Biodiversity in Southern Africa

Vol. 3 Vol. LanduseImplications for and Management Biodiversity inBiodiversity AfricaSouthern

KLAUS HESS for

and PUBLISHERS 3 ment Biodiversity in Southern Africa Impli- cations Landuse Manage- ISBN-Germany Germany ISBN 9 783933 117441 9 783933 117458 9 783933 117472 9 783933 117465 Germany ISBN 9 783933 117441 9 783933 117458 9 783933 117472 9 783933 117465 Namibia ISBN 9 789991 657301 9 789991 657325 9 789991 657332 9 789991 657318 Namibia ISBN ISBN-Namibia Biodiversity is important for sustaining life on Earth yet it is threatened globally. The BIOTA BIOTA The globally. Biodiversity is important for sustaining life on Earth yet it is threatened of change in biodiversity in Africa project analysed the causes, trends, and processes Southern This book, until 2010. 2001 from decade, full a nearly Africa over and western South Namibia a summary of the results from the many and diverse which is comprised of three volumes, offers rst period subprojects of during long-term this observation ﬁ and related research, at both local region. the for options management land sustainable on focus a with and scales, regional and 9 789991 657301 9 789991 657325 9 789991 657332 9 789991 657318 © University of Hamburg 2010 All rights reserved

Klaus Hess Publishers www.k-hess-verlag.de

ISBN all volumes: 978-3-933117-44-1 (Germany), 978-99916-57-30-1 (Namibia) ISBN this volume: 978-3-933117-47-2 (Germany), 978-99916-57-33-2 (Namibia)

Printed in Germany

Suggestion for citations:

Volume: Hoffman, M.T., Schmiedel, U., Jürgens, N. (eds.) (2010): Biodiversity in southern Africa 3: Implications for landuse and management. Göttingen & Windhoek: Klaus Hess Publishers.

Chapter (example): Schmiedel, U., Linke, T., Christiaan, A.R., Falk, T., Gröngröft, A., Haarmeyer, D.H., Hanke, W., Henstock, R., Hoffman, M.T., Kunz, N., Labitzky, T., Luther-Mosebach, J., Lutsch, N., Meyer, S., Petersen, A., Röwer, I.U., van der Merwe, H., van Rooyen, M.W., Vollan, B., Weber, B. (2010): Environmental and socio-economic patterns and processes in the Succulent Karoo—frame conditions for the management of this biodiversity hotspot. – In: Hoffman, M.T., Schmiedel, U., Jürgens, N. (eds.): Biodiversity in southern Africa 3: Implications for landuse and management: 109–150. Göttingen & Windhoek: Klaus Hess Publishers.

Subchapter: If you want to refer to a Subchapter you should insert the citation to the Subchapter as follows (example): “... see Subchapter by Hanke & Schmiedel on Restoring degraded rangelands in Schmiedel et al. (2010) ....”

Corrections brought to our attention will be published at the following location: http://www.biota-africa.org/biotabook/

Cover photograph: A farmer and a researcher walking and talking in the veld in the northern Succulent Karoo, South Africa. Photo: Imke Oncken, Hamburg/Germany. Cover Design: Ria Henning

IV Article IV.4 – Author’s copy –

Please cite this article as follows:

Schmiedel, U., Linke, T., Christiaan, R. A., Falk, T., Gröngröft, A., Haarmeyer, D. H., Hanke, W., Henstock, R., Hoffman, M. T., Kunz, N., Labitzky, T., Luther-Mosebach, J., Lutsch, N., Meyer, S., Petersen, A., Röwer, I. U., van der Merwe, H., van Rooyen, M. W., Vollan, B., Weber, B. (2010): Environmental and socio-economic patterns and processes in the Succulent Karoo - frame conditions for the management of this biodiversity hotspot. – In: Hoffman, M. T., Schmiedel, U., Jürgens, N. [Eds.]: Biodiversity in southern Africa. Volume 3: Implications for landuse and management: pp. 109–150, Klaus Hess Publishers, Göttingen & Windhoek. Succulent Karoo

108 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT Succulent Karoo

Flower display on fallow land in Namaqualand. Photo: J. Dengler.

Part IV

IV.4 Environmental and socio-economic patterns and processes in the Succulent Karoo—frame conditions for the management of this biodiversity hotspot

IV.4.1 Introduction ...... 111 IV.4.2 Interdependence of soils and vascular plant vegetation in the Succulent Karoo ...... 114 Introduction • Drivers of vegetation diversity at the landscape scale • Medium-scale drivers of vegetation patterns • Plant-soil-interaction at the micro-scale IV.4.3 Old field succession in the Namaqua National Park, Succulent Karoo, South Africa ...... 119 Introduction • Methods • Results and discussion • Conclusions IV.4.4 Restoring degraded rangelands in the Succulent Karoo: lessons learnt from our trials ...... 122 Introduction • Description of the BIOTA restoration experiments • Ecological concepts and results of the tested treatments • Understanding the processes of vegetation recovery • Implications for further research • Implications for restoration practice IV.4.5 Land management in the biodiversity hotspot of the Succulent Karoo. A case study from Soebatsfontein, South Africa ...... 133 Introduction • Material and methods • Results and discussion • Consequences for land management in Soebatsfontein IV.4.6 Conclusions and further research needs ...... 145

SUCCULENT KAROO 109 Environmental and socio-economic patterns and processes in the Succulent Karoo— frame conditions for the management of this biodiversity hotspot

UTE SCHMIEDEL*, THERESA LINKE, REGINALD A. CHRISTIAAN, THOMAS FALK, ALEXANDER GRÖNGRÖFT, DANIELA H. HAARMEYER, WIEBKE HANKE, RONEL HENSTOCK, M. TIMM HOFFMAN, NATALIE KUNZ, TIMO LABITZKY, JONA LUTHER-MOSEBACH, NICOLE LUTSCH, SARAH M EYER, ANDREAS PETERSEN, INGA UTE RÖWER, HELGA VAN DER MERWE, MARGARETHA W. VAN ROOYEN, BJÖRN VOLLAN & BETTINA WEBER Succulent Karoo Succulent Karoo Summary: The Succulent Karoo, the arid winter rainfall area of southern Africa, is one of the most diverse biomes of the region. BIOTA Southern Africa’s research activities in the Succulent Karoo Biome focussed on Namaqualand, a 50,000 km2 region in the western part of the biome. In order to describe the environmental and socio-economic frame conditions of the land management of the Namaqualand, this chapter has been subdivided into six Subchapters. The introduction (IV.4.1) provides general background information on the natural environment and history of the area and the resulting challenges for the landuse management in Namaqualand. In the subsequent Subchapter (IV.4.2), we describe the BIOTA researchers’ fi ndings on soil patterns at different spatial scales, i.e. landscape- (km²), habitat- (hectare) and micro-scales (< 1 m²), and their effect on vegetation patterns and plant species turnover. We discuss the consequences of these fi ndings for land management practices that aim to maintain the unique species richness of the area. We continue (Subchapter IV.4.3) by describing the effect of cultivation on the species richness of annual and perennial plants and species composition of the vegetation. Results showed that over the past 15 years total species richness has increased with time since abandonment at all monitoring sites. Species composition of the perennial species showed a clear directional trend over the monitored years, whereas annual species composition was dictated by the timing and the amount of rainfall in a growing season. The fi ndings have consequences for the restoration of old lands but also for their management for the best fl owering display to attract tourists. The best fl owering displays were associated with the old fi elds that were tilled approximately every four years. In order to restore disturbed and transformed ecosystems within a practical time scale, BIOTA implemented and tested active restoration measures (i.e. brushpacks, dung, microcatchment, functional plant transplants, stone cover), in four major ecological regions of Namaqualand (i.e. Richtersveld, Coastal plain, Namakwaland Klipkoppe, Knersvlakte). The results are compared and the underlying processes that infl uence critical factors like soil hydrology, chemical soil properties, life history traits of plants, and plant interactions, are discussed in the context of ecological dynamics as well as implications for restoration practices (Subchapter IV.4.4). Based on our interdisciplinary insight into the socio-economic environment of the Soebatsfontein settlement and its new commonage, in the second last Subchapter (IV.4.5) we describe the challenging and facilitating frame conditions for sustainable land management in that community. The study provided some practical recommendations but also revealed that successful implementation of the recommendations would require an integrative process comprising a broad range of local stakeholders. Finally, we conclude (Subchapter IV.4.6) that land management recommendations alone are not suffi cient to help the small-scale farming communities like Soebatsfontein to improve their livelihood and to make their farm management practices more sustainable. Beyond sound recommendations, other support is needed. Among these are transdisciplinary, participatory action research that aims to solve the socio-economic challenges of the community, access to more farmland for poor farming communities and in the long-run, more capacity development in order to reach more diverse livelihood options.

110 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT 4.1 Introduction [M.T. Hoffman & U. Schmiedel]

The Succulent Karoo Biome covers a wide geographic range extending from southern Namibia in the north into the intermontane valleys of the Little Karoo in the south (Mucina et al. 2006). The focus of this chapter, however, is only on a portion of the Succulent Karoo Biome called Namaqualand (Cowling & Pierce 1999), where the BIOTA Observatories are located and where the work of several BIOTA researchers has been concentrated. Namaqualand itself is about Succulent Karoo 50,000 km2 in extent and comprises about a quarter of the Succulent Karoo Biome. It is located in the extreme northwestern part of South Africa and is quite unique in the southern African context. It is bound- Photo 1: The mountainous granite-derived Kamiesberg uplands are characterised by nar- ed in the north by the Gariep River and in row valleys, which, in the communal areas, are frequently occupied by individual stockposts the south by the Olifants River and incor- such as in this photograph taken near Garies. Photo: Timm Hoffman. porates several distinct bioregions (Des- met 2007). These include the Knersvlakte in the south, the Kamiesberg Mountains in the central parts, and the Richtersveld in the north, with the coastal plain and Bushmanland bordering the region in the west and east respectively. Namaqualand supports a varied topography with elevations up to 1,700 m in places. This topography is underlain by a relatively complex geology, which can change rapidly over short distances. For example, ancient igneous sediments of the Kamiesberg Mountains (Pho- tos 1 & 2) are juxtaposed with the more recent marine-derived sands of the coastal plains in the west, the Kalahari sediments of Bushmanland in the east and the complex assemblage of metamorphic rocks of the Richtersveld to the north. The unique and eroded quartz-dominated basin of the Knersvlakte (Photo 3), which Photo 2: An example of the gently rolling hills characteristic of the sandy coastal plain is situated south of the Kamiesberg, abuts with the foothills of the granite-derived Kamiesberg uplands in the distance. Photo: Pippin the dolerite escarpment to the west while Anderson. the sandstone mountains of the Cape Su- pergroup sediments occur further to the south. This geological complexity has sity of the region (Desmet 2007, Article fl ora thus adding to the diversity of the resulted in a diversity of soil textures, III.3.8, see also Subchapter IV.4.2). Ad- region (Esler & Cowling 1995, Rahlao et depths, and salinity contents (Francis ditional infl uences from burrowing ani- al. 2008). et al. 2007), which is quite unlike any mals and especially from termites adds to The climate regime in Namaqualand other region explored along the BIOTA this pedological diversity. These termite is decidedly less complex and although South transect (Petersen et al. 2010, Ar- mounds or ‘heuweltijes’ are a distinc- annual rainfall can vary from less than ticle III.3.3) and is thought to have an tive feature of the Succulent Karoo Bi- 50 mm in the Richtersveld to more than important infl uence on the biotic diver- ome landscape and support a distinctive 300 mm in the Kamiesberg uplands it

SUCCULENT KAROO 111 thousand individuals were still present in Namaqualand when early European travellers to the region encountered them settled amongst the western foothills of the Kamiesberg in the late 17th century (Valentyn 1971). They were highly mo- bile in the landscape and traversed large areas with their herds (Webley 2007). However, a series of smallpox epidem- ics, which spread from European settlements in the Cape during the 18th century decimated the populations of indigenous people in the region who also lost much of their wealth to the combined impact Succulent Karoo of cattle and land dispossession. Settler expansion from the late 18th century brought agriculture to Namaqualand and soils were cultivated for subsistence grain production for the fi rst time (Hoff- man & Rohde 2007). Photo 3: The plains of the southern Knersvlakte, seen from the escarpment in the east. The famous quartz fi elds of the Knersvlakte are situated further to the north and west and not Fearing complete dispossession of visible on this picture. Photo: Ute Schmiedel. their land by European settler farmers, Nama-speaking pastoralists approached the church for protection in the early falls largely in the winter months from quartzite mountain complexes or quartz 19th century and secured a number of May to September particularly in the pebble, lag-gravel plains such as the key pastoral areas for their exclusive west. The variability in annual rainfall, as Knersvlakte (Desmet 2007). While there use. These relatively large tracts of land determined by the coeffi cient of variation have been some recent advances in our form the basis for the communal areas (CV), is also low (CV < 35%) compared understanding of the exceptional diver- of Namaqualand today (Rohde & Hoff- to most other parts of southern Africa sity of this region much remains to be man 2008) which in 2001 comprised where summer rainfall regimes predomi- learnt, particularly the role that quartz en- about 37% of the area of Namaqualand nate. Temperatures in Namaqualand are vironments might have played in promot- (Desmet 2007). They have been added also relatively benign (average sum- ing the recent diversifi cation of succulent to considerably under the South African mer maximum temperature of < 30ºC, plant lineages. Our basic understanding government’s post-1994 land reform pro- Desmet 2007) compared to many other of insect diversifi cation in the region is gramme (May & Lahiff 2007) and have deserts of the world. also in its infancy. recently been incorporated into the lo- This combination of high soil variabil- The archaeological and pre-colonial cal municipality’s governance structures ity together with the relatively predict- history of human occupation of the re- which are today responsible for the man- able winter rainfall and benign tempera- gion is equally interesting and quite agement and well-being of people who ture regime is thought to be the reason unique to southern Africa. Although live in these areas (see also Subchapter for the high biotic diversity in the region. Namaqualand supports a relatively small IV.4.5). Namaqualand is promoted as the ‘richest population of only about 60,000 people, Despite these recent additions to the desert in the world’ (Myers 2003) and today the region has been occupied by land holdings of communal area farmers contains exceptional levels of diversity humans for millennia. Archaeological as well as the improvement in village in- and endemism within a number of plant, (Webley 2007) and historical (Raper & frastructure (e.g. through the installation insect, and reptile families. Furthermore, Boucher 1988) evidence of modern hu- of electricity, water and sanitation) un- radiation within some groups, such as mans who lived a hunter-gatherer life- employment and consequently, poverty the largely succulent plant family, the style is found throughout Namaqualand levels, remain high (Rohde et al. 2003). Aizoaceae, has occurred relatively re- and attests to the long history of people in This has not been helped by the recent cently (i.e. in the last 5 million years) the region which continued into the early collapse of the mining sector in the re- and is coincident with the onset of the colonial period. However, a fundamental gion. While the substantial increase in winter rainfall regime in southern Africa shift in landuse practices occurred about welfare grants from the state over the (Klak et al. 2004). This diversity is not 2,000 years B.P. when Khoekhoen pas- last 10 years has helped to support the uniformly distributed across Namaqua- toralists are thought to have entered the region’s poorest inhabitants there are a land, however, and appears concentrated region from the north. Small groups of number of key challenges for the region in regional centres, often associated with Nama-speaking herders of perhaps a few today which are centred on improving the

112 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT well-being of a large number of previously-marginalised people living primarily in the communal areas of Namaqualand. Interventions, which facilitate greater access to land for interested farmers and which expand people’s livelihood opportunities beyond agriculture are needed. Although the conservation sector holds some promise in this regard (James et al. 2007) the region is far from the major tourist centres of southern Africa and sustainable job opportunities in conservation will take some time to develop. Furthermore, it is likely that most of the land in Namaqualand will continue Succulent Karoo to be used for small stock production for several decades. It is here that the greatest challenge for the region lies. A fundamental question faced by those who use the land is how to improve economic op- Photo 4: The spectacular fl ower display of the Namaqualand attracts many tourists portunities for livestock farmers without every year. The photo shows a site near Kamieskroon in the year 2006. Photo: Ute compromising the natural ecosystems and Schmiedel. the services they provide in the process (see for instance Article III.5.7). Several studies have shown that high stocking environments in southern Africa it will be however, contribute to the attractiveness rates transform the lowland environ- the Namaqualand environment that will of Namaqualand for tourists, by support- ments in particular, which change from be most affected by the impact of climate ing a spectacular display of mass-fl ow- a diverse mix of palatable and semi-pal- change in the 21st century (Midgley & ering annuals during springtime (Photo atable, perennial shrubs to a relatively Thuiller 2007, Article III.3.1). Therefore, 4), which attracts thousands of tourists to uniform mix of unpalatable perennial understanding the basic biology of the the area every year. The economic poten- shrubs and annual herbaceous species, region in response to climate change, as tial of the landscape and the spring dis- which erupt in abundance in good rainfall well as the critical management problems play for the small rural settlements has years but which remain largely dormant faced by landusers requires on-going in- not been fully explored and may provide under drought conditions (Anderson & vestigation. one of the options for alternative liveli- Hoffman 2007, Todd & Hoffman 1999, In the following subchapters, we out- hoods. In order to manage and maintain 2009). This has signifi cant implications line the fi ndings of ecological and so- the fl ower display at sites and to restore for the livestock industry, particularly cio-economic research activities within the natural vegetation of the old lands at for resource-poor farmers who are un- BIOTA in the Namaqualand region of others, the successional trajectories and able to purchase emergency fodder to the Succulent Karoo that are critical for the processes behind them need to be sustain their animals over the drought land management decisions. In Subchap- understood. Processes that support or period. Understanding how best to man- ter IV.4.2 we describe the role of the soil hamper the restoration of rangelands that age rangelands given the diverse tenure patterns in the Succulent Karoo as drivers have been transformed through overgraz- regimes and biological environments of for the plant diversity at different spatial ing, trampling, or mechanical disturbance Namaqualand remains a key challenge for scales. The understanding and protection of soil structure by ploughing or infra- researchers in the region. Increasing mo- of the resulting spatial heterogeneity is structure installation are described and bility is one option (Cousins et al. 2007, one important criterion for land manage- discussed in Subchapter IV.4.4. Insights Samuels et al. 2008) although changes in ment that aims to maintain the exception- into ecological processes that drive the this regard remain to be implemented. In ally high species richness in the land- succession of disturbed lands as provided addition, many local environments have scape. Disruptive landuse practices such by the Subchapters IV.4.3 and IV.4.4 are already been transformed and rangeland as repeated ploughing has a homogenis- important for future management deci- restoration has become increasingly im- ing effect on the soils. Subchapter IV.4.3 sions and for the successful restoration portant in recent years (e.g. Gabriels et al. shows that it takes more than eight dec- of transformed rangelands. This kind of 2003, Simons & Allsopp 2007, Subchap- ades for the vegetation of old fi elds to re- information is required by the local land- ter IV.4.4) although the fi nancial and lo- cover and to reach a composition of spe- users and conservationists. gistic costs need to be properly evaluated. cies and plant life forms that is similar to The challenges that managers of com- Finally, it has been predicted that of all the the pre-ploughing condition. Old lands, munal lands in Namaqualand have to face

SUCCULENT KAROO 113 landscape and also what role the plants themselves play regarding the maintenance of patchiness in the landscape.

