Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Priority areas for the conservation of South African vegetation: a coarse‐filter approach

Priority areas for the conservation of South African vegetation: a coarse‐filter approach Blackwell Science, Ltd B. REYERS*1, D. H. K. FAIRBANKS 1,2, A. S. VAN JAARSVELD 1,3 and M. THOMPSON 4 1Conservation Planning Unit, Department of Zoology and Entomology, University of Pretoria, Pretoria, 0002, South Africa, *Correspondence. E-mail: breyers@zoology.up.ac.za, 2Present address: Fitzpatrick Institute, University of Cape Town, Rondebosch, 7701, South Africa, 3Centre for Environmental Studies, University of Pretoria, Pretoria, 0002, South Africa, 4CSIR Division of Water, Environment and Forestry Technology, PO Box 395, Pretoria, 0001, South Africa Abstract. South Africa has an important responsibility to global biodiversity conservation, but a largely inadequate conservation area network for addressing this responsibility. This study employs a coarse-filter approach based on 68 potential vegetation units to identify areas that are largely transformed, degraded or impacted upon by roadeffects. The assessment highlights broad vegetation types that face high biodiversity losses currently or in the near future due to human impacts. Most vegetation types contain large tracts of natural vegetation, with little degradation, transformation or impacts from road networks. Regions in the grasslands, fynbos and forest biomes are worst affected. Very few of the vegetation types are adequately protected according to the IUCN’s 10% protected area conservation target, with the fynbos and savanna biomes containing a few vegetation types that do achieve this arbitrary goal. This investigation identifies areas where limited conservation resources should be concentrated by identifying vegetation types with high levels of anthropogenic land use threats and associated current and potential biodiversity loss. Key words. Coarse-filter, biodiversity conservation, land-cover, vegetation types, road-effects. INTRODUCTION South Africa contains a wealth of biodiversity within its borders, unequalled by other temperate regions, earning a place in the top 25 most biodiverse nations ( WCMC, 1992; Conservation International, 1998). In addition South Africa harbours the fifth highest number of plant species in the world, with the Cape Floristic Region being recognized as one of the six floral kingdoms of the world. This region contains 8200 plant species, of which 5682, are endemic and has lost approximately 30.3% of its primary vegetation (Fairbanks et al., 2000; Myers et al., 2000). Although its responsibility towards global biodiversity conservation is large South Africa, with only 4.8% ( DEAT, 1996) ( Fig. 1a) of its land surface under formal protection, falls far short of the IUCN’s nominal recommendation of 10% protected area coverage. This coverage also lags behind the 10% average attained by the rest of subSaharan Africa, with Botswana reaching 18.5%, Mozambique 12.7% and Namibia 12.4% ( McNeely, 1994; WRI, 1994; Siegfried et al., 1998). A moderately expanding human population (Central Statistical Survey, 1998) and associated land transformation in South Africa (mainly urbanization, cultivation and afforestation (Hoffmann, 1997) ) leaves 79% of the country covered with natural woody and grassland vegetation communities ( Fig. 1b) (Fairbanks et al., 2000). Water bodies and wetlands cover less than 1% of the land surface area, with human land uses making up the remaining 20% ( Fairbanks et al., © 2001 Blackwell Science Ltd. http://www.blackwell-science.com /ddi B. Reyers et al. Fig. 1 Maps of: (a) South African national and provincial protected areas (DEAT, 1996); (b) transformed, degraded and natural land-cover; (c) biomes (Low & Rebelo, 1996); and (d) road network buffered according to Stoms (2000). © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Fig. 1 continued. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. aggregate, the so-called ‘coarse-filter’ approach ( Noss, 1987, 1990). This approach focuses on protecting higher levels of the biodiversity hierarchy (e.g. land-classes and land-types) rather than species, assuming that these broad-scale biodiversity surrogates represent the finer scale aspects of biodiversity ( Pressey & Logan, 1994; Pressey, 1994; Williams & Humphries, 1996; Wessels et al., 1999; Fairbanks & Benn, 2000). However, as Pressey (1994) points out, the assumed relationship between environmental classes and species distribution and abundance is unclear and seldom investigated. In addition, certain species, especially rare species confined to small patches of habitat which are not recognized as distinct environmental classes, may ‘fall through the coarse filter’ when using broad-scale environmental classes ( Noss, 1983; Bedward et al., 1992; Panzer & Schwartz, 1998). Despite the shortcomings associated with a species-based approach to conservation planning, these higher order biodiversity surrogates may well fail to identify the composition, configuration and quantity of elements necessary for biodiversity retention, making species data a necessary component of the conservation planning process (Lambeck, 1997). The shortcomings of species distribution data and the limitations of environmental surrogate measures in the selection of priority conservation areas suggest that perhaps a combination of the two approaches in conservation planning may be advisable ( Maddock & du Plessis, 1999). At a national scale South Africa has a few databases of broader surrogates for biodiversity, including Acocks’ Veld Types (Acocks, 1988) and the more recent Vegetation of South Africa, Lesotho and Swaziland ( Low & Rebelo, 1996; McDonald, 1997). Acocks (1988) defined biological resources from a purely agricultural potential perspective, while Low & Rebelo (1996) looked at the definition of these resources from a management and potential use angle. These vegetation units were defined as having: ‘… similar vegetation structure, sharing important plant species, and having similar ecological processes’. Thus, these are units that would have potentially occurred today were it not for all the major human-made transformations, e.g. agriculture and urbanization. Therefore the Low & Rebelo (1996) vegetation map contains significant potential for acting as a broad scale surrogate of South African biodiversity 2000). Fairbanks et al. (2000) demonstrate that along with the approximately 30% transformation in the fynbos biome, the savanna and grassland biomes are about 10% and 26% transformed and degraded by human land uses, respectively ( Fig. 1c) (see also Thompson et al., 2001). In addition to this there are a total of 1176 species presently recognized as threatened ( WRI, 1994; van Jaarsveld, 2000). Thus with these valuable and often endemic biodiversity resources facing ever-increasing threats from human-induced land transformation, and mostly inadequate conservation efforts to stem these threats, South Africa has an obvious responsibility to do more towards the conservation of biodiversity (van Jaarsveld, 2000). Most of South Africa’s existing protected areas were proclaimed in an ad hoc fashion, usually because they contained areas with high scenic or tourism potential, contained endemic diseases and did not conflict with other forms of land use ( Pringle, 1982; Pressey et al., 1993; Freitag et al., 1996). Because this form of land allocation to conservation is highly inefficient and fails to effectively conserve biodiversity, several techniques have been developed for the systematic selection of land with a high conservation value, i.e. with high levels of biodiversity and large anthropogenic threats facing that biodiversity (for reviews see Williams, 1998; Margules & Pressey, 2000). However, these techniques require data on the distribution of biodiversity and threats facing biodiversity in order to identify areas important to conservation. Because the biodiversity of a region can never be fully observed and inventoried, species distribution data are often used as a surrogate or substitute measure of biodiversity. This form of data, however, has a large number of shortcomings associated with it. These include inadequate taxonomical knowledge of the groups employed, biased sampling efforts and lack of spatial congruency between areas of conservation importance to different taxa (van Jaarsveld et al., 1998; Maddock & du Plessis, 1999; Fairbanks & Benn, 2000; Reyers et al., 2000). Broad-scale biodiversity surrogates In recent years, the focus for conservation has shifted, with recommendations towards a more holistic approach of protecting biodiversity in the © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Table 1 Land-cover classes reclassified into broad categories Land-cover categories Natural land-cover Degraded land-cover Transformed land-cover % area 73.4% 10.1% 16.5% Land-cover class Wetlands, grassland, shrubland, bushland, thicket, woodland, forest Degraded land, erosion scars, water bodies Cultivated lands, urban/built-up areas, mines and quarries, forestry plantations and for identifying land important to biodiversity conservation. METHODS Current land-cover data Before the Low & Rebelo (1996) map can be used one has to differentiate between the potential vegetation cover of regions (as defined by Low & Rebelo, 1996) and that which is in reality found in the region. In other words, one needs an indication of current natural vegetation pattern, degree of transformation and amount of protection afforded each vegetation type before one can decide if it constitutes a conservation priority ( Rebelo, 1997). As Low & Rebelo (1996) point out: ‘there is little point in setting aside more of a vegetation type with vast expanses in pristine condition, while ignoring the last patches of a type which is not yet conserved’. Low & Rebelo (1996) provide some estimates of protection and transformation data, however, as they admit: ‘these are woefully incomplete’. Thus, some indication of current land-cover (the suite of natural and human-made features that cover the earth’s immediate surface) at a national scale is required for effective land-use planning, sustainable resource management, environmental research and in this instance conservation planning ( Rebelo, 1997; Fairbanks et al., 2000). To this end the advent of the National Landcover ( NLC) database is of extreme relevance. This national database was derived using manual photo-interpretation techniques from a series of 1 : 250 000 scale geo-rectified hardcopy satellite imagery maps, based on seasonally standardized, single date Landsat Thematic Mapper ( TM) satellite imagery captured principally during the period 1994– 95 ( Fairbanks & Thompson, 1996). It provides the first single standardized database of current land-cover information for the whole of South Africa, Lesotho and Swaziland (Fairbanks et al., 2000). For the purpose of the present study