Konrad J Wessels

The use of land facets as biodiversity surrogates during reserve selection at a local scale

Abstract Where species distribution data are inadequate reserve selection procedures have to rely on surrogate measures of biodiversity. The informativeness of land facets (the simplest units of a landscape with uniform slope, soils and hydrological conditions) as a local scale environmental surrogate was investigated in the Venetia-Limpopo Nature Reserve, South Africa. Multivariate analysis (MDS, ANOSIM) revealed that the land facets adequately represent distinct bird and dung beetle assemblages and are therefore useful surrogates. These land facets/assemblages were subsequently used as attributes in the following reserve selection procedures: (i) Percentage Area Representation (PAR—represent a nominated percentage area of each assemblage); (ii) Species-Assemblage Representation (SAR—represent each species within the smallest number of assemblages); (iii) Assemblage Diversity (AD—maximising diversity by first selecting areas containing most dissimilar assemblages). The influence of grid cell size, target representation percentages and an over-representation constraint on the efficiency of the algorithms were illustrated. The SAR procedure did not represent assemblages lacking distinguishing species and were thus more efficient in terms of total area selected. The AD procedure selected a slightly larger area than the PAR procedure, but was highly effective at rapidly increasing the diversity of the reserve network.

Relationship between herbaceous biomass and 1‐km2 Advanced Very High Resolution Radiometer (AVHRR) NDVI in Kruger National Park, South Africa

The relationship between multi‐year (1989–2003), herbaceous biomass and 1‐km2 Advanced Very High Resolution Radiometer (AVHRR) Normalized Difference Vegetation Index (NDVI) data in Kruger National Park (KNP), South Africa is considered. The objectives were: (1) to analyse the underlying relationship between NDVI summed for the growth season (ΣNDVI) and herbaceous biomass in field sites (n = 533) through time and (2) to investigate the possibility of producing reliable herbaceous biomass maps for each growth season from the satellite ΣNDVI observations. Landsat Enhanced Thematic Mapper Plus (ETM+) and Thematic Mapper (TM) data were used to identify highly heterogeneous field sites and exclude them from the analyses. The average R 2 for the ΣNDVI–biomass relationship at individual sites was 0.42. The growth season mean biomass and ΣNDVI of most landscape groups were strongly correlated with rainfall and each other. Although measured tree cover and MODIS estimates of tree cover did not have a detectable effect on the ΣNDVI–biomass relationship, other observations suggest that tree cover should not be ignored. The ΣNDVI was successful at estimating inter‐annual variations in the biomass at single sites, but on an annual basis the relationship derived from all the sites was not strong enough (average R 2 = 0.36) to produce reliable growth season biomass maps. This was mainly attributed to the fact that the biomass data were sampled from very small field sites that were not fully representative of 1‐km2 AVHRR pixels. Supplementary field surveys that sample a larger area for each field site (e.g. 1 km2 or larger) should account for the variability in biomass and may improve the strength of ΣNDVI–biomass relationships observed in a single growth season.

Flooding Patterns of the Okavango Wetland in Botswana between 1972 and 2000

Abstract The inundated area of the Okavango Delta changes annually and interannually. The variability relates to regional precipitation over the catchment area in the Angolan highlands, and to local rainfall. The patterns of the wetland were captured using more than 3000 satellite images for the period 1972 to 2000, near daily NOAA AVHRR data for 1985–2000, and less frequent images of the Landsat sensors from 1972 onwards. One AVHRR image for every 10-day period was classified into land and water using an unsupervised classification method. Evaluation against Landsat TM and ERS2-ATSR data indicate an agreement of 89% for the size of estimated inundation area. Results show that the wetland area has varied between approximately 2450 km2 and 11 400 km2 during the last 30 years.

An evaluation of the gradsect biological survey method

Biological surveys are necessary to gather species distribution data for the identification of priority conservation areas. The rationale of the gradsect method is that sampling (transects) oriented along the steepest environmental gradient should detect the maximum number of species in an area. The efficiency of the gradsect survey method was evaluated by comparing it to random, systematic and habitat-specific survey methods, during faunal field surveys (target groups: birds and dung beetles). Three gradsects were positioned within the study area to follow the major physiographical characteristics, incorporate all environmental strata (land facets) and yet be as logistically convenient as possible. The efficiency of survey methods was expressed as the number of species recorded per sampling unit effort and illustrated using bootstrap estimations to plot species accumulation curves. The gradsect method proved to be as efficient as the habitat-specific survey method and consiste ntly more efficient than the systematic and random surveys for both taxa sampled. The present study therefore illustrates that the gradsect survey method provides a cost-effective and swift representative sample of regional fauna. Moreover, the results indicate that land-form sequences, specifically ‘land facets’, are useful surrogates when sampling environmental diversity where distinct environmental gradients such as altitude and rainfall are absent.

RELEVANCE OF RANGELAND DEGRADATION IN SEMIARID NORTHEASTERN SOUTH AFRICA TO THE NONEQUILIBRIUM THEORY

According to the nonequilibrium theory, livestock grazing has a limited effect on long-term vegetation productivity of semiarid rangelands, which is largely determined by rainfall. The communal lands in northeastern South Africa contain extensive degraded areas which have been mapped by the National Land Cover (NLC) program. Much evidence suggests that long-term heavy grazing is the cause of this degradation. In order to test for the prevalence of nonequilibrium dynamics, we determined the relative effects of rainfall- and grazing-induced degradation on vegetation productivity. The vegetation production in the NLC degraded areas was estimated using growth-season sums of the Normalized Difference Vegetation Index (sigmaNDVI), calculated using data from both the Advanced Very High Resolution Radiometer (AVHRR) (1985-2003) and Moderate-resolution Imaging Spectroradiometer (MODIS) (2000-2005). On average, rainfall and degradation accounted for 38% and 20% of the AVHRR sigmaNDVI variance and 50% and 33% of the MODIS sigmaNDVI variance, respectively. Thus, degradation had a significant influence on long-term vegetation productivity, and therefore the rangelands did not behave according to the nonequilibrium model, in which grazing is predicted to have a negligible long-term impact.