Drivers of vegetation diversity at the landscape scale The studies along the BIOTA Southern African transect revealed a high heterogeneity of soils in the Succulent Karoo. The variance of soil chemical parameters of the BIOTA Observatories in the Suc- culent Karoo (i.e. S21–S28; 25 samples per 1 km²) by far exceeded those of any other BIOTA Observatory in southern Succulent Karoo Africa (Petersen 2008, Article III.3.3). Due to the high diversity of soil features, which has been corroborated by many studies (see Cowling & Hilton-Taylor 1999, Watkeys 1999, Francis et al. 2007), Photo 5: Heuweltjies in the landscape near Soebatsfontein. Photo: Jürgen Deckert. the Succulent Karoo has been identifi ed as a centre of pedodiversity (i.e. centre of soil diversity, Petersen 2008). Within are described in Subchapter IV.4.5 by drivers. The climatic conditions during the 15,000 hectare commonage of Soe- using the example of the Soebatsfontein the main growing season in winter and batsfontein in the Namaqualand lowlands community in the Lowland Succulent Ka- spring, which are characterised by mild (around the Soebatsfontein Observatory, roo. Over nine years, BIOTA researchers temperatures and low but highly predict- S22), eight different soil reference groups worked closely with members of the Soe- able winter rainfall (Cowling et al. 1999), (according to the World Reference Base batsfontein community and on the com- are often referred to as an important for Soil Resources 2006) were identifi ed. munal lands, in order to better understand driver of the species richness. Klak et al. This represents 25% of the 32 defi ned the ecosystems, their processes and dy- (2004) identifi ed the onset of the winter soil reference groups worldwide and namics, the ecosystem services that they rainfall regime approximately 5 mya as 66% of the 12 groups identifi ed for the provide as well as the frame conditions the main trigger of the diversifi cation of entire BIOTA Southern Africa transect for management decisions. In Subchapter the Aizoaceae, one of the most species- (Article III.3.3). Geology, topography, IV.4.5 we summarise our interdiscipli- rich and dominant plant families in the and biological activity (e.g. heuweltjies) nary fi ndings and discuss the emerging Succulent Karoo. are the main factors responsible for the consequences for landuse management. Besides climatic factors, geologi- high variability in these soils (Francis et Finally, we summarise the conclusions cal and soil diversity have often been al. 2007, Petersen 2008). resulting from our research and suggest discussed as other important drivers This high pedodiversity corresponds future needs for research and for the im- of species richness at various spatial well with the very high species richness plementation of sustainable land manage- scales in the Succulent Karoo (Mucina of vascular plants in the Succulent Karoo ment practices in the Succulent Karoo. et al. 2006, Desmet 2007, Francis et (Mucina et al. 2006, Desmet 2007, Arti- al. 2007). Based on the interdiscipli- cle III.3.8). The high diversity of vegeta- nary assessment of the nested plots on tion patterns in the Succulent Karoo are 4.2 Interdependence the BIOTA Observatories (Subchapter closely related to patterns of soil types. In of soils and vascular II.1.2), BIOTA researchers analysed the particular, soil water conditions, salinity, plant vegetation in the interdependence of soils and vegetation and soil acidity are the main drivers of Succulent Karoo patterns in the Succulent Karoo, at the the distribution of the various vegetation [U. Schmiedel, N. Lutsch, D.H. Haarmey- landscape- (km²), habitat- (hectare), and units occurring in the landscape around er, I.U. Röwer, A. Gröngröft, J. Luther- micro-scales (< 1 m²). In this article we Soebatsfontein (Part II, Observatory S22; Mosebach & A. Petersen] summarise our fi ndings regarding the Luther-Mosebach 2009). Plains with spatial patterns of three characteristic deep coarse sandy soils, for instance, Introduction features in the Succulent Karoo land- provide relatively favourable soil water The extraordinarily high biodiversity scape, namely heuweltjies (zoogenic conditions due to the restricted capillary of the Succulent Karoo is the subject of earth mounds), quartz fi elds and bio- rise potential that prevents evaporation studies from various disciplines and has logical soil crusts. We discuss how these of soil water stored in deeper soil layers. been ascribed to several environmental patterns affect the plant diversity in the The Stipagrostis ciliata-Othonna sedi-

114 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT folia unit that is typical of these sandy soils is dominated by the palatable perennial grass Stipagrostis ciliata (Lang- beenboesmangras). Another habitat type, which provides relatively favourable soil water conditions occurs between large rocky outcrops, which are typical for the gneissic mountains of Namaqualand. These soils receive additional run-off water from the rock faces and possibly also additional nitrogen provided from nitrogen-fi xing cyanobacteria and cyanobacterial lichens covering the rocks (Dojani et al. 2007). These mountains and rocky areas are inhabited by the Rhus incisa- Succulent Karoo Rhus undulata unit, which consists of a very dense cover of tall shrubs and trees (i.e. chamaephytes and phanerophytes), such as Rhus species (Taaibos), which are Fig. 1: Schematic representation of the heuweltjie transect in Soebatsfontein showing typi- accompanied by various annuals during cal soil units and their position. Durinodes and Duripans are highlighted in reddish colour. spring time. Other communities, which (A/B/C represents the main soil horizons, suffi x k = accumulation of calcium-carbonates; are typical of the loamy-sandy soils of the qc = accumulation of silica concretions; qm = massive duripan). valleys and gentle slopes in the area are often dominated by Lebeckia multifl ora (Fluitjiesbos) shrubs and are thus part of Table 1: Analyses of a microtransect with fi ve profi les across a heuweltjie structure at the BIOTA Observatory S22 (results are means of all horizons in each profi le) the Lebeckia multifl ora-Galenia sarco- phylla unit. These plant communities in the low-lying areas are driven by the Median strong gradients of soil salinity and pH Parameter Unit Matrix Slope Centre Slope Matrix of region that are typical of the Succulent Karoo distance m 18 8 0 –8 –20 (Petersen 2008, Herpel 2008, Labitzky Clay % 11.7 11.2 9.7 10.8 12.2 n.a. 2009). Silt % 22.5 27.0 30.1 30.9 17.8 n.a. Besides topography and geology, the Sand % 65.7 61.8 58.9 58.4 69.7 n.a.

Succulent Karoo possesses additional pH (CaCl2) pH 5.6 7.7 8.1 8.1 5.1 6.0 characteristic features that increase soil CaCO3 % 0.0 0.2 1.5 3.2 0.0 < 0.1 heterogeneity. Heuweltjies (Afrikaans: ECe mS/cm 17.5 7.4 40.0 26.6 1.2 2.3 small hills), for instance, are dominant Ca g/kg 3.2 5.6 20.7 14.1 3.9 4.5 structures (Photo 5) that contribute to the Mg g/kg 3.2 6.6 10.3 12.3 2.2 1.7 habitat heterogeneity of the landscape Crusts — Duripan Durinodes DurinodesDurinodes Duripan — forming a mosaic of distinct circular features on slopes and plains. They are fossil termitaria, which form earth mounds 0.5–3.0 m in height and 15–50 m in di- may be sustained by rodent burrowing scale habitat diversity and therefore phy- ameter. They occupy 14–25% of the land activity. Their old age means that they todiversity (Knight et al. 1989, Milton surface of the western Succulent Karoo have persisted through times of funda- & Dean 1990, Francis et al. 2007). In in the south western part of southern Af- mental climate change (Turner 2003). particular, the accumulation of salts and rica (Ellis 2002, Petersen 2008, Picker Heuweltjies studied at Soebatsfontein calcium carbonate generates relatively et al. 2007). The age and origin of heu- (in the surroundings of Observatory S22) high pH values in the soil proﬁ les and weltjies remains a matter of contention, by an interdisciplinary team of BIOTA topsoil horizons of the heuweltjies. Thus, although most authors agree that they researchers, exhibit high variance in soil heuweltjies form a pattern of moderately are fossil termitaria of the species Micro- properties that change within a distance alkaline patches in landscapes with other- hodotermes viator (Lovegrove 1991, of less than 10 m (Table 1; for further wise carbonate free, neutral to moderately Picker et al. 2007, Petersen 2008) dating results see Petersen 2008, Herpel 2008, acidic soils. Nitrogen and total organic back more than 4,000 years (Moore & Röwer 2009). carbon contents were also highest at the Picker 1991) to more than 20,000 years The physical and chemical soil prop- centre of these heuweltjies. The reasons ago (Midgley et al. 2002), and that they erties of the heuweltjies enhance small- for the nutrient enrichment at the centre of

SUCCULENT KAROO 115 Table 2: Differences in vegetation parameters between heuweltjie zones and other pooled categories (ANOVA)

Surrounding matrix Fringe Heuweltjie centre Zones vegetation zone vegetation vegetation zone Transformation Parameter Mean SDMean SD Mean SD p Type p Soil parameters pH 5.97a ± 0.856.87b ± 0.89 8.10c ± 0.28 < 0.001 Rank < 0.001 EC [μS/cm] 987 ± 666 880 ± 678 685 ± 947 0.392 a a b CaCO3 [%] 0.04 ± 0.050.31 ± 0.84 3.00 ± 2.38 < 0.001 Log < 0.001 TOC [%] 0.87a,b ± 0.380.71a ± 0.24 1.06b ± 0.51 < 0.001 Log < 0.001 N [%] 0.07a ± 0.030.06a ± 0.02 0.10b ± 0.03 < 0.001 C/N ratio 13.3a ± 1.6112.2b ± 2.75 13.7a ± 1.73 0.002 Vegetation parameters Species richness [per 100 m2] 23.7a ± 5.9720.3b ± 5.70 15.1c ± 6.37 < 0.001* Succulent Karoo Shannon index 1.702 ± 0.514 1.519 ± 0.468 1.288 ± 0.503 0.008 Evenness 0.539 ± 0.148 0.508 ± 0.139 0.491 ± 0.169 0.501 Simpson index 0.686 ± 0.191 0.646 ± 0.167 0.581 ± 0.198 0.093 Average palatability 27 ± 12 25 ± 20 23 ± 14 0.724 Cover annuals [%] 28.7a ± 11.8926.6a ± 11.88 20.2b ± 9.32 0.013 Log 0.013 Vegetation cover [%] 23.7a ± 5.9720.3b ± 5.70 15.1c ± 6.37 < 0.001* Cover life form [%] Chamaephytes 0.9 a ± 2.02 0.68 ± 1.33 4.86 ± 6.47 < 0.001* Log < 0.001 Geophytes 25.2a ± 12.5424.2a ± 12.32 14.7b ± 10.79 < 0.001 Hemicryptophytes 0.1 ± 0.10 0.11 ± 0.26 0.17 ± 0.59 0.75 Phanerophytes 0.7a ± 1.760.2b ± 0.44 0.0b ± 0.06 0.009 Log 0.007 Structural parameters Cover fine material [%] 95.7 ± 5.84 97.6 ± 3.81 99.1 ± 2.38 0.011* Cover biotic crust [%] 58.9 ± 30.98 60.8 ± 32.22 56.2 ± 32.68 0.811 Cover dead wood [%] 2.2 ± 2.00 2.1 ± 2.06 3.7 ± 3.45 0.009 Rank 0.073 Cover soft litter [%] 2.3 ± 3.14 2.6 ± 3.23 1.3 ± 1.74 0.15 Cover dung [%] 0.06 ± 0.11 0.04 ± 0.09 0.04 ± 0.05 0.562 Cover bioturbation [%] 0.5a ± 1.121.1a ± 2.86 14.8b ± 26.38 < 0.001 Rank < 0.001 Cover sheet erosion [%] 0.8 ± 0.83 1 ± 0.74 0.5 ± 0.68 0.018* Cover rill erosion [%] 0.7a,b ± 0.91 0.8a ± 0.85 0.3b ± 0.55 0.023

Df residual = 118, N = 121 (NMatrix = 23, NFringe = 68, NCentre = 30) Asterisks (*) indicate application of significance level p = 0.01 due to heterogeneity of variances and skewed or non-existent normal distribution. Otherwise significance level was p = 0.05. Significant p-values are bold. Superscript letters (a) and (b) indicate homogenous groups according to Tukey’s HSD at p = 0.05 or p = 0.01 respectively.

heuweltjies is presumed to be of zoogenic Eberlanzia cyathiformis unit, which scribed for the quartz fi elds, which are origin (i.e. accumulation of nutrients by is dominated by the mat-forming leaf- another azonal and unique habitat type termites; Midgley & Musil 1990, Pe- succulent shrub Eberlanzia cyathiformis typical of the Succulent Karoo. The soil tersen 2008) and the continued habitation (Vygie) and the annual Asteraceae Fo- conditions and vegetation differ signifi - of the heuweltjies by burrowing animals. veolina dichotoma (Stinkkruid) (Röwer cantly from surrounding zonal habitats Heuweltjies are dominated by oppor- 2009). These vegetation patches on heu- (Schmiedel & Jürgens 1999, Schmiedel tunistic plant species, which can cope weltjies are embedded in a matrix of low 2002, Haarmeyer et al. 2010). The plots with bioturbation (Esler & Cowling vegetation, which is often dominated by on quartz fi elds in the Knersvlakte have 1995, Röwer 2009, compare also Ta- the dwarf shrub Galenia fruticosa (Porse- lower and more variable soil pH values as ble 2). These species are able to complete lein) and the creeping Cephalophyllum well as higher conductivity than the zon- their life cycles within a short time pe- inaequale (Rankvye). The combination al soils, whereas the latter typically pos- riod (Anderson & Hoffman 2007) and the of the patches and matrix belongs to the sess a higher carbon content (Schmiedel heuweltjie soils are able to satisfy their Zygophyllum cordifolium-Cephalophyl- 2002, Haarmeyer et al. 2010). These dif- relatively high nutrient requirements lum inaequale unit. ferences in soil features between azonal (Knight et al. 1989). In the Lowland Suc- Similarly strong gradients in soil quartz fi elds and zonal soils drive species culent Karoo, for instance, they are often salinity and soil pH that occur at dif- turnover in the landscape and contribute covered by the Foveolina dichotomoa- ferent spatial scales have also been de- signifi cantly to the local endemism,

116 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT which is characteristic of the Succulent erogeneity are the above-mentioned heu- by continuous disturbance caused by Karoo. In 2007, we compared species weltjies. the burrowing activities of small mam- numbers and the abundance of individual Soils of quartz fi elds do not only differ mals (bioturbation) and—depending on plants for plots (100 subplots of 400 cm² from surrounding zonal soils as described topography and osmotic potential—also nested within a 20 m x 50 m plot) inside above, and abrupt changes in chemical poor water availability on heuweltjies (N = 27) and outside (N = 24) of quartz and physical soil features are also found (Francis et al. 2007, Petersen 2008). fi elds in the Knersvlakte (around Ob- between different types of quartz fi elds, However, although the alpha diversity servatories S26–S28). Although less in- resulting in a high turnover of quartz-fi eld (i.e. diversity within a plant community) dividuals and species were recorded for plant communities (Schmiedel & Jürgens is low, heuweltjies contribute strongly the quartz fi eld plots (quartz: 186 ± 116; 1999, Schmiedel & Mucina 2006). Gra- to the habitat diversity and thus to beta zonal soil: 330 ± 197), they possessed dients of soil salinity within quartz fi elds diversity (species turnover between plant higher numbers of endemic individu- in the Knersvlakte were negatively relat- communities) in the landscape. Although als (quartz: 75 ± 42, zonal soil: 34 ± 12) ed to mean species richness per subplot heuweltjies cover only about 12% of the and species (quartz 8 ± 3; zonal 6 ± 2) (400 cm²) and plot (1000 m²) as well as land surface in the study area, the vari- (Haarmeyer et al. 2010). This supports to the abundance of endemic species. As ation of soil and structural parameters, Succulent Karoo the account of the southern African salinity intensifi es the effects of drought diversity, and vegetation was about the quartz fi eld fl ora by Schmiedel (2004), by increasing the osmotic pressure in the same magnitude as found for the entire which recorded an extraordinarily high soil (Campbell & Reece 2005), it can be communal land area (Röwer 2009). number of local endemics. Out of the assumed that soil water availability in 67 obligate quartz fi eld taxa, 63 taxa are combination with salinity are the main Plant-soil-interaction at endemic to the Knersvlakte (Schmiedel drivers of the abundance and diversity of the micro-scale 2004). The obligate quartz fi eld fl ora plant species on quartz (Schmiedel 2002, Soil patterns at the micro-scale scale thus contributes 40% of the approxi- Haarmeyer 2009). This medium-scale (< 1 m2) are typically caused by local in- mately 150 endemic Knersvlakte taxa soil pattern can easily be destroyed by teractions between plants or animals and recorded by Hilton-Taylor (1994). mechanical disturbance through tram- the soils. In this context, we describe the The described diversity at the land- pling, ripping, or ploughing of the soil. effects of abiotic soil surface features, the scape scale can be explained by interac- Plots on disturbed quartz fi elds in the role of biological soil crusts, as well as the tions between soil features at different Knersvlakte showed signifi cantly lower fertile island effects of vascular plants. spatial scales ranging from the medium- numbers of local endemics compared to a) Infl uence of soil surface features (app. 100 m²) to micro-scale (< 1 m²). In undisturbed plots (Etzold 2006). on plant species composition: Soil sur- the following we focus on typical pat- The soil transect across a heuweltjie on face characteristics on quartz fi elds may terns of such soil features that drive plant a slope in Soebatsfontein (Fig. 1) showed also infl uence species composition at the species composition at different spatial strong changes of soil features from the micro-scale. The density of quartz gravel scales, but that may also be infl uenced by matrix soil to the soil at the centre of a cover on the soil surface (i.e. sparse to plants at the same time, within the Suc- heuweltjie (Petersen 2008). Fig. 1 re- complete cover) in the Knersvlakte, for culent Karoo landscape. veals the high occurrence of durinodes instance, infl uenced species composition (soil fragments cemented by silica) at the of seedlings (Haarmeyer et al. 2010). Medium-scale drivers of centre of the heuweltjie and towards the We found signifi cantly more seedlings vegetation patterns matrix as well as massive duripans (soil on sub-plots (400 m²) that were covered Medium-scale soil patterns (ca. 100 m²) horizons cemented by silica) outside the with more than 66% quartz gravel than are driven by abiotic and biotic factors. heuweltjie. A general trend of higher val- on those with sparser quartz cover. This The abiotic factors are topography, lo- ues of silt, soil pH, calcium carbonate, effect was even more pronounced when cal variations in parent rock material that electrical conductivity, and total nutrient considering only Aizoaceae seedlings, infl uences the content of coarse frag- content is evident at the centre of heu- whereas no differences in the numbers ments, soil depth, soil salinity, acidity weltjies compared to surrounding soils of non-Aizoaceae seedlings could be de- and the nutrient content at a given site. (Table 1). Additionally, soluble salts such tected (Table 3). Small variations in topography result in as sodium chloride (NaCl) accumulate in This fi nding corresponded with the run-off (leaching) or run-on (accumula- the mounds, causing increased electrical generally higher abundance of Aizoaceae tion of water-dissolved ions) of surface conductivity (up to 40 mS/cm) and high seedlings recorded in quartz fi eld plots water and result in remarkable patchiness osmotic pressure, which result in water than in the zonal habitat plots (Haarmey- of soil types which again drives species stress for plants. These soil conditions er et al. 2010) and suggests a strong in- turnover within a short distance. Typical on heuweltjies infl uence the total veg- fl uence of quartz cover on recruitment phenomena of medium scale abiotic pat- etation cover as well as species richness of Aizoaceae seedlings. A reason for terns will be described using the quartz and Shannon diversity indices, which this could be that the soil between the fi eld habitats as an example. The biotic decrease towards the centre of heu- quartz stones is less exposed to solar ra- factors that drive medium scale soil het- weltjies (Table 2). This can be explained diation and is therefore generally cooler

