The effects on vegetation of three small mammalian herbivores were determined experimentally in a subalpine meadow in southwestern Colorado. Pikas (Ochotona princeps) inhabit talus and forage on surrounding vegetation. Their foraging was consistent with predictions of central place foraging theory. Amount of foraging decreased with distance from talus, whereas selectivity increased. These patterns are expected to influence the resources used by pikas, and plant abundance and species richness increased with distance from talus, as predicted. Experimental exclusion of pikas demonstrated that a significant portion of the lower vegetational cover and species richness was due to pikas. A model postulating that the effects of pikas on vegetation resulted from a gradient in herbivory, and a corresponding gradient in the relative importance of herbivory and competition among plants, was generally supported. Total vegetational abundance and species richness increased in the absence of pikas. The increase in pika exclosures compared to controls was greatest near talus and decreased with distance. The initial effect of pika exclusion on cushion plants (assumed to be the poorest competitors among the plants in the absence of herbivory) was positive near talus, but was insignificant and tended to be negative at greater distance from talus. The prediction of eventual decline of cushion plant populations in the absence of pikas was supported in two of three sites after three years. The effects on vegetation of two co-occurring herbivores, pocket gophers (Thomomys talpoides) and voles (Microtus montanus), were evaluated using a factorial exclusion experiment. Non-independence of effects was assessed as presence of a significant interaction term in 2-way ANOVA. Gopher x vole interaction terms were at least marginally significant for many analyses. Results only weakly suggested direct foraging interactions between gophers and voles, but showed considerable importance of indirect, plant-plant interactions. Exclusion of either gophers or voles significantly increased vegetational abundance, however, no additional change in abundance resulted when both were excluded simultaneously. Exclusion of either increased forb abundance, and exclusion of gophers increased grass abundance. Only simultaneous exclusion of both, however, altered proportional abundances of plant groups.

Herbivores have a significant influence on their environment. Factors that influence herbivore distribution in a landscape are important for conservation. Artificial water provision is one such factor, with water sources being focal points of herbivore activity. Variation between herbivore utilisation of different waterhole types and habitat integrity surrounding the different waterholes is assessed in this study. Correlations are drawn between herbivore utilisation and habitat integrity to quantify the impact that artificial waterholes have on the landscape. A scoring system is devised to investigate the various factors affecting vegetation around artificial waterholes. Results show that there are significant variations between herbivore utilisation for different artificial waterhole types, as well as significant variation in habitat integrity surrounding the different waterhole types. Distance between waterholes and drainage lines, and utilisation by specific herbivore species have a significant impact on habitat integrity - specifically affecting veld condition and disturbance observed on woody plant species. A habitat score that was created by combining data from both the herbaceous and woody layers is not affected by waterhole type, distance from waterholes, or the different herbivore species utilising the different waterhole types. Earth dams have the greatest impact on surrounding vegetation and are the most utilised waterhole type. Closures of earth dams are not recommended due to their importance to herbivores. Less utilised waterhole types are also important, mitigating the impact of herbivore damage to vegetation at earth dams. / Environmental Sciences / M. Sc. (Nature Conservation)