the 31 land-cover classes were reclassified into three categories: natural, degraded and transformed land-cover ( Table 1). Natural land-cover included all untransformed vegetation, e.g. forest, woodland, thicket and grassland. The degraded land-cover category was dominated by degraded classes of land-cover. These areas have a very low vegetation cover in comparison with the surrounding natural vegetation cover and were typically associated with rural population centres and subsistence level farming, where fuel-wood removal, over-grazing and subsequent soil erosion were excessive ( Thompson, 1996). The transformed category consisted of areas where the structure and species composition were completely or almost completely altered, which includes all areas under crop cultivation, forestry plantations, urbanized areas and mines/quarries. The databases of potential vegetation cover and current land-cover were overlaid in a geographical information system (GIS) to determine the extent of natural, degraded and transformed area within each of the 68 vegetation types identified in Low & Rebelo (1996). These values could then be used to highlight areas of high current and future vulnerability to biodiversity loss through land use impacts. Levels of transformation were compared against the transformation thresholds predicted by a geometric model developed by Franklin & Forman (1987). This work suggested that the most critical time for land planning and conservation is when between 10 and 40% of the landscape has been transformed or impacted upon. Specifically, most of the rapid ecological changes (e.g. loss of interior species) can be expected when this level increases from 20 © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Table 2 Buffer widths assigned to road classes for calculating road effect zone (after Stoms, 2000) South African Surveyor General Description National route Freeway Arterial Main Secondary (connecting and magisterial roads) Other (rural road) Vehicular trail (4-wheel drive route) to 40%. Regions showing greater than 40% loss of natural habitat have already undergone significant ecological disruptions. Furthermore, the NLC database could be used to identify the major broad categories of current threat (e.g. cultivation, forestry) facing the vegetation types. An additional GIS layer of protected area coverage for the country ( DEAT, 1996) was also employed to determine the extent of conservation areas existing within the vegetation types. Buffer width (m) 1000 1000 500 250 100 50 25 Patterns of roads In addition to these land use threats, one of the most widespread forms of alteration of natural habitats and landscapes over the last century has been the construction and maintenance of roads ( Trombulak & Frissell, 2000). Road networks affect landscapes and biodiversity in seven general ways: (1) increased mortality from road construction; (2) increased mortality from vehicle collisions; (3) animal behaviour modification; (4) alteration of the physical environment; (5) alteration of the chemical environment; (6) spread of exotic species and (7) increased alteration and use of habitats by humans (from Trombulak & Frissell, 2000). These networks cover 0.9% of Britain and 1.0% of the United States (Forman & Alexandra, 1998); however, the road-effect zone, the area over which significant ecological effects extend outward from the road, is usually much wider than the road and roadside. This road effect zone can thus provide an additional estimate of areas with a high vulnerability to biodiversity loss through changing land uses and increased human impacts. Some evidence on the size of the road-effect zone is available from studies in Europe and North America. Reijnen et al. (1995) estimated that road-effect zones cover between 12 and 20% of the Netherlands, while Forman (2000) illustrated that 19% of the United States is affected ecologically by roads and associated traffic. The road-effect zone for South Africa was determined using a similar method to that used by Stoms (2000), in which the spatial extent of road effects can be used as an ecological indicator that directly represents impacts on biodiversity. For this, the road-effect zone was used as a measure of the area potentially affected by roads. The affected distances were estimated from the reviews mentioned above, as well as from local studies ( Milton & MacDonald, 1988). Therefore national routes and freeways were assumed to affect biodiversity for a greater distance from the roadway (1 km on each side) than farm roads (100 m, Table 2). Road segments from the South African Surveyor General 1993 1 : 500 000 scale map series files (SA Surveyor General, 1993) were buffered using a standard GIS operation to the distance related to its class (Fig. 1d). Although the roads in protected areas do have an impact on biodiversity within these areas, they were excluded from this analysis as by and large protected areas overwhelmingly contribute to biodiversity conservation. A road-effect zone was calculated for the remaining untransformed areas within each vegetation type by summing the total area within the road effect zone surrounding roads in each vegetation type and converting to a percentage of the total remaining untransformed area in that vegetation type. However, the road-effect zone used here does not consider the spatial pattern of roads. Therefore, although roads clearly have a significant impact on many species, meaningful indicators of road-effects on landscapes await the attention of landscape ecologists and other scientists (Forman, 1998). As articulated by Stoms (2000), many aspects of roads affect biodiversity: road width, traffic volume, traffic speed, vehicle miles travelled, road network structure or its spatial configuration, management of the right-of-way, noise levels, light disturbance and chemical pollution. Most of these factors also vary over daily, weekly and annual cycles, which may interfere with critical behavioural periods such as breeding or © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation migration. As such, the road-effect zone can represent only a first order approximation attempt to capture more of the multidimensional nature of road network effects. RESULTS AND DISCUSSION Vulnerability assessment of vegetation types The majority of vegetation types of South Africa are not largely degraded or transformed ( Table 3). Of the 68 vegetation types, 61 contain more than 50% natural vegetation cover with a median value of 81.1% natural vegetation cover across all vegetation types. The vegetation types show low levels of degradation with a median value of 2.8%, with all but one (Afro Mountain Grassland ) being less than 20% degraded ( Table 3). Only five of the vegetation types are more than 50% transformed by anthropogenic land uses, with a median of 10% being transformed within vegetation types. Figure 2 provides a diagrammatic representation of the current levels of transformation, degradation Table 3 Percentage of natural, degraded, transformed and protected area of each of the vegetation types, as well as the percentage of each vegetation type exposed to road-effect zones Code Vegetation type 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 % natural % degraded % transformed 1.2 2.9 15.6 8.5 13.0 2.0 7.0 4.2 0.0 0.9 10.2 0.8 0.1 7.1 12.6 13.8 12.0 14.1 9.9 1.4 9.6 12.3 15.9 8.2 9.9 11.9 0.0 3.1 1.2 18.9 0.2 9.3 29.2 5.8 27.6 14.8 3.0 14.5 2.6 0.0 6.6 6.0 % protected % road-effect Coastal forest 89.3 Afromontane forest 67.9 Sand forest 72.3 Dune thicket 62.2 Valley thicket 72.1 Xeric succulent thicket 95.0 Mesic succulent thicket 78.5 Spekboom succulent thicket 93.1 Mopane hrubveld 100.0 Mopane bushveld 92.4 Soutpansberg arid 83.8 mountain bushveld Waterberg moist 90.2 mountain bushveld Lebombo arid 90.2 mountain bushveld Clay thorn bushveld 58.7 Subarid thorn bushveld 78.7 Eastern thorn bushveld 69.7 Sweet bushveld 78.3 Mixed bushveld 69.3 Mixed lowveld bushveld 70.4 Sweet lowveld bushveld 85.1 Sour lowveld bushveld 54.4 Subhumid lowveld bushveld 84.1 Coastal bushveld–grassland 43.5 Coast–hinterland bushveld 56.7 Natal central bushveld 72.2 Natal lowveld bushveld 72.5 Thorny Kalahari 83.5 dune bushveld Shrubby Kalahari 96.0 dune bushveld Karroid Kalahari bushveld 98.8 Kalahari plains 73.6 thorn bushveld Kalahari mountain bushveld 99.5 (43) 1.3 (9.5) 6.5 (44) 16.1 (17.6) 6.4 (45) 46.7 (44.6) 1.7 (25) 10.6 (14.5) 11.2 (51) 1.5 (2.1) 6.1 (51) 4.6 (8.0) 6.4 (51) 4.0 (5.3) 14.2 (unknown) 1.2 (1.8) 4.9 (0) 100.0 (100.0) 0.0 (8) 34.0 (38.3) 3.0 (65) 10.1 (12.6) 4.3 6.2 (8.6) 37.1 (38.0) 1.0 0.0 0.2 1.8 3.6 22.5 62.2 7.0 20.9 13.5 2.1 1.3 14.1 99.6 (0.9) (0.2) (0.5) (2.3) (3.1) (28.3) (67.3) (9.7) (21.5) (14.0) (3.6) (1.6) (17.8) (99.8) 3.2 1.0 5.1 8.2 11.1 4.5 5.3 3.1 1.1 4.7 1.1 5.9 4.4 7.2 5.3 0.0 2.2 3.3 3.9 4.6 9.0 (28) 9.1 (unknown) 34.1 8.7 16.5 9.5 16.6 19.8 13.5 36.0 3.6 39.8 35.0 18.0 15.6 0.0 (60) (unknown) (unknown) (27) (60) (30) (30) (76) (36) (unknown) (87) (80) (35) (unknown) 0.0 (55) 0.0 (55) 7.1 (55) 0.3 (25) 19.4 (19.5) 0.1 (0.1) 0.5 (0.5) 0.0 (0.0) © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Table 3 continued. Code Vegetation type 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 Kimberley thorn bushveld Kalahari plateau bushveld Rocky highveld grassland Moist clay highveld grassland Dry clay highveld grassland Dry sandy highveld grassland Moist sandy highveld grassland Moist cool highveld grassland Moist cold highveld grassland Wet cold highveld grassland Moist upland grassland North-eastern mountain grassland South-eastern mountain grassland Afro mountain grassland Alti mountain grassland Short mistbelt grassland Coastal grassland Bushmanland Nama Karoo Upper Nama Karoo Orange River Nama Karoo Eastern mixed Nama Karoo Great Nama Karoo Central lower Nama Karoo Strandveld succulent Karoo Upland succulent Karoo Lowland succulent Karoo Little succulent Karoo North-western mountain renosterveld Escarpment mountain renosterveld Central mountain renosterveld West coast renosterveld South and south-west coast renosterveld Mountain fynbos Grassy fynbos Laterite fynbos Limestone fynbos Sand plain fynbos % natural % degraded % transformed 76.1 92.7 66.3 68.2 34.9 63.5 67.6 60.4 46.8 88.0 61.4 67.6 94.5 51.9 87.5 38.5 81.7 99.7 99.0 98.1 94.9 99.1 90.2 86.3 97.1 94.2 89.0 94.0 98.9 80.4 9.0 39.4 88.5 88.7 64.8 87.2 34.4 4.4 3.0 0.1 0.4 0.1 0.8 0.7 1.6 11.3 2.4 17.0 7.1 4.0 36.7 8.8 4.6 5.1 0.2 0.9 0.1 1.8 0.8 9.0 2.0 0.7 2.6 2.6 0.0 0.3 1.8 1.1 1.9 0.7 0.8 1.1 7.6 8.5 19.5 4.2 33.6 31.4 (55) (55) (65) (79) % protected 1.8 0.0 0.8 0.0 (3.1) (0.0) (1.4) (0.0) % road-effect 6.8 5.5 10.2 11.3 9.0 9.1 9.4 9.6 6.7 4.1 5.5 4.8 5.7 0.8 1.2 7.6 7.0 3.4 5.8 4.6 7.4 5.4 6.0 4.0 4.4 3.9 7.7 3.0 2.4 5.4 8.1 8.8 2.9 6.0 8.6 4.0 7.1 65.1 (67) 35.8 (65) 31.6 (55) 38.0 (72) 41.8 (70) 9.7 (60) 21.6 (60) 25.3 (45) 1.5 (32) 11.4 3.6 56.9 12.9 0.1 0.1 1.6 3.3 0.2 0.8 9.5 1.7 3.2 8.4 6.0 (32) (32) (89) (unknown) (unknown) (unknown) (unknown) (unknown) (unknown) (unknown) (24) (unknown) (unknown) (unknown) (unknown) 0.0 (0.0) 0.3 (0.3) 0.0 (0.7) 0.7 (0.3) 0.8 (0.6) 9.4 (6.7) 2.3 (2.5) 3.3 (7.4) 0.6 (0.3) 0.0 11.7 0.9 0.1 0.0 0.0 0.1 1.6 0.7 0.1 0.4 4.2 0.9 3.2 0.0 (0.0) (12.5) (2.4) (1.1) (0.0) (0.0) (1.5) (1.1) (0.2) (0.0) (0.4) (4.4) (1.3) (2.3) (0.0) 0.8 (unknown) 17.8 (11) 89.8 (97) 58.7 (32) 10.8 (11) 10.3 (3) 34.1 (50) 5.2 (40) 57.1 (50) 0.0 (0.1) 5.1 (3.6) 0.7 (1.8) 1.5 (1.4) 26.4 15.5 0.0 13.6 1.2 (26.1) (16.1) (0.5) (13.8) (1.1) Values in brackets indicate estimates from Low & Rebelo (1996). Vegetation types with more than 10% protected area coverage are indicated in bold type. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation and protection across all vegetation types. Similar to the findings of the coarse-scale species-based approach used by Rebelo (1997), the grasslands and fynbos have experienced the most transformation (see Fairbanks et al., 2000), with the coastal indigenous forests having been subjected to extensive transformation for its size (Fig. 2a,b). Although degradation levels are generally low, a few regions in the grasslands biome as well as a few in the savanna biome show the highest levels of degradation, ranging from 10 to 36% of the vegetation extent ( Fig. 2c). The average amount of vegetation type currently under protection is 9.6% (median value of 1.5%) with only 18 vegetation types conforming to the IUCN’s nominal recommendation of 10% protected area coverage ( Table 3). However, this well-cited protected area recommendation of 10% is widely criticized as too little to guarantee the persistence of biodiversity within the region. Soulé & Sanjayan (1998) illustrate that up to 50% of land area may be required to successfully represent all biodiversity elements. Therefore, perhaps even these 18 supposedly well-protected vegetation types are inadequately protected ( Fig. 2d). The road-effect zone impacts on an average of 5.5% (with a median value of 6) of the remaining natural land-cover in all vegetation types ( Table 3). Five vegetation types (mesic succulent thicket, moist clay highveld grassland, dune thicket, eastern thorn bushveld, rocky highveld grassland) contain between 10 and 14.2% road-effect zones ( Table 3). The rest of the vegetation types lie under this 10% level, with the Mopane Shrubveld containing no road-effect due to the fact that it all falls entirely within the boundaries of the Kruger National Park ( Table 3). In Table 4 the areas within each vegetation type that are transformed, degraded or exposed to road-effects are summed to provide an indication of vegetation that has been disturbed or affected by these human land uses. Types with large areas affected face a high risk of biodiversity loss due to a combination of extensively degraded and transformed areas with a large road network. The west coast renosterveld, sand plain fynbos, dry clay highveld grassland, south and south-west renosterveld, short mistbelt grassland, coastal bushveld-grassland, moist cold highveld grassland, sour lowveld bushveld, afro mountain grassland, coast–hinterland bushveld, moist cool highveld grassland, clay thorn bushveld, dune thicket, moist upland grassland, dry sandy highveld grassland, rocky highveld grassland and laterite fynbos are all areas of concern due to the fact that over 40% of their extent is impacted upon by land use threats. This level of land use impact corresponds with the threshold determined by Franklin & Forman (1987), indicating extreme ecological disruption within these vegetation types. All these vegetation types are also poorly protected ( Table 3), with the coastal bushveld– grassland and dune thicket being the only types to reach the IUCN’s recommended 10% protected area coverage. However, as stated previously this level of protection is inadequate, especially in the case of these two vegetation types where it would not be sufficient to stem the biodiversity loss associated with such high levels of land use change. Of the 68 vegetation types 38 (56%) fall within the 10 – 40% category of land use impact determined by Franklin & Forman (1987) and are thus at a critical time for land use planning and conservation. Table 5 provides a list of the land-cover types within each of the top 10 priority conservation vegetation types drawn from Table 4. The Afro mountain grassland and moist cold highveld grassland contain large areas of degraded vegetation. These same vegetation types, along with the west coast renosterveld, sand plain fynbos, dry clay highveld grassland, south and southwest coast renosterveld, short mistbelt grassland, coastal bushveld–grassland, sour lowveld bushveld and coast–hinterland bushveld, contain extensive areas of commercial, semicommercial and subsistence dryland cultivation ( Table 5). The short mistbelt grassland, coastal bushveld–grassland, sour lowveld bushveld and coast–hinterland bushveld contain large areas of exotic forestry plantations and, with the exception of the sour lowveld bushveld, commercial sugarcane cultivation ( Table 5). Of all these priority vegetation types only the coastal bushveld–grassland has more than 10% protected area coverage at 13.5%, but high levels of degradation as well as high levels of transformation still make it an area of concern along its entire latitudinal distribution. The rest of these top 10 priority vegetation types all fall below 5% protected area coverage ( Table 3). This land use analysis is an example of a potential © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Fig. 2 Diagrammatic representation of levels of percentage (a) transformed (b) degraded (c) natural and (d) protected vegetation cover within each of Low & Rebelo’s (1996) vegetation types. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Fig. 2 continued. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Table 4 Percentage area of vegetation type exposed to the combined land-cover threats of degradation, transformation and road effects Code 62 68 36 63 47 23 40 21 45 24 39 14 4 42 37 34 66 35 38 16 2 43 18 7 25 5 19 26 32 30 15 61 17 48 3 11 58 65 67 22 1 41 20 54 55 64 46 12 33 52 Vegetation type West coast renosterveld Sand plain fynbos Dry clay highveld grassland South and south-west coast renosterveld Short mistbelt grassland Coastal bushveld–grassland Moist cold highveld grassland Sour lowveld bushveld Afro mountain grassland Coast–hinterland bushveld Moist cool highveld grassland Clay thorn bushveld Dune thicket Moist upland grassland Dry sandy highveld grassland Rocky highveld grassland Laterite fynbos Moist clay highveld grassland Moist sandy highveld grassland Eastern thorn bushveld Afromontane forest North-eastern mountain grassland Mixed bushveld Mesic succulent thicket Natal central bushveld Valley thicket Mixed lowveld bushveld Natal lowveld bushveld Kimberley thorn bushveld Kalahari plains thorn bushveld Subarid thorn bushveld Central mountain renosterveld Sweet bushveld Coastal grassland Sand forest Soutpansberg arid mountain bushveld Little succulent Karoo Grassy fynbos Limestone fynbos Subhumid lowveld bushveld Coastal forest Wet cold highveld grassland Sweet lowveld bushveld Central lower Nama Karoo Strandveld succulent Karoo Mountain fynbos Alti mountain grassland Waterberg moist mountain bushveld Kalahari plateau bushveld Eastern mixed Nama Karoo Affected area (%) 92.3 69.5 67.8 65.4 64.8 60.3 56.7 49.1 48.6 47.0 45.8 45.1 43.9 42.5 42.3 42.2 40.8 39.6 39.3 38.2 37.9 36.2 34.8 34.0 33.3 32.9 32.0 31.6 29.4 29.0 28.0 25.9 25.2 23.8 22.8 20.0 18.7 17.9 17.2 16.9 16.8 16.2 16.0 15.2 15.1 14.8 13.5 12.9 12.4 12.3 © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Table 4 continued. Code 8 6 44 13 10 57 59 56 50 53 51 28 31 29 49 60 27 9 Vegetation type Spekboom succulent thicket Xeric succulent thicket South-eastern mountain grassland Lebombo arid mountain bushveld Mopane bushveld Lowland succulent Karoo North-western mountain renosterveld Upland succulent Karoo Upper Nama Karoo Great Nama Karoo Orange River Nama Karoo Shrubby Kalahari dune bushveld Kalahari mountain bushveld Karroid Kalahari bushveld Bushmanland Nama Karoo Escarpment mountain renosterveld Thorny Kalahari dune bushveld Mopane shrubveld Affected area (%) 11.8 11.3 11.1 10.3 10.3 9.5 9.1 6.8 6.7 6.3 6.3 5.2 5.1 4.5 3.6 3.5 0.0 0.0 management tool for vulnerable areas, and is not limited to these top 10 vegetation types. Other vegetation types, although not as affected as these 10, are none the less also impacted upon by land use changes and should therefore also be considered and monitored in a conservation plan. Table 5 is an example of what can be done and similar analyses can be performed on all vegetation types in order to investigate the land use impacts and management parameters within each area. The vegetation types listed at the bottom of Table 4 are less impacted upon by land uses, and are generally better protected ( Table 3), with the Mopane shrubveld and thorny Kalahari dune bushveld including 100 and 99.6% protected area, respectively. These areas also contain extensive tracts of natural vegetation ranging from 83.5% for the thorny Kalahari dune bushveld to 100% for the Mopane shrubveld (Table 3). This, however, does not preclude them from further analysis and the tools developed in this study have a potential role to play in the monitoring and future management of these currently less impacted areas. Comparison of vulnerability status Low & Rebelo (1996) also provided an estimate of threat status of the vegetation types. This included a measure of land transformed by agriculture and other uses, based on ‘scant information for some of the Acocks Veld Types and should be cautiously interpreted as a rough index of habitat loss’ ( Low & Rebelo, 1996). They also include an estimate of the proportion of each vegetation type falling within conserved areas, based on an approximation of conservation area boundaries which still require confirmation (Low & Rebelo, 1996). Following a similar methodology to Thompson et al. (2001), we evaluate these estimates from Low & Rebelo (1996) as well as the calculations of protected and transformed land obtained from this study using the National Land-cover database and the Department Environmental Affairs and Tourism (DEAT, 1996) protected area database ( Table 3). Top conservation priority vegetation types identified based on Low & Rebelo’s (1996) estimates of transformed area in Table 3 highlight the west coast renosterveld, short mistbelt grassland, coast– hinterland bushveld, Natal central bushveld and the moist clay highveld grassland as areas of conservation concern due to large areas transformed. The Mopane shrubveld, grassy fynbos, Mopane bushveld, central mountain renosterveld and mountain fynbos are estimated to be areas of low priority for conservation as they are little transformed according to Low & Rebelo’s © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Table 5 Description and percentage area coverage of land-cover threats facing conservation priority vegetation types South and Sand Dry clay south-west Short Coastal Moist Sour Afro West coast plain highveld coast mistbelt bushveld– cold highveld lowveld mountain renosterveld fynbos grassland renosterveld grassland grassland grassland bushveld grassland 34.64 0.14 0.05 0.00 0.63 0.00 7.66 5.20 0.00 0.00 2.78 39.53 0.00 4.88 7.11 0.00 0.00 1.03 0.00 0.20 0.56 0.00 34.89 0.