SUCCULENT KAROO 117 Table 3: Numbers of seedlings per 400 cm² sub-plots at low, medium and high quartz cover This was valid for all sites sampled at the densities in the Knersvlakte micro-scale within the Soebatsfontein Observatory although the sample sites

Low Medium High were located on strongly differing parent Microhabitat (quartz cover) mean ± SD) (mean ± SD) (mean ± SD) p-value materials, ranging from calcareous heu- N = 46 N = 39 N = 30 weltjie substrates to acidic matrix soils Number of seedlings 0.46 ± 0.57ab 0.34 ± 0.58a 0.87 ± 0.98b 0.013 on the slopes. The underlying process of Number of Aizoaceae this phenomenon is assumed to be bio- seedlings 0.26 ± 0.32a 0.21± 0.39a 0.74 ± 0.85b < 0.001 alkalisation triggered by photosynthetic Number of non-Aizoaceae seedlings 0.18 ± 0.40 0.13 ± 0.32 0.13 ± 0.31 0.642 activity of the organisms within the crusts

Different superscript letters indicate significant differences among levels. (Büdel et al. 2004). The comparison of biological soil crusts and non-crusted open soils also showed higher silt content within the ﬁ rst Succulent Karoo Table 4: Differences in soil properties between “open soil” and the soil under plant canopies few centimetres of soils with crusts and a (“plant canopy soil”) for three topsoil layers corresponding increase in the content of particular elements in the soils (e.g. Alu- minium, Zinc, and Manganese). Zhang 0–1 cm 1–5 cm 5–10 cm Parameter N mean ± SD N mean ± SD N mean ± SD et al. (2006) ascribed this to the sticky surfaces of the biological soil crusts pHH2O 50 0.37 * ± 0.54 50 0.40 * ± 0.48 49 0.70 * ± 0.57 that capture silty dusts, which leads to pHCaCl 50 0.51 * ± 0.51 50 0.56 * ± 0.72 49 0.65 * ± 0.82 the enrichment and patchy distribution Cinorg [%] 20 –0.18 * ± 0.12 20 –0.16 * ± 0.13 20 –0.14 * ± 0.16

Corg [%] 50 1.00 * ± 2.06 50 0.02 ± 0.58 49 –0.12 ± 0.64 of correlated total elements. Biological

Nt [%] 50 0.047 * ± 0.070 50 0.005 ± 0.029 49 –0.008 ± 0.030 soil crusts also reduced water inﬁ ltra- C/N-ratio 50 0.96 ± 3.8 50 –0.7 ± 1.7 49 –0.84 ± 2.5 tion rates into the soils at Soebatsfontein.

Pdl [g/kg] 19 –0.07 ± 0.09 – – Soils covered by biological soil crusts Kdl [g/kg] 19 0.22 * ± 0.31 – – were always among those that had the The difference between means after the dataset was standardised are presented. Negative values lowest infi ltration rates and, in most cas- indicate lower contents for “plant canopy soil” sites compared to “open soil” sites. * = significant es, exhibited lower infi ltration rates than differences between means; pHH2O = pH measured in water; pHCaCl = pH measured in 0.01 M CaCl2- solution; Cinorg = inorganic carbon; Corg = organic carbon; Nt = total nitrogen; Pdl = double lactate adjacent open soil sites (Herpel 2008). extractable (plant available) phosphorus; Kdl = double lactate extractable (plant available) potassium (Herpel 2008) c) Effects of vascular plants on soil features: Our study at Soebatsfontein also showed that leaf-succulent dwarf shrubs have small-scale effects on their (Schmiedel & Jürgens 2004) and moister b) Infl uence of biological soil crusts abiotic environment by creating fertil- (C. Musil, pers. comm.) than soils on soil features: Biological soil crusts ity islands underneath their canopies. without quartz cover. As water uptake (i.e. cyanobacteria, algae, lichens, bryo- We found that the topsoil below plant and moderate daily maximum tempera- phytes, see Büdel et al. 2009, Article crowns relative to adjacent open soil tures are essential for seedling survival III.3.4, Subchapter IV.4.5) can also infl u- was enriched by organic matter, had in- of succulents that grow close to the soil ence their environment in various ways. creased pH-values and higher coarse soil surface (Nobel 1984), the quartz habitat For example, they promote the accumu- fraction contents, and was also partly seems to better fulfi l germination relation of organic carbon (Evans & Lange enriched by nutrients (Table 4). The un- quirements for Aizoaceae seedlings than 2001) and nitrogen (Belnap 2002b), sta- derlying processes of the fertility islands, the zonal soils without quartz cover. The bilise the soil surface, and reduce water where litter fall leads to the accumula- importance of quartz cover for Aizoaceae infi ltration into the soil (Belnap 2006, tion of organic matter below the crowns species found by Haarmeyer et al. (2010) Warren 2001). We studied soil patterns of shrubs, is well described by Hook et is in line with a study on disturbed quartz with regard to these micro-scale features al. (1991). The redistribution of topsoil fi elds (due to installation of water pipe- (Herpel 2008). material by wind and water as well as the lines) in the Knersvlakte. Etzold (2006) Biological soil crusts only occur on trapping of fi ne material below canopies compared species richness per plant fam- the upper few millimetres of soils. None- increase this effect. In addition, nutri- ily inside and outside the disturbed quartz theless, our scientifi c soil study in the ents taken up by roots from inter-canopy fi elds and found approximately 20% less Lowland Succulent Karoo (Observatory areas are accumulated in plant tissues Aizoaceae species on the disturbed quartz S22) revealed signifi cant increases in (e.g. leaves) and deposited again as lit- fi elds, while the number of (mainly non- soil pH values by up to 0.5 pH-units in ter beneath the plants. Animals seeking succulent) Asteraceae species increased three different soil depth layers (0–1 cm, shelter under plants contribute to the by approximately 100%. 1–5 cm, and 5–10 cm) (Herpel 2008). formation of fertility islands by leav-

118 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT ing behind excrement and unconsumed food components. This fertile island effect, which also seems to depend on the growth-form and average lifespan of the plant species (Stock et al. 1999), results in horizontal patchiness of soil fertility and structure at the micro-scale.

Conclusions Soil patterns in the Succulent Karoo occur at various spatial scales and are caused by different processes. Patterns already occur at the micro-scale where they arise from mutual interactions between abiotic soil surface features, biological soil crusts, Succulent Karoo and vascular plants. The occurrence of biological soil crusts, vascular plants, and animal activity depends on the soil type Fig. 2: Diagrammatic representation of micro-features in an idealised dryland landscape. but these factors, in turn, also change a) dwarf-shrub, b) larger shrubs, c) tussock grasses, d) open soil patch, e) mineral soil the soils conditions by accumulating or- crust, f) biological soil crust, g) pebble covered areas, h) disturbance spot caused by ganic material and nutrients, creating digging animals, and i) small depressions or channels. Source: Herpel 2008. micro-disturbances by digging, and alter- ing soil surface structure (Fig. 2). These micro-scale patterns of soil heterogene- tion. In particular, our studies on quartz provide fodder for livestock during dry ity again are nested within the medium- fi elds in the Knersvlakte showed that lo- periods. Some farmers, especially in the scale pattern of soil types that is mainly cal and habitat endemics seem to be most higher rainfall parts of Namaqualand, driven by topography (e.g. run-on, run-off strongly affected by such homogenising later started clearing large tracts of land of surface water), geology, and the heu- impacts of trampling and ripping of soil for commercial crop production. weltjies of zoogenic origin. The resulting surface (Etzold 2006, Haarmeyer et al. The normal farming practice in the steep gradients in the chemical and physi- 2010). Protection of small- and medium- region is to cultivate a fi eld for one year cal features of the soils are characteristic scale heterogeneity of soil patterns should and then to let it lie fallow for a year of the Succulent Karoo. The mosaic of be the highest priority of management or two before cultivating it again. These medium-scale patterns accumulate at the measures in order to maintain the extra- fi elds lying fallow between cultivation macro-scale, contributing to the land- ordinarily high soil and plant diversity of often produce the breathtaking mass scape diversity of the area, which has a the Succulent Karoo landscape. Once de- displays of spring fl owers for which strong infl uence on the biodiversity of the stroyed, the ecosystem may take decades Namaqualand has become world re- Succulent Karoo. The predominance of to recover (Subchapters IV.4.3, IV.4.4). nowned (van Rooyen 1999). The spring- gentle rain showers in the Succulent Ka- time fl oral spectacle draws thousands roo (Desmet 2007), which are typical of of tourists annually and is a valuable the winter rainfall regime, have less of a 4.3 Old fi eld succession source of income to the region (James leaching or erosive effect on the soils than in the Namaqua National et al. 2007). In 1988, the South African summer rainfall events and certainly also Park, Succulent Karoo, Nature Foundation (later incorporated contribute to the maintenance of these South Africa into WWF-South Africa) bought the small-scale patterns. [M.W. van Rooyen, R. Henstock & 930 ha farm Skilpad as a wildfl ower re- In summary, micro- and medium-scale H. van der Merwe] serve as the farm was well-known for features in the Succulent Karoo play a its springtime fl ower displays on the crucial role in the maintenance of bio- Introduction old fi elds. When the reserve was later diversity at the landscape scale. Unsus- Private land ownership in Namaqualand donated to SANParks to become part of tainable land management may have a began during the late 19th century and the Namaqua National Park, it was on negative effect on these patterns: Severe brought an end to the nomadic lifestyle condition that the old fi eld vegetation trampling, ploughing, or mechanical of both the KhoiKhoi and European would still be managed to provide annu- destruction due to infrastructure devel- ‘Trekboere’. Permanent human settle- al shows of wild fl owers. Although the opment destroys the small-scale hetero- ment at the same time necessitated the mass effect on the abandoned croplands geneity in soils, impacts biological soil cultivation of crops. Crop cultivation was is created by the dominance of a few crusts (Subchapter IV.4.5) and thus ho- initially primarily to provide food for pioneer species and is associated with a mogenises the plant species composi- daily living, although it was also used to low diversity, SANParks are honouring

SUCCULENT KAROO 119 67 Succulent Karoo

Photos 6–9: The four monitoring sites in 2009, 6) abandoned 18 years previously; 7) abandoned 19 years previously; 8) abandoned approximately 25 years previously; and 9) abandoned approximately 55 years previously. Photo: Noel van Rooyen.

this agreement and a small number of In general, studies on secondary suc- to detect changes in species composition the abandoned fi elds on the Skilpad sec- cession on old fi elds in arid regions are and diversity. When the fi rst surveys were tion of the park are still ploughed from scarce (Otto et al. 2006), and little has conducted in 1994, the times since last time to time to produce mass fl ower been reported for the Succulent Karoo or cultivation for these fi elds were 3, 4, ~10, displays. These disturbed fi elds cover a Namaqualand (van Rooyen 2002, Wit- and ~40 years respectively (Photos 6–9). minute portion of the park and the bulk booi & Esler 2003, Simons & Allsopp A step-point survey of 1,000 points was of the abandoned fi elds are left to re- 2007). The objectives of this study were conducted annually in springtime at each cover naturally. to investigate natural vegetation changes monitoring site to determine the frequen- Despite Namaqualand being a margin- on abandoned fi elds in the Namaqua Na- cy of species. At each point, the nearest al environment for crop production, it is tional Park over time in order to aid fu- annual as well as the nearest perennial estimated that croplands covered nearly ture planning and effective management species were recorded, noting which of 30,000 ha in the early 1970s (Hoffman of the many old fi elds scattered through- the two had in fact been the closest. In & Rohde 2007). As a result of increased out this extensive conservation area. this way the annual species could be ana- production costs since then, farmers have Understanding the process of natural lysed separately from the perennial spe- been forced to cultivate fewer fi elds (van vegetation change, and the rate at which cies, but a combined analysis, ignoring der Merwe 2009), and the area under cul- it occurs could also assist in developing life duration, could also be undertaken. tivation has declined by approximately techniques that could enhance restoration The signifi cance of a linear regression of two-thirds. Many abandoned croplands possibilities of old fi elds. species richness against time was ana- lie scattered throughout the Namaqua- lysed in Graphpad Prism 4.03 for Win- land landscape as a consequence. A simi- Methods dows (GraphPad software, San Diego, lar trend of increasing land abandonment California, USA, www.graphpad.com). has been reported worldwide as a result Chronosequence of environmental and socio-economic Vegetation recovery on four abandoned Floristic analysis changes (Cramer & Hobbs 2007, Cramer fi elds, differing in time since last culti- The Braun-Blanquet method of vegetation et al. 2007). vation was examined from 1994 to 2009 sampling was used to conduct a fl oristic

120 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT analysis at 62 sites on the abandoned fi elds in and around the Namaqua National Park 35 during 2006 and 2007. This paper will only 30 report on the results of 31 sampling sites 25 on the Skilpad section of the park as well 20 as on the adjacent farmland. An analysis 15 of the fl oristic data was conducted using the TURBOVEG and MEGATAB com- 10 puter packages (Hennekens & Schamineé 5 2001) and refi ned using Braun-Blanquet Species richness 0 procedures (Werger 1974).

The SYN-TAX computer program (a) 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Years (Podani 2001) was used to ordinate the standardised (natural logarithm stand- 40 Succulent Karoo ardisation) ﬂ oristic data of the relevés 35 and monitoring sites using Principal Co- 30 ordinate Analysis (PCoA) and the Bray 25 Curtis distance measure. 20 Results and discussion 15 10 Chronosequence 5 Species richness 0 Over the past 15 years, total species rich-

ness increased with time since abandon- 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 ment at all four monitoring sites (Fig. 3a). (b) Years This increase was mainly due to the sig- 70 nifi cant increase in perennial species 60 (Fig. 3b). Only at monitoring sites 1 and 50 2, where monitoring was commenced soon after abandonment, was there a sig- 40 nifi cant increase in annual species rich- 30 ness with time. After 16 years of moni- 20 toring, the annual species richness at all 10 four sites were quite similar, but did not Species richness show any signs of decreasing even after 0 > 50 years of abandonment (Fig. 3c). Spe- 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 cies composition of the perennial species Years showed a clear directional trend over the (c) Site 1 Site 2 Site 3 Site 4 monitored years (Fig. 4), whereas annual species composition did not show a simi- Fig. 3: Changes in species richness of (a) annual species (b) perennial species and (c) total lar directional trend. The annual species species at the four monitoring sites from 1994 to 2009. composition as well as richness was dictated by the timing and amount of rainfall in a growing season. drained soils that are less than 300 mm The four sub-communities represented deep. The vegetation of the old fi elds a gradient in time since abandonment, Floristic analysis around the farm Skilpad belonged to a with sub-community 1.1 being in an ear- The natural vegetation in the Skilpad single vegetation community, the Ursinia ly successional stage and sub-community section of the Namaqua National Park cakilefolia old fi eld vegetation, and could 1.4 in a late successional stage. The fi rst falls within the Namaqualand Klipkoppe be subdivided into four sub-communi- sub-community (Leysera gnaphalodes- (SKn1) of Mucina et al. (2006). The area ties. In general, the vegetation was char- Ursinia cakilefolia old fi eld vegetation) lies on the escarpment with the altitude acterised by the presence and high cover represented those fi elds in the Namaqua ranging from 620 to 750 m above sea of Ursinia cakilefolia, Gazania leiopoda, National Park that were still regularly level and the predominant land type in Felicia dubia, Heliophila variabilis, Ley- disturbed to maintain a fl owering display which the croplands were established is sera gnaphalodes, Pentaschistis tomen- for tourists. The second sub-community Ag94 (Agricultural Research Council, tella, Conicosia elongata, and Dimor- (Heliophila variabilis-Ursinia cakile- undated) with red-yellow apedal, freely photheca sinuata. folia old fi eld vegetation) was found

SUCCULENT KAROO 121 Most studies on vegetation succession on abandoned croplands show a sequence in dominance of life forms from annuals, through herbaceous perennials to woody perennials (Debussche et al. 1996, Baz- zaz 2000). An investigation of the contribution of the different life forms to the vegetation cover in this study revealed the same pattern. However, an analysis of species richness did not show the an- ticipated decrease in species richness of the annual species. As time progressed after abandonment of a fi eld, dominance changes occurred in Succulent Karoo the species composition, with the annual species assemblages changing to include fewer species producing showy fl owers. Fig. 4: Directional changes in species composition at monitoring site 1 from 1994 to 2009. Furthermore, the visual mass display effect also depended on the cover of the perennial species present. towards the eastern part of the Skilpad 73 ± 8% in sub-communities 1.2, 1.3, and The spectacular displays of fl owering section. Many of these old fi elds were ac- 1.4 respectively. Conversely, the relative annuals in spring each year is the founda- tively cultivated before the Skilpad Wild- contribution of the annual species to the tion of the tourism industry in Namaqua- fl ower Reserve was established, but have vegetation cover showed a decrease from land (James et al. 2007). Historically, the not been disturbed since then. In general, 57 ± 3% to 46 ± 8% to 46 ± 5% and fi nally croplands on Skilpad were intentionally the old fi elds in this sub-community were to 23 ± 7% from sub-community 1.1 to disturbed to promote good fl ower dis- from 12 to 15 years old. Sub-community 1.4. Furthermore, within the annual spe- plays to attract visitors. Although de- 1.3 (Felicia bergeriana-Ursinia cakilefo- cies it was noted that the contribution of liberate disturbance is not desirable in a lia old fi eld vegetation) was transitional the species producing mass displays de- national park, the results of the current between sub-communities 2 and 4, and creased, whereas the contribution made study confi rm that some disturbance does included sites in the park as well as in the by non-showy annual species increased. seem essential to maintain a spectacular adjacent farmland. Some of the fi elds in The relative contribution made by the fl owering display. The practice of tilling the park were abandoned only ~20 years geophytes remained fairly constant and a few selected fi elds along the circular ago, whereas others had already been ranged from 3% to 7%. tourist route in the Namaqua National abandoned for more than 40 years be- Park approximately every four years will fore the park was proclaimed and were Conclusions therefore have to be continued. Although therefore at least 55 years old at the time Overall, results show that secondary suc- vegetation change is slow on the aban- of the study. In contrast, the fi elds on the cession on the old fi elds in Namaqualand doned croplands, natural recovery does adjacent farmland were estimated to be proceeds very slowly. The vegetation on occur and it is recommended that all oth- only ~15–20 years old at the time of the many old fi elds abandoned more than er abandoned croplands in the park are study. Sub-community 1.4 (Stipagrostis 50 years ago was still clearly distinguish- allowed to recover naturally. zeyheri-Ursinia cakilefolia old fi eld veg- able from the surrounding vegetation, etation) represents those fi elds that had although the vegetation on the fi eld that been cultivated more than 50 to 80 years had not been cultivated for more than 4.4 Restoring degraded ago. Large perennial shrub species such 80 years could hardly be distinguished rangelands in the as Searsia horrida, Wiborgia mucronata, from the natural vegetation. In spite of all Succulent Karoo: lessons Nylandtia spinosa, and Dodonaea vis- sites having a ploughing history in com- learnt from four trials cosa were prominent in this sub-commu- mon, recovery of the vegetation depends [W. Hanke & U. Schmiedel] nity. upon other farming practices on the old The results of the fl oristic analysis fi elds such as grazing regime (Bonet Introduction were used to analyse the contributions of 2004, Cadenasso et al. 2002), number the different life forms towards the veg- of times ploughed, crops planted (Mys- Aim of the study etation cover. The relative contribution ter & Pickett 1990, Bonet 2004), and the In this Subchapter we collate the fi ndings of the perennial species increased from rainfall amount and temporal distribution of four restoration experiments, which 40 ± 3% (mean ± standard error) in sub- in the year of abandonment (Myster & were conducted within the context of the community 1.1 to 47 ± 10%, 48 ± 6% and Pickett 1990, Pickett et al. 2008). BIOTA project on degraded rangelands

