Plenary: Communal rangeland dynamics

Land deterioration or degradation is pervasive in African rangelands utilized under continuous grazing system. Restoration efforts should, therefore, be of a great concern to the land users as well as to help mitigate CO2 emissions. Grazing exclusion can be an effective strategy in southern Africa, but its effect to restore rangelands is not well understood. We tested for differences in vegetation and soil characteristics between exclosures and adjacent open-grazed lands over three year periods. Treatments were established in six semi-arid communal lands located in three soil types: shallow, red stony-ground (SRSG), shallow, dark sandy-loam (SDSL) and deep, dark clay-loam (DDSL soil). Average estimates of herbage removed by grazing animals were: 47% (SRSG soil), 59% (DDSL), and 62% (SDSL soil) of the total amount produced during the growing season. Compared to the grazed sites, grazing exclusion increased grass biomass by 98%, 138%, and 152% in SRSG, SDSL, and DDSL soil, respectively. Most forage nutrients of samples harvested from the exclosure and the grazed sites were adequate to meet ruminants' requirements. In all soil types, soil magnesium (Mg), organic carbon (OC), nitrogen (N), phosphorus (P), and manganese (Mn) were higher (p ≤ 0.05) in the exclosure than the grazed sites. Results suggested that continuous grazing was detrimental to soil and vegetation in the semi-arid rangelands, but this can be restored with the application of grazing abandonment practice. We conclude that short-term exclosures may enhance C and N sequestration, and improve aboveground forage biomass without compromising forage quality.

Rangelands are an important part of the ecosystem to support livelihoods, biodiversity, and livestock production. In the past, rangelands were managed in a semi-nomadic manner; where pastoralists would distribute livestock to different parts of the rangeland dependent on forage availability. It is important to understand livestock grazing distribution in a communal rangeland system, because livestock try to make use of the patchy mosaic of available forage in time and space in order to maximize intake. This study made use of Global Positioning Systems (GPS) to assess livestock distribution in a communal rangeland of the Eastern Cape province, South Africa. Six animals, which included three sheep and three cattle, were selected randomly from participating households and fitted with a neck GPS collar over a three month period for both wet (November 2016–January 2017) and dry (July 2017–September 2017) seasons. Geographic positions of the animals were recorded at five-minute intervals during both the wet and dry season of 2016/17 in order to assess rangeland use by livestock. Animal daily weight gain was assessed by weighing animals in the mornings on the day of collaring and on the day of removing collars for both dry and wet seasons. The difference between initial and final weight was divided by the number of days to get the daily animal weight gain. Grass species were surveyed from three production domains where livestock grazing was concentrated. Lastly, normalised difference vegetation index (NDVI) for the area was downloaded from Landsat imagery using Google Earth Engine (GEE) to determine the most productive parts of the rangeland. Livestock grazing distribution was analysed using time local convex hull (T-LoCoH) installed in R. Hull sets were created to identify the areas where livestock spent most of their time grazing during the day, which showed that livestock spent most of their time grazing in areas that are associated with human features. The grass species survey revealed that Hyperrenia hirta, a robust perennial C4 grass that is moderately resistant to continuous grazing, was the dominant grass species. This was followed by Eragrostis plana, suggesting that the rangeland had been disturbed. Furthermore, areas around the homesteads had a higher NDVI, which could possibly be another reason for the strong association of livestock with human features. In this study, it is assumed that lack of livestock herding might have been the reason why livestock spent most of their time grazing around the homesteads, together with the fact that the animals are saving energy and that this area is close to the drinking water. These results suggest a need for livestock owners to use herders so that the animals can use rangelands optimally. This is hoped to improve the grazing distribution, species composition and the forage quality and quantity and reduce soil erosion.

Provision of access to land and its resources in South Africa is considered one of the great challenges facing the post-apartheid South African government. The Municipal Commonage Programme has been one of the means of addressing this matter. South African government policy emphasises municipal commonage in improving the livelihoods of poor urban residents by providing a grazing resource for livestock owned by the urban poor. However, increased urbanisation, heavy stocking rates and ineffective management have raised concerns regarding the degradation of these rangelands. This paper observes the rangeland condition of the Inxuba Yethemba (Cradock and Middelburg) municipal commonages to assess the sustainability of this rangeland in improving livelihoods. Veld types on the commonages and on the farms within commonages were identified and their condition assessed separately by examining plant species composition. Visual assessment was used to collect both ecological and livestock data. Lease costs for tenants were also calculated to estimate municipal commonage revenue and assess if it had an impact on commonage management negligence and range degradation. The analysis revealed that both commonages were highly overstocked, plant cover was poor, and surface condition was prone to degradation. A substantial amount of revenue was received by the municipality, which could trigger negligence in monitoring stock numbers based on the carrying capacity on commonages. Rotational grazing and adaptive management such as reduction of stock during times of drought appear to be ideal measures to ensure the sustainability of the resource. Interviews with commonage users indicated that they believe it is the municipality's role to maintain infrastructure such as fencing at these commonages as they pay to keep livestock.

