Wilma Matheson

The use of integrated remotely sensed and GIS data to determine causes of vegetation cover change in southern Botswana

Abstract The characteristics and dynamics of dry savanna vegetation cover are receiving considerable attention from the perspectives of both global change and range degradation studies. Problems include the establishment of major savanna determinants and the floristic response of vegetation cover to given stimuli. Basic work on determinants is required to assess the nature and causes of natural resource depletion, particularly in the Kalahari region. Use of image processing techniques involving the association of pixel values and field data have resulted in the development of a vegetation map indicating floristic content and structure. Results indicate that a clear distinction can be made between classes containing high proportions of taller woodland species and those that contain mainly woody weeds. Degraded areas with sparse vegetation cover and large areas of bare soil were also identified. The GIS technique of buffer analysis was applied to determine the extent to which herbivory (livestock) and the gathering of bush products by the local population were directly involved in the spatial distribution of savanna types. Results indicate that most of the degraded areas are within 2 km of villages and boreholes. Most of the woody weed areas fall within a 2–4-km zone around boreholes. Spatial association indicates that uncontrolled bush product harvesting and goat grazing are primarily responsible for village-centred degradation, while cattle grazing around numerous boreholes is a primary cause of woody weed development. These kinds of savanna adaptive responses are difficult to reverse in rural Botswana because of increasing population pressure and concomitant poverty.

Spectral assessment of indicators of range degradation in the Botswana hardveld environment

Abstract The literature suggests that two different approaches have been applied to problems of rangeland monitoring using MSS data. These are referred to as the near infrared over red ratio which has been successfully applied to areas of relatively dense vegetation in the humid zone, and the darkening effect which is applicable in the sparsely vegetated semiarid zone. Data from Botswana suggests that neither of these is singularly appropriate in the savanna woodland zone of southern Africa. In the Botswana hardveld the measured vegetation cover consists of green vegetation which generally occupies less than 60% of the cover in a given area. The soil component is dominant. This, in addition to other vegetation components which produce a darkening effect, results in high reflectance values in the red and infrared parts of the spectrum for areas with a low vegetation cover and low reflectance values in both wavebands for areas with a high vegetation cover. In savanna woodland environments which contains elements of both the near infrared to red ratio and the darkening approach, the most suitable indicators of range condition and degrees of desertification can be obtained by directly applying spectral ranges from the red band. The range of values used is heavily ecosystem, therefore soil-type-dependent, and is referred to as the savanna woodland model.

The darkening effect in drought affected savanna woodland environments relative to soil reflectance in Landsat and SPOT wavebands

Abstract Results of soil tests and correlation analyses on 94 soil samples showed that calcium carbonate content was relatively significant in influencing soil reflectance in the Botswana semiarid savanna woodland environment. Low vegetation density reflectance (Acacia and Terminalia species) was measured in situ using Landsat and SPOT wavebands against a background of typical Kalahari soils. The soils were grouped according to Munsell Color notation. Results showed that combined soil plus vegetation reflectance was dependent on the reflectance of the soil background in the case of moderately dark to light soils. Low density vegetation cover lowered the combined reflectance entirely in red wavebands. Reflectance in the near infrared was also lowered by Acacia species when the vegetation cover was lower than Leaf Area Index = 2.5. This lowering of reflectance is referred to as “darkening.” Analysis of darkening variables in the case of a moderately dark soil showed that green leaf cover contributed 16.3% (red band) and 11.8% (near infrared band) to the overall darkening effect. Savanna woodland shrubs induce this effect by internal and external drought adaptive mechanisms which decrease near infrared scatter and enhance through shadow the efficacy of red absorption as a predictor of green vegetation cover. The rate of red absorption varies as a function of soil reflectance (relative soil brightness). This is shown on twelve predictive regression curves which were developed for both Acacia and Terminalia species.

A Landsat analysis of range conditions in the Botswana Kalahari drought

Abstract An analysis was undertaken of range conditions in the Botswana Kalahari during the 1980-1987 drought. Information from Landsat MSS print and tape data was field-checked both aerially and on the ground through two wet seasons. Comparison of 327 field and digitized aerial photograph sites with spectral data on MSS4, 5 and 7 in relation to hues on Landsat colour composite print data revealed three main spectral-vegetation types. Type 1 darkened areas comprise green tree and shrub vegetation, shown as varying tones of blue-green hues. The rate of darkening varied with the soil substrate but regression analysis showed that higher green vegetation densities occurred at lower MSS4 and 5 reflectance values. Type 2 areas combine darkened areas with areas of high near-infrared reflectance. Vegetation indices do not predict green leaf cover in type 1 or type 2 areas. This is best achieved through density slicing on MSS5. Type 3 areas show high near-infrared and low red reflectance. Green leaf cover can be p...