122 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT Succulent Karoo 10 11

Photos 10 and 11: Comparison of rangelands the Richtersveld, Succulent Karoo. 10) Brownanthus pseudoschlichtianus-dominated rangeland in good condition and 11) degraded rangeland with sparse vegetation consisting of annual and pioneer species. Photos: Wiebke Hanke.

in the Succulent Karoo. The aim of the ralists were constrained within relatively takes decades for the transformed range- experiments was to identify resource over-crowded, communally-managed lands to recover (Hoffman & Rohde manipulations that are able to restore areas where their livestock movements 2007, Richardson et al. 2007), because functionality of the ecosystem and the were restricted to smaller areas than had poor water supply (Beukes & Cowling vegetation cover. The experiments were historically been the case (Hoffman et 2003, Botha et al. 2008) and poor repre- located at different regions and moni- al. 2007). Inappropriate management sentation of perennial species in the seed tored over several years, which allows an strategies and overstocking have led to a bank (de Villiers et al. 2003) restrict au- assessment of their general applicability signifi cant transformation of these com- togenic vegetation recovery. Active res- for the Succulent Karoo. In this compila- munally-managed areas and a decline toration measures in combination with tion we attempt to answer the following in rangeland quality (Todd & Hoffman improved management strategies may questions: 2009). Under heavy grazing, the diverse induce rehabilitation within shorter and 1. What were the effects of different dwarf shrub communities tend to be re- practically relevant time-scales. treatments on soil and vegetation pa- placed by few, often unpalatable species. rameters? Heavy grazing also results in the replace- Environmental context for restoration 2. Which processes determined the suc- ment of perennial shrubs by ephemeral in the Succulent Karoo cess or failure of the treatments? species (Todd & Hoffman 1999). Since The characteristic features of the Suc- 3. What are the practical implications for the abundance of the ephemerals is high- culent Karoo climate result in specifi c rangeland restoration in the Succulent ly dependent on rainfall, farmers have environmental and ecological processes, Karoo? to deal with large fl uctuations in forage which should be taken into considera- resources (Vetter 2005). Where peren- tion during the development of restora- Rangeland degradation in the nial vegetation has been reduced, bare tion strategies (Carrick & Krüger 2007). Succulent Karoo patches predominate during the dry pe- Most of the area receives winter rainfall The Succulent Karoo Biome is a winter riod (Photos 10 & 11). This affects the of less than 150 mm per annum (Cowling rainfall semi-desert, favouring a succu- soil quality negatively, which is further et al. 1999), but compared to other desert lent dwarf shrub vegetation of excep- compounded by livestock trampling. regions rainfall is reliable and prolonged tionally high diversity (Cowling et al. The rich biological soil crusts (see also droughts rarely occur (Desmet & Cowling 1999, Subchapters IV.4.1, IV.4.2). The Article III.3.4) are disturbed (Rutherford 1999). Precipitation during the growing majority of the area is used for livestock & Powrie 2010, Subchapter IV.4.5) and season usually occurs as a ‘drizzle’ of low farming with sheep and goats and, to a soil stability decreases (Petersen et al. to moderate intensity or in the form of fog lesser extent, also cattle and donkeys 2004). Further, Beukes & Ellis (2003) or dew (Desmet & Cowling 1999) which (Anonymous 1986). With the arrival found that degraded soils are shallower wets the soil only to a shallow depth. The of European settlers the transhumant due to erosion and therefore have a poor- Succulent Karoo differs from most other strategies of the indigenous pastoral- er water holding capacity. (semi-) arid environments also in terms of ists who moved their sheep and goats Rangeland transformation in the Suc- the mild temperatures experienced in the between summer and winter rainfall ar- culent Karoo may disrupt crucial ecologi- growing season during winter and early eas came to an end. Instead, the pasto- cal processes and landscape functions. It spring (Cowling et al. 1999).

SUCCULENT KAROO 123 Succulent Karoo

Photos 12–14: Restoration plots shortly after treatment application in Knersvlakte quartz ﬁ elds disturbed by the pipeline construction. 12) Level- ling (control), 13) levelling and planting of C. spissum individuals, and 14) levelling and scattering of quartz stones. Photos: Sophia Etzold.

Both water and wind erosion and con- They all improved soil properties and of- of the Succulent Karoo: at two farms on sequently deposition also play an impor- fered considerable potential to enhance the Soebatsfontein commonage (Coast- tant role in habitat degradation and re- the re-establishment of vegetation. The al plain) (Meyer 2009), at sites on the covery processes. Water erosion is often studies showed that functional transplants Paulshoek commonage (Namakwaland associated with the heavy thunderstorms and organic mulch should be further in- Klipkoppe) (Simons & Allsopp 2007), and subsequent surface run-off which can vestigated, whereas gypsum did not prove at sites on the Eksteenfontein common- occur in the region during the summer to be cost-effective due to high transport age (Richtersveld) (Hanke et al. 2010, months while the strong winds, which costs. The fi ndings from an experiment by submitted), and on the farm Ratelgat prevail throughout the year in most areas Simons & Allsopp (2007) will be reported ( Knersvlakte) (Etzold 2006, Meyer 2009). can result in signifi cant levels of wind in the present compilation together with In this article we integrate the fi ndings erosion (Desmet & Cowling 1999). the BIOTA experiments, as it is situated in of all treatments applied at more than one the direct vicinity of a BIOTA observatory of these sites: i.e. livestock exclusion, in- Research on rangeland restoration and its monitoring was partly conducted troduction of brushpack, dung or stones, has only just begun within the BIOTA project. soil surface structuring by microcatch- Degradation of (semi-) arid rangelands Further restoration treatments have ments, and the use of plants with a special poses a serious problem worldwide and been tested in informal trials by local functional role. The treatments aimed to their restoration has become an impor- farmers. These experiences should be reverse degradation caused by overstock- tant fi eld of research, which is gaining in captured and incorporated into the resto- ing, except for the site in the Knersvlakte prominence. Various effi cient methods ration research (Botha et al. 2008, see Ar- (Photos 12–14) where disturbance was have been described for practice (e.g. ticle III.8.2). Also, the experiences from caused by the construction of a subterra- Dregne 1992, Coetzee 2005, Bainbridge mine spoil restoration projects (Blignaut nean water pipeline. 2007), but up to now, only a few of them & Milton 2005, Carrick & Krüger 2007) The key data and references to the have been tested under the unique envi- can be included in the development of res- experiments are listed in Table 5. The ronmental conditions of the Succulent toration methods for rangelands. These evaluation at all four sites included the Karoo. Apart from the studies in the studies have shown that fl oristic features response of natural plant recruitment BIOTA context, to our knowledge only like the occurrence of succulence and the with respect to growth form. Further- three types of active rangeland restora- high drought tolerance of many seedlings more, the evaluation of each experiment tion treatments have been scientifi cally present good opportunities for the suc- is characterised by a special focus. The evaluated and published for this biome, cess of species re-introductions. most important site specifi c installation namely: functional plants (i.e. plants with and evaluation characteristics for the ex- a functional role such as facilitation of Description of the BIOTA periments are noted below. other plants or reduction of soil erosion) restoration experiments Soebatsfontein: The evaluation fo- (Anderson et al. 2004, see also Gabriels The restoration experiments were con- cussed on soil chemical parameters. Effects et al. 2003), application of organic mulch ducted in degraded rangelands in the vi- on the total amount of nitrogen and carbon, (Beukes & Cowling 2003), and gypsum cinity of four BIOTA Observatories (see C/N ratio, pH, and electrical conductivity treatment (Beukes & Cowling 2003). Table 5) in the major ecological regions were assessed in mixed soil samples from

124 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT

(100% cover), applied (100% Galenia africana of 6 per treatment of 6 per treatment N Paulshoek (Namakwaland Klipkoppe) Paulshoek (Namakwaland 36 plots with each 3 blocks, Block: (49 m²), seedling survival, Seedling emergence, natural recruitment of plant growth forms 10 cm deep) per plot, applied alone and in combination with brushpacking individuals of Galenia africana with alone and in combination microcatchments 8 1999–2001 1999, revisits monthly Start Succulent Karoo — Already existing, naturally established

of 10 per treatment Block: 10 blocks, 20 m²-sized plots, 20 m²-sized plots, 10 blocks, Block: N Plant growth forms (species richness, cover, abundances), Etzold 2006, Meyer 2009 Simons & Allsopp 2007 (90% cover) rica, during 1999–2009 Farm Ratelgat ( Knersvlakte) Galenia africana Galenia 005, 2008 Start 2004, revisits 2005, 200 — on Pipeline installation Overgrazing on Pipeline Cephalophyllum Cephalophyllum 27) Ratelgat (S27) (S24), Reemhoogte Paulshoek (S25) 20 individuals per plot 20 individuals of 10 per treatment Quartz Zonal Quartz Block: 10 blocks, 3 m²-sized plots, 3 m²-sized plots, 10 blocks, Block: N Plant growth forms, characteristic species of the quartz field, opportunists cover, abundances) (species richness, spissum, in prep. Etzold 2006, Meyer 2009, et al. — Transplants of — — Transplants — — 12–16 triangular pits, (75 x 75 cm, ts annually 2005–2008 Start 2004, revisits 2

. sp atsfontein (Coastal plain) of 10 per treatment combination Split-plot (fenced/unfenced): 2 whole plots, each with subplots of 25 m², N Plant growth forms cover, (species richness, soil properties (pH, chemical abundances), electrical conductivity, C/N ratio, total amount of nitrogen and carbon) Salsola (90% cover) 20 individuals per — 6 stone heaps (25 x 25 cm) per plot cover Scattered quartz stones, 10% — — Brownanthus

of 5 per treatment combination Overgrazing Overgrazing Pipeline installati Eksteenfontein (Richtersveld) Koeroegap Vlakte (S18) Soeb Start 2007, revisit 2008 Overgrazing Pipeline Split-plot (fenced/unfenced): 5 whole plots, each with 10 subplots (100 m²), Soebatsfontein (S22) N Start 2004, revisi Soil hydrological parameters (water content after a rainfall event at 10 cm, water storage at 5 cm depth), plant growth forms (preliminary assessment) Hanke et al. 2010, submitted Ratelgat (S Meyer 2009 ha 5 exclosures, each 1 Brownanthus pseudoschlichtianus (70% cover) Sheep and goat dung ha 2 exclosures, each 0.25 pits (10 cm deep, 25 half-moon-shaped 40 cm wide and 60 long) per plot Transplants of pseudoschlichtianus, with palm Sheep and goat dung fonds on top subplot — — — — ha 3 exclosures, each 0.25 —

Related Observatory frame Time cause Degradation designExperimental variables Response Detailed description Livestock exclusion Brushpacking Dung mulching Microcatchments plants Functional Stone applications Tested treatments

Table 5: Key data of the rangeland restoration experiments conducted in the framework of BIOTA in the Succulent Karoo, South Af in the Succulent Karoo, South 5: Key data of the rangeland restoration experiments conducted in framework BIOTA Table

SUCCULENT KAROO 125 groups, namely: quartz fi eld fl ora and op- portunist fl ora (Meyer 2009, Meyer et al., in prep.).

Ecological concepts and results of the tested treatments

Livestock exclusion Within (semi-) arid summer rainfall ecosystems, rest from grazing may induce recovery of perennial grass cover within less than a decade (Yayneshet et al. 2009, Kaouthar & Chaieb 2010). In the win- Succulent Karoo ter rainfall Succulent Karoo, however, grasses play a minor role and the time Fig. 5: Split-plot design at Soebatsfontein. The experimental site is divided into two plots of which one is fenced, while the other is grazed by livestock. Each plot contains subplots period needed for the recovery of shrub within which the active treatments were randomly allocated (modifi ed from Meyer 2009). vegetation may be longer. According to recent studies from the Succulent Ka- roo, conducted after 40 (Haarmeyer et al. 2010) and 67 years (Rahlao et al. 2008) of resting, vegetation is able to recover to a pre-disturbance stage in the long-term. However, the triggering processes of au- togenic succession, important for management decisions, still remain poorly understood. If livestock is removed, when and how is recovery induced? We therefore included livestock exclusion plots (Fig. 5) at Eksteenfontein and on the two farms at Soebatsfontein (Patrysegat and Quaggasfontein) in the restoration experiments (Table 5). After one year of livestock exclusion in Eksteenfontein, soil water content at a depth of 10 cm was not signifi cantly affect- Photo 15: Increased cover of the annual species Foveolina dichotoma in an exclosure one year after fence establishment in the communal rangelands of the Eksteenfontein village. ed after the winter rainfall event (Hanke et Photo: Wiebke Hanke. al. 2010, submitted). This result was expected considering the short timeframe. Chemical soil parameters were evalu- 1–10 cm depth four years after the treat- cloudy, windy, and cool weather condi- ated four years after fence establishment ments were installed (Meyer 2009). tions). Additionally, soil water storage in the two exclosures at Soebatsfontein. Paulshoek: In addition to the monitor- (at a depth of 5 cm) was monitored over The only parameters signifi cantly affected ing of natural plant recruitment, seeds 11 days after this rainfall event. Vegeta- were carbon and nitrogen. Both total car- and seedlings of palatable key species tion has not yet been statistically evalu- bon and nitrogen contents had increased were introduced as indicators under the ated, but some obvious effects will be by 20% in the exclosures on Patrysegat. different treatments and their emergence presented in a descriptive way (Hanke et The same trend was found at Quaggas- and survival monitored monthly (Simons al., submitted). fontein, but the difference between grazed & Allsopp 2007). Knersvlakte: Prior to treatment appli- and ungrazed plots was not signifi cant. Eksteenfontein: The evaluation fo- cation, the plots were cleared of invasive The mean C/N ratio on the farm Patryseg- cussed on soil hydrological parameters. Atriplex lindleyi subsp. infl ata, and the at was signifi cantly higher in the exclosure The volumetric soil water content (at a soil surface, which was heavily ripped (C/N ratio = 10.5) than outside (C/N ratio depth of 10 cm) was recorded one year from the pipeline construction work, was = 9.93), whereas the exclosure on Quag- after the establishment of the treatment, levelled (Photos 12–14). The evaluation gasfontein showed the opposite trend, just after a typical winter rainfall event of vegetation response included analyses although this was not signifi cant (Meyer (5.2 mm of fi ne drizzle, accompanied by with respect to two different plant species 2009).

126 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT In both experiments, vegetation was affected by the exclusion of livestock and annual plant cover responded immediately to the release from grazing. The benefi t that annual plant cover receives from being rested from grazing even for one year is clearly evident at the Eksteenfontein site (Photo 15). Also in one exclosure at Soebatsfontein (i.e. Quaggasfontein) four years after treatment application, the mean cover of annuals was more than twice as high inside the exclosure (7.84%) compared to outside (3.88%) where grazing continued with high intensity. The cover of perennial plants remained unaffected. Succulent Karoo Species richness of annuals and perenni- Fig. 6: Mean (+1 SE) volumetric soil water content after a rainfall event of 5.2 mm at als increased signifi cantly by 25% inside Eksteenfontein. Asterisks indicate signifi cant differences from the control derived from the exclosure. Compared to Quaggasfon- pairwise a priori comparisons following the analyses using linear models (*p < 0.05; tein, on the farm Patrysegat grazing inten- ***p < 0.001), N = 10; Micro = microcatchment, Brush = brushpack. sity adjacent to the exclosure was only of moderate intensity (Meyer 2009). Here the cover of annuals remained unchanged world (Coetzee 2005, Ludwig & Tong- was covered with branches of indigenous in the exclosure. The cover of perennials way 1996). Brushpacks protect the soil shrub species covering 70–100% of the decreased signifi cantly by 25% and spe- against erosion and solar radiation, and plots (Table 5). cies richness by 27% inside this exclo- also improve soil moisture conditions. Soil moisture measurements shortly sure. Unlike living vegetation, brushpacks after the winter rainfall event in Eksteen- do not compete for resources with other fontein (Fig. 6), showed that volumet- Brushpacking plants, but like them they may generate ric soil water content under brushpacks Brushpacks, which simulate the pro- fertile islands by trapping soil particles (7.8%) was higher than in the control tective effect of the natural plant cov- or organic material and attract soil living plots (4.5%), and even exceeded the er, have been successfully applied in organisms. Brushpacks were applied at amount of water provided by the preced- (semi-)arid environments throughout the all four restoration sites. The soil surface ing rainfall event (5.2 mm). Prolonged

Control -1 Brushpack* Dung * -3 Microcatchment Planting **

-5

-7

-9

Soil water potential (bar) potential water Soil -11

-13

-15

1234567891011 Day after rainfall

Fig. 7: Soil water depletion curves after a rainfall event of 5.2 mm. Data are medians of water potential over 11 days at 5 cm soil depth from gypsum block sensors (GB-1, Delmhorst Instrument Co.). Asterisks in the legend indicate signiﬁ cant differences from the control of treatment by time interaction (*p < 0.05, **p < 0.01), derived from nonparametric repeated measures analyses (Brunner et al. 2002). N = 5.

