Aim The search for possible factors influencing the spatial variation of grass quality is an important step towards understanding the distribution of herbivores, as well as a step towards identifying crucial areas for conservation and restoration. A number of studies have shown that grass quality at a regional scale is influenced by climatic variables. At a local scale, site factors and their interaction are considered important. In this study, we aimed at examining environmental correlates of grass quality at a local scale. The study also sought to establish if biotic factors interact significantly with abiotic factors in influencing a variation in grass quality.Location The study area is located in the Kruger National Park of South Africa. The study area stretches from west (22°49¢ S and 31°01¢ E) to east, (22°44¢ S and 31°22¢ E) covering an area of about 25 · 6 km in the far northern region of the Kruger National Park.Methods We collected environmental data such as soil texture, percentage grass cover and biomass as well as grass samples for chemical analysis from specific locations in the study area. In addition, a digital elevation model (DEM) with a resolution of 5 m was used to derive elevation, slope and aspect using a geographic information system (GIS), which were related to grass quality. We used correlation analysis and anova to relate environmental variables to grass quality. Multivariate analysis techniques were used to simultaneously analyse and explore the complex interactions between variables. Results and conclusionsOur results indicate that there is a significant relationship between grass quality parameters and site-specific factors such as slope, altitude, percentage grass cover, aspect and soil texture. Relatively, percentage grass cover and soil texture were more critical in explaining a variation in grass quality. Plant characteristics such as species type interact significantly with slope, altitude and geology in influencing nutrient distribution. The results of this study may give a better insight on foliar nutrient distribution patterns at a landscape scale in savanna rangelands. Furthermore, the results of this study may help in the selection of ancillary information, which could be used in conjunction with other data such as remotely sensed data to map grass quality -an important step towards understanding the distribution and feeding patterns of wildlife. However, we acknowledge that this study is based on one seasonal snapshot, therefore some slightly different findings may be obtained during other times of the year. Nevertheless, the study has revealed that under the conditions experienced during the study period, nutrient distribution varies with varying biotic and abiotic factors.

No abstract

Information on household responses to fuelwood scarcity was collected in two villages, one being categorized as large (Shambamuto) and another as small (Svova 'B'). Fuelwood was found to be scarcer in Svova 'B' where about 93 per cent of the respondents indicated that they find it difficult to collect compared to 76 per cent in Shambamuto. The proportion of households collecting fuelwood from non-woodlands was 60 per cent in Shambamuto and 63 per cent in Svova 'B'. In both villages, significant alternative sources of fuelwood were found to be cultivated lands, dilapidated or unwanted wooden structures, woodlots and riverbanks. Fuelwood that could be sourced from wooden structures was estimated at 8Á4 t household À1 in Shambamuto and 7Á2 t household À1 in Svova 'B'. Tree stumps in arable lands were found to be coppicing well with about 84 per cent of the coppicing stems having lengths greater than mean fuelwood stick length of 174 cm. Due to fuelwood shortage, non-traditional fuels were used as supplements and their consumption was estimated at 4Á0 t household À1 y À1 in Shambamuto and 4Á3 t household À1 y À1 in Svova 'B' respectively.

No abstract

No abstract