Mapping ecological conditions in the Okavango delta, Botswana using fine and coarse resolution systems including simulated SPOT vegetation imagery

Aspects of wetland monitoring in southern Africa are significant in the light of the Ramsar Convention, which has recently been signed by Botswana. The present work initially indicated that most ecological units in the Okavango delta can be identified and mapped using fine resolution satellite imagery and aerial photography. A second step involved assessing which of these units could also be identified using coarse resolution imagery. Contrasting results were obtained from a relatively high quality solid state sensor system (simulated SPOT 4-VEGETATION) and a relatively low quality system (NOAA-AVHRR). The extent of the water surface in the flooded areas could not be accurately mapped using NOAA-AVHRR data because of pixel saturation effects. However the presence of water was generally located using AVHRR data. The ecologically significant riparian woodlands were found to saturate pixel response on Thematic Mapper (TM) imagery. As a result of detailed comparisons, it was found that simulated SPOT VEGETATION data could be used for mapping and monitoring major changes in grassed floodplains and wooded peripheral drylands which are not spectrally separable using NOAA-AVHRR data. Most ecological change in the Okavango delta which involves such small-scale transformations as thicket development (shrub growth) and minor land-use changes related to tourism, requires the use of finer resolution systems for instance high resolution false colour aerial photography in conjunction with fieldwork. In terms of overall needs for mapping and monitoring wetland conditions in semi-arid areas it appears that major changes in flood extent and related floodplain drying vegetation changes are best detected accurately using fine resolution (TM) satellite data. Large-scale ecological changes can be semi-quantified using high quality imagery data at coarse resolutions similar to SPOT VEGETATION data.

Use of image processing and GIS techniques to determine the extent and possible causes of land management/fenceline induced degradation problems in the Okavango area, northern Botswana

Abstract Attention worldwide has been focused on the need to assess the appropriateness of land management strategies especially where these occur near sensitive areas of wildlife habitat. This work considers the use of mainly Thematic Mapper data in providing an assessment of the relative impact of different land management strategies on the natural vegetation cover in part of the sensitive Okavango area in Botswana. Supervised classification (maximum likelihood) techniques when used on six-band TM imagery showed that differential degradation was prevalent in land management areas, especially where these are separated by fencelines with an overall accuracy 72 per cent. Marginally more degradation is evident in a controlled hunting area adjacent to the Game Reserve, relative to a communal grazing area. Band transform analyses indicate that distinctive changes in cover type and density frequently take place over boundaries or fencelines separating land management areas. Some degradation in the controlled h...

Evaluation of vegetative criteria for near-surface groundwater detection using multispectral mapping and GIS techniques in semi-arid Botswana

Abstract The critical need to consider all options in the search for groundwater in semi-arid areas has promoted work on the possible association of near-surface groundwater and vegetation characteristics using a combination of remote-sensing data and geographic information systems (GIS) techniques. Two vegetative criteria (dense woody cover and abundance of deep-rooting species) are identified as being indicative of near-surface groundwater, and their spatial distribution is tested against the location of aquifers in southeast Botswana. Vegetative criteria classes were combined in a GIS environment with the distribution of geomorphic units and bedrock geology to determine the degree of coincidence with assumed or known aquifers. Results indicate that the distribution of dense woody vegetation as mapped from Thematic Mapper imagery has some potential in identifying especially surficial but also bedrock near-surface groundwater sources in mostly naturally vegetated semi-arid areas. Dense woody cover classes tend to select aquifers in topographically higher areas while classes comprising some deep-rooting species tend to select low-lying aquifers such as those occurring in fossil valleys. Deep-rooting species, however, are less successful as a vegetative criterion. Although various technical refinements are suggested, this work shows that vegetative criteria mapping can however be used in conjunction with conventional geological/geophysical techniques to enhance the prospects for groundwater location in relatively undisturbed semi-arid areas.

The use of Landsat mss imagery to determine plant characteristics and community spectral separability from sub‐tropical to arid zones in northern Australia

Abstract Plant communities along a vegetation gradient comprising sub‐tropical, semi‐arid and arid regions differed in terms of their green canopy cover, percent count (number of species present) and species richness per survey site. Analysis of green vegetation cover relative to background on Landsat MSS resulted in least distinctiveness for the subtropical region, where higher green vegetation percentages and a senesced herbaceous layer tended to homogenise the mostly near‐infrared reflective response. In the arid zone, plant spectral distinctiveness was minimised in areas of stony or bedrock cover but enhanced due to darkening in the dune areas despite low cover percentages. Due to vegetation darkening in the semi‐arid area, the spectral separability of green plant communities was enhanced within the 25% cover range. This range corresponded mostly to upper threshold levels of darkened vegetation prevalent globally in semi‐arid rangelands and in sandy areas of arid environments.

The use of Landsat MSS imagery to determine the aerial extent of woody vegetation cover change in the west-central Sahel

Residual vegetation cover was mapped over an extensive area of the west-central Sahel using Landsat MSS imagery from the 1986/87 and 1989/90 dry seasons. Results showed that major impacts decreasing the availability of range were the increase in area of windblown sand and the development of erosional features. Increases in millet and sorghum cultivation also severely decreased range availability. These factors combined appeared to be reducing the residual range cover at a rate of 5-9% per year.

Characterization of Woody Vegetation Cover in the South-East Botswana Kalahari

Analysis of three sub-scene areas (850 km2) showing natural rangeland response to drought in south-east Botswana was undertaken using Landsat MSS imagery. A plot through MSS7 and MSS5 data space revealed that small zones with high species diversity showed high NIR (near infra red) and low red band values. Samples taken from the sub-