SUCCULENT KAROO 127 The dung cover had pronounced effects on soil water status, chemical soil properties and vegetation cover. The soil under dung had a lower water content than in the control directly after the rainfall event at Eksteenfontein (Fig. 7). It appears that some of the rainwater was absorbed by the dung before percolating into the soil. However, the 5.2 mm of rainfall was suffi cient to saturate the dung and wet the soil, although soil-wetting was retarded. Once the water reached the soil, it was stored longer than in control plots, indicating the shading effect of the mulch Succulent Karoo (Fig. 7) (Hanke et al. 2010, submitted). The chemical analyses of soil samples from Soebatsfontein revealed that the mean pH value was elevated from pH 6.7 Fig. 8: Means (+SD) of the total number of perennials at fenced and unfenced subplots at Soebatsfontein on the farm Patrysegat in 2005, one year after installation of treatments. in control plots to pH 7.3 under the dung on Different letters above bars indicate signifi cant differences between treatments according to both farms. Mean electrical conductivity LSH with Bonferroni-Holm correction following Kruskal-Wallis tests. was increased on both farms by the dung, especially on Quaggasfontein, where the mean was fi ve times higher under dung water storage compounded this positive annuals remained unaffected (Meyer (237 μS/cm) compared to the control effect on soil water content. While soils 2009). Perennial plant cover was not (52 μS/cm). Dung also signifi cantly in- of the control plots were desiccated to the infl uenced positively at any site. At creased total nitrogen (up to 4 times) and permanent wilting point (–15 bar) at 5 cm Paulshoek, brushpacks neither infl u- carbon content (up to 5 times) compared soil depth within 5 days after the rainfall enced the emergence of the introduced to the control. C/N ratio decreased sig- event, soils under the brushpacks did not seeds, nor was it effective in promoting nifi cantly on one farm (Quaggasfontein), reach the permanent wilting point within the survival of planted seedlings (Si- whereas it was not affected on the other the 11-day measurement period (Fig. 7) mons & Allsopp 2007). On one farm at farm (Patrysegat) (Meyer 2009). (Hanke et al. 2010, submitted). Soebatsfontein, the abundance (Fig. 8) The cover of annuals increased at Ek- Soil parameters signifi cantly affected and species richness of perennial plants steenfontein, especially the invasive spe- by brushpacks in Soebatsfontein were was actually impacted negatively by cies Atriplex spp. (Photo 16), of which soil pH and the total amount of carbon. brushpacks (Meyer 2009). We assume the seeds were brought in by the dung On both farms, the mean soil pH un- that during treatment installation plant (W. Hanke, University of Hamburg, under brushpacks (pH 6.98) was signifi - individuals were damaged through the publ. data). There was a similar trend at cantly higher than for the control plots spreading of the brushpacks. Soebatsfontein, but the increase of an- (pH 6.70). Electrical conductivity values nual cover was only signifi cant during tended to be increased under brushpacks, Dung mulching the high rainfall year in 2006 (175 mm but this trend was not signifi cant. The Positive responses of soil properties and annual rainfall compared to a mean of total amount of carbon was signifi cantly plant cover have been demonstrated after 130 mm from 2001–2008). Species rich- higher (up to 20%) under brushpacks on mulching with different organic materi- ness of perennial and annual plants was one of the two farms in Soebatsfontein als in various (semi-) arid environments negatively impacted by the dung treat- (Patrysegat), refl ecting the input of cel- (Zink & Allen 1998, Querejeta et al. 2000, ment in most of the years at Soebatsfon- lulose. On the other farm (Quaggasfon- van den Berg & Kellner 2005) and also in tein (Meyer 2009). tein), carbon values under brushpacks the Succulent Karoo using thatching reed were also higher, but this trend was not (Beukes & Cowling 2003). We applied Microcatchments signifi cant (Meyer 2009). mulching with dung in the restoration tri- Soil surface structuring by microcatch- The cover of ephemeral plants in the als at Soebatsfontein and Eksteenfontein ments is widely and successfully ap- Knersvlakte one year after the installa- (Table 5). Old sheep and goat dung from plied for water and soil conservation tion of brushpacks was twice as high as nearby stock post was spread onto the throughout Africa (Critchley et al. 1994). in control plots (Meyer 2009). The cov- plots, so that 70–90% of the soil surface Microcatchments were applied at the er of ephemerals was also increased by was covered by a thin layer. In Soebats- experimental sites at Paulshoek and Ek- brushpacking at Paulshoek (Simons & fontein the dung cover was stabilised us- steenfontein (Table 5) in order to test Allsopp 2007), while in Soebatsfontein ing locally available palm fronds. whether they would improve the local

128 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT accumulation of resources such as seeds, organic material and water, and thus enhance plant recruitment. Pits of 10 cm depth were dug at both sites, perpendicu- lar to the general direction of water runoff. The pits were triangular (75 cm x 75 cm x 75 cm) at the Paulshoek sites, whereas they were halfmoon-shaped (40 cm wide and 60 cm long) at Eksteenfontein. Half of the microcatchments were combined with brushpacks at Paulshoek. Soil water content in the pits shortly after the rainfall event was not signifi cantly different from the control plots at Eksteen- fontein (Fig. 6). The intensity of the typi- Succulent Karoo cal winter rainfall event was apparently too low for the development of runoff. Soil water storage was not improved either Photo 16: Increased cover of the annual Atriplex semibaccata in a dung-treated plot at (Fig. 7) (Hanke et al. 2010, submitted). Eksteenfontein. Photo: Wiebke Hanke. Microcatchments also did not improve natural recruitment or the establishment of perennial seedlings and resulted in only schlichtianus at Eksteenfontein and Ce- phalophyllum spissum transplant plots. a minor increase in the cover of ephemer- phalophyllum spissum on quartz fi elds Among these were opportunists as well als at Paulshoek. Controls showed better in the Knersvlakte (Table 5). At the as typical species of undisturbed quartz seedling survival than microcatchments. Paulshoek study site naturally established fi eld vegetation. The planted seedlings The poor seedling survival in microcatch- Galenia africana shrubs were tested for underneath the canopies of naturally ments may be attributed to the removal their potential as nurse plants. established Galenia africana shrubs at of fertile topsoil when digging the mi- After the 5.2 mm rainfall event in Ek- Paulshoek showed higher survival rates crocatchments. Cover of geophytes was steenfontein, the mean volumetric soil at two out of three restoration blocks negatively affected, possibly because the water content (7.8%) under the planted compared to the other treatments and the digging of microcatchments uprooted the Brownanthus pseudoschlichtianus in- control (Simons & Allsopp 2007). bulbs. The performance of microcatch- dividuals was approximately one third ments was not improved in combination higher than in the control plots (4.8%) Stone applications with brushpack (Simons & Allsopp 2007). (Fig. 6). In accordance with the result Stones are normally used for restora- received for brushpacks, the water con- tion purposes in gullies or on slopes as a Functional plants tent exceeded the amount provided by barrier against erosion (Critchley 1994, The fact that vegetation cover may posi- the preceding rainfall event and this posi- Coetzee 2005), but landusers in the tively affect soil properties and often fa- tive impact on soil water was enhanced Succulent Karoo have also observed cilitates the growth of other plants in dry- by prolonged storage of soil water due to additional positive effects of stones lands has been considered in the design of shading (Fig. 7). (Botha et al. 2008). Soils that have restoration measures (Padilla & Pugnaire Perennial cover was directly improved small stones or rocks on the surface 2006, King & Stanton 2008). Shrubs that by planting, demonstrating the short- are less prone to drought-induced plant crucially infl uence their surrounding en- term success of this treatment. One year mortality. Stones further reduce wind vironment are called ecosystem engineers after transplanting, the Brownanthus speed and catch wind-blown seeds, soil (Jones et al. 1994). In the Succulent Ka- pseudoschlichtianus transplants showed particles, and organic material. Thus, roo they create fertile islands by trapping a survival rate of 71% and the Cepha- stones might fulfi l similar functions as wind-blown soil particles, seeds and or- lophyllum spissum transplants showed a brushpacks, but require no harvesting ganic matter (Stock et al. 1999, Anderson survival rate of 55%. Further monitor- of plant material. et al. 2004). Some species also function ing revealed that although the number of Stone treatments were applied at Soe- as nurse plants (Callaway & Walker 1997) Cephalophyllum spissum individuals de- batsfontein and in the Knersvlakte (Ta- in grazed systems, e.g. Tripteris sinuatum creased towards 2008, there were still on ble 5). At the Soebatsfontein experiments which forms positive associations with the average nine times more Cephalophyllum six stone heaps (approximately 25 cm x refuge plant Ruschia robusta (Todd 2000). spissum individuals present on the trans- 25 cm x 25 cm) per plot were installed. On Key species of the natural plant com- plant plots compared to the control. The the quartz fi eld site in the Knersvlakte, we munity were transplanted at two of the abundances of other plant species also tried to reconstitute the situation prior to restoration sites: Brownanthus pseudo- increased signifi cantly (Fig. 9) on the Ce- the disturbance. Quartz stones (6–20 cm

SUCCULENT KAROO 129 a small strip, whereas the rangelands have been disturbed over an extensive area. Our results thus confi rm that small- scale degradation is easier to restore than large-scale degradation. In a review on mine spoil restoration in the Succulent Karoo, Carrick & Krüger (2007) mention several reasons for this. Most important- ly, large-scale disturbances only have a small contact zone with the surrounding vegetation and thus lower probability of seed input. Furthermore, the lack of vegetation on large areas causes higher wind speeds. The few seeds that blow into the Succulent Karoo area are not readily trapped, and therefore the chances of establishment are lower. Higher wind speeds also inhibit plant re- Fig. 9: Means (+SD) of the number of all plant individuals on restoration plots in quartz fi elds (Cephalophyllum spissum community) in the Knersvlakte. A treatment effect was cruitment because they result in greater indicated by Kruskal-Wallis tests for the mean number of individuals in 2005. Different erosion rates of fertile topsoil. letters above bars indicate signifi cant differences between treatments according to LSH In addition to the spatial scale of degra- with Bonferroni-Holm correction. dation, the temporal scale might also have infl uenced the success of restoration, particularly in the Knersvlakte. The pipeline in size) were evenly scattered, covering Growth form restoration experiment in the Knersvlakte about 10% of the soil surface. Annuals responded more strongly to the started only four years after the initial dis- The soil and vegetation variables ex- treatments than perennials. Only two turbance, whereas for the other sites, over- amined at the stone treatment plots did of the restoration treatments had an un- grazing had occurred for decades prior to not differ from those at the control plots ambiguously positive infl uence on the the studies being conducted. The funda- at the Soebatsfontein site (Meyer 2009). perennial vegetation cover: the scatter- mental soil properties were probably less Positive effects on the vegetation were ing of quartz stones on disturbed quartz affected in the Knersvlakte and the soil observed four years after the introduction fi elds; and planting, where perennial seed bank may still have contained some of the quartz stones on the quartz fi elds cover was directly improved through the of the species, which were present in the of the Knersvlakte. The number of plant transplants. However, the cover of annual former, undisturbed vegetation. individuals, including both opportunists species was promoted by four treatments: and obligate quartz fi eld taxa, increased livestock exclusion, dung mulching, scat- Soil hydrology signifi cantly when compared to the con- tering of quartz stones, and to a certain Degradation of (semi-) arid rangelands trol plots in 2005 (Fig. 9). However, the degree also by brushpacking. The strong- increases water stress for plants in an signifi cant effect ceased towards 2008 er response of annuals compared to per- environment where water is the limiting (Meyer et al., in prep.). ennials may be attributed to their oppor- factor for plant growth even under undis- tunistic nature (Lavorel & Garnier 2002), turbed conditions. The increase of water Understanding the processes their higher abundance in the seed bank supply for plants by improving the soil of vegetation recovery (de Villiers et al. 2003) and, in the case of water status is therefore of major concern mulching, to the abundance of their seeds for restoration success (Thurow 2000). Background in the dung. The assumption of de Villiers The Eksteenfontein experiment showed We collated the results of the four resto- et al. (2003), who studied soil seed banks that soil water content directly after a ration experiments described above. The in the Strandveld Succulent Karoo, that typical rainfall event was considerably fact that the experiments were conduct- successful restoration will often require increased under planting and the applica- ed at different places and over different the reintroduction of perennial plants is tion of brushpacks. Storage of soil water years offers the opportunity to identify confi rmed by our results. was improved by planting, brushpacks, some principle restoration requirements and dung. For microcatchments no effect for degraded Succulent Karoo range- Spatial and temporal scale on soil water was detected. lands. The variety of measured soil and of degradation In dry ecosystems, the net input of vegetation parameters provides insights Restoration efforts on the pipeline in rainwater into the soil is determined by into the processes, which determined the Knersvlakte showed clearer success canopy interception and the partitioning treatment performance. than those in the overgrazed rangelands. of infi ltration rate and runoff (Loik et al. The pipeline installation only disturbed 2004). Soil water under the transplants

130 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT and brushpacks even exceeded the wa- rainfall outside the growing season is not et al. 2007a). Nitrogen was elevated after amount provided by the preceding likely to promote seedling germination or ter livestock exclusion and dung input, rainfall event. The rainfall followed a six establishment of perennials (Esler 1999), and dung also decreased the C/N ratio. week dry period and the upper soil was it could enhance the survival and growth The increase of nitrogen after livestock probably similarly desiccated under all of established plant individuals. exclusion could have partly arisen from treatments prior to the event. It is hence All treatments that shaded the soil sur- the accumulated litter. However, a sec- likely that brushpacks and planting some- face and protected the soil surface from ond process, which can be explained by how increased the water input during the wind (i.e. transplants, brushpacks, and the nitrogen loss hypothesis, might have rainfall. Since the rainfall event was as- dung) had a positive effect on soil water contributed to the increase of nitrogen sociated with strong winds, we suggest storage. After winter rainfall events, day in the exclosures. Grazing increases the that the branches ‘combed out’ raindrops temperatures may increase considerably removal of nitrogen out of the system from the air and water was thus concen- within a few days, leading to rapid evapo- because dung and urine concentrate the trated and accumulated in the soil below. ration of soil water. This is compounded nitrogen on small patches around stock The capturing of raindrops may lead to by the shallow percolation of rainwater posts and water points where it is partly evaporation losses of the intercepted wa- after low-intensity rainfall events (Fran- lost through volatilisation or leaching (Pi- Succulent Karoo ter from the canopies (Dunkerley 2008), cis et al. 2007). Under these conditions, neiro et al. 2009, Hiernaux et al. 1999). unless air temperatures are suffi ciently protective cover of the soil surface is a The increases of carbon and nitrogen in low to retard evaporation rates (Loik et crucial factor for water retention. The im- response to dung treatment was stronger al. 2004). In winter in the Succulent Ka- proved soil water conditions under dung than in response to the accumulated litter roo, evaporation losses from the canopies mulch, brushpack, and functional plant in the exclosures or to brushpack, which are probably of minor relevance, because treatments may be one factor responsible refl ects the different properties of the input the winter rainfall derives from coastal for the enhanced plant growth recorded materials (compare Mokolobate & Haynes lows (Desmet & Cowling 1999) with under these treatments. However, rain- 2002). Dung possesses a higher mineral weather conditions during rain gaps and fall amount remained the limiting factor. content and a lower C/N ratio. Dung also shortly after the rainfall event usually be- Dung promoted growth of annuals only increased pH and electrical conductivity ing cloudy, cool, and humid. The amount in good rainfall years, which could be ex- considerably, while the effects of brushpack of gathered rainwater channelled to the plained by the observation that the rain- on these variables were less pronounced. soil below also depends on the canopy water fi rst had to saturate the dung layer The increase of pH after treatment with architecture (Mauchamp & Janeau 1993, before entering the soil. dung or plant residues concurs with stud- Domingo et al. 1998, Dunkerley 2008). Low rainfall was probably the main ies from other parts of the world (Mbagwu Apparently the succulent, cylindrical in- reason that only 90 seedlings out of the 1992, Tongway & Ludwig 1996, Wong et ternodes of Brownanthus pseudoschlich- 20,000 seeds sown at the Paulshoek res- al. 1999, Whalen et al. 2000, Mokolobate tianus channelled the water effectively. toration trial emerged, irrespective of the & Haynes 2002). This can be explained by

In regions where intense, erosive accompanying treatments (Simons & All- alkalizing components, such as CaCO3 or rainfall events are prominent during the sopp 2007). Planting was more successful Na and by the proton consumption capacity growing season, manipulations to pre- than sowing. Apparently the transplants, of the humic material present in the dung vent or capture runoff are key measures in both cases succulent shrubs, could and plant residues (compare Mokolobate & for restoration processes (Critchley et cope well with little rainfall. One year Haynes 2002). al. 1994). However, microcatchments after transplanting, the survival rates of The increase of electrical conductiv- did not impact soil water content show- Cephalophyllum spissum and Brownan- ity after dung treatment also concurs ing that during this low-intensity rainfall thus pseudoschlichtianus were 55% and with fi ndings of Mokolobate & Haynes event, runoff was of minor relevance 71% respectively, which is remarkable (2002), who showed increased levels of (compare Li et al. 2005). considering that poor rainfall conditions exchangeable Ca, Mg, K, and Na after the In general, the results suggest that the that followed transplantation. application of poultry dung. In (semi-) spatial distribution of water input into the arid rangelands, the forage resource in- soil was infl uenced by processes related to Chemical soil properties cludes halophytic plants (Ward 2009) and canopy interception rather than by the par- Signifi cant effects regarding chemical the salt content of dung is therefore often titioning of runoff and infi ltration during soil properties were detected for those particularly high. An increase of soluble the typical winter rainfall event. However, treatments, which were associated with salts in the soil is of potential concern although a large portion of water input the introduction of organic matter. The in arid environments and dung should during the growing season in the Succu- organic input from dung and brushpacks not be applied to soils prone to salinisa- lent Karoo results from small, low-inten- was refl ected by an increase of carbon tion. Areas of concern in the Succulent sity rainfall events (Desmet & Cowling in the soil. The fact that livestock exclu- Karoo are, for instance, fi ne-textured 1999), microcatchments could be effec- sion also increased the amount of carbon soils typical for the Knersvlakte (Fran- tive during strong rainfall events typical of indicates that litter accumulated after cis et al. 2007) or calcareous and gyp- thunderstorms during summer. Although grazing was stopped (compare Allsopp sum-rich pans. Soils of the study site at

