Project topics and materials
Get Complete Project Material File(s) Now! »
The Maputaland Hotspot
South Africa’s political and segregationist history under colonial and Apartheid eras shaped much of the current conservation landscape in the country (Barrow & Fabricius 2002; Jones 2005). Discriminatory laws segregated Africans into overcrowded and marginal semi-autonomous communal ‘homelands’ (Cousins and Claasens 2003). Conservation via protected areas was pursued on state and communal land, sometimes by forced removals of local people (Kepe et al. 2003). After democracy in 1994, ‘homelands’ were amalgamated with provincial governments while retaining local tribal structures and communal tenure. Today, areas that comprise the former KwaZulu ‘homeland’ are known to contain some of South Africa’s most important biodiversity, much of which lies in protected areas demarcated during authoritarian Apartheid rule (CI 2005). Due to historic laws, practices, and cultural attitudes of those from without, communal land outside of protected areas is perceived to suffer from overpopulation, environmental degradation, and unsustainable resource practices.
The present study focuses on the Umkhanyakude District Municipality in Northeast KwaZulu-Natal, part of the recently designated Maputaland-Pondoland-Albany Hotspot (CI 2005). The district municipality encompasses the region commonly referred to as the Maputaland section of the hotspot, covering approximately 12,772 km², stretching from the southern boundary of the Greater St Lucia Park to the Mozambican border. Formal protected areas managed by the provincial conservation agency, Ezemvelo KwaZulu-Natal Wildlife, cover 27% of the municipality. Communal land tenure accounts for more than half of the total municipality and one fifth of communal land is found inside fenced protected areas. The municipality contains 553,702 people in 56 civil wards, the lowest level of local government, outside of parks. The region is characterized by extreme underdevelopment and impoverished subsistence livelihoods, with 70% to 80% of people below the minimum living income level (Fenske 2004). Demographic and spatial data were collected and stored as polygons in ArcView shapefiles (ESRI 1999), including population censuses, political and administrative boundaries, land tenure, protected areas, land cover, and biodiversity. Population data were obtained for Umkhanyakude District Municipality from the South African Censuses of 1996 and 2001 at two different scales, sub place and ward. Although enumeration areas were the finest spatial scale at which population data were collected for censuses, their boundaries changed between 1996 and 2001, rendering them unsuitable for temporal change analysis. Thus, sub places were used as the finest demographic scale while wards, conglomerates of sub places, were used as the lowest level of local government planning and management. For land cover, the latest available data were obtained from the 1996 National Land Cover project, produced by the South African Centre for Scientific and Industrial and the Agricultural Research Council (Thompson 1996). ‘Intrinsic biodiversity importance’ scores were obtained from Ezemvelo KwaZulu-Natal Wildlife and are a measure of conservation importance based on endemism, rarity, vulnerability, threat status, conservation importance, and level of protection. These data (29,108 ArcView polygons) were created based on landscape, ecosystem/community (wetlands, grasslands, forests, vegetation communities), and species level attributes (85 threatened plants, nine economically important medicinal plants, four endemic mammals, 29 threatened birds, four endemic amphibians, 10 endemic reptiles, 21 threatened fish, five freshwater crustaceans, and 99 endemic insects birds) (Goodman 2000). The scores are not measures of raw biodiversity richness, nor do they include measures of irreplaceability or complementarity and are not intended for systematic conservation planning (Margules & Pressey 2000; Pressey & Cowling 2001; Reyers et al. 2002; Balmford 2003). Other conservation planning research in Africa has focused on species distribution, particularly birds and mammals, due to data availability. Biodiversity importance data used here, of which species richness is only one component, proved to be a suitable surrogate with a strong correlation between measures of overall biodiversity importance and species richness extracted from the overall biodiversity importance score (r = 0.976). Weighted biodiversity scores were calculated by scaling up the intrinsic biodiversity polygonal scores to each independent level, multiplying this value by the polygon’s percent area of the total area for that scale, and then summing all polygons for an area. The weighed scores were calculated at three different spatial scales of analyses:
quarter degree grid square (QDS) (mean area = 683 km²), census ward (mean area = 162 km²), and census sub place (mean area = 40 km²). Spearman rank correlations and Kruskal-Wallis analyses of variance (ANOVA) were conducted (p<0.05 significance level) at each spatial scale. The focus was on land outside of existing, and mostly uninhabited, conservation areas in an effort to identify demographic trends and anthropogenic threats facing unprotected biodiversity, sites for future conservation, and areas of potential resource conflict. The South African National Spatial Biodiversity Assessment identified local communities as an important scale for biodiversity conservation (Driver et al. 2005). Thus, discussions focus on ward-level results as they provide a practical spatial scale for civic participation, local decision-making, policy enactment, service delivery, and micro planning and management.
Parks Are Not Always People Magnets
A negative, although weak, correlation was found between human population density and biodiversity for wards outside of protected areas (Table 1). These results are contrary to previous findings of a positive correlation between population density and biodiversity at broader national (quarter degree) and African (one degree) scales (Balmford et al. 2001; Chown et al. 2003; Janse van Rensburg et al. 2004). The negative results were initially thought to be a scale issue or influenced by the exclusion of protected areas in the analyses. However, correlations for Maputaland at both the sub place and QDS, with and without protected areas, yielded similar negative relationships (Table 1). Regarding previous findings of positive correlations at broad scales, Chown et al. (2003) note how both humans and biodiversity respond positively to rainfall and net primary productivity; thus the overlap in their locations. The negative relationships in Maputaland, while possibly a fine-scale phenomenon, suggests the relative importance of other socio-economic drivers of population distribution and change at the local level. Population density and biodiversity were related to land use as expected. Areas with high biodiversity scores had high levels of natural land cover and low levels of transformed and degraded land; areas with high human population density had high levels of transformed and degraded land.
Another misconception is that all protected areas are magnets for local rural poor people, resulting in higher densities surrounding parks. This notion is partly due to the previously described broad relationships between biodiversity and population density, but also supposedly because local people are attracted to parks in search of benefits or resources. In Maputaland, wards bordering protected areas had significantly lower population densities (ANOVA p=0.012) than those that do not share a border with a protected area. Accordingly, land around parks had higher levels of natural land (ANOVA p=0.023) and lower levels of transformed land (ANOVA p=0.040). Low population densities appear to buffer parks in Maputaland, a favorable condition for their biodiversity conservation. Again, this is contrary to popular perceptions that parks are threatened by proximate human encroachment and related anthropogenic degradation. While many protected areas are capable of directly contributing to local livelihoods, in reality access to park resources remains limited, benefits to communities overstated, and costs disproportionately absorbed by local people (Barrow & Fabricius 2002; Ferraro 2002; Scherl et al. 2004; Jones, 2005).
Chapter 1: General introduction
Chapter 2: Using community-integrated geographic information systems to study land reform in South Africa
Chapter 3: Who benefits from ‘Conservation and Development’? The political ecology of biodiversity in KwaZulu-Natal, South Africa
Chapter 4: Land tenure, HIV/AIDS, and population dynamics in the Maputaland Conservation Hotspot
Chapter 5: Modeling conservation-induced resettlement in Mbangweni, South Africa
Chapter 6: Transboundary conservation: development implications for communities in KwaZulu-Natal, South Africa
Chapter 7: General discussion
