Land management practices may affect the herbaceous and woody vegetation of savannah rangelands in different seasons. This study, therefore, determined the herbaceous and woody vegetation distribution and pattern under three land management systems (commercial ranch, game reserve and communal) over two seasons (summer and winter) in the central northern Namibia. For herbaceous sampling, a total of five 0.25 m2 quadrats were randomly laid in each 250 m2 belt transect in each of the three camps in each land management system. For woody sampling, a total of three 250 m2 belt transect were randomly laid in each of the three camps in each land management system. A total of 23 grass species were identified in all three land management systems, of which 48 % were perennials and 52 % were annuals. In this study, Aristida meridionalis, Eragrostis trichophora, Enteropogon macrostachyus, Eragrostis viscosa, Pogonarthria fleckii and Urochloa trichopus were classified as the most commonly occurring species. The abundance of A. meridionalis responded significantly (P < 0.001) to land management systems being greater in the game reserve than the ranch. Enteropogon macrostachyus was recorded only in summer showing greatest (P < 0.05) and least abundance in the ranch and the communal land, respectively. The abundance of E. trichophora was high (P < 0.001) in summer in the game reserve followed by the communal and ranch sites, whereas in winter, the abundance was still greater in the game reserve followed by the ranch. The abundance of E. viscosa in summer was greater in the communal lands than the game reserve, but in winter this species was recorded in greatest frequency in the game reserve only. The occurrence of Pogonarthria fleckii was greater (P < 0.001) in the ranch than the game reserve in summer. Urochloa trichopus was rarely recorded in the ranch in summer, but the other land management systems showed similar abundance, though in winter Urochloa trichopus was recorded in greater abundance in the ranch sites only. Herbaceous plant tuft density was different (P < 0.001) between the land management systems, only in summer season being greatest in the game reserve and lowest in the communal area. Herbaceous dry matter yield was significantly different (P < 0.001) between the land management systems, in summer being greater in the game reserve than the ranch and communal areas. The total woody plant density was significantly higher (P < 0.05) in the ranch (542 TE ha-1) than in the game reserve (449 TE ha-1) and communal area (324 TE ha-1). The percentage of woody cover did not differ significantly (P > 0.05) between the land management systems. Woody plant density in the height classes of >0–1 m, >1-2 m and >2-3 m were considerably greater (P < 0.05) in the ranch than the game reserve and communal area. The height class >3 m had the greatest density in the game reserve than the ranch and communal area. Therefore, this study concludes that land management systems may greatly affect the herbaceous and woody vegetation of savannah rangelands in different seasons.

The savanna biome is characterized by a continuous herbaceous layer and a discontinuous layer of trees. The root systems of trees and grasses enable co-existence, with grasses utilizing resources in the top layers while trees access water and nutrients from deeper layers of the profile. Although several studies have been carried out to assess the effect of trees on diversity and biomass of herbaceous community, little is known on the effect of similar tree species occurring on different positions of the catena on herbaceous biomass and diversity. In order to understand how soils across a topo-sequence influences tree-grass interaction, we sampled grass species underneath Vachellia nilotica canopies as well as nearby open savanna matrix control plots. We hypothesized that diversity and biomass of herbaceous plants is higher in vertisols and under tree canopies. In order to visualise differences in grass species community composition between tree and savanna plots, and between the two soil types, non-metric multidimensional scaling (NMDS) was carried out using the vegan package in R statistical software. Generalized linear mixed effects models were applied using the lme4 package in to determine the influence of soils on grass on aerial cover, basal cover and biomass. There were significant differences in aerial cover and basal cover between open savannas in both soils (p0.05). Biomass varied significantly between open savannas in both soils (p0.05). Grass species composition was highly dissimilar between open savanna plots and under tree canopy plots in fersiallitic soils only. On vertisols, under tree canopy plots and open savanna plots had no effect on grass species composition. As expected, there were higher species diversity in fersiallitic soils and more pronounced under tree canopies than in open savannas. This suggests that the shading effect and leaf litter from the large trees have noticable influence on vegetation in different soils. In conclusion, it is evident that composition and diversity of grasses are greatly increased through tree canopy effects.
 