SUCCULENT KAROO 131 Soebatsfontein are less prone to salinisa- enhanced performance of the treatments However, a study in the Succulent Karoo tion, because the high sand content guar- in the Knersvlakte could be related to this demonstrated that facilitation appears to antees good drainage and thus prevents issue. Here, all plots were cleared of the play a less important role (Carrick 2003). accumulation in the upper soil layer. The opportunistic species Atriplex lindleyi Future revisits of the experiments could values after dung treatment did not in- subsp. infl ata before treatment establish- help to better understand the interplay dicate strong soil salinisation and total ment. This may be a reason why key between competition and facilitation in plant cover was, in fact, greatest for the perennial species of the quartz fi eld fl ora Succulent Karoo rangelands. dung treatment in our experiments. How- could recover well. Competition from an- ever, the elevated electrical conductivity nuals has been shown to increase the mor- Implications for further research might have contributed to the negative tality of seedlings of perennial plants un- Restoration of degraded rangelands in infl uence on species richness. It seems der arid conditions (van Epps & McKell the Namaqualand, as part of the Succu- that salt-tolerant ephemerals, especially 1983). By contrast, annuals had no detect- lent Karoo Biome, is still in its pioneer Atriplex spp., benefi tted and had an ad- able effect on the seedlings of perennials phase. The broad range of basic ecologi- vantage over other species (compare de in a clearing experiment in the southern cal research as well as the various res- Succulent Karoo Villiers et al. 1992). Also, the nitrogen Karoo, which borders the Succulent Ka- toration trials conducted over the last fertilisation from the dung might have roo (Milton 1995). However, we assume decade have signifi cantly contributed to enhanced the growth of Atriplex spp. and that in the case of opportunistic Atriplex an understanding of the ecosystems of other fast-growing ephemerals, as high spp., which tend to form a relatively dense the Succulent Karoo. However, there are nutrient availability supports species cover, the recovery of the shrubby vegeta- many important aspects that are still not with rapid resource allocation and rapid tion is hampered. On the other hand, the clearly understood, as has been shown by turnover of organs (Grime 1979). establishment of annuals or fast growing, the integrative analysis of the four resto- opportunistic perennials could protect the ration studies in this Subchapter. Plant interactions: competition exposed soil, and may in time add to the This analysis that went beyond the and facilitation organic matter of the soil and thus trigger scope of a single case study identifi ed Besides the manipulation of the grow- a succession towards long-living species. some abiotic and biotic processes that ing conditions, treatments also induced Continued monitoring of the experiments seem to drive the dynamics in restora- changes in the vegetation cover, which will show if annual cover facilitates an tion trials. The responses of opportun- appeared to have further impact on plant increase in perennial plant cover or not. istic plant species, for instance, have growth, involving both, competition and The experiments also revealed several been related to treatment effects on soil facilitation. Competitive interactions facilitative effects of perennial plant spe- properties. However, the responses of between the species were probably ma- cies themselves. The abundance of other long-lived shrubs, that characterise in- nipulated by the exclosures in Soebats- plant species increased signifi cantly tact Succulent Karoo rangelands, were fontein. Species richness was increased along with the planting of the perennial less pronounced. Continued monitoring by rest from heavy grazing, but decreased Cephalophyllum spissum individuals. will show whether the enhanced emer- in response to rest from moderate grazing. The nurse plant experiments with Ga- gence of opportunistic species, which This is in line with the intermediate dis- lenia africana also showed that condi- especially followed the dung treatment, turbance hypothesis (Grime 1973, Con- tions under the shrubs supported seedling will support or inhibit the recovery of nell 1978). According to this hypothesis, establishment and improved the survival key, perennial shrubs. The data will moderate disturbance increases the com- of perennial plants (also compare Gabri- thus provide insights into the important petitive interactions between species, els et al. 2003) since it possibly provides competition and facilitation processes which affects species richness positively. nutrient-enriched soils (Allsopp 1999b), among different plant strategy types Only under severe disturbance is biodi- protection from climatic extremes and en- and contribute to a general understand- versity expected to decrease. hanced soil moisture. On the other hand, ing of ecological processes in the Suc- Among the active treatments, dung Galenia africana might have competitive culent Karoo Biome. The monitoring of appeared to have the greatest impact on impacts on the seedlings that it nurses as the exclosures would also help to iden- competitive interactions between growth they grow. Several studies in arid eco- tify the appropriate resting periods re- forms. Although soil properties were systems have demonstrated that seedling quired for vegetation recovery, which is enhanced in terms of nitrogen and wa- establishment and survival were greater important information required for the ter supply at the dung plots in Soebats- underneath the canopies of shrubs than development of appropriate and effec- fontein, the establishment of perennial in the open spaces between shrubs (Cal- tive rangeland management strategies. species was not supported and their rich- laway & Walker 1997). According to the Future research should also test some ness even reduced. This may be due to stress gradient hypothesis it is assumed modifi cations of the trails that have competition with invasive Atriplex spp., that competition becomes less important been described in this Subchapter. The which spread in response to dung treat- and facilitation relatively more important brushpacking technique, for instance, ment and may have negatively affected in drylands, since abiotic conditions are should be further refi ned by compar- the establishment of perennials. Also, the harsh (e.g. Callaway & Walker 1997). ing different densities and techniques of

132 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT packing (e.g. brushpack islands of vary- tiveness and long-term effects. Peren- which allows enough light to penetrate ing sizes). nial plants may spread into disturbed and leaves more space for plants to For the application of restoration in areas from the surrounding vegetation establish. The spatial arrangement of practice at larger spatial scale (i.e. at by themselves. brushpacks should consider that the stock post-, camp- or even farm-scale), • In contrast, the restoration of large- lee side might receive less precipita- quantitative aspects of practicability as scale degradation will often require the tion due to a ‘rain-shadow’ effect. well as cost and benefi t have to be in- reintroduction of key perennial species. cluded in future evaluations. The aspects Transplanting of mature individuals of practicability and costs comprise mon- proved to be more successful than sow- 4.5 Land management in etary expenses as well as availability of ing. Transplants with succulent leaves the biodiversity hotspot natural and human resources. The benefi t have a good chance of survival, since of the Succulent Karoo. aspects comprise the immediate mon- they are drought-adapted. Transplants A case study from etary benefi t of restoration through im- might be the only palatable species Soebatsfontein, proved rangeland but should also consid- within a vast area and protection from South Africa er the improved or enhanced ecosystem selective grazing using fences would [U. Schmiedel, T. Linke, R. Christiaan, Succulent Karoo services that are provided by the restored therefore aid in restoration success. T. Falk, A. Gröngröft, N. Kunz, T. Labitz- ecosystems. The socio-economic benefi ts Habitat characteristics of the focal ky, J. Luther-Mosebach, S. Meyer, I.U. of ecological restoration are critical for system Röwer, B. Vollan & B. Weber] their societal and political appreciation • Considering the characteristics of the and support. However, they are hardly winter rainfall regime in the Succu- Introduction adequately quantifi ed in respective stud- lent Karoo, it is recommended to ma- Soebatsfontein is a small rural settlement ies (Aronson et al. 2010). nipulate canopy water ‘combing’ and in the Namaqualand lowlands of the Every future ecological-economic soil shading rather than runoff related Succulent Karoo biome, an area, which evaluation of restoration trials and processes for the improvement of wa- is widely acknowledged as a centre of emerging recommendations should be ter supply during the growing season. biodiversity and endemism (Myers et aware that besides the described particu- • Treatments that tend to increase salt al. 2000, Mucina et al. 2006). The Soe- larities of the Succulent Karoo, the biome content of the soil (i.e. dung), should batsfontein commonage has experienced is extraordinarily heterogeneous. As has be applied only after ensuring that the a transformation of tenure and manage- also been shown by this study, the vari- respective soils are not prone to salini- ment over the course of recent politi- ous habitats, driven by a high diversity of sation. cal changes in South Africa. As part of topography, soil types, rainfall patterns, Degradation level and restoration post-apartheid land and agrarian reform, fl oristic inventories, historical and recent objectives 15,069 hectares of farmland were trans- landuse, etc. also respond differently to • In cases where the target system is in ferred from the private De Beers min- restoration measures. a less degraded state, in which veg- ing company to the rural Soebatsfontein etation cover of an early successional community, which until the year 2000 Implications for restoration state still occurs, restoration measures did not have any access to farmland. The practice would aim to reconstitute a diverse communal farming area is referred to as The evaluation of the restoration experi- succulent shrub community. If fi nan- “new-commons” and is characterised ments outlined in this analysis has opened cial and labour resources are available, by a subdivision of the land into several new insights into the treatment effects on the introduction of the respective key camps and the allocation of these camps soil properties and vegetation. But what perennial species is recommended. to farmers within the community. While are the practical implications for resto- • The situation is different when it comes the community members have usage ration projects in degraded rangelands to restoring rangelands where bare soil rights, the land is formally owned by the of the Succulent Karoo? What is the ap- patches predominate on a large scale. local municipality of Kamiesberg. propriate treatment and how can restora- In such cases it may already be a sat- One of the central aims of this land re- tion success be optimised in the fi eld? isfactory achievement to cover the soil form was to support the subsistence live- Beside resource availability in terms of with vegetation, even if it is only by op- lihoods of the community by enabling manpower and materials for treatment in- portunistic species with an annual life them to farm with small livestock. The stallation and maintenance, the fi ndings cycle. Dung could be an appropriate, farmland, comprising 15,069 hectares of this article suggest several factors that cost-effective treatment, which is easy of land, is relatively small compared should be considered: to install in such a situation. The effects to the number of households that are in Scale of disturbance of brushpack were less pronounced but need of additional sources of income (the • In case of the small-scale disturbance it can also increase ephemeral plant size of an average commercial farm that due to the pipeline construction on cover. We suggest avoiding destruc- is owned by one family is about 5,000 quartz fi elds, the application of stones tive installation effects by creating a hectare). When the community received is suffi cient, considering cost-effec- thinner, patchier layer of brushpacks, the farmland, some of the camps were in

SUCCULENT KAROO 133 relatively poor condition and they were tation patterns may, in turn, infl uence the tions for resource management. Based on therefore left to rest for some years in density, structure, height and thus quality an interview survey, he further analysed order to enable vegetation recovery. This and quantity of plant biomass in the area. the framework of formal and informal re- management decision was economically Biological soil crusts, formed by an source management institutions. Vollan unproblematic for the farmers as the association of the uppermost layer of (2009) examined norms of trust and trust- stocking density on the rangelands was the soil with cyanobacteria, algae, bryo- worthiness (i.e. reciprocity) in southern comparatively low during the fi rst two phytes, lichens, microfungi, and hetero- Namibia and Namaqualand. Norms of years due to a lack of fi nancial capital and trophic bacteria in varying proportions, reciprocity to a large degree explain why the time required for the herds to grow. occur worldwide in arid and semi-arid people co-operate (i.e. they only co-op- From the outset, the farmers in Soebats- regions and wherever an arid microcli- erate if others also co-operate). To elicit fontein were aware of the widely held mate occurs (Belnap et al. 2003a, Article the crucial norms of trust and reciprocity, prejudice that communal farmers always III.3.4). Due to the agglutination of soil a standard experimental procedure was let their farmland degrade. Several of the particles by the gelatinous sheaths of cy- followed (Berg et al. 1995). Soebatsfontein farmers therefore aimed anobacteria they effectively prevent soil Succulent Karoo to disprove this stereotype by trying to erosion by wind and water (Belnap & Sampling and analysis of soil and manage the land sustainably. The exist- Gardner 1993, Belnap 2002a). They also vascular plant patterns ence of camps and their clear allocation enrich the soils with nutrients and in- The communal land of the Soebatsfon- to individual farmers seemed to be in fa- crease their water-holding capacity (Bel- tein region is subdivided into 17 camps, vour of this aim. However, management nap 1995, 2002a, Belnap et al. 2003b, Ja- which are separated by fences. We classi- of the Soebatsfontein commonage during fari et al. 2004), thereby promoting plant fi ed the camps into three categories (low, the fi rst decade faced social and econom- establishment and growth. However, bio- intermediate, high, Fig. 10) according to ic challenges, which were exacerbated by logical soil crusts are very susceptible to their grazing intensities between 1986 the arid environment. disturbances caused by different kinds of and 1999 during management by the De Socio-economic studies within BIOTA landuse (Belnap & Gardner 1993, Belnap Beers Company (hereafter referred to as Southern Africa aimed to understand the 1995) and can therefore be employed as historic grazing intensity). We overlaid socio-economic framework under which sensitive indicators of the effects of cur- high resolution satellite images (Dig- the community members of Soebatsfon- rent landuse on the soil. ital Globe, 11/2003, provided by Google tein manage their farmland. The mapping The fi rst part of this section provides in- Earth) of the commonage and the adja- of soils, biological soil crusts, and veg- sights into the socio-economic conditions cent farms with a 300 m x 300 m grid. etation aimed to: a) provide baseline data that infl uence management decisions on The more than 4,000 resulting grid points for farm management in Soebatsfontein, the Soebatsfontein commonage. The were classifi ed into ten categories accord- which could also be used as benchmark second part describes the diverse natural ing to geomorphological, geological and data for future assessments, and b) assess environment and tries to disentangle the substrate properties, as well as historical the impacts of the historic (during De impact of present versus pre-communal grazing intensity classes per camp. Beers ownership) versus current grazing grazing management on the biodiversity A random selection was drawn from management. of the farmland. To conclude, we discuss among the 4,000 grid points with four Soil patterns, distribution, and states the perspective for future land manage- replicates for each historical grazing in- are, together with the topography, the ment in Soebatsfontein. tensity class for soil-profi le and vegeta- main drivers of vegetation composition tion sampling (90 profi les and relevés), in a given climatic zone. In Namaqua- Material and methods 10 replicates for vegetation and soil-sur- land, the patchiness of soils at different face sampling (in total 107 relevés), and spatial scales has been shown to strongly Socio-economic studies 30 replicates (10 replicates per historic infl uence species composition of the BIOTA Southern Africa carried out sev- grazing intensity class) for vegetation vegetation (Herpel 2008, Schmiedel & eral socio-economic studies within the and soil sampling on heuweltjies (total Jürgens 1999, Subchapter IV.4.2), thus Soebatsfontein community (Falk 2008, of 121 relevés). The study area was as- contributing to the exceptionally high Linke 2009, Vollan 2008, 2009). Linke sessed in terms of historical grazing in- species richness of the Succulent Karoo. (2009) conducted an ethnographic cen- tensity classes using the following data Soils, in return, are also infl uenced by sus, gathered qualitative data on local and calculations (Labitzky 2009): biotic interactions, i.e. burrowing insects knowledge and small stock farming, and • ‘Total cumulative sheep grazing days’ and mammals (Lovegrove 1991, Picker monitored landuse activities over a pe- were calculated based on the cumula- et al. 2007, Petersen 2008) and, less obvi- riod of fi ve months. In her case study she tive number of animal days per camp ously so, by the plants themselves (Hook analysed the links between features of the (data kindly provided by Mr. F. Brandt, et al. 1991, Stock et al. 1999). The water- community and the pastoral land man- the former farm manager for De Beers). holding capacity, nutrient content, and agement system. Falk (2008) assessed the • For the ‘actual sheep grazing days’ the other chemical and physical features of availability of natural, human, fi nancial, ‘total cumulative sheep grazing days’ soils, which may be altered by local vege- physical, and social capital as precondi- were averaged over the recorded years.

134 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT • The ‘potential sheep grazing days’ ! were calculated based on an assumed ! ! carrying capacity of 9 ha per SSU (i.e. ! ! k one adult sheep or goat) by applying the following equation: (Area/Carry- # ! ! ing capacity) x 365 days. ik i i ! ii • Camps that were used for over 120% of # ! #! # ! the ‘potential sheep grazing days’ were ! k ! ! ! k ! ! k! ! ! # ! classifi ed as grazing intensity class ‘high’ ! ! # ! # # k = 3. Those used for 80–120% ‘potential ! ! ! ! ! ! ! ! # ! sheep grazing days’ were classifi ed as ! ! # ‘intermediate’ = 2, and those camps used ! ! ! ! for less than 80% ‘potential sheep graz- ! # ! ing days’ were classifi ed as ‘low’ = 1. i! !

! ! i ! ! Succulent Karoo The recent landuse intensity classes k ! # k were classifi ed according to the ‘actual ii i ! Vegetation relevé ! sheep grazing days’ recorded for 2007 #! # Sampled heuweltjie # (R. Christiaan, unpubl. data) and relating ! ! # k Stockpost # them to ‘potential sheep grazing days’ per ! ! ! i Waterpoint ! ! ! # ! camp (Röwer 2009). ! ! Soil profile and vegetation relevé ! ! ! ! Each soil profi le was sampled and ana- ! ! ik # Biological soil crust study area ! ! # lysed according to the KA5 (AG Boden # Camps # 2005). A soil sample (about 600 g) of # ! Historic grazing intensity ! ! each horizon and the topsoil was taken # no data for laboratory analysis and the soil was ! # low ! ! classifi ed according to the World Refer- # ! # # medium ± ## ! ! ! ! ence Base for Soil Resources (2006). Us- ! high k! ! ! i ! ! ing SPSS for Windows 15.0 (SPSS Inc.), ! ! ! 0 2,500 5,000 7,500 10,000 the infl uence of landuse and land type on meter soil properties was analysed. A detailed description is given by Labitzky (2009). All occurring vascular plant taxa were Fig. 10: Historical grazing intensity classes per camp and the distribution of plots for vascu- listed and their percentage cover estimat- lar plant relevés on the Soebatsfontein commonage. ed for each plot. The nomenclature of vascular plants follows Hartmann (2002a, b) for the Aizoaceae and Germishuizen & ent grazing intensities. We sampled the 25 cm x 25 cm grid was randomly placed Meyer (2003) for the remaining taxa. abundance and projected cover of vascu- at six points and mean BSC coverage val- Total cover of mosses and biotic crusts, lar plant species at 20 plots (5 m x 5 m) ues in each plot were determined using including other bryophytes, were also each outside and inside the grazing exclo- the point-intercept method described by estimated. We clustered the relevés into sures four years after installation. For this Belnap et al. (2001). We distinguished be- distinct vegetation units using numeri- study, we analysed various vegetation tween initial BSC, formed by cyanobac- cal classifi cation methods (Luther-Mo- parameters (i.e. annual cover, perennial teria that cause a patchy discoloration of sebach 2009). Vegetation relevés of the cover, total vegetation cover, and species the soil (type 1; Büdel et al. 2009 and Ar- heuweltjie centres were analysed using richness) in the control plots that were ticle III.3.4), and climax BSC, consisting multiple linear regressions with histori- not treated with restoration measures of well established cyanobacterial mats, cal and recent grazing intensities (square inside and outside the exclosure (Meyer as well as bryophyte (moss) and lichen root transformed) and topography as 2009). Differences between values were dominated BSC (types 2–5, Büdel et al. predictors and a set of soil chemical, analysed using ANOVA. 2009, Article III.3.4). Local utilisation of vegetation and structural parameters as each plot was estimated by counting sheep response variables (Appendix 1). Sampling and analysis of and goat faecal pellets within each grid. BIOTA installed two grazing exclosure biological soil crusts As a second grazing related parameter, the fences (each 50 m x 50 m) for restoration Sampling of biological soil crusts (BSC) distance of each plot to the next stock post experiments (Subchapter IV.4.4) in two was conducted on 60 plots (5 m x 5 m) lo- or watering point was measured using a different camps of the Soebatsfontein cated in the exclosures and on four different Geographical Information System (GIS; commonage. The areas surroundings the camps within the commonage (Kateklip, ArcView 3.2, ESRI, Redlands, USA). two exclosures were subjected to differ- Kruisvlei, Rooshogte, and Langkamp). A For this, the route most likely taken by

