- Title
- Assessment of the impact of land use practices on vegetation , soil and carbon-nitrogen sequestration potential in Mopane rangelands of Central Northern Namibia
- Creator
- Kahumba, Absalom
- Subject
- Rural development Land use Land tenure
- Date
- 2018
- Type
- Thesis
- Type
- Doctoral
- Type
- Pasture Science
- Identifier
- http://hdl.handle.net/10353/11718
- Identifier
- vital:39100
- Description
- This study was conducted in Ogongo district, situated in Omusati Region of the central northern Namibia. The objectives of this study were to assess the impact of land use practices on 1) vegetation distribution and status; 2) soil properties; 3) carbon-nitrogen sequestration potential, and 4) investigate the indigenous pastoralists’ knowledge and perceptions on livestock production, rangeland degradation and its indicators in three villages (Omaandi, Epukunoyana and Oshitutuma) in Mopane savannah rangelands of central northern Namibia. For vegetation and soil assessment, three grazing land management systems (communal, game reserve and cattle ranch) were identified adjacent to each other. In each grazing system, three sites (for game reserve and cattle ranch, these were camps) were selected. In each site, three belt transects of 50 m x 5 m2 were marked to record vegetation and soil data as well as to collect vegetation samples for analysis. Grass species composition was recorded using step point methods from 250 points per belt transect. A total of five 0.25 m2 quadrats were randomly laid within each 250 m2 belt transect to record the herbaceous plant density, tuft diameter, tuft distance and biomass. All woody plants in the belt transect (250 m2) were measured and recorded for species plant height and canopy diameters. Three soil samples per belt transect were collected at the time of forage sampling down to a depth of 20 cm using a soil auger. Soil and plant samples were analysed for macro and micro minerals, total organic carbon, total nitrogen, carbon-nitrogen ratios, and carbon and nitrogen isotopes. For bulk density, soil samples were collected using soil core samplers. In the social study, a total of 90 households were randomly selected from three villages and a total of 48 elders were purposefully selected based on the age, time spent in the village and livestock ownership. A combination of structured and semi-structured questionnaires were used to gather household information, while a structured questionnaire and open ended discussion were used to interview elders on the constraints of livestock production, rangeland degradation related issues and their consequences on livelihood. Field vegetation and soil data were analyzed using a General Linear Model (GLM) in SAS (2007), whereas the social data were analyzed using SPSS version 21 (2014). A total of 23 grass species were identified in all three land management systems, of which 48% were perennials and the remaining 52% were annuals. Aristida meridionalis dominated the game reserve and the ranch, whereas Eragrostis trichophora was dominant in all land management systems. Forbs occurred commonly in the game reserve and the ranch, but were dominant in the communal areas. The abundance of A. meridionalis responded significantly (P < 0.001) to land management systems with the game reserve showing remarkably greater abundance than the ranch. In summer, the abundance of E. trichophora was high (P < 0.001) in the game reserve followed by the communal and ranch sites, whereas in winter, the species had still greater abundance in the game reserve followed by the ranch. In summer, the mean grass density was greatest (P < 0.001) in the game reserve and least in the communal lands, but vice versa for the tuft distance. In summer, the average distance between tufts showed great variations (P < 0.001) with the communal areas having greater values than the ranch and game reserve. In winter, both the game reserve and the ranch showed records of similar grass tuft density and distance between tufts, whereas these values were nil for communal areas, because, there were no grasses recorded. Herbaceous plant dry matter was significantly greater (P < 0.05) in the game reserve followed by the ranch and the least in the communal areas. Results on biomass showed remarkably higher values (P < 0.001) in summer than winter in all land management systems. The total woody plant density was significantly higher (P < 0.001) in the ranch (542 TE ha-1) than the game reserve (449 TE ha-1) and communal area (324 TE ha-1). However, percentage of canopy cover did not differ significantly (P > 0.05) between the three land management systems. Woody plant density of height classes were significantly different (P < 0.05) between three land management systems. The abundance of seedlings, saplings and young shrubs were significantly greater (P < 0.05) in the ranch than the game reserve and communal area. The mature trees had a significantly greater density (P < 0.05) in the game reserve than the ranch and communal area. Soil analysis results showed that Magnesium (Mg), Potassium (K), Manganese (Mn) and Copper (Cu) had similar (P > 0.05) concentrations in soils obtained from all land management systems in both seasons. Soil Calcium (Ca) levels showed differences between land management systems in winter only being greater (P < 0.001) in the game reserve than the ranch and the communal lands. Soil Phosphorus (P) and Iron (Fe) also showed variations between land management systems in winter being significantly greater (P < 0.001) in the ranch than the other land management systems. The concentration of soil K differed between the two seasons in soils collected from the ranch and the communal lands, being greater (P < 0.001) in winter than summer seasons. Soil Zinc (Zn) also showed differences between land management systems in summer being greater (P < 0.001) in the game reserve than in the ranch and communal land. Seasonal variations were observed in soil P level of samples collected from the ranch being greater (P < 0.001) in winter than summer season. The concentration of soil Zn was higher (P < 0.001) in the soil collected from the game reserve and the ranch in summer than winter season. Woody forage analysis results showed that browse samples harvested from the ranch had greater (P < 0.001) Nitrogen (N) level than the other land management systems in both seasons. The browse samples harvested from the game reserve had greater (P < 0.001) Ca level than the other land management systems in winter season only. The level of P showed variations (P < 0.05) between land management systems