Vegetation on and adjacent to mesas in the Nama-Karoo, South Africa : characteristics and comparisons

[摘要] ENGLISH ABSTRACT:Arid lands comprise some 47.2 % of the world's land surface. Between 32 % and 35 %of these drylands are subject to some form of desertification and land degradation. SouthAfrica is an arid country with water being a major limiting resource. The Nama-Karoobiome is an extensive tract of semi-arid vegetation comprising some 22.7 % of SouthAfrica, characterised by low rainfall and high temperatures. Grazing is the most commonform of land use in the country, and particularly in this biome. Aridity and grazing aretwo factors that make large areas of South Africa (including the Nama-karoo) susceptibleto land degradation. The Nama-Karoo biome has been neglected in past research and isan understudied part of South Africa's vegetation.This study is a component of a bigger umbrella project entitled Restoration of degradedNama-karoo rangelands: the role of conservation islands. The aim of this project was toassess the role played by isolated hills and mountains (mesas, inselbergs, etc.) inconserving the remnant biological diversity of the Nama-karoo rangelands of Namibiaand South Africa. Vegetation composition, seed banks, seedling recruitment, pollinationand grazing patterns and intensity on and off isolated mesas in the eastern parts of theNama-karoo were investigated in the South African component of this project.I report on patterns of 1) plant species diversity and 2) plant communities across theMiddelburg District, Eastern Cape, South Africa. Three isolated mesas (Tafelberg,Folminkskop and Buffelskop) and their surrounding plains were selected for this study.Permanent plots were established in broad transects extending from the plains to southeasternslopes, plateaux, north-western slopes and north-western plains of each mesa.Plots were 25 m2 in size, and ten 1 m2 blocks within the 25 m2 plot were randomlyselected and sampled. Species composition was recorded and cover values determinedfor all plants in every subplot.In the eastern Nama-karoo mesas were not found to be higher in Alpha Diversity thantheir surrounding plains. Plains habitats were mostly equally as diverse as mesa habitats,with some plains habitats being higher in diversity than mesa habitats. Greaterdifferences between mesas and their surroundings were found in a parallel study in thenorthern parts of the Nama-karoo in Namibia, indicating that mesas are more distinctislands of diversity with an increase in aridity. In the Middelburg District, a greaterpresence of woody phanerophytes occurred on mesa habitats in comparison with plainshabitats, a possible result of the more moist microhabitats on mesas. Cooler, moisterconditions on the south-eastern slopes of the mesas led to higher Alpha Diversity there incomparison with the warmer and drier north-western slopes, a phenomenon commonlyfound in arid areas. In terms of Alpha Diversity, the largest mesa sampled, Tafelberg(450 m above surroundings), was not significantly more diverse than the two smallermesas (Folminkskop and Buffelskop both being 200 m above surroundings).Tafelberg and Buffelskop were equally high in Beta Diversity, with Folminkskop beingmuch less diverse. Increased Beta Diversity for Tafelberg was explained by the increasedsize of the mesa (providing a more diverse microhabitat) relative to Folminkskop, whichhas the same shape and geology but is much smaller in size. High Beta Diversity valuesfor Buffelskop were explained by the presence of degraded communities on the northwesternslopes and plains, while the south-eastern slopes were some of the most diversehabitats sampled in the landscape. Mesas and plains shared few species. Proportionallyfewer species were shared between the plateau and slopes of the larger Tafelberg incomparison with the smaller two mesas. Isolation of the plateau could possibly becomemore distinct with an increase in size of the mesa. However, Tafelberg had more speciesin common with the surrounding plains than did the smaller mesas. It is suggested thatbigger mesas support a more diverse array of microhabitats enabling plains species tooccupy selective sites on the slopes of the mesa.Average Shannon-Wiener Alpha Diversity Indices for Middelburg were higher than thatof the Nama-karoo in general. This could be explained by the higher precipitation in theeastern parts of the Nama-karoo relative to the rest. Diversity in Middelburg comparedfavourably with that in other arid lands of North America and Australia and was found tobe similar in patterns and determinants of diversity compared to arid lands such as thoseof the New World.Regarding plant community composition, mesas were found to be distinctly differentfrom their surrounding plains, with no shared communities between mesas and plains.The distribution of communities across the landscape was mainly attributed to a soilmoisturegradient. Mesas, compared to plains, have shallower, rockier soils coupled withvery little run-off and naturally higher precipitation due to elevation. This results in amore mesic habitat. Plains, compared with mesas, have little rock cover, highpercentages of bare soil and higher run-off rates than mesas, resulting in a more xerichabitat. Habitat differences such as these probably existed before the impact of domesticlivestock but overgrazing has probably exacerbated the differences. The dominancestructure of plant communities on the plains has probably changed in favour of toxic,spinescent or unpalatable plant species due to selective grazing by livestock. For the twodolerite-capped mesas (Tafelberg and Folminkskop) aspect and the expected cooler,moister conditions on south-eastern slopes as factors determining communitycomposition were overridden by soil type and associated nutrient status. On Buffelskop(sandstone mesa), aspect and slope overrode soil type and associated nutrient status asdeterminants of community compositions.The potential to use mesas as a source of seeds and propagules to restore degraded plainshabitats is low. Approximately 28 % of species were shared between mesas and theirsurroundings, and not all of these species could be regarded as good colonisers.Generalist, palatable species occurring across the landscape, such as Eragrostis obtusa,Felicia muricata, F. ovata, Fingerhuthia africana, Jamesbrittenia tysonii, Limeumaethiopicum, Pentzia incana and Selago albida could have some potential for futurerestoration attempts.Mesas are distinct in composition and can be regarded as islands of one vegetation type(mesas) in a sea of a different vegetation type (plains). Mesa habitats are not generallyused for grazing by livestock, due to their general inaccessibility, and are not threatenedby development. Plant communities in the Middelburg area were very similar Incomposition to communities identified in other parts of the Nama-karoo.Results from this study suggest that the plains are at present more degraded than the mesahabitats, but it is impossible to conclude whether or not these rangelands have stabilisedgiven current stocking rates and climatic conditions.The Braun-Blanquet classification system proved an effective method to describe plantcommunities in the semi-arid Nama-karoo. All the plots used during the duration of thestudy were marked by metal stakes and are therefore of a permanent nature, so that theycan be resampled in future. This might possibly shed some light on questions related toresilience, stability and degradation of the karoo. Restoring rangeland in the future is animportant option, however, few farmers would, in the short term, be able to afford costlyrestoration techniques. It is recommended that farmers inspect the condition of their veldon a regular basis, and adjust stocking rates accordingly. Veld should be rested on aregular basis, and assessment should precede the movement of stock to a camp. Mostfarmers plough denuded areas of veld to encourage restoration, and it is recommendedthat this practice be combined with reseeding, followed by rest during and shortly aftergermination of the seed.

[发布机构] Stellenbosch University