Thesis (MFor)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: The factors determining habitat selection of large herbivores, on a private wildlife reserve in semi-arid southeastern Zimbabwe, were investigated. Gross vegetative structure, herbaceous composition and topographic features thought to determine herbivore distribution were measured. Seasonal variation in resource distribution was considered, and research therefore extended over an entire year. Herbivore distribution and ecological niche separation was explained through several important environmental variables, and potential for inter-specific competition inferred. Additionally, predictive habitat suitability models were designed for each of the grazing species in the critical dry season. Herbivores showed a large degree of niche overlap in both the hotwet season and the cool-dry season, when food resources were more plentiful. Niche separation between grazers was pronounced in the hot-dry season. Herbivore distribution was associated most closely with distance to water, grass sward height, time since burn, woody plant density and by the presence of predominant grasses, these being Urochloa mossambicensis, Panicum maximum, Heteropogon contortus and Digitaria eriantha. Ecological separation of herbivores by the grasses P. maximum, H. contortus and D. eriantha was more indicative of associated environmental variables than feeding niche separation. A GIS-based analysis, using species presence data and quantitative coverages of environmental variables, produced maps of gradations of habitat suitability for grazing species during the dry season. Results of both analyses were integrated and gave a better understanding of ecological separation, and possible competitive interactions, among the large herbivore community on Malilangwe Estate. Recommendations were made based on the interpretation of findings, within the context of available management options. / AFRIKAANSE OPSOMMING: Die bepalende faktore van habitat seleksie by groot herbivore op ‘n privaat wildreservaat in semi-dorre suidoos Zimbabwe is ondersoek. Globale plantegroei stuktuur, kruidagtige samestelling en topografiese eienskappe wat glo herbivoor verspreiding bepaal, is gemeet. Seisoenale variasie in hulpbronverspreiding is in ag geneem en dus het navorsing oor ‘n hele jaar gestrek. Herbivoor verspreiding en ekologiese nisskeiding is verduidelik deur verskeie belangrike omgewingsveranderlikes en die potensiaal vir interspesifieke kompetisie is afgelei. Boonop is voorspellende habitat-geskiktheidsmodelle ontwerp vir elk van die weidingspesies in die krities droë seisoen. Herbivore toon ‘n hoë graad van nis oorvleueling in sowel die warm, nat seisoen as die koel, droë seisoen wanneer voedingsbronne meer volop is. Nis verdeling tussen weidiere was duidelik herkenbaar in die warm, droë seisoen. Herbivoor verspreiding is meestal geassosieer met die afstand na die water, die grasveld hoogte, tydperk sedert ‘n brand, digtheid van houtagtige plantsoorte en met die teenwoordigheid van die oorheersende grasse, Urochloa mossambicensis, Panicum maximum, Heteropogon contortus en Digitaria eriantha. Die ekologiese skeiding van herbivore deur die grasse P. maximum, H. contortus en D. eriantha het meer gedui op geassosieerde omgewingsveranderlikes as op skeiding van voedingsnisse. ‘n GIS-gebaseerde analise wat spesie-teenwoordigheidsdata en kwantitatiewe dekking van omgewingsveranderlikes gebruik, het klassifikasiekaarte geproduseer van habitatgeskiktheid vir weidiere tydens die droë seisoen. Resultate van beide analises is geïntegreer en het ‘n beter begrip van ekologiese skeiding en moontlike kompeterende interaksies tussen die groot herbivore gemeenskap op Malilangwe Estate tot gevolg gehad. Aanbevelings is gemaak, gebaseer op die interpretasie van bevindinge, binne die konteks van beskikbare bestuursopsies.

The present study compared Sand Forest bird assemblages found in a communal land area with that of the Tembe Elephant Park, and determined the habitat preference and status of selected herbivore species within the park. The study forms part of the Maputaland Conservation-based and Integrated Rural Development Programme of the Centre for Wildlife Management from the University of Pretoria and is linked to the activities of the Lubombo Transfrontier Conservation Area (LTFCA). The main purpose of the study was to compare Sand Forest bird assemblages found in the Tshanini Community Conservation Area, which is characterised by low levels of human utilisation, with that of the Tembe Elephant Park, which is characterised by wildlife utilisation. This approach was used to determine the biological importance of this communal land area in contributing towards the conservation of the rare Sand Forest habitat. Visual and auditory bird surveys revealed that the communal land area contains unique Sand Forest bird assemblages, which demonstrated the biological importance of the communal land for Sand Forest conservation, especially from an avian perspective. The second purpose of the study was to identify possible competition between selected herbivore species within the Tembe Elephant Park and/or a decrease in numbers of rare species. Herbivores that might be adversely affected by the destruction of the Sand Forest, or who may themselves have a destructive effect on the Sand Forest were also identified. Target herbivores included the nyala Tragelaphus angasii, impala Aepyceros melampus, Burchell’s zebra Equus burchellii, greater kudu Tragelaphus strepsiceros, red duiker Cephalophus natalensis and suni Neotragus moschatus. Road transects were used to record the spatial distribution of the target herbivores, and the vegetation types that were used more or less often than expected were subsequently determined. None of the target herbivores showed a preference for the Sand Forest or appeared to have a destructive effect on the Sand Forest. The suni, however, reached its highest density within the Sand Forest and the destruction of this habitat will therefore negatively affect the suni population. In several parks and reserves that aim to conserve a variety of species, it has been necessary to control the populations of highly competitive species. Both the nyala and the impala are highly competitive and occur in relatively high numbers within the Tembe Elephant Park, and consequently their population numbers should be kept sufficiently low in order not to have a negative influence on the vegetation or the survival of less competitive ungulates. Total aerial counts and transect distance sampling counts indicated an increase in the numbers of all the target herbivores. It is important to protect a viable portion of the preferred habitat of every target species within a reserve, and to keep competition with rare species to a minimum for the long-term survival of the regional biodiversity. Key aspects of wildlife and their habitat should be monitored so that trends are noted in time, and management adjustments can be made accordingly. / Dissertation (MSc (Wildlife Management))--University of Pretoria, 2007. / Centre for Wildlife Management / unrestricted