanic Garden Edinburgh 45 : 503525. Ratter JA, Ribeiro JF. 1996. Biodiversity of the ora of the cerrado. In : Pereira RC, Nasser LCB, eds. Anais\Proceedings of the VIII SimpoT sio sobre o cerrado\1st Internat. Symposium on Tropical Sa annas. Planaltina, DF, Brazil : EMBRAPA\CPAC : 36. Ribeiro JF. 1983. Comparaa g o da concentraa g o de nutrientes na