in samples harvested during winter only being greater (P < 0.001) in samples from the ranch than the other land management systems. Similarly, browse harvested during winter showed significant differences (P < 0.05) in K levels being lowest in samples collected from game reserves, where the Ca level was greatest. Browse samples collected from the communal land and the ranch had the lowest Mg levels in summer and winter seasons, respectively. In all management systems, browse harvested during summer had greater (P < 0.001) N and P levels than those harvested in winter, and the reverse holds true for K and Ca levels. Browse samples harvested during winter had greater (P < 0.001) Mn levels in samples from the communal area and game reserve than the ranch, but in summer, the game reserve had higher Mn level than the other two sites. Browse samples harvested during winter only showed differences (P < 0.05) in Cu and Zn levels being greatest (P < 0.001) in samples collected from ranch and game reserves. In all management systems, browse samples harvested during winter had greater (P < 0.001) Fe, Mn, Zn and Cu (except for the game reserve and communal land) levels than those harvested in summer. Herbaceous analysis results showed that samples harvested in summer from the game reserve and ranch had greater (P < 0.001) TOC levels than the communal area, but in winter, samples from the game reserve and ranch had similar TOC levels, except for communal areas that had no herbaceous vegetation. Herbaceous samples harvested in winter from the ranch had greater (P < 0.001) TN levels than the game reserve, and that made the C:N ratio higher (P < 0.001) in the game reserve. Woody foliar samples harvested in winter from the game reserve and communal lands had greater (P < 0.001) TOC than the ranch. Values on C:N ratio showed that in winter, the communal and ranch had greater (P < 0.001) C:N than the game reserve, but in summer, the ranch had higher (P < 0.001) C:N than the game reserve and communal area. The vegetation TOC (kg ha-1) showed the order of the game reserve > ranch > communal area and the trend was similar in both seasons. In summer, the game reserve had the greatest (P < 0.001) TN (kg ha-1) followed by the ranch and the lowest in the communal area, but in winter the game reserve and ranch had similar and higher (P < 0.001) TN (kg ha-1) than the communal areas. In summer, both the game reserve and ranch recorded greater (P < 0.001) herbaceous δ13C and lowest δ15N than the communal area, but in winter, the game reserve had higher (P < 0.001) δ13C than the ranch, though both the game reserve and ranch had similar (P > 0.001) δ15N values. In summer, both the communal and ranch had slightly higher (P < 0.001) woody foliar δ13C than the game reserve, but in winter the ranch had greater (P < 0.001) δ13C than game reserve and communal areas. Woody foliar samples harvested from the ranch showed variations (P < 0.05) in δ15N between seasons being higher (P < 0.001) in summer than winter. The TOC was greatest in summer, in soils collected from the game reserve followed by the ranch and least in the communal land, but in winter, the lowest TOC was recorded in soils collected from the ranch. The soil C:N ratio varied between land management systems in soils collected during summer only being greater (P < 0.001) in the game and ranch than the communal area. The soil TOC (kg ha-1) and TN (kg ha-1) did not show (P > 0.05) differences between land management systems. Soil δ13C level was remarkably (P < 0.05) affected by land management systems, but did not interact with seasons. In both seasons, the communal land had slightly higher (P < 0.001) δ13C level than the other two land management systems. In both seasons, δ15N levels showed great differences (P < 0.001) between land management systems, with highest soil δ15N levels observed in the communal area than the ranch and game reserve. Results on social survey showed that the majority of the interviewed households in Omaandi (63percent), Oshitutuma (53percent) and Epukunoyana (50percent) were female headed families and the remaining were male headed. The mean population of goats and cattle were significantly higher (P < 0.05) in Omaandi and Epukunoyana than in Oshitutuma village, respectively. In this study, household respondents in all villages indicated that all livestock species were primarily kept for wealth status and provision of meat for consumption. Respondents in all villages also indicated that lack of grazing lands and shortage of feed were the primary constraints of livestock production followed by water scarcity and recurring droughts. Elder respondents in all villages perceived overgrazing and climate change as the primary cause of rangeland degradation. Elder interviewees in Omaandi and Oshitutuma villages regarded low grass cover and dominance of poor quality grass (Epukunoyana only) as the primary vegetation indicators of rangeland degradation. The animal related indicators such as stunted animal body as a result of mineral deficiency, high livestock mortality, emaciated body and animals traveling long distance for grazing and water (Oshitutuma only) were regarded in all villages as the primary livestock related indicators of rangeland degradation. Elders from Epukunoyana village ranked low rainfall, high soil and atmospheric temperatures as the primary climate related indicators of rangeland degradation, although in Omaandi and Oshitutuma, high soil and atmospheric temperatures were considered as the primary climate related indicators of rangeland degradation. Based on the findings, this study concludes that land management systems had great influences on the distribution and production of herbaceous and woody vegetation in summer and winter seasons. This study also observed that land management systems interacted strongly with season to influence most vegetation and soil variables. Lastly, based on the study findings, the study concludes that all three communal rangelands are degraded and degradation led to a gradual declining trend in the livestock population over the past 10 to 20 years in the communal areas as a result of feed shortage.
- Format
- 224 leaves
- Format
- Publisher
- University of Fort Hare
- Publisher
- Faculty of Science and Agriculture
- Language
- English
- Rights
- University of Fort Hare
- Hits: 614
- Visitors: 714
- Downloads: 204
Thumbnail | File | Description | Size | Format | |||
---|---|---|---|---|---|---|---|
View Details | SOURCE1 | PhD (Pasture Science) KAHUMBA, A - (Oct Grad 2018).pdf | 5 MB | Adobe Acrobat PDF | View Details |