SUCCULENT KAROO 135 should be noted that there are also some exceptions to this pattern. Living in nuclear families, with the majority of household members belong- ing to the core family, is the norm in Soe- batsfontein. New households are usually formed after marriage. Quite a few children are born out of wedlock and are raised in the parental home of the mother until she marries. The average household size is 4.6 persons. Livelihoods in Soebatsfontein are typically diversifi ed and household members are involved in different economic ac- Succulent Karoo tivities. This has also been described for other rural communities in Namaqualand (Berzborn 2007) and Namibia (Schnegg 2009). Wage labour, state pensions and retirement benefi ts from the De Beers pen- sion fund are the main sources of income, Photo 17: Soebatsfontein settlement, 2004. Photo: Ute Schmiedel. whereas livestock farming contributes mostly to subsistence. In about 60% of all cases in Soebatsfontein, monthly house- the animals was considered. For plots lo- fertility decline, even though the com- hold income is low (< 120 US$/house- cated in the exclosures, the distance to the munity increased in numbers, especially hold) and insecure. The majority of house- next stock post or watering point was set during the 1970s and 1980s. As the ob- hold members are temporarily employed to 4,000 m (the greatest distance within a served growth cannot be just a natural and pensions or retirement benefi ts are grazed camp was 3,262 m). All statisti- one, immigration must be taken into often the only regular sources of income. cal analyses were carried out using SPSS account. During the 1970s and 1980s, Households that are dependent solely on (SPSS 15.0 for Windows, SPSS Inc., Chi- two larger family groups, amongst oth- short-term work are the most vulnerable cago, USA). ers, immigrated to the area due to job and poorest in the community (about two- opportunities on the surrounding farms. thirds of low-income households). Results and discussion Nowadays, a rapid increase in the Soe- One of the few opportunities for com- batsfontein population is highly un- munity members to obtain permanent Social and economic factors and likely, and neither natural population employment is to work at one of the dia- norms of trust and cooperation increase nor immigration can been seen mond mines along the west coast of South as framework conditions for as crucial factors affecting land man- Africa (e.g. Kleinsee or Koingnaas). In management decisions in agement. Nonetheless, immigration has 2005, about 40% of the households in Soebatsfontein signifi cantly shaped the social organisa- Soebatsfontein derived income from a Social and economic factors: The Soe- tion of the community. The two above- permanent work position or relied on batsfontein community has approxi- mentioned families are distinguished a combination of regular income from mately 270 inhabitants that live in 60 from kin groups who have lived within state pensions and retirement benefi ts. households (unless stated otherwise, data the community for several generations These households are economically sta- in this chapter are based on Linke 2009). (the “original” community). Although ble and the average per capita income is Today, the community comprises a few incoming families are nowadays well at least three times higher compared to large family groups, which are closely re- linked through marriage relations with those with an insecure work status. About lated through marriage. Before the apart- long-established families, there are still one third of households own small stock heid era, the ‘coloured’ families worked tensions between them. A major reason (sheep and goats), which provides them and lived dispersed on the surrounding for this is that the incoming families are with milk and particularly meat. Only a farms. During the 1950s, these fami- perceived to be economically more suc- few farmers earn cash from larger live- lies had to move to a small area owned cessful. The ethnographic census showed stock sales. However, livestock farming by the church, which was declared the that households with higher and more is becoming an increasingly important ‘coloured’ settlement of Soebatsfontein stable livelihood income sources are of- activity in the area as mines are closing (Photo 17). ten those of immigrant families, whereas down in the region. Since then the population of Soebats- most of the low income households are Norms of trust and cooperation: fontein has experienced a substantial ‘originally’ from the village. However, it In Soebatsfontein, farmers collective-

136 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT ly own infrastructure, such as fences, fontein under similar situations called the community to commonage resources, pipes, and wind pumps that have to be for external regulation (i.e. a rule where in particular. Farmers earning more than maintained, and there is also the daily players who overharvested would get approximately US$ 180 per month may need of borrowing tools and vehicles for sanctioned by an external agency). In be excluded from using the land. transportation. There is therefore a need contrast to the participants’ expecta- By signing a contract with the munici- for co-operation between community tions, we found that this rule fosters co- pality, farmers agree to the rules of the members in order to cope with the re- operation when levels of trust are low commonage management plan and the stricted, jointly used resources. In order but might lead to more selfi sh decisions grazing regulations, as well as the regu- to assess whether these and other inter- when prior trust is high. In the case of lar payment of grazing fees per head of actions in their everyday lives are based Soebatsfontein, the choice of the penalty livestock. The grazing contract specifi es on mutual trust and cooperation, several rule highlights the greater acceptance of the camp to be used and the period of economic fi eld experiments were carried outside structures and their capacity to the contract. Farmers gain access to two out (Article III.6.5). The experiments re- help to build a functioning community. camps, which they are supposed to use, vealed that levels of trust are extremely Governance of land management: together with the other camp occupiers, low in Soebatsfontein, and in many Until 2000, the majority of the Soebats- for rotational grazing (RSA 2002). Succulent Karoo other more traditional farming com- fontein population had no direct access to An elected commonage committee is munities of Namaqualand, compared the natural resources of Namaqualand. A installed as the advisory and executive in particular to the results of the same few community members, who worked panel, which liaises with local govern- experiment in southern Namibia, where on the farms of De Beers, were allowed ment but does not have the power to cultural norms are still more prevalent. to hold small herds of sheep and goats. make, change, or abolish institutions. It The general interpretation from trust Besides that, a few households used oth- is supposed to implement the common- games is that the weak performance of er communal areas in the region to keep age management plan and grazing regu- trust can be attributed to weak local in- some livestock (Surplus Peoples Project lations (Falk 2008). stitutions that do not prevent free-riding 1995, 1997). In 2000, the South African The key problem with commonage and untrustworthy behaviour, so that government acquired 15,069 hectare of management at Soebatsfontein is that this behaviour prevails in a society. Ac- farmland from the De Beers mining com- neither the municipality nor the com- cording to this view, local institutions in pany and made the land available as commonage committee effectively regulate Namaqualand do not punish untrustwor- monage to the people of Soebatsfontein and coordinate resource use. Sophisticat- thy behaviour as much as they should (Falk 2008) as part of the post-apartheid ed rules regarding stocking rates, water in order to establish trust. Therefore, it land and agrarian reform programme. rights, payment of grazing fees, main- pays people to break contracts and cheat The formal owner of the communal farm- tenance of infrastructure, and fi rewood on others. As a result, people learn not to land is the local municipality of Kamies- collection are ignored because they are let others take advantage of them and are berg, which is obliged to use the land for not enforced. Commonage management thus less trusting in social relationships. the benefi t of Soebatsfontein residents. In at Soebatsfontein is therefore a good However, in other experiments and sur- the year 2002, the Kamiesberg Munici- example of an institutional challenge veys described in more detail by Vollan pality introduced grazing regulations on typical within developing countries. The (2009), there are a couple of fi ndings the commonage. In addition, it developed constraining factors are the degree and that highlight how the village’s specifi c various regulations in cooperation with intensity of monitoring and enforcement history has led to a different perception the community and national NGOs such of rules and norms rather than the formal- of norms, rules, and regulations. For ex- as the Legal Resource Centre and Surplus isation of rights (O’Riordan 2002). ample, the results from our social capital Peoples Project. Examples include the The municipality, which formally mo- survey showed that people in Soebats- Soebatsfontein Management Plan and nopolises key management functions fontein stated that they have problems individual contracts between farmers and does, in fact, not have the capacity to ful- with decision-making in meetings and the municipality. fi l its own self-given duties. One example committees more often in comparison In terms of the regulations regarding is the maintenance of infrastructure (RSA with other villages of the Namaqualand. access to land, the grazing regulations 2002). If fencing or water installations One interpretation of the results is that give the municipality the ultimate formal break down, the farmers have to inform the lack of prior experience with col- power to allow and deny access to the the municipality. The municipality is ei- lective farmland management makes it commonage. The main selection criteria ther obliged to repair the damage or must more diffi cult to manage the commons are residency in Soebatsfontein, the car- agree that funds will be made available under the new co-management legisla- rying capacity of the land, the production for the material used by the community to tion. Not surprisingly, in another experi- capacity of the farmer, the neediness of repair damages. In 2004, tensions arose ment conducted by Vollan (2008), it was the farmer and her/his history of com- when urgent maintenance measures were found that people in the old commons monage use. The municipality is obliged required, but the municipality did not preferred communication to solve a so- to improve the access of poor, previously even respond to requests by the farm- cial dilemma, while people in Soebats- disadvantaged, and female members of ers. As a result, more and more farmers

SUCCULENT KAROO 137 16 precipitation 300 stocking rate in sheep/km² 14 recommended carrying capacity 250 12 200 10 8 150 6 100 4

50 precipitation in mm 2 stocking rate sheep/km² 0 0 `01 `02 `86/`87 `87/`88 `88/`89 `89/`90 `90/`91 `91/`92 `92/`93 `93/`94 `94/`95 `95/`96 `96/`97 `97/`98 Succulent Karoo

Fig. 11: The ﬂ uctuation of stocking rates (livestock use-intensity) in relation to annual precipitation on the Soebatsfontein commonage. The land was transferred to the Soebatsfontein community in 2000. The data previous to 2000 reﬂ ect the number of livestock maintained by De Beers on the area Sources: rainfall and stocking rates 1986–1998: De Beers; rainfall 2001/02: BIOTA; stocking rates 2001/02: U. Schneiderat, unpubl. data; based on Falk 2008.

stopped paying grazing fees because cal government shows little awareness of stocking rate was 10.7 ha/SSU (Fig. 11), maintenance is supposed to be fi nanced the power of endogenous institutions and indicating that on average, the recom- by such fees. In the spiral of tension, their community-based enforcement (cf. mended stocking density was not ex- the municipality also failed to enforce Vollan 2009). The perceived distribution ceeded. Nonetheless, livestock numbers the payment of these fees, which further of property rights is becoming increas- decreased signifi cantly with the transfer highlighted its weakness (Falk 2008). ingly blurred, creating an institutional of the land to the Soebatsfontein commu- Some farmers started to repair the infra- vacuum, which is a serious threat to sus- nity. The main reason for this is the lack structure themselves. This increased their tainable natural resource management. of capital within the community: they feelings of ownership of the commonage Infrastructure and stocking density: were not able to buy as many livestock but undermined formal regulations. The The Soebatsfontein commonage is cur- as the land could carry immediately. The farmers increasingly perceived that they rently (as of 2009) subdivided into 17 Soebatsfontein residents believe the re- owned their camps privately and were fenced camps ranging between 360 and duction of use-intensity has led to specifi c not willing to comply with any externally 1,280 ha in size. One or two camps are al- improvements, such as the disappearance derived rules. located to each livestock owner, who typi- of bare soil patches, the return of specifi c The commonage committee as an al- cally herds his own livestock together with grass species (e.g. the annual Schmid- ternative management body is hardly that of other family members. Permitted tia kalahariensis, Kalahari Suurgraas), accepted within the community and is stock numbers are tied to camp size and and the fact that bushes grow larger and considered to be functioning poorly. On vary between 30 SSU (Small Stock Units) grasses grow taller. In 2002, livestock the one hand, this is the result of the mo- and 109 SSU. According to the manage- numbers had already increased notably, nopolisation of management authority by ment plan, larger herds are supposed to be which was aided by the subsidised gov- the municipality and the de-facto priva- grazed in rotation, and thus should regu- ernment loans made available for the pur- tisation of portions of the commonage, larly be moved between the two camps chase of livestock. At this time, approxi- which decreases the sense of ownership allocated to the farmer. Four wind pumps mately 1,100 SSU grazed and browsed by the community. On the other hand, provide the communal land with brackish on the commonage (Schneiderat, unpubl. it can be seen as a confi rmation of the ground water for the livestock, and the wa- data, Falk 2008) and this number did not afore mentioned weakness of community ter is stored in tanks that are spread across change until 2005, when approximately norms. The question remains whether the the camps. Farmers that have direct access 1,140 SSU grazed on the land (Linke commonage committee could play a more to a wind pump are advantaged, especially 2009). Although the total number of ani- important role in commonage manage- in the hot summer months, when stored mals on the commonage is still far below ment if it were given more authority and water becomes scarce (Linke 2009). the recommended stocking rate, grazing effective power to enforce its decisions. Based on the experience of the agricul- pressure varies considerably from camp In summary, subsidiary rules are weak- tural extension offi ce, the previous owner to camp. In 2005 stocking rates ranged ly implemented in Soebatsfontein, which De Beers defi ned a recommended stock- between 5 and 25 ha/SU (Linke 2009). strongly increases the transaction costs ing rate of 9 ha/SSU. The records show Pastoral production and herd manage- of land management (Falk 2008). The lo- that between 1996 and 1998 the mean ment: In most cases, livestock is owned

138 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT Succulent Karoo

Photo 18: Farmer on horse back herding his goats, 2007. Photo: Photo 19: Stock post at Soebatsfontein commonage. Photo: Bettina Ute Schmiedel. Weber.

by the male head of the household and among those that graze unobserved, thus when lambs are born. Instead, to take some herds include animals of several increasing the risk of livestock loss. Only advantage of the second camp, some of owners, who are usually close relatives a few of the households with insecure them subdivide their herds into differ- (e.g. brothers). Total herd size varies income sources watch over their animals ent management units while keeping the from 30 to 200 animals and only larger regularly, and in most cases these farm- newborn lambs and pregnant ewes close herds contribute signifi cantly to the cash ers then draw on the assistance of their to the stock post (Photo 19). Splitting up income of households (Linke 2009). adult sons who are still living with them. the herds is suboptimal compared to rota- Owners usually take care of the livestock In contrast, owners of large herds belong tional grazing (i.e. moving the entire herd themselves (Photo 18), and in fewer cas- to economically stable households. They between camps) as the latter would allow es, a herder is employed. are less dependent on livestock sales, the camps to rest for some time. Never- The small herds turned out to be more and can use them as an extra source of theless, compared to camps that do not get vulnerable to hazards such as epidemic income, and are able to provide mainte- any rest, reducing the grazing pressure for diseases and predators. Beyond this, the nance of the herd either through social or some time seems to be the better option. risk of (total) animal loss depends on the economic capital (Linke 2009). However, observation has shown (Linke herd management practices, which dif- Grazing strategies: Most of the farm- 2009) that, in most of the cases, the entire fer in various respects. One of the most ers with low livestock numbers let their herd is kept close to the stock post, result- evident differences concerns the tending animals graze unobserved on one single ing in high grazing pressure in the main of the animals: some farmers keep their camp, whereas farmers with large herds camp. Splitting the herds is not favoured herds unobserved in the camps and herd (> 100 SSU) have two camps available. by the farmers because they cannot ob- them only twice to four times a month Nevertheless, most of the farmers hardly serve the different management units at to the kraal; other herders stay continu- use the available pasture but keep their the same time and animals therefore run ously at the stock post and herds are herds near the stock posts. Several reasons the risk of being caught by predators. Fur- observed on a daily basis (Linke 2009). can be considered to explain this practice, thermore, splitting the herd substantially Herd size and management are to some which often causes local overstocking increases the amount of work. Herders extent dependent on farmer preferences, with up to 5 ha/SSU (Linke 2009). have to stay near the stock post (to take but available household resources are The construction of infrastructure fa- care of lambs, water, etc.) and addition- also constraining factors. Only a few of cilities, such as stock posts with shelter ally watch over the animals on the second the numerous low-income households for the herder and the lambs, is restricted camp, which is often reached on foot. own livestock and, if they do, they man- to one main camp. The stock post is typi- These diffi culties could be addressed age smaller herds (up to 50 SSU) which cally built near a wind pump or other wa- by closer cooperation between the farm- were usually purchased on a land bank ter source, which have to be monitored ers, which could improve the provision loan. Since most of them rely on insecure above all during the hot summer months. and maintenance of infrastructure, trans- income sources (casual work), they de- Farmers are reluctant to move their en- port on the farms, effi ciency of herd pend on livestock sales to pay back their tire herd to the additional camps, as they management, and grazing strategies. loans. Moreover, most of these herds are depend on this infrastructure, especially However, the above mentioned tensions

SUCCULENT KAROO 139 typically grazed with low intensity under the recent management in order to allow for the vegetation to recover, and vice versa. Our research attempted to detect the effects of the long-term impacts of the historic management and, the relatively short-term impacts of the recent management on the soil features and vegetation patterns on the Soebats- fontein commonage. Soils: Soils on the Soebatsfontein . commonage are characterised by high heterogeneity at different spatial scales (see Subchapter IV.4.2). Eight different

Succulent Karoo Legend soil types were identiﬁ ed according to the WRB Classification World Reference Base for Soil Resources Soil type (2006). The dominant soils are Haplic Arenosol Leptosols, which are very shallow soils Calcisol above continuous bedrock, and Epipetric Cambisol Durisols, which are soils with a strongly Durisol cemented horizon of secondary silica Gypsisol

Leptosol (SiO2) (‘dorbank’) within 50 cm from the Regosol soil surface, covering about 30% of the Solonchak sampled sites (Fig. 12). The shallow soils farm camps are unfavourable habitats for perennial Historic Grazing intensity plants, as their water storage potential unknown and rooting space is limited. The soils in low the study area have a sandy texture ex- medium cept for the hardpans. The heterogeneity high 063km of the soils on the Soebatsfontein commonage is mainly due to high variance in soil pH, electrical conductivity, carbon and nitrogen contents, as well as nutrient availability (as derived from the amount of water-soluble ions). Parent material, Fig. 12: Distribution of soil types on the Soebatsfontein commonage. topography, and biological activity in the form of heuweltjies (i.e. termitaria, see Subchapter IV.4.2) are the main factors within the community might not favour are frequently used as negative exam- responsible for this variability (Petersen the creation of cooperative solutions. ples (Allsopp et al. 2007b). This second 2008), with the latter especially infl uenc- part of the chapter provides an account ing small-scale heterogeneity (Herpel Effect of past and recent landuse of the biodiversity patterns on the Soe- 2008). Hence, the Soebatsfontein area on soil and vegetation batsfontein commonage and investigates is not only a hotspot of biodiversity but The Soebatsfontein commonage is char- whether these biodiversity patterns are also of pedodiversity, i.e. diversity of soil acterised by an extraordinarily high di- infl uenced by land management. The types (Petersen 2008). versity of soils, biological soil crusts, and Soebatsfontein commonage has experi- Comparison of the soil features vascular plants. This diversity is typical enced two types of land management in within the three historic grazing classes of the Succulent Karoo Biome (Mucina recent history: a) the pre-communal land showed a signifi cant decrease of soil pH et al. 2006) in general, and the Namaqua- management (until 1999) by the previous (Fig. 13) and plant-available phospho- land in particular (Desmet 2007). How- land owner, the De Beer Company, and rous content, and an increase of C/N- ever, the potentially negative impacts of b) the recent land management by the ratio (i.e. the ecologically relevant ratio unsustainable landuse on biodiversity is Soebatsfontein community (since 2000). between carbon and nitrogen), nitrogen undisputed (Riginos & Hoffman 2003, The two land management types have and organic carbon contents (Table 6) Mayer et al. 2006, Anderson & Hoffman largely been characterised by reverse with increasing historic grazing inten- 2007, Pufal et al. 2008, Todd & Hoffman patterns of use-intensity: those camps sity (Labitzky 2009). However, due to 1999, 2009) and communal farmlands that were heavily grazed in the past are the generally high heterogeneity of soils

140 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT in the area, it is extremely diffi cult to separate landuse from other factors that may have an infl uence on soil properties. Since general landuse effects on all soil types cannot be proved, we compared only those soil types occurring in suffi - cient amounts in all grazing classes (i.e. Leptosols and Durisols). We found a signifi cant correlation between plant-available phosphorous and historic landuse intensity for all soil types. For C/N ratio, signifi cant differences between landuse classes were restricted to Durisols and for soil pH to Leptosols, respectively (Labitzky 2009). Succulent Karoo However, the relationship between nutrient enrichment or soil pH and historic grazing intensity as indicated by our analysis above is unlikely to be causal as they contradict the fi ndings from Fig. 13: Topsoil pH under different historical grazing intensities: low (N = 26), moderate (N = other studies. In contrast to our fi ndings, 25), severe (N = 24), and unspecifi ed (N = 16). most other studies showed that degraded soils along grazing gradients tend to be more alkaline (Allsopp 1999a, Carrick Table 6: Correlation of topsoil properties to historic grazing intensity 2003, Mills & Fey 2003, Gebremeskel & Pieterse 2007) due to a loss of vegetation cover, which in turn reduces the amounts Median for low Median for high of basic cations absorbed and acids se- grazing intensity grazing intensity creted by plants (Mills & Fey 2003). Parameter (N = 26) (N = 24) Significance of t-test

Hence, we assumed that other factors may pHH2O 7.7 6.7 0.001 be responsible for these trends and there- Corganic (%) 0.64 0.84 0.009 fore also compared our soil data with re- Ntotal (%) 0.065 0.076 0.018 cent grazing intensities, which are partly C/N-ratio 10.3 11.2 0.024 a reversal of historic grazing trends (see Ppv (g/kg) 0.042 0.013 0.013 above). The results of this analysis were independent two-way t-test, pv = plant available more in line with previous studies, with pH increasing and C/N-ratio decreasing signifi cantly from low to intermediate grazing intensities (the differences due to high grazing intensity was not signifi cant due to the small sample size). Although there is a statistically signifi - cant relationship evident, this does not necessarily indicate a causal relationship. The soil heterogeneity is extremely high and soil and historic landuse are not randomly spread across the commonage but follow a north-south gradient (Fig. 10), which seems to be related to a gradient of increasing fog and rainfall precipitation towards the south (B. Weber, unpubl. data). Hence, the observed trend in soil pH may be driven either by rainfall or grazing pressure or by both. Vascular plants: The vegetation patterns Photo 20: Soebatsfontein commonage after good seasonal rainfall in August 2006. Photo: of vascular plants on the Soebatsfontein Ute Schmiedel.