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 64.65 0.00 0.00 0.36 0.00 0.00 0.00 0.00 0.00 0.02 0.01 39.87 0.83 0.00 0.00 0.00 0.00 1.05 1.77 0.00 0.00 2.17 53.07 0.00 0.31 0.78 0.00 0.00 0.02 0.00 0.00 0.07 0.00 39.32 0.24 0.00 0.00 0.61 3.73 0.00 0.03 0.39 10.79 1.67 4.74 7.02 30.86 0.83 0.00 0.14 0.00 0.00 0.02 0.00 0.00 43.56 4.69 0.00 0.87 7.50 2.82 0.00 0.00 0.01 15.39 0.02 0.00 10.18 9.31 3.10 0.00 0.90 0.00 0.00 0.13 0.33 0.06 46.85 0.21 0.09 0.00 0.02 11.02 0.00 0.00 1.78 0.00 0.05 19.58 21.27 0.06 0.79 0.00 0.00 0.00 0.04 0.00 0.01 0.00 54.44 0.11 0.00 5.88 3.12 0.49 0.00 1.55 0.00 0.34 2.55 1.30 11.80 15.29 1.30 0.00 0.00 0.00 0.00 0.02 0.01 0.03 51.92 0.01 0.00 0.00 36.65 0.00 0.00 0.00 0.00 0.00 0.00 0.00 11.40 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.01 92 B. Reyers et al. Coast– hinterland bushveld 56.87 0.12 0.00 0.42 4.77 2.93 0.00 0.01 0.32 8.91 0.23 0.49 13.75 9.11 1.98 0.00 0.04 0.00 0.15 0.06 0.15 0.02 Description Natural land-cover 9.01 Water bodies 0.24 Dongas and sheet erosion scars 0.00 Degraded: forest and woodland 0.00 Degraded: thicket and bushland (etc.) 0.11 Degraded: unimproved grassland 0.00 Degraded: shrubland and low fynbos 0.76 Cultivated: permanent–commercial 11.70 irrigated Cultivated: permanent–commercial 0.05 dryland Cultivated: permanent–commercial 0.00 sugarcane Cultivated: temporary–commercial 0.15 irrigated Cultivated: temporary–commercial 74.78 dryland Cultivated: temporary– 0.00 semicommercial/subsistence dryland Forest plantations 0.60 Urban/built-up land: residential 1.59 Urban/built-up land: residential 0.00 (small holdings: woodland) Urban/built-up land: residential 0.00 (small holdings: bushland) Urban/built-up land: residential 0.45 (small holdings: shrubland) Urban/built-up land: residential 0.00 (small holdings: grassland) Urban/built-up land: commercial 0.06 Urban/built-up land: 0.03 industrial/transport Mines and quarries 0.07 © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Bold values indicate main land uses in the vegetation type. Conservation of South African vegetation (1996) estimates ( Table 3). Once again the areas of high threat are estimated by Low & Rebelo (1996) to be poorly protected with less than 4% of their surface area protected and those that are low priorities are seen to be generally well protected. As found in Thompson et al. (2001), there is some degree of similarity in the rank orders of vegetation types according to threat status found in this study (i.e. affected area) and in Low & Rebelo’s (1996) (i.e. areas estimated to be transformed) (rs = 0.55; P < 0.001). However, as Table 3 illustrates, there are differences between these estimates of transformation and protection from Low & Rebelo (1996) and values generated in this study. The Low & Rebelo (1996) estimates for land transformation and protection being consistently and significantly higher (paired t-test for levels of transformation, t = 9.00, degrees of freedom = 49, P < 0.0001; paired t-test for levels of protection, t = 3.8, degrees of freedom = 67, P < 0.01). This could however, be explained by the fact that the estimates of transformation in Low & Rebelo (1996) included grazed areas, while the NLC transformation category does not (Thompson et al., 2001). The grazed areas (especially overgrazed area) are included in the degraded category of the NLC database and as such are included in the present study in the measure of affected areas ( Table 4). CONCLUSION South Africa, with its large biodiversity conservation responsibility, faces the additional problems of limited resources for conservation as well as pressing land reform initiatives. The land tenure system is a problem for conservation throughout Africa and is now becoming an increasingly demanding problem in South Africa. The almost total transfer of land in most regions of South Africa, from government to private ownership, is possibly unique in the annals of European colonization. The state by the mid 1930s had lost control over resources which in countries such as Australia or the United States were retained by the authorities because of their unsuitability for agriculture (Christopher, 1982). In effect the absence of state interest in land through a leasehold system has led to a strong demand for land and an attempt to make a living in areas highly unsuitable for the purposes of farming. Demand for land has further driven land prices to levels far in excess of its value as an agricultural commodity. Therefore the limited resources of available government land and funding need to be efficiently applied in order to ensure effective conservation as well as development opportunities. This investigation provides an important first approximation towards identifying areas where these limited resources should be concentrated by identifying vegetation types with high levels of current and potential anthropogenic land use and inadequate conservation efforts in order to constrain future spreading of transformation. As Rebelo (1997) points out, few vegetation units are spatially uniform in terms of species composition and ecosystem processes, thus further study within these priority areas is required to identify representative conservation sites within these types. Although Low & Rebelo (1996) provided rough estimates of areas considered to be facing high threats, the value of timely land-cover information on the decision making ability for planning is evident from the present study. The advent of the National Land-cover database has provided a much-needed standardized dataset of current land-cover to significantly improve South African land use and conservation planning. Further issues relevant to the identification of priority conservation areas are the scale of conservation priority setting, and the effects of global climate change on southern African vegetation. Rebelo (1997) points out that generally vegetation types shared with other neighbouring nations are more adequately conserved than vegetation endemic to South Africa. Thus a classification of vegetation types across political boundaries, as well as international co-operation, are urgent requirements for future priority setting. In addition to this, future conservation strategies will have to consider the effects of climate change on biodiversity (Rutherford et al., 2000). Not much is known on what these climate changes or their biological impacts will be, but recent work has highlighted a general eastward shift in South African species distributions as areas in South Africa dry out and warm up (Rutherford et al., 2000; van Jaarsveld & Chown, 2000; van Jaarsveld et al., 2000). It has also been shown that premier flagship conservation areas in South Africa © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Fairbanks, D.H.K. & Thompson, M.W. (1996) Assessing land-cover map accuracy for the South African land-cover database. South African Journal of Science 92, 465 –470. Fairbanks, D.H.K., Thompson, M.W.M., Vink, D.E., Newby, T.S., van den Berg, H.M. & Everard, D.A. (2000) The South African land-cover characteristics database: a synopsis of the landscape. South African Journal of Science 96, 69 –82. Forman, R.T.T. (1998) Road ecology: a solution for the giant embracing us. Landscape Ecology 13, iii–v. Forman, R.T.T. (2000) Estimate of the area affected ecologically by the road system in the United States. Conservation Biology 14, 31– 35. Forman, R.T.T. & Alexandra, L.E. (1998) Roads and their major ecological effects. Annual Reviews of Ecology and Systematics 29, 207– 231. Franklin, J.F. & Forman, R.T.T. (1987) Creating landscape patterns by forest cutting: ecological consequences and principles. Landscape Ecology 1, 5 –18. Freitag, S., Nicholls, A.O. & van Jaarsveld, A.S. (1996) Nature reserve selection in the Transvaal, South Africa: what data should we be using? Biodiversity and Conservation 5, 685– 698. Hoffmann, M.T. (1997) Human impacts on vegetation. Vegetation of Southern Africa (ed. by R.M. Cowling, D.M. Richardson and S.M. Pierce), pp. 507 –534. Cambridge University Press, Cambridge. Lambeck, R.J. (1997) Focal species a muti-species umbrella for nature conservation. Conservation Biology 11, 849 –856. Low, A.B. & Rebelo, A.G. (1996) Vegetation map of South Africa, Lesotho and Swaziland. Department of Environmental Affairs, Pretoria, South Africa. Maddock, A. & Du Plessis, M.A. (1999) Can species data only be appropriately used to conserve biodiversity? Biodiversity and Conservation 8, 603 – 615. Margules, C.R. & Pressey, R.L. (2000) Systematic conservation planning [Review]. Nature 405, 243– 253. McDonald, D.J. (1997) VEGMAP: a collaborative project for a new vegetation map of southern Africa. South African Journal of Science 93, 424– 426. McNeely, J.A. (1994) Protected areas for the 21st century: working to provide benefits to society. Biodiversity and Conservation 3, 390– 405. Milton, S.J. & MacDonald, I.A.W. (1988) Tree deaths near tar roads in the Northern Transvaal. South African Journal of Science 84, 164–165. Myers, N., Mittermeier, R.A., Mittermeier, C.G., de Fonesca, G.A.B. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature 403, 853– 858. Noss, R.F. (1983) A regional landscape approach to maintaining diversity. Bioscience 33, 700 –706. are not likely to meet their conservation goals due to an inability to track climate induced species (especially vulnerable species) range shifts (van Jaarsveld et al., 2000). This is of obvious importance in any conservation-planning scenario. In many respects ‘lines conquer’, and the South African landscape is a testament to their power. Compasses and plumblines, more than a force of arms, subdue landscapes and henceforth demarcate control and change. If current development policies (i.e. Spatial Development Initiatives, unstructured land reform) continue without proper equity towards conserving the most threatened vegetation communities, in a few decades not only will the remaining ‘natural’ areas be gone, but the people will be even poorer for it. ACKNOWLEDGMENTS We thank the Mellon Foundation, University of Pretoria, the National Research Foundation and the South African Biodiversity Monitoring and Assessment Programme for financial assistance, as well as ESRI and GIMS® for GIS software and support. We thank Bob Pressey and an anonymous referee for their helpful comments on the MS. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diversity and Distributions Wiley

Priority areas for the conservation of South African vegetation: a coarse‐filter approach

Loading next page...