Grazing intensity can induce changes in the relative abundance of herbaceous vegetation on savannah rangelands. The induced changes can be used for evaluating rangeland management regimes. In this study the change in abundance of herbaceous species with grazing management regime was investigated at Mahikeng, South Africa, in two parallel transects from game reserve to private cattle ranch and then onto communal grazing. Along the transect the numbers of palatable and unpalatable species were enumerated during March 2017, in 10 m × 10 m plots at sampling intervals of 100 m. Significance of differences in the relative proportions were assessed using χ2 analysis. Spatial trends were mapped by interpolation in a Geographic Information System (GIS). The χ2 analysis revealed significant differences in the relative abundances (p > 0.05) between the communal rangeland and both the game reserve and private cattle ranch rangelands. There were significantly higher proportions of unpalatable species on the communal rangeland. Indicative palatable species were Brachiaria serrate, Digitaria eriantha, and Eragrostis lehmanniana, with the unpalatable species being Aristida congesta, Cymbopogon pospischilii, and Eragrostis gummiflua. Spatial interpolation revealed a gradient towards the communal rangeland. Since the rangelands were adjacent (with similar soil and rainfall) and had similar types of ungulate grazers which included cattle and donkeys on the communal and private ranch, and zebra and buffalo on the game reserve, the differences were attributable to differences in grazing management. In comparison with the sedentary grazing on the communal rangeland, grazing was rotated on the private cattle ranch. The relatively large size of the game reserve ameliorated the effects of high grazing intensity since the ungulate grazers could roam over a large grazing area.

 
It is perceived that Karoo shrublands of South Africa are generally stocked in excess of the recommended stocking rate to increase profit margins under extensive farming conditions. This leds to the establishment of a long-term stocking rate trial west of the town, Carnarvon in the Northern Cape Province of South Africa. Since 1988, four stocking rates (8, 7, 5.5 and 4 ha small stock unit-1 (SSU)) were applied in a rotational three camp grazing management system using Afrino sheep. The aim of this study was to quantify the effect of four different stocking density treatments on sheep production by making use of 16 years of historical weight data. The sheep in the grazing trial were weighed every 14 days and were replaced in the month of October each year. From this long-term study, animal performance did not vary markedly between treatments over the 16 years, except for gains per hectare when comparing the stocking densities during drought and wet seasons. Animal performance had a curvilinear relationship with seasonal rainfall, an indication that a rainfall year of 240 mm is optimal for sheep production in the Western Upper Karoo region. The average long term annual rainfall measured at the stocking rate trial was 219 mm. Production potential is further influenced by rainfall, where the meat grade quality was found to be lower (p < 0.05) during the wet seasons compared with the drier seasons. The mean group weight for the very light stocking density (VLSD, 8 ha SSU-1) treatment (911.8 ± 122.8 kg) was similar to the measured group weight for the heavy stocking density (HSD, 4 ha SSU-1) treatment (851.4 ± 118.0 kg). The mean for the daily live weight gain per animal calculated was unaffected by treatments. The mean live weight gain found under the HSD treatment was 48.9 ± 34.0 g per day. The highest (p < 0.05) recorded mean live weight gain was found under the light stocking density (LSD, 7 ha SSU-1) treatment with 62.7 ± 30.7 g per day. Lower performance due to higher energy outputs towards grazing and finding high-quality forage was expected to take place under HSD. However, the weight gain and overall group weight results of the HSD treatment did not confirm this. Differences in sheep performance were only found under drought conditions. From these results, it is evident that on condition that a well-managed three-camp rotational grazing system is implemented, conservative increases in stocking rate can improve profitability in a sustainable manner during non-drought situations in the Karoo rangelands of South Africa.
 

Grazing lawns underpin some of Africa’s most productive natural grazing systems, but yet are largely undervalued in a rangeland context. Consequently, the potential for higher grazer productivity that is also ecologically sustainable may be foregone in some rangeland ecosystems. However, establishing and maintaining grazing lawns requires frequent grazing, which under some environmental conditions can instead result in overgrazing, and substantial declines in primary and secondary productivity. Thus, before advocating widespread management for grazing lawns in rangeland ecosystems, a number of important questions need to be addressed. These include: 1) how do we recognise a ‘valuable’ grazing lawn, 2) where does frequent grazing produce lawns, and where it will result in overgrazing, and 3) what are the benefits and limitations of grazing lawns, and how could they best be integrated into rangeland management plans? Here I provide an emerging perspective on the potential role of grazing lawns in rangeland ecosystems, based on a synthesis of recent ecological literature, preliminary analyses of frequently grazed grass communities in South Africa and Tanzania, and a new trait-based framework for understanding grass responses to grazing. This synopsis aims to stimulate further discussion and research on the opportunities and risks associated with managing for grazing lawns in rangeland ecosystems.