SUCCULENT KAROO 141 Table 7: Grazing response of heuweltjie centres (multiple linear regressions)

StandardiVed coefficient SEM Estimate SEM Tolerance t p

pH, Adjusted R² = 0.084, F (1,21) = 3.006, p = 0.098 Intercept 8.317 ±0.153 54.232 < 0.001 Relative historic grazing intensity –0.354 ±0.204 –0.003 ±0.002 1.000 –1.734 0.098

th CaCO3 (4 root), Adjusted R² = 0.034, F (1,21) = 1.763, p = 0.199 Intercept 1.387 ±0.191 7.257 < 0.001 Relative historic grazing intensity –0.278 ±0.210 –0.003 ±0.002 1.000 –1.328 0.199

Succulent Karoo Vegetation cover (log10), Adjusted R² = 0.427, F (2,20) = 9.181, p = 0.001 Intercept 1.330 ±0.113 11.725 < 0.001 Relative recent grazing intensity –0.578 ±0.174 –0.027 ±0.008 0.864 –3.328 0.003 Relative historic grazing intensity 0.223 ±0.174 0.001 ±0.001 0.864 1.281 0.215

Cumulative palatability, Adjusted R² = 0.022, F (1,21) = 1.505, p = 0.233 Intercept 13.500 ±6.703 2.014 0.057 Relative historic grazing intensity 0.259 ±0.211 0.082 ±0.067 1.000 1.227 0.233

Species richness, Adjusted R² = 0.137, F (2,20) = 2.740, p = 0.089 Intercept 13.323 ±5.046 2.640 0.016 Relative historic grazing intensity 0.322 ±0.213 0.054 ±0.036 0.864 1.510 0.147 Relative recent grazing intensity –0.236 ±0.213 –0.406 ±0.367 0.864 –1.105 0.282

Shannon index, Adjusted R² = 0.027, F (1,21) = 1.615, p = 0.218 Intercept 1.030 ±0.273 3.767 0.001 Relative historic grazing intensity 0.267 ±0.210 0.003 ±0.003 1.000 1.271 0.218

Cover annuals (log10), Adjusted R² = 0.007, F (1,21) = 1.162, p = 0.293 Intercept 0.135 ±0.157 0.860 0.400 Northing 0.229 ±0.212 0.260 ±0.242 1.000 1.078 0.293

Chamaephyte cover, Adjusted R² = 0.302, F (2,20) = 5.768, p = 0.011 Intercept 15.256 ±7.160 2.131 0.046 Relative recent grazing intensity –0.421 ±0.192 –1.147 ±0.521 0.864 –2.200 0.040 Relative historic grazing intensity 0.305 ±0.192 0.081 ±0.051 0.864 1.594 0.127

Bioturbation (4th root), Adjusted R² = 0.070, F (1,21) = 2.652, p = 0.118 Intercept 2.061 ±0.468 4.402 < 0.001 Relative historic grazing intensity –0.335 ±0.206 –0.008 ±0.005 1.000 –1.628 0.118

N = 23 Model selection via forward selection, significant partial estimates with bold p, F-to-enter = 1.0 Full models include predictor variables recent and historic grazing intensity, northing.

commonage (Photo 20) are strongly driven These strong abiotic drivers of vegeta- features on the vegetation, we restricted by the aforementioned soil heterogeneity. tion patterns seem to mask the historic our analysis to vegetation and soil vari- In total, 17 different plant communities grazing effects on the scale of the present ables of plots at the centre of heuweltjies. have been identiﬁ ed on the commonage, vegetation units in Soebatsfontein Heuweltjie centres are clearly distinct with the main driving factors being soil (Luther-Mosebach 2009). To distinguish from their surroundings in terms of soil water availability, salinity and soil pH the effect of historic or recent grazing and vegetation features (Röwer 2009) (Luther-Mosebach 2009, see Subchapter intensity from the effects of the above- and are preferred grazing areas due to IV.4.2). mentioned latitudinal gradient in soil the presence of higher concentrations of

142 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT ions (Armstrong & Siegfried 1990) and can thus be employed as sensitive indicators of grazing impacts. We used historic and recent grazing classes as well as orientation towards north (northing) as a substitute for exposition to the four main points of the compass as predictor variables for the model (Table 7). The analyses showed no correlation of any of the predictor variables with any of the soil variables. The only signifi cant correlation was the negative correlation of recent grazing intensity with total vegetation and chamaephyte (i.e. shrub) cover. Historic grazing intensity did not Succulent Karoo show any signifi cant results. The fact that we did not fi nd any response of the soil variables with grazing intensity supports our earlier fi nding that soil variables on Fig. 14: Cover of the different types of biological soil crusts (BSC; N = 60) on the Soebatsfontein heuweltjies are rather unique (Röwer commonage. 2009, Subchapter IV.4.2). They also did not seem to be affected by the climatic north-south gradient on the Soebatsfon- Table 8: Response of biological soil crusts (BSC) to two grazing related parameters tein commonage. It can thus be assumed that the observed response of the total Local utiliVation (number of Distance to stock post or vegetation and chamaephytes cover to faecal pellets/0.375 m²) watering point (m) recent grazing intensity is not caused by r p r p collinearity with any of these factors. A Total BSC cover (%) –0.340 0.0079 0.257 0.0475 decrease in chamaephyte (shrub) cover Initial BSC cover (%) –0.099 0.4510 0.024 0.8575 in favour of annual plant cover has been Climax BSC cover (%) –0.340 0.0079 0.275 0.0337 described as a characteristic response of Spearman rank test, N = 60 Namaqualand vegetation to high grazing pressure (Todd & Hoffman 1999, 2009, Anderson & Hoffman 2007). The cover of annual plants on sites with high graz- the exclosure, but the cover and species highly variable (0–60%, median 25%; ing pressure varies with the timing and richness of perennials inside the exclosure Fig. 14), which complies with our remote total amount of seasonal rainfall (Todd & decreased by 25% and 27%, respectively sensing assessments (see Article III.4.1) Hoffman 1999), which could also explain (see also Subchapter IV.4.4). The results as well as with fi ndings from the Kala- the absence of any signifi cant response to indicate that increasing current grazing hari (Dougill & Thomas 2004, Thomas recent grazing intensity. pressure is having a negative impact on & Dougill 2006). Coverage of initial Based on the analysis of the exclosure the vegetation cover. Grazing exclosures and well developed (climax) BSC in the camps at Quaggafontein and Patrysegat, facilitated an increase of annual plant cov- study area was similar (median 12.7% we found a measurable effect of recent er. Moderate grazing, however, allowed and 9.7%, respectively; Fig. 14) but that grazing on the vegetation (Meyer 2009). an increase in perennial vegetation with of climax BSC varied more strongly. In the camp with intensive recent grazing the annual plant cover remaining the same However, both the amount and composi- (Quaggafontein), the vegetation of the inside and outside the exclosure. tion of BSC depended on current landuse control plots (plots without active res- Biological soil crusts: Our study of patterns. The total coverage of biological toration measures) inside the exclosure the soil and biological soil crusts (BSC) soil crusts was found to be negatively showed a more signifi cant increase in showed measurable effects of current correlated with local utilisation intensity the cover of annual plant species (mean landuse. The soils of Soebatsfontein are (according to density of livestock faecal = 7.48%) than grazed plots outside (mean not only extraordinary diverse, but their pellets) and positively correlated with the = 3.88%). Also, total species richness in- surfaces also contain an extraordinarily distance to the next stock post or water- creased by 25% inside the exclosure. Un- high cover and diversity of biological soil ing point (Table 8). The response of BSC der moderate recent grazing at Patrysegat, crusts (Büdel et al. 2009). to landuse intensity depended on their the annual vegetation cover of the control Total cover of biological soil crust successional stage: whereas the cover plots was unaffected by the resting inside on the Soebatsfontein commonage was of well-developed climax BSC and total

SUCCULENT KAROO 143 BSC responded very similarly, the cover- occur homogenously within each camp. and unequal utilisation of the camp. age of initial BSC showed no response Our studies on disturbance-sensitive eco- Many farmers see “kraaling” as criti- to grazing (Table 8). The results suggest system components such as biological cal to protect their livestock against a strong infl uence of grazing both on the soil crusts and vegetation at the centre of predators (mainly jackals and caracal). absolute coverage and the composition heuweltjie, as well as the comparison of However, alternative strategies such as of BSC. The high sensitivity of BSC to- fence-line contrasts (inside and outside the introduction of Andalusian sheep- wards disturbances, e.g. grazing, is well- of exclosures), revealed local effects of dogs as guards for the livestock (cur- described in the literature (Belnap 1995, current grazing practices on the Soebats- rently tested by South African National Belnap & Gillette 1998, Warren & El- fontein ecosystem. Parks) or other strategies recommend- dridge 2003) and especially well-devel- With regard to biological soil crusts, ed by The Cape Leopard Trust (Quin- oped BSC (climax BSC) are known to be both their abundance and composition ton Martins, personal communication) susceptible to grazing because they are was found to be strongly correlated with should be explored. easily destroyed and grow slowly (Bel- recent local landuse intensity. Since 3. More effi cient use of the available nap 2002a, Thomas & Dougill 2006). biological soil crusts play a major role camps could also be attained through Succulent Karoo Initial biological soil crusts, on the other in semi-arid and arid ecosystems by ef- the installation of more, or strategi- hand, can recolonise open soil within fectively reducing soil erosion and pro- cally better distributed, water points eight months or one rainfall season and moting the growth and nutrient status of within the camps. A denser network of are therefore less affected by landuse im- vascular plants (Belnap & Eldridge 2003, water points could potentially reduce pacts (Dojani et al., submitted). In areas Belnap et al. 2003c), their large-scale ex- the local impacts experienced around with low grazing and trampling intensi- istence has to be ensured. Only if landuse the current water points. Such measure ties, climax BSC start to replace initial intensity is kept within a moderate range could be further enhanced if only small BSC within 20 months, whereas in areas are biological soil crusts able to provide amounts of water were supplied at each with high livestock utilisation intensities, their invaluable ecosystem services that water point, as is applied by farmers constant disturbance probably prevents promote sustainable land management. in Australia (B. Büdel, pers. comm.). succession (Dojani et al., submitted). However, it is likely that landuse pres- This would support the movement of sure on the Soebatsfontein commonage herds within the camps. Consequences for land will increase in the future. The main rea- 4. The condition of fences also needs to management in Soebatsfontein son for this is the closure of the surround- be improved in order to keep livestock The Soebatsfontein commonage is an ing mines, with the result that former in the delineated camps. Future land extraordinarily (bio-) diverse farmland, a mine workers have the fi nancial means management strategies should be ac- fact that is also appreciated by most of the and economic vision to become fulltime companied by a simple but effective members of the Soebatsfontein commu- farmers for their livelihood (Anseeuw & monitoring programme, which com- nity who are the current custodians of the Laurent 2007)—a general trend that also bines landuser experiences and criteria land. Our study indicates that heteroge- applies to the Soebatsfontein community (Desert Research Foundation of Na- neity in geomorphology, soil water avail- (U. Schmiedel, own observations). Con- mibia 2005) with quantitative, scientif- ability, bioturbation (heuweltjies), and certed efforts towards sustainable land ically sound methods. The monitoring soil characteristics are probably the main management will therefore become in- activities should assess the effects of drivers of this high biological diversity. creasingly important. management on the natural resources The soil and therefore vegetation patterns Our data and observations allow us to and livestock condition, allowing for are not evenly distributed across the com- suggest some landuse management meas- the evaluation of management impacts monage but largely follow a north-south ures that should benefi t both landusers against management objectives. gradient, which may also be related to the and biodiversity by improving utilisation The mentioned recommendations to- frequency and amount of rainfall and fog of the existing resources without requiring wards more effi cient use of the natural re- from the coast. The high heterogeneity additional land and fi nancial resources. sources seem to be straight forward. How- of abiotic factors, varying durations of 1. The available grazing resources could ever, they need to be considered within the historic and recent landuse intensities, be used more effi ciently if the graz- local context. Any intervention interacts and the lack of benchmark sites without ing and resting periods per camp were with local settings, which have been shaped any grazing history, make the analysis guaranteed through proper rotational by the complex interrelationship between of the impacts of anthropogenic factors grazing and the sharing of manpower historical, social, and economic factors, diffi cult. Therefore, our data on soils and for herding. This may require tempo- and have to be adapted to the community vascular plants on the Soebatsfontein rary sharing of camps in order to allow context. For example, cooperative manage- commonage does not show unambiguous other camps to rest. ment to improve grazing and herding strat- signs of an overall impact by either his- 2. The gathering of animals overnight egies is a diffi cult task to undertake. This is torical or current landuse patterns. at the stock post (“kraaling”) requires particularly true at Soebatsfontein, where However, grazing and trampling by much movement of the animals and farmers and community members cannot livestock is typically patchy and does not therefore causes additional trampling build on past experience of joint manage-

144 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT ment, as the farmland was only transferred in a low recommended carrying capac- be adapted to the specifi c conditions of the to the community in 2000. This may also ity (10 ha/SSU, Baker & Hoffman 2006) respective communities. The case of Soe- be one of the reasons why socio-economic and relatively low agricultural potential of batsfontein showed that even if camps are analysis revealed that social tensions are the area (Desmet 2007). This is particu- allocated to single farmers, co-operative apparent and trust is comparatively low larly problematic for communal farmers management strategies could lead to a within this community. Furthermore, the that have restricted fi nancial and natural more effective use of the available resourc- management of the commonage is con- resources. The lack of alternative liveli- es. However, “young” communal farming strained by economic problems, as the un- hood options (Anseeuw & Laurent 2007, communities like Soebatsfontein received employment rate is high, employment op- Subchapter IV.3.5) and lack of appropriate access to communal land relatively re- portunities are restricted, and the farmland technical skills and training make agricul- cently and cannot draw on long-standing is too small to serve as an additional source ture an important source of income for a experiences regarding joint and coordi- of income for the entire community. signifi cant number of people living in the nated pasture management. Furthermore, Therefore, the formulation of an ap- communal areas. interpersonal tensions as well as social proach to improve land management on However, the biodiversity contained and economical problems in the commu- the Soebatsfontein commonage should within the over-crowded, heavily-grazed nal farming communities in Namaqualand Succulent Karoo be developed following a participatory communal areas has often been trans- have to be taken into account. process and should be based on several formed in terms of both its species and Most of the communities are not ho- steps, which are outlined in the overall life form composition (Todd & Hoff- mogeneous entities but are rather char- conclusion. man 1999, 2009, Anderson & Hoffman acterised through groups and actors with 2007, Rutherford & Powrie 2010) with different interests and capacities. Such knock-on effects for other animal groups challenges could be overcome though 4.6 Conclusions and (Article III.5.7) and the abiotic environ- a strongly transdisciplinary (Max- further research needs ment (Article III.3.3, Subchapter IV.4.3). Neef 2005) and participatory process [U. Schmiedel & T. Linke] In particular, our studies on quartz fi elds (Wadsworth 1998, Kemmis & McTag- in the Knersvlakte showed that local and gart 2000) as part of participatory action The Succulent Karoo Biome in general habitat endemics seem to be most strongly research. Such an approach, which can be and Namaqualand (i.e. the western part affected by such homogenising impacts of taken as a fi rst immediate step to improve of the Succulent Karoo Biome) in partic- trampling and ripping of the soil surface the social and natural environment, has to ular, has been widely acknowledged for (Etzold 2006, Haarmeyer et al. 2010). be focussed around the communal farm- its high biodiversity (Myers et al. 2000). However, ploughing, mainly applied by ing communities but also involve vari- The BIOTA Observatories in the Succu- commercial farmers to extensive areas ous other local stakeholder groups (e.g. lent Karoo harbour the highest density of may have a much more devastating and municipality, extension services, nature vascular plant species per surface area at long-lasting effect (Subchapter IV.4.3) conservation authorities, rural develop- various spatial scales along the arid part than overgrazing (Rahlao et al. 2008). ment initiatives but also researchers and of the BIOTA transects (Article III.3.8). Our socio-economic studies from Soe- well-trained para-ecologists; see Article The plant species richness of the Succu- batsfontein have shown that the solutions III.8.3). The aim of such an approach is lent Karoo is only exceeded by the Fyn- for the challenges of communal land to assess the current challenges that are bos Biome, which receives much higher management, which have been described experienced by the farmers and commu- annual rainfall. The high biodiversity of in Subchapter IV.4.5, have to go beyond nity members and to jointly explore, test, the Succulent Karoo has been related to scientifi c recommendations concerning and evaluate new approaches towards an various factors (for overview on recently improvement in rangeland and herd man- adapted and cooperative management of discussed drivers of biodiversity in the agement and changes in infrastructure. ecosystem services. The participatory ac- Succulent Karoo see also Mucina et al. In the long-term, the economically- and tion research process would also require 2006) related to current and historic cli- socially-disadvantaged communities on the involvement of competent community matic features (Desmet & Cowling 1999) the communal lands of Namaqualand also workers that can help to resolve the social but also the habitat and soil diversity of need perspectives that go beyond the cur- and economic challenges within the small the area has been identifi ed as an im- rent landuse and livelihood options. Thus communities as a precondition for fruitful portant driver (Jürgens 1986, Schmiedel the approach towards sustainable manage- cooperation. Cooperative management & Jürgens 1999, Herpel 2008, Petersen ment of the communal lands in Namaqua- can improve the economic situation and 2009, Petersen et al. 2010, Articles land should follow several steps that re- strengthen empowerment of small-scale III.3.3, III.3.8, Subchapter IV.4.2). quire different timelines, structured in farmers as experiences from other projects Low annual rainfall (app. 50–250 mm a-1, immediate steps and medium-term (dec- in Namaqualand have shown (Nel et al. Desmet 2007, see Part II.2) and the domi- ade) to long-term (several decades) strate- 2007, Oettlé 2005, Oettlé et al. 2009). nance of slow-growing dwarf shrubs, re- gies, which are outlined in the following. Participatory approaches should also fa- sults in a comparatively low primary pro- The implementation of a sustainable cilitate access to farmland and to farming duction (Article III.3.10), which results management plan, as a fi rst step, has to resources for economically disadvantaged

SUCCULENT KