Land use and land transformation are major threats to biodiversity. Only a small percentage of land and thus biodiversity is protected within reserves. The majority of biodiversity lies in the hands of private and communal farmers and in order to protect biodiversity they must perceive it as having some value and have the means and incentive to conserve it. This study examined two things: (i) the relationship between biodiversity and measures of ecosystem health, range condition, primary production and presence of useful plants that would be expected to be of relevant use to land users, (ii) the perceptions of farmers of vegetation states that differ in the abovementioned attributes. Within the Eastern Province Thornveld of the Smaldeel area, four different vegetation states were selected for the study, namely park-like grassland with scattered Acacia karroo (“savanna”), heavily infested Acacia karroo grassland (“acacia”), thicket-grassland mosaic (“thicket”) and heavily utilised thicketgrassland in communal lands (“communal”). The four states are a consequence of different patterns of browsing and fire, in both pre-colonial and recent times. Different ecological attributes were assessed and compared for each state and the relationships between the different attributes determined. Ecosystem health, in terms of stability or resistance to erosion, infiltration/water-holding capacity, and nutrient cycling were compared using Landscape Function Analysis. The agricultural value of the different landscapes was measured using range condition assessment techniques. Plant species richness and other measures of diversity, along with their conservation and usefulness values, were compared between states. Plant productivity and biomass were compared using satellite data. The thicket state was found to be the most functional due to the added habitat complexity provided by the vegetation. It was the most biodiverse, the most useful and contained many, but not all of the important conservation species. The communal state had high biodiversity and was be fairly resilient to heavy usage, not showing the expected signs of land degradation. The savanna state, although thought of as the optimum state for cattle production, was not significantly different from the other states in terms of agricultural potential, but had the lowest values for plant diversity, browse potential, abundance of useful plant species and biomass. The acacia state had the highest and least stable values in terms of biomass production, however it was found to contain species of conservation importance. The preference that four groups of land users, namely men and women from commercial and communal farming areas, expressed for the four vegetation states was assessed using semi-structured interviews in conjunction with A3 colour photographs. The men and women from the communal areas and the men from the commercial areas valued the thicket state highly for its farming potential. The male commercial farmers also valued the savanna state highly as they perceived it as being most productive for cattle farming. Both the thicket state and the savanna state were found aesthetically pleasing by all the user groups, in line with theories of preference for modified savannas and for familiar environments. There appears to be potential for preserving biodiversity on farmland. The farmers in this study, with their very utilitarian perspective, were found to intuitively gauge the health of the landscape and recognise biodiversity as indicating good farming land. The commercial farmers were strongly influenced by economic motives and thus attracted to the savanna state, but also recognised the opportunity for a wider variety of farming activities in the thicket state. The communal farmers have a high dependence on the land for their livelihoods and preferred the thicket state for its overall usefulness. The views of the farmers indicate that the opportunity for conservation is good, however much of the thicket state lies in the communal land, where the openaccess land use system makes managing for biodiversity difficult.

The objective of the study was to determine sheep growth performance, fleece weights and wool quality characteristics in the sweet and sourveld communal rangelands of the Eastern Cape Province of South Africa. Data were obtained in 80 ewes aged two to three years, selected in four ecologically different communities. Two communities (Roxeni and Tyabane) were in a sweetveld area and the other two communities (Luzi I and Luzi II) were in a sourveld area. The ewes from Tyabane were nondescript, whereas ewes from Roxeni, Luzi I and Luzi II were crossbred between Merinos and non-descript sheep. Selected ewes were monitored for live weight, body condition scoring and fleece weight during shearing. Wool samples were collected from each ewe and were analysed for fibre diameter, comfort factor, clean yield, wool crimps, staple length, staple strength and the Position of break. Sheep from Roxeni were heavier (P <0.05) (39.9 ± 1.1kg) than sheep from Tyabane (29.8 ± 1.1kg). Sheep from Tyabane had the lightest (P <0.05) fleece weight (1.0 ± 1.1kg) and sheep from Roxeni had the heaviest (3.1 ± 1.1kg). There were no significant differences (P ≥0.05) between body condition scores of sheep in Roxeni and the Luzi communities. The live weights of the two-year old ewes were similar (P ≥0.05) to the three-year old ewes’ live weight. However, the three year old ewes had higher (P <0.05) body condition scores than the two year old ewes. Two year old ewes also had lower (P <0.05) fleece weight than the three years old ewes. Wool from Roxeni, Luzi I and Luzi II had lower fibre diameter (P <0.05) than wool from Tyabane. Roxeni sheep had the highest clean yield (P <0.05) of wool, followed by Luzi I and Luzi II; Tyabane had lowest clean yield percentage of wool. Wool from sheep grazing on sweetveld had a higher (P <0.05) clean yield percentage (74.3 ± 0.7%) than those grazing on sourveld (71.7 ± 0.7%). There were higher staple mid-breaks (69.5 ± 5.2) and lower base-breaks (30.5 ± 5.2) in sweetveld (P <0.05) than in sourveld, 7.5 ± 5.2 and 92.5 ± 5.2 for mid-breaks and lower base-breaks, respectively. It was concluded that sheep raised in the sweetveld and bred with purebred Merino rams had higher performance than the ones reared in the sourveld. The young upgraded ewes had a higher wool quality than the older ewes raised on communal rangelands.