Identification of potential conflict areas between land transformation and biodiversity conservation in north-eastern South Africa

Abstract Transformation of natural vegetation to other land-uses, such as crop cultivation and urban development, presents the most important threat to biodiversity. Plant and animal species distribution data were employed to identify areas of high biodiversity value in the major summer crop production region in north-eastern South Africa. These areas of biodiversity conservation importance were then evaluated in terms of their (1) potential overlap with areas currently transformed by land-uses in the region and (2) potential co-occurrence with areas of natural vegetation cover likely to become cultivated. Integrating species distribution, land-cover and land capability data allowed for potential conflict areas, i.e. areas with a high biodiversity value facing large current or future land transformation threats to be identified. Areas of potential conflict appear to be central Gauteng, the KwaZulu-Natal coastline, Maputuland and the escarpment of Mpumalanga. Most of the arable areas, that are not currently under some form of land-use, are marginal lands where the physical land characteristics demand high input costs, give rise to low yields and are thus not suitable for full scale commercial cultivation. As the results indicate some of these areas have a high biodiversity value, land reform programs should therefore refrain from promoting cultivation on marginal lands in these conflict areas, as they provide the last safe havens for many species. The proportion of bird, butterfly, mammal and plant species’ ranges remaining in an untransformed state was quantified. Animal species with less than 60% of their natural range remaining, referred to as impacted species, comprised 63 bird, 207 butterfly and 15 mammal species. The grid cells containing these impacted species were identified as additional potential conflict areas. This study presents evidence that there is significant overlap between areas of biodiversity conservation interest and transformed or arable land in this region of South Africa and that there is an urgent need for the formulation of appropriate policies to promote biodiversity conservation on private farmland.

Detecting Land Cover Change Using an Extended Kalman Filter on MODIS NDVI Time-Series Data

A method for detecting land cover change using NDVI time-series data derived from 500-m MODIS satellite data is proposed. The algorithm acts as a per-pixel change alarm and takes the NDVI time series of a 3 × 3 grid of MODIS pixels as the input. The NDVI time series for each of these pixels was modeled as a triply (mean, phase, and amplitude) modulated cosine function, and an extended Kalman filter was used to estimate the parameters of the modulated cosine function through time. A spatial comparison between the center pixel of the 3 × 3 grid and each of its neighboring pixels mean and amplitude parameter sequence was done to calculate a change metric which yields a change or no-change decision after thresholding. Although the development of new settlements is the most prevalent form of land cover change in South Africa, it is rarely mapped, and known examples amount to a limited number of changed MODIS pixels. Therefore, simulated change data were generated and used for the preliminary optimization of the change detection method. After optimization, the method was evaluated on examples of known land cover change in the study area, and experimental results indicate an 89% change detection accuracy while a traditional annual NDVI differencing method could only achieve a 63% change detection accuracy.

Unsupervised Land Cover Change Detection: Meaningful Sequential Time Series Analysis

An automated land cover change detection method is proposed that uses coarse spatial resolution hyper-temporal earth observation satellite time series data. The study compared three different unsupervised clustering approaches that operate on short term Fourier transform coefficients computed over subsequences of 8-day composite MODerate-resolution Imaging Spectroradiometer (MODIS) surface reflectance data that were extracted with a temporal sliding window. The method uses a feature extraction process that creates meaningful sequential time series that can be analyzed and processed for change detection. The method was evaluated on real and simulated land cover change examples and obtained a change detection accuracy exceeding 76% on real land cover conversion and more than 70% on simulated land cover conversion.

Unsustainable fuelwood extraction from South African savannas.

Wood and charcoal supply the majority of sub-Saharan Africa’s rural energy needs. The long-term supply of fuelwood is in jeopardy given high consumption rates. Using airborne light detection and ranging (LiDAR), we mapped and investigated savanna aboveground biomass across contrasting land uses, ranging from densely populated communal areas to highly protected areas in the Lowveld savannas of South Africa. We combined the LiDAR observations with socio-economic data, biomass production rates and fuelwood consumption rates in a supply‐demand model to predict future fuelwood availability. LiDAR-based biomass maps revealed disturbance gradients around settlements up to 1.5 km, corresponding to the maximum distance walked to collect fuelwood. At current levels of fuelwood consumption (67% of households use fuelwood exclusively, with a 2% annual reduction), we calculate that biomass in the study area will be exhausted within thirteen years. We also show that it will require a 15% annual reduction in consumption for eight years to a level of 20% of households using fuelwood before the reduction in biomass appears to stabilize to sustainable levels. The severity of dwindling fuelwood reserves in African savannas underscores the importance of providing affordable energy for rural economic development.

Improving Land Cover Class Separation Using an Extended Kalman Filter on MODIS NDVI Time-Series Data

It is proposed that the normalized difference vegetation index time series derived from Moderate Resolution Imaging Spectroradiometer satellite data can be modeled as a triply (mean, phase, and amplitude) modulated cosine function. Second, a nonlinear extended Kalman filter is developed to estimate the parameters of the modulated cosine function as a function of time. It is shown that the maximum separability of the parameters for natural vegetation and settlement land cover types is better than that of methods based on the fast Fourier transform using data from two study areas in South Africa.