The Eastern Cape has vast communal grazing lands with the potential to sustain large herds of livestock. However, large parts of these rangelands are degraded to varying extents. Understanding the vegetation and soil dynamics in response to rangeland degradation is therefore essential to plan land use and management interventions, and recommend sustainable restoration efforts. This study assessed grass species composition, biomass production, forage nutrients, soil chemical properties, and blood serum mineral concentration of cattle in two degraded communal rangelands (less degraded and severely degraded) located in the Amathole district of the Eastern Cape. Three communal grazing lands each were identified in the two degraded conditions. A total of nine 100 × 20 m plots were established at each communal grazing area surrounding a homestead to record the soil and vegetation data. Blood samples were collected for dry and wet seasons from a total of eight Nguni cattle (2.5–3 years of age) per degradation level to determine the serum macro and micro-element concentrations. All data were analysed using the General Linear Model (GLM) procedure of SAS (2007). Results showed that Themeda triandra, Cynodon dactylon, Digitaria eriantha, Aristida congesta and Eragrostis chloromelas were recorded as dominant species in the study areas. Generally, decreaser species were dominant in less severely degraded land, whereas increaser species dominated the severely degraded land. Grass biomass was higher (p < 0.05) in less severely degraded land. Forage macro- and micro mineral elements did not show a difference between degradation levels. However, soil nitrogen (N), copper (Cu), and manganese (Mn) concentration were greater (p < 0.05) in less degraded (0.18 ± 0.011, 4.17 ± 0.268 and 330.21 ± 23.712, respectively), than severely degraded areas (0.13 ± 0.011, 2.07 ± 0.268 and 231.35 ± 23.712, respectively). The interaction between soil and forage in severely degraded areas was not significant in all minerals except Mn, whereas in less severely degraded areas, a strong correlation (P < 0.05) between soil and forage was observed for N, P, and Ca. The blood mineral concentration of the cattle showed variation between dry and wet season for both severely and less degraded rangelands, with winter recording more deficient nutrients. Forage and blood levels seemed to show correlations for some observed minerals. The study concluded that ruminants may suffer intake of minerals in both degraded areas, but could be worse under severely degraded land conditions. Animals could also experience deficiencies in P, Fe, and Cu in winter season in both severely and less degraded rangeland, though the gap in deficiency may be wider under the latter than the former.

Measuring and confirming the net primary production (NPP) that is achieved by semi-arid ecosystems is a vital part of the sustainable management of rangeland resources. There are several innovative in-field sampling techniques available for measuring aboveground production, and when compared with modelled NPP, the results can provide informed guidelines to management about the quantity of the service that the ecosystem provides. Using GIS and earth observation products, we stratified the Riemvasmaak Rural Area prior to field surveys. We defined a gradient of biomass and elevation classes within the study area, and selected sample sites, taking into account the elevation and active green biomass signals from the vegetation. The former were derived from a high-resolution digital terrain model of the study area (ASTER DEM at 30 m spatial resolution) and the latter was derived from the Landsat TM image of January 2010, as well as MODIS LAI images for the study area. The cadastral boundaries of the study site were used to ensure that samples sites fell within the farms that make up the Riemvasmaak Rural Area. Based on a 100 m transect at each site, we used the nearest plant method to identify 100 plants along each transect. The species name and the distance from the point to the plant were measured (point to tuft distance or PTD). Plant samples were taken to ensure correct identification. We have defined a scale of veld condition based on the following five categories: very poor, poor, moderate, good, very good. In the "very poor" class are samples with high cover of woody shrubs (Senegalia mellifera, Rhigozum trichotomum), and low cover of grasses. PTD > 30 cm. "Poor" class comprises moderate invasion of S. mellifera or R. trichotomum. Grass species are present, but usually, comprise the undesirable Schmidtia kalahariensis or Stipagrostis uniplumis. PTD i= 20–30 cm. "Moderate" condition had Euphorbia gregaria dominating the biomass of the site. Woody species such as S. mellifera and R. trichotomum are less abundant or even absent. PTD = 15–20 cm. "Good" condition comprises those sites where woody encroachment is less obvious, with higher diversity and the occurrence of other woody species. Grass species present include the desirable grasses such as Stipagrostis amabilis and Enneapogon cenchroides. There were no samples in the "very good" condition class. The condition of the "poor" class is likely driven by intensive wet season herbivory, which would have promoted the success of phreatic shrubs and less palatable grasses. Future research should consider reversing this trend by testing management techniques that promote grass cover.

Communal rangelands in South Africa are perceived to be on the brink of ecological collapse because they are vastly overstocked compared to those rangelands used by commercial agricultural enterprises. Livestock production plays a significant role in satisfying human needs such as beneficial economic, social and cultural services, and these may be converted into cash through sales. This paper assesses the performance of households engaged in livestock production in the north Eastern Cape communal areas of South Africa. Survey data from 120 households from Mgwalana and Mahlungulu villages collected in 2015 and 2016 were used in a stochastic frontier model to estimate technical efficiency scores and evaluate determinants among households in a communal production environment where rangelands are the cheapest source of fodder for livestock. Livestock water productivity (LWP) for various households in rural villages was described and the livestock goods and services that benefit rural households were assessed. Further, the study focuses on the water used by the rangelands, which is calculated using MODIS ET.  The findings of the study revealed that households use available resources sub-optimally and produce less output compared to the theoretical average. The average technical efficiency score of 0.79 was obtained in this study.  These results show that households have a high ability to attain livestock beneficial goods and services. These findings suggest that there is significant potential to improve outputs using existing inputs and to address the wide variation in the performance among households. Female-headed households were found to perform better with the help of hired labour. Furthermore, slight variations in LWP among households in different wealth categories were found, as better off (0.34 USD.m-3) attain a high LWP followed by middle-wealth and poor households (0.29 USD m-3). These results could be explained by the differences in livestock ownership, availability of labour, dwelling type, and net beneficial goods and services obtained by different household. Livestock holdings, the gender of the household head, and labour have a positive impact on improving LWP. The study provides essential information for understanding the productive performance among households in both villages, and thus provides important policy directions and possible interventions to improve production efficiency and reduce the livestock water footprint.