 
/lp/wiley/priority-areas-for-the-conservation-of-south-african-vegetation-a-u2AwsJUdgA

References (58)

Publisher
Wiley
Copyright
Copyright © 2001 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1366-9516
eISSN
1472-4642
DOI
10.1046/j.1472-4642.2001.00098.x
Publisher site
See Article on Publisher Site

Abstract

Blackwell Science, Ltd B. REYERS*1, D. H. K. FAIRBANKS 1,2, A. S. VAN JAARSVELD 1,3 and M. THOMPSON 4 1Conservation Planning Unit, Department of Zoology and Entomology, University of Pretoria, Pretoria, 0002, South Africa, *Correspondence. E-mail: breyers@zoology.up.ac.za, 2Present address: Fitzpatrick Institute, University of Cape Town, Rondebosch, 7701, South Africa, 3Centre for Environmental Studies, University of Pretoria, Pretoria, 0002, South Africa, 4CSIR Division of Water, Environment and Forestry Technology, PO Box 395, Pretoria, 0001, South Africa Abstract. South Africa has an important responsibility to global biodiversity conservation, but a largely inadequate conservation area network for addressing this responsibility. This study employs a coarse-filter approach based on 68 potential vegetation units to identify areas that are largely transformed, degraded or impacted upon by roadeffects. The assessment highlights broad vegetation types that face high biodiversity losses currently or in the near future due to human impacts. Most vegetation types contain large tracts of natural vegetation, with little degradation, transformation or impacts from road networks. Regions in the grasslands, fynbos and forest biomes are worst affected. Very few of the vegetation types are adequately protected according to the IUCN’s 10% protected area conservation target, with the fynbos and savanna biomes containing a few vegetation types that do achieve this arbitrary goal. This investigation identifies areas where limited conservation resources should be concentrated by identifying vegetation types with high levels of anthropogenic land use threats and associated current and potential biodiversity loss. Key words. Coarse-filter, biodiversity conservation, land-cover, vegetation types, road-effects. INTRODUCTION South Africa contains a wealth of biodiversity within its borders, unequalled by other temperate regions, earning a place in the top 25 most biodiverse nations ( WCMC, 1992; Conservation International, 1998). In addition South Africa harbours the fifth highest number of plant species in the world, with the Cape Floristic Region being recognized as one of the six floral kingdoms of the world. This region contains 8200 plant species, of which 5682, are endemic and has lost approximately 30.3% of its primary vegetation (Fairbanks et al., 2000; Myers et al., 2000). Although its responsibility towards global biodiversity conservation is large South Africa, with only 4.8% ( DEAT, 1996) ( Fig. 1a) of its land surface under formal protection, falls far short of the IUCN’s nominal recommendation of 10% protected area coverage. This coverage also lags behind the 10% average attained by the rest of subSaharan Africa, with Botswana reaching 18.5%, Mozambique 12.7% and Namibia 12.4% ( McNeely, 1994; WRI, 1994; Siegfried et al., 1998). A moderately expanding human population (Central Statistical Survey, 1998) and associated land transformation in South Africa (mainly urbanization, cultivation and afforestation (Hoffmann, 1997) ) leaves 79% of the country covered with natural woody and grassland vegetation communities ( Fig. 1b) (Fairbanks et al., 2000). Water bodies and wetlands cover less than 1% of the land surface area, with human land uses making up the remaining 20% ( Fairbanks et al., © 2001 Blackwell Science Ltd. http://www.blackwell-science.com /ddi B. Reyers et al. Fig. 1 Maps of: (a) South African national and provincial protected areas (DEAT, 1996); (b) transformed, degraded and natural land-cover; (c) biomes (Low & Rebelo, 1996); and (d) road network buffered according to Stoms (2000). © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Fig. 1 continued. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. aggregate, the so-called ‘coarse-filter’ approach ( Noss, 1987, 1990). This approach focuses on protecting higher levels of the biodiversity hierarchy (e.g. land-classes and land-types) rather than species, assuming that these broad-scale biodiversity surrogates represent the finer scale aspects of biodiversity ( Pressey & Logan, 1994; Pressey, 1994; Williams & Humphries, 1996; Wessels et al., 1999; Fairbanks & Benn, 2000). However, as Pressey (1994) points out, the assumed relationship between environmental classes and species distribution and abundance is unclear and seldom investigated. In addition, certain species, especially rare species confined to small patches of habitat which are not recognized as distinct environmental classes, may ‘fall through the coarse filter’ when using broad-scale environmental classes ( Noss, 1983; Bedward et al., 1992; Panzer & Schwartz, 1998). Despite the shortcomings associated with a species-based approach to conservation planning, these higher order biodiversity surrogates may well fail to identify the composition, configuration and quantity of elements necessary for biodiversity retention, making species data a necessary component of the conservation planning process (Lambeck, 1997). The shortcomings of species distribution data and the limitations of environmental surrogate measures in the selection of priority conservation areas suggest that perhaps a combination of the two approaches in conservation planning may be advisable ( Maddock & du Plessis, 1999). At a national scale South Africa has a few databases of broader surrogates for biodiversity, including Acocks’ Veld Types (Acocks, 1988) and the more recent Vegetation of South Africa, Lesotho and Swaziland ( Low & Rebelo, 1996; McDonald, 1997). Acocks (1988) defined biological resources from a purely agricultural potential perspective, while Low & Rebelo (1996) looked at the definition of these resources from a management and potential use angle. These vegetation units were defined as having: ‘… similar vegetation structure, sharing important plant species, and having similar ecological processes’. Thus, these are units that would have potentially occurred today were it not for all the major human-made transformations, e.g. agriculture and urbanization. Therefore the Low & Rebelo (1996) vegetation map contains significant potential for acting as a broad scale surrogate of South African biodiversity 2000). Fairbanks et al. (2000) demonstrate that along with the approximately 30% transformation in the fynbos biome, the savanna and grassland biomes are about 10% and 26% transformed and degraded by human land uses, respectively ( Fig. 1c) (see also Thompson et al., 2001). In addition to this there are a total of 1176 species presently recognized as threatened ( WRI, 1994; van Jaarsveld, 2000). Thus with these valuable and often endemic biodiversity resources facing ever-increasing threats from human-induced land transformation, and mostly inadequate conservation efforts to stem these threats, South Africa has an obvious responsibility to do more towards the conservation of biodiversity (van Jaarsveld, 2000). Most of South Africa’s existing protected areas were proclaimed in an ad hoc fashion, usually because they contained areas with high scenic or tourism potential, contained endemic diseases and did not conflict with other forms of land use ( Pringle, 1982; Pressey et al., 1993; Freitag et al., 1996). Because this form of land allocation to conservation is highly inefficient and fails to effectively conserve biodiversity, several techniques have been developed for the systematic selection of land with a high conservation value, i.e. with high levels of biodiversity and large anthropogenic threats facing that biodiversity (for reviews see Williams, 1998; Margules & Pressey, 2000). However, these techniques require data on the distribution of biodiversity and threats facing biodiversity in order to identify areas important to conservation. Because the biodiversity of a region can never be fully observed and inventoried, species distribution data are often used as a surrogate or substitute measure of biodiversity. This form of data, however, has a large number of shortcomings associated with it. These include inadequate taxonomical knowledge of the groups employed, biased sampling efforts and lack of spatial congruency between areas of conservation importance to different taxa (van Jaarsveld et al., 1998; Maddock & du Plessis, 1999; Fairbanks & Benn, 2000; Reyers et al., 2000). Broad-scale biodiversity surrogates In recent years, the focus for conservation has shifted, with recommendations towards a more holistic approach of protecting biodiversity in the © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Table 1 Land-cover classes reclassified into broad categories Land-cover categories Natural land-cover Degraded land-cover Transformed land-cover % area 73.4% 10.1% 16.5% Land-cover class Wetlands, grassland, shrubland, bushland, thicket, woodland, forest Degraded land, erosion scars, water bodies Cultivated lands, urban/built-up areas, mines and quarries, forestry plantations and for identifying land important to biodiversity conservation. METHODS Current land-cover data Before the Low & Rebelo (1996) map can be used one has to differentiate between the potential vegetation cover of regions (as defined by Low & Rebelo, 1996) and that which is in reality found in the region. In other words, one needs an indication of current natural vegetation pattern, degree of transformation and amount of protection afforded each vegetation type before one can decide if it constitutes a conservation priority ( Rebelo, 1997). As Low & Rebelo (1996) point out: ‘there is little point in setting aside more of a vegetation type with vast expanses in pristine condition, while ignoring the last patches of a type which is not yet conserved’. Low & Rebelo (1996) provide some estimates of protection and transformation data, however, as they admit: ‘these are woefully incomplete’. Thus, some indication of current land-cover (the suite of natural and human-made features that cover the earth’s immediate surface) at a national scale is required for effective land-use planning, sustainable resource management, environmental research and in this instance conservation planning ( Rebelo, 1997; Fairbanks et al., 2000). To this end the advent of the National Landcover ( NLC) database is of extreme relevance. This national database was derived using manual photo-interpretation techniques from a series of 1 : 250 000 scale geo-rectified hardcopy satellite imagery maps, based on seasonally standardized, single date Landsat Thematic Mapper ( TM) satellite imagery captured principally during the period 1994– 95 ( Fairbanks & Thompson, 1996). It provides the first single standardized database of current land-cover information for the whole of South Africa, Lesotho and Swaziland (Fairbanks et al., 2000). For the purpose of the present study the 31 land-cover classes were reclassified into three categories: natural, degraded and transformed land-cover ( Table 1). Natural land-cover included all untransformed vegetation, e.g. forest, woodland, thicket and grassland. The degraded land-cover category was dominated by degraded classes of land-cover. These areas have a very low vegetation cover in comparison with the surrounding natural vegetation cover and were typically associated with rural population centres and subsistence level farming, where fuel-wood removal, over-grazing and subsequent soil erosion were excessive ( Thompson, 1996). The transformed category consisted of areas where the structure and species composition were completely or almost completely altered, which includes all areas under crop cultivation, forestry plantations, urbanized areas and mines/quarries. The databases of potential vegetation cover and current land-cover were overlaid in a geographical information system (GIS) to determine the extent of natural, degraded and transformed area within each of the 68 vegetation types identified in Low & Rebelo (1996). These values could then be used to highlight areas of high current and future vulnerability to biodiversity loss through land use impacts. Levels of transformation were compared against the transformation thresholds predicted by a geometric model developed by Franklin & Forman (1987). This work suggested that the most critical time for land planning and conservation is when between 10 and 40% of the landscape has been transformed or impacted upon. Specifically, most of the rapid ecological changes (e.g. loss of interior species) can be expected when this level increases from 20 © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Table 2 Buffer widths assigned to road classes for calculating road effect zone (after Stoms, 2000) South African Surveyor General Description National route Freeway Arterial Main Secondary (connecting and magisterial roads) Other (rural road) Vehicular trail (4-wheel drive route) to 40%. Regions showing greater than 40% loss of natural habitat have already undergone significant ecological disruptions. Furthermore, the NLC database could be used to identify the major broad categories of current threat (e.g. cultivation, forestry) facing the vegetation types. An additional GIS layer of protected area coverage for the country ( DEAT, 1996) was also employed to determine the extent of conservation areas existing within the vegetation types. Buffer width (m) 1000 1000 500 250 100 50 25 Patterns of roads In addition to these land use threats, one of the most widespread forms of alteration of natural habitats and landscapes over the last century has been the construction and maintenance of roads ( Trombulak & Frissell, 2000). Road networks affect landscapes and biodiversity in seven general ways: (1) increased mortality from road construction; (2) increased mortality from vehicle collisions; (3) animal behaviour modification; (4) alteration of the physical environment; (5) alteration of the chemical environment; (6) spread of exotic species and (7) increased alteration and use of habitats by humans (from Trombulak & Frissell, 2000). These networks cover 0.9% of Britain and 1.0% of the United States (Forman & Alexandra, 1998); however, the road-effect zone, the area over which significant ecological effects extend outward from the road, is usually much wider than the road and roadside. This road effect zone can thus provide an additional estimate of areas with a high vulnerability to biodiversity loss through changing land uses and increased human impacts. Some evidence on the size of the road-effect zone is available from studies in Europe and North America. Reijnen et al. (1995) estimated that road-effect zones cover between 12 and 20% of the Netherlands, while Forman (2000) illustrated that 19% of the United States is affected ecologically by roads and associated traffic. The road-effect zone for South Africa was determined using a similar method to that used by Stoms (2000), in which the spatial extent of road effects can be used as an ecological indicator that directly represents impacts on biodiversity. For this, the road-effect zone was used as a measure of the area potentially affected by roads. The affected distances were estimated from the reviews mentioned above, as well as from local studies ( Milton & MacDonald, 1988). Therefore national routes and freeways were assumed to affect biodiversity for a greater distance from the roadway (1 km on each side) than farm roads (100 m, Table 2). Road segments from the South African Surveyor General 1993 1 : 500 000 scale map series files (SA Surveyor General, 1993) were buffered using a standard GIS operation to the distance related to its class (Fig. 1d). Although the roads in protected areas do have an impact on biodiversity within these areas, they were excluded from this analysis as by and large protected areas overwhelmingly contribute to biodiversity conservation. A road-effect zone was calculated for the remaining untransformed areas within each vegetation type by summing the total area within the road effect zone surrounding roads in each vegetation type and converting to a percentage of the total remaining untransformed area in that vegetation type. However, the road-effect zone used here does not consider the spatial pattern of roads. Therefore, although roads clearly have a significant impact on many species, meaningful indicators of road-effects on landscapes await the attention of landscape ecologists and other scientists (Forman, 1998). As articulated by Stoms (2000), many aspects of roads affect biodiversity: road width, traffic volume, traffic speed, vehicle miles travelled, road network structure or its spatial configuration, management of the right-of-way, noise levels, light disturbance and chemical pollution. Most of these factors also vary over daily, weekly and annual cycles, which may interfere with critical behavioural periods such as breeding or © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation migration. As such, the road-effect zone can represent only a first order approximation attempt to capture more of the multidimensional nature of road network effects. RESULTS AND DISCUSSION Vulnerability assessment of vegetation types The majority of vegetation types of South Africa are not largely degraded or transformed ( Table 3). Of the 68 vegetation types, 61 contain more than 50% natural vegetation cover with a median value of 81.1% natural vegetation cover across all vegetation types. The vegetation types show low levels of degradation with a median value of 2.8%, with all but one (Afro Mountain Grassland ) being less than 20% degraded ( Table 3). Only five of the vegetation types are more than 50% transformed by anthropogenic land uses, with a median of 10% being transformed within vegetation types. Figure 2 provides a diagrammatic representation of the current levels of transformation, degradation Table 3 Percentage of natural, degraded, transformed and protected area of each of the vegetation types, as well as the percentage of each vegetation type exposed to road-effect zones Code Vegetation type 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 % natural % degraded % transformed 1.2 2.9 15.6 8.5 13.0 2.0 7.0 4.2 0.0 0.9 10.2 0.8 0.1 7.1 12.6 13.8 12.0 14.1 9.9 1.4 9.6 12.3 15.9 8.2 9.9 11.9 0.0 3.1 1.2 18.9 0.2 9.3 29.2 5.8 27.6 14.8 3.0 14.5 2.6 0.0 6.6 6.0 % protected % road-effect Coastal forest 89.3 Afromontane forest 67.9 Sand forest 72.3 Dune thicket 62.2 Valley thicket 72.1 Xeric succulent thicket 95.0 Mesic succulent thicket 78.5 Spekboom succulent thicket 93.1 Mopane hrubveld 100.0 Mopane bushveld 92.4 Soutpansberg arid 83.8 mountain bushveld Waterberg moist 90.2 mountain bushveld Lebombo arid 90.2 mountain bushveld Clay thorn bushveld 58.7 Subarid thorn bushveld 78.7 Eastern thorn bushveld 69.7 Sweet bushveld 78.3 Mixed bushveld 69.3 Mixed lowveld bushveld 70.4 Sweet lowveld bushveld 85.1 Sour lowveld bushveld 54.4 Subhumid lowveld bushveld 84.1 Coastal bushveld–grassland 43.5 Coast–hinterland bushveld 56.7 Natal central bushveld 72.2 Natal lowveld bushveld 72.5 Thorny Kalahari 83.5 dune bushveld Shrubby Kalahari 96.0 dune bushveld Karroid Kalahari bushveld 98.8 Kalahari plains 73.6 thorn bushveld Kalahari mountain bushveld 99.5 (43) 1.3 (9.5) 6.5 (44) 16.1 (17.6) 6.4 (45) 46.7 (44.6) 1.7 (25) 10.6 (14.5) 11.2 (51) 1.5 (2.1) 6.1 (51) 4.6 (8.0) 6.4 (51) 4.0 (5.3) 14.2 (unknown) 1.2 (1.8) 4.9 (0) 100.0 (100.0) 0.0 (8) 34.0 (38.3) 3.0 (65) 10.1 (12.6) 4.3 6.2 (8.6) 37.1 (38.0) 1.0 0.0 0.2 1.8 3.6 22.5 62.2 7.0 20.9 13.5 2.1 1.3 14.1 99.6 (0.9) (0.2) (0.5) (2.3) (3.1) (28.3) (67.3) (9.7) (21.5) (14.0) (3.6) (1.6) (17.8) (99.8) 3.2 1.0 5.1 8.2 11.1 4.5 5.3 3.1 1.1 4.7 1.1 5.9 4.4 7.2 5.3 0.0 2.2 3.3 3.9 4.6 9.0 (28) 9.1 (unknown) 34.1 8.7 16.5 9.5 16.6 19.8 13.5 36.0 3.6 39.8 35.0 18.0 15.6 0.0 (60) (unknown) (unknown) (27) (60) (30) (30) (76) (36) (unknown) (87) (80) (35) (unknown) 0.0 (55) 0.0 (55) 7.1 (55) 0.3 (25) 19.4 (19.5) 0.1 (0.1) 0.5 (0.5) 0.0 (0.0) © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Table 3 continued. Code Vegetation type 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 Kimberley thorn bushveld Kalahari plateau bushveld Rocky highveld grassland Moist clay highveld grassland Dry clay highveld grassland Dry sandy highveld grassland Moist sandy highveld grassland Moist cool highveld grassland Moist cold highveld grassland Wet cold highveld grassland Moist upland grassland North-eastern mountain grassland South-eastern mountain grassland Afro mountain grassland Alti mountain grassland Short mistbelt grassland Coastal grassland Bushmanland Nama Karoo Upper Nama Karoo Orange River Nama Karoo Eastern mixed Nama Karoo Great Nama Karoo Central lower Nama Karoo Strandveld succulent Karoo Upland succulent Karoo Lowland succulent Karoo Little succulent Karoo North-western mountain renosterveld Escarpment mountain renosterveld Central mountain renosterveld West coast renosterveld South and south-west coast renosterveld Mountain fynbos Grassy fynbos Laterite fynbos Limestone fynbos Sand plain fynbos % natural % degraded % transformed 76.1 92.7 66.3 68.2 34.9 63.5 67.6 60.4 46.8 88.0 61.4 67.6 94.5 51.9 87.5 38.5 81.7 99.7 99.0 98.1 94.9 99.1 90.2 86.3 97.1 94.2 89.0 94.0 98.9 80.4 9.0 39.4 88.5 88.7 64.8 87.2 34.4 4.4 3.0 0.1 0.4 0.1 0.8 0.7 1.6 11.3 2.4 17.0 7.1 4.0 36.7 8.8 4.6 5.1 0.2 0.9 0.1 1.8 0.8 9.0 2.0 0.7 2.6 2.6 0.0 0.3 1.8 1.1 1.9 0.7 0.8 1.1 7.6 8.5 19.5 4.2 33.6 31.4 (55) (55) (65) (79) % protected 1.8 0.0 0.8 0.0 (3.1) (0.0) (1.4) (0.0) % road-effect 6.8 5.5 10.2 11.3 9.0 9.1 9.4 9.6 6.7 4.1 5.5 4.8 5.7 0.8 1.2 7.6 7.0 3.4 5.8 4.6 7.4 5.4 6.0 4.0 4.4 3.9 7.7 3.0 2.4 5.4 8.1 8.8 2.9 6.0 8.6 4.0 7.1 65.1 (67) 35.8 (65) 31.6 (55) 38.0 (72) 41.8 (70) 9.7 (60) 21.6 (60) 25.3 (45) 1.5 (32) 11.4 3.6 56.9 12.9 0.1 0.1 1.6 3.3 0.2 0.8 9.5 1.7 3.2 8.4 6.0 (32) (32) (89) (unknown) (unknown) (unknown) (unknown) (unknown) (unknown) (unknown) (24) (unknown) (unknown) (unknown) (unknown) 0.0 (0.0) 0.3 (0.3) 0.0 (0.7) 0.7 (0.3) 0.8 (0.6) 9.4 (6.7) 2.3 (2.5) 3.3 (7.4) 0.6 (0.3) 0.0 11.7 0.9 0.1 0.0 0.0 0.1 1.6 0.7 0.1 0.4 4.2 0.9 3.2 0.0 (0.0) (12.5) (2.4) (1.1) (0.0) (0.0) (1.5) (1.1) (0.2) (0.0) (0.4) (4.4) (1.3) (2.3) (0.0) 0.8 (unknown) 17.8 (11) 89.8 (97) 58.7 (32) 10.8 (11) 10.3 (3) 34.1 (50) 5.2 (40) 57.1 (50) 0.0 (0.1) 5.1 (3.6) 0.7 (1.8) 1.5 (1.4) 26.4 15.5 0.0 13.6 1.2 (26.1) (16.1) (0.5) (13.8) (1.1) Values in brackets indicate estimates from Low & Rebelo (1996). Vegetation types with more than 10% protected area coverage are indicated in bold type. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation and protection across all vegetation types. Similar to the findings of the coarse-scale species-based approach used by Rebelo (1997), the grasslands and fynbos have experienced the most transformation (see Fairbanks et al., 2000), with the coastal indigenous forests having been subjected to extensive transformation for its size (Fig. 2a,b). Although degradation levels are generally low, a few regions in the grasslands biome as well as a few in the savanna biome show the highest levels of degradation, ranging from 10 to 36% of the vegetation extent ( Fig. 2c). The average amount of vegetation type currently under protection is 9.6% (median value of 1.5%) with only 18 vegetation types conforming to the IUCN’s nominal recommendation of 10% protected area coverage ( Table 3). However, this well-cited protected area recommendation of 10% is widely criticized as too little to guarantee the persistence of biodiversity within the region. Soulé & Sanjayan (1998) illustrate that up to 50% of land area may be required to successfully represent all biodiversity elements. Therefore, perhaps even these 18 supposedly well-protected vegetation types are inadequately protected ( Fig. 2d). The road-effect zone impacts on an average of 5.5% (with a median value of 6) of the remaining natural land-cover in all vegetation types ( Table 3). Five vegetation types (mesic succulent thicket, moist clay highveld grassland, dune thicket, eastern thorn bushveld, rocky highveld grassland) contain between 10 and 14.2% road-effect zones ( Table 3). The rest of the vegetation types lie under this 10% level, with the Mopane Shrubveld containing no road-effect due to the fact that it all falls entirely within the boundaries of the Kruger National Park ( Table 3). In Table 4 the areas within each vegetation type that are transformed, degraded or exposed to road-effects are summed to provide an indication of vegetation that has been disturbed or affected by these human land uses. Types with large areas affected face a high risk of biodiversity loss due to a combination of extensively degraded and transformed areas with a large road network. The west coast renosterveld, sand plain fynbos, dry clay highveld grassland, south and south-west renosterveld, short mistbelt grassland, coastal bushveld-grassland, moist cold highveld grassland, sour lowveld bushveld, afro mountain grassland, coast–hinterland bushveld, moist cool highveld grassland, clay thorn bushveld, dune thicket, moist upland grassland, dry sandy highveld grassland, rocky highveld grassland and laterite fynbos are all areas of concern due to the fact that over 40% of their extent is impacted upon by land use threats. This level of land use impact corresponds with the threshold determined by Franklin & Forman (1987), indicating extreme ecological disruption within these vegetation types. All these vegetation types are also poorly protected ( Table 3), with the coastal bushveld– grassland and dune thicket being the only types to reach the IUCN’s recommended 10% protected area coverage. However, as stated previously this level of protection is inadequate, especially in the case of these two vegetation types where it would not be sufficient to stem the biodiversity loss associated with such high levels of land use change. Of the 68 vegetation types 38 (56%) fall within the 10 – 40% category of land use impact determined by Franklin & Forman (1987) and are thus at a critical time for land use planning and conservation. Table 5 provides a list of the land-cover types within each of the top 10 priority conservation vegetation types drawn from Table 4. The Afro mountain grassland and moist cold highveld grassland contain large areas of degraded vegetation. These same vegetation types, along with the west coast renosterveld, sand plain fynbos, dry clay highveld grassland, south and southwest coast renosterveld, short mistbelt grassland, coastal bushveld–grassland, sour lowveld bushveld and coast–hinterland bushveld, contain extensive areas of commercial, semicommercial and subsistence dryland cultivation ( Table 5). The short mistbelt grassland, coastal bushveld–grassland, sour lowveld bushveld and coast–hinterland bushveld contain large areas of exotic forestry plantations and, with the exception of the sour lowveld bushveld, commercial sugarcane cultivation ( Table 5). Of all these priority vegetation types only the coastal bushveld–grassland has more than 10% protected area coverage at 13.5%, but high levels of degradation as well as high levels of transformation still make it an area of concern along its entire latitudinal distribution. The rest of these top 10 priority vegetation types all fall below 5% protected area coverage ( Table 3). This land use analysis is an example of a potential © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Fig. 2 Diagrammatic representation of levels of percentage (a) transformed (b) degraded (c) natural and (d) protected vegetation cover within each of Low & Rebelo’s (1996) vegetation types. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Fig. 2 continued. © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Table 4 Percentage area of vegetation type exposed to the combined land-cover threats of degradation, transformation and road effects Code 62 68 36 63 47 23 40 21 45 24 39 14 4 42 37 34 66 35 38 16 2 43 18 7 25 5 19 26 32 30 15 61 17 48 3 11 58 65 67 22 1 41 20 54 55 64 46 12 33 52 Vegetation type West coast renosterveld Sand plain fynbos Dry clay highveld grassland South and south-west coast renosterveld Short mistbelt grassland Coastal bushveld–grassland Moist cold highveld grassland Sour lowveld bushveld Afro mountain grassland Coast–hinterland bushveld Moist cool highveld grassland Clay thorn bushveld Dune thicket Moist upland grassland Dry sandy highveld grassland Rocky highveld grassland Laterite fynbos Moist clay highveld grassland Moist sandy highveld grassland Eastern thorn bushveld Afromontane forest North-eastern mountain grassland Mixed bushveld Mesic succulent thicket Natal central bushveld Valley thicket Mixed lowveld bushveld Natal lowveld bushveld Kimberley thorn bushveld Kalahari plains thorn bushveld Subarid thorn bushveld Central mountain renosterveld Sweet bushveld Coastal grassland Sand forest Soutpansberg arid mountain bushveld Little succulent Karoo Grassy fynbos Limestone fynbos Subhumid lowveld bushveld Coastal forest Wet cold highveld grassland Sweet lowveld bushveld Central lower Nama Karoo Strandveld succulent Karoo Mountain fynbos Alti mountain grassland Waterberg moist mountain bushveld Kalahari plateau bushveld Eastern mixed Nama Karoo Affected area (%) 92.3 69.5 67.8 65.4 64.8 60.3 56.7 49.1 48.6 47.0 45.8 45.1 43.9 42.5 42.3 42.2 40.8 39.6 39.3 38.2 37.9 36.2 34.8 34.0 33.3 32.9 32.0 31.6 29.4 29.0 28.0 25.9 25.2 23.8 22.8 20.0 18.7 17.9 17.2 16.9 16.8 16.2 16.0 15.2 15.1 14.8 13.5 12.9 12.4 12.3 © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Conservation of South African vegetation Table 4 continued. Code 8 6 44 13 10 57 59 56 50 53 51 28 31 29 49 60 27 9 Vegetation type Spekboom succulent thicket Xeric succulent thicket South-eastern mountain grassland Lebombo arid mountain bushveld Mopane bushveld Lowland succulent Karoo North-western mountain renosterveld Upland succulent Karoo Upper Nama Karoo Great Nama Karoo Orange River Nama Karoo Shrubby Kalahari dune bushveld Kalahari mountain bushveld Karroid Kalahari bushveld Bushmanland Nama Karoo Escarpment mountain renosterveld Thorny Kalahari dune bushveld Mopane shrubveld Affected area (%) 11.8 11.3 11.1 10.3 10.3 9.5 9.1 6.8 6.7 6.3 6.3 5.2 5.1 4.5 3.6 3.5 0.0 0.0 management tool for vulnerable areas, and is not limited to these top 10 vegetation types. Other vegetation types, although not as affected as these 10, are none the less also impacted upon by land use changes and should therefore also be considered and monitored in a conservation plan. Table 5 is an example of what can be done and similar analyses can be performed on all vegetation types in order to investigate the land use impacts and management parameters within each area. The vegetation types listed at the bottom of Table 4 are less impacted upon by land uses, and are generally better protected ( Table 3), with the Mopane shrubveld and thorny Kalahari dune bushveld including 100 and 99.6% protected area, respectively. These areas also contain extensive tracts of natural vegetation ranging from 83.5% for the thorny Kalahari dune bushveld to 100% for the Mopane shrubveld (Table 3). This, however, does not preclude them from further analysis and the tools developed in this study have a potential role to play in the monitoring and future management of these currently less impacted areas. Comparison of vulnerability status Low & Rebelo (1996) also provided an estimate of threat status of the vegetation types. This included a measure of land transformed by agriculture and other uses, based on ‘scant information for some of the Acocks Veld Types and should be cautiously interpreted as a rough index of habitat loss’ ( Low & Rebelo, 1996). They also include an estimate of the proportion of each vegetation type falling within conserved areas, based on an approximation of conservation area boundaries which still require confirmation (Low & Rebelo, 1996). Following a similar methodology to Thompson et al. (2001), we evaluate these estimates from Low & Rebelo (1996) as well as the calculations of protected and transformed land obtained from this study using the National Land-cover database and the Department Environmental Affairs and Tourism (DEAT, 1996) protected area database ( Table 3). Top conservation priority vegetation types identified based on Low & Rebelo’s (1996) estimates of transformed area in Table 3 highlight the west coast renosterveld, short mistbelt grassland, coast– hinterland bushveld, Natal central bushveld and the moist clay highveld grassland as areas of conservation concern due to large areas transformed. The Mopane shrubveld, grassy fynbos, Mopane bushveld, central mountain renosterveld and mountain fynbos are estimated to be areas of low priority for conservation as they are little transformed according to Low & Rebelo’s © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Table 5 Description and percentage area coverage of land-cover threats facing conservation priority vegetation types South and Sand Dry clay south-west Short Coastal Moist Sour Afro West coast plain highveld coast mistbelt bushveld– cold highveld lowveld mountain renosterveld fynbos grassland renosterveld grassland grassland grassland bushveld grassland 34.64 0.14 0.05 0.00 0.63 0.00 7.66 5.20 0.00 0.00 2.78 39.53 0.00 4.88 7.11 0.00 0.00 1.03 0.00 0.20 0.56 0.00 34.89 0.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 64.65 0.00 0.00 0.36 0.00 0.00 0.00 0.00 0.00 0.02 0.01 39.87 0.83 0.00 0.00 0.00 0.00 1.05 1.77 0.00 0.00 2.17 53.07 0.00 0.31 0.78 0.00 0.00 0.02 0.00 0.00 0.07 0.00 39.32 0.24 0.00 0.00 0.61 3.73 0.00 0.03 0.39 10.79 1.67 4.74 7.02 30.86 0.83 0.00 0.14 0.00 0.00 0.02 0.00 0.00 43.56 4.69 0.00 0.87 7.50 2.82 0.00 0.00 0.01 15.39 0.02 0.00 10.18 9.31 3.10 0.00 0.90 0.00 0.00 0.13 0.33 0.06 46.85 0.21 0.09 0.00 0.02 11.02 0.00 0.00 1.78 0.00 0.05 19.58 21.27 0.06 0.79 0.00 0.00 0.00 0.04 0.00 0.01 0.00 54.44 0.11 0.00 5.88 3.12 0.49 0.00 1.55 0.00 0.34 2.55 1.30 11.80 15.29 1.30 0.00 0.00 0.00 0.00 0.02 0.01 0.03 51.92 0.01 0.00 0.00 36.65 0.00 0.00 0.00 0.00 0.00 0.00 0.00 11.40 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.01 92 B. Reyers et al. Coast– hinterland bushveld 56.87 0.12 0.00 0.42 4.77 2.93 0.00 0.01 0.32 8.91 0.23 0.49 13.75 9.11 1.98 0.00 0.04 0.00 0.15 0.06 0.15 0.02 Description Natural land-cover 9.01 Water bodies 0.24 Dongas and sheet erosion scars 0.00 Degraded: forest and woodland 0.00 Degraded: thicket and bushland (etc.) 0.11 Degraded: unimproved grassland 0.00 Degraded: shrubland and low fynbos 0.76 Cultivated: permanent–commercial 11.70 irrigated Cultivated: permanent–commercial 0.05 dryland Cultivated: permanent–commercial 0.00 sugarcane Cultivated: temporary–commercial 0.15 irrigated Cultivated: temporary–commercial 74.78 dryland Cultivated: temporary– 0.00 semicommercial/subsistence dryland Forest plantations 0.60 Urban/built-up land: residential 1.59 Urban/built-up land: residential 0.00 (small holdings: woodland) Urban/built-up land: residential 0.00 (small holdings: bushland) Urban/built-up land: residential 0.45 (small holdings: shrubland) Urban/built-up land: residential 0.00 (small holdings: grassland) Urban/built-up land: commercial 0.06 Urban/built-up land: 0.03 industrial/transport Mines and quarries 0.07 © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 Bold values indicate main land uses in the vegetation type. Conservation of South African vegetation (1996) estimates ( Table 3). Once again the areas of high threat are estimated by Low & Rebelo (1996) to be poorly protected with less than 4% of their surface area protected and those that are low priorities are seen to be generally well protected. As found in Thompson et al. (2001), there is some degree of similarity in the rank orders of vegetation types according to threat status found in this study (i.e. affected area) and in Low & Rebelo’s (1996) (i.e. areas estimated to be transformed) (rs = 0.55; P < 0.001). However, as Table 3 illustrates, there are differences between these estimates of transformation and protection from Low & Rebelo (1996) and values generated in this study. The Low & Rebelo (1996) estimates for land transformation and protection being consistently and significantly higher (paired t-test for levels of transformation, t = 9.00, degrees of freedom = 49, P < 0.0001; paired t-test for levels of protection, t = 3.8, degrees of freedom = 67, P < 0.01). This could however, be explained by the fact that the estimates of transformation in Low & Rebelo (1996) included grazed areas, while the NLC transformation category does not (Thompson et al., 2001). The grazed areas (especially overgrazed area) are included in the degraded category of the NLC database and as such are included in the present study in the measure of affected areas ( Table 4). CONCLUSION South Africa, with its large biodiversity conservation responsibility, faces the additional problems of limited resources for conservation as well as pressing land reform initiatives. The land tenure system is a problem for conservation throughout Africa and is now becoming an increasingly demanding problem in South Africa. The almost total transfer of land in most regions of South Africa, from government to private ownership, is possibly unique in the annals of European colonization. The state by the mid 1930s had lost control over resources which in countries such as Australia or the United States were retained by the authorities because of their unsuitability for agriculture (Christopher, 1982). In effect the absence of state interest in land through a leasehold system has led to a strong demand for land and an attempt to make a living in areas highly unsuitable for the purposes of farming. Demand for land has further driven land prices to levels far in excess of its value as an agricultural commodity. Therefore the limited resources of available government land and funding need to be efficiently applied in order to ensure effective conservation as well as development opportunities. This investigation provides an important first approximation towards identifying areas where these limited resources should be concentrated by identifying vegetation types with high levels of current and potential anthropogenic land use and inadequate conservation efforts in order to constrain future spreading of transformation. As Rebelo (1997) points out, few vegetation units are spatially uniform in terms of species composition and ecosystem processes, thus further study within these priority areas is required to identify representative conservation sites within these types. Although Low & Rebelo (1996) provided rough estimates of areas considered to be facing high threats, the value of timely land-cover information on the decision making ability for planning is evident from the present study. The advent of the National Land-cover database has provided a much-needed standardized dataset of current land-cover to significantly improve South African land use and conservation planning. Further issues relevant to the identification of priority conservation areas are the scale of conservation priority setting, and the effects of global climate change on southern African vegetation. Rebelo (1997) points out that generally vegetation types shared with other neighbouring nations are more adequately conserved than vegetation endemic to South Africa. Thus a classification of vegetation types across political boundaries, as well as international co-operation, are urgent requirements for future priority setting. In addition to this, future conservation strategies will have to consider the effects of climate change on biodiversity (Rutherford et al., 2000). Not much is known on what these climate changes or their biological impacts will be, but recent work has highlighted a general eastward shift in South African species distributions as areas in South Africa dry out and warm up (Rutherford et al., 2000; van Jaarsveld & Chown, 2000; van Jaarsveld et al., 2000). It has also been shown that premier flagship conservation areas in South Africa © 2001 Blackwell Science Ltd, Diversity and Distributions, 7, 79 – 95 B. Reyers et al. Fairbanks, D.H.K. & Thompson, M.W. (1996) Assessing land-cover map accuracy for the South African land-cover database. South African Journal of Science 92, 465 –470. Fairbanks, D.H.K., Thompson, M.W.M., Vink, D.E., Newby, T.S., van den Berg, H.M. & Everard, D.A. (2000) The South African land-cover characteristics database: a synopsis of the landscape. South African Journal of Science 96, 69 –82. Forman, R.T.T. (1998) Road ecology: a solution for the giant embracing us. Landscape Ecology 13, iii–v. Forman, R.T.T. (2000) Estimate of the area affected ecologically by the road system in the United States. Conservation Biology 14, 31– 35. Forman, R.T.T. & Alexandra, L.E. (1998) Roads and their major ecological effects. Annual Reviews of Ecology and Systematics 29, 207– 231. Franklin, J.F. & Forman, R.T.T. (1987) Creating landscape patterns by forest cutting: ecological consequences and principles. Landscape Ecology 1, 5 –18. Freitag, S., Nicholls, A.O. & van Jaarsveld, A.S. (1996) Nature reserve selection in the Transvaal, South Africa: what data should we be using? Biodiversity and Conservation 5, 685– 698. Hoffmann, M.T. (1997) Human impacts on vegetation. Vegetation of Southern Africa (ed. by R.M. Cowling, D.M. Richardson and S.M. Pierce), pp. 507 –534. Cambridge University Press, Cambridge. Lambeck, R.J. (1997) Focal species a muti-species umbrella for nature conservation. Conservation Biology 11, 849 –856. Low, A.B. & Rebelo, A.G. (1996) Vegetation map of South Africa, Lesotho and Swaziland. Department of Environmental Affairs, Pretoria, South Africa. Maddock, A. & Du Plessis, M.A. (1999) Can species data only be appropriately used to conserve biodiversity? Biodiversity and Conservation 8, 603 – 615. Margules, C.R. & Pressey, R.L. (2000) Systematic conservation planning [Review]. Nature 405, 243– 253. McDonald, D.J. (1997) VEGMAP: a collaborative project for a new vegetation map of southern Africa. South African Journal of Science 93, 424– 426. McNeely, J.A. (1994) Protected areas for the 21st century: working to provide benefits to society. Biodiversity and Conservation 3, 390– 405. Milton, S.J. & MacDonald, I.A.W. (1988) Tree deaths near tar roads in the Northern Transvaal. South African Journal of Science 84, 164–165. Myers, N., Mittermeier, R.A., Mittermeier, C.G., de Fonesca, G.A.B. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature 403, 853– 858. Noss, R.F. (1983) A regional landscape approach to maintaining diversity. Bioscience 33, 700 –706. are not likely to meet their conservation goals due to an inability to track climate induced species (especially vulnerable species) range shifts (van Jaarsveld et al., 2000). This is of obvious importance in any conservation-planning scenario. In many respects ‘lines conquer’, and the South African landscape is a testament to their power. Compasses and plumblines, more than a force of arms, subdue landscapes and henceforth demarcate control and change. If current development policies (i.e. Spatial Development Initiatives, unstructured land reform) continue without proper equity towards conserving the most threatened vegetation communities, in a few decades not only will the remaining ‘natural’ areas be gone, but the people will be even poorer for it. ACKNOWLEDGMENTS We thank the Mellon Foundation, University of Pretoria, the National Research Foundation and the South African Biodiversity Monitoring and Assessment Programme for financial assistance, as well as ESRI and GIMS® for GIS software and support. We thank Bob Pressey and an anonymous referee for their helpful comments on the MS.

Journal

Diversity and DistributionsWiley

Published: Jan 1, 2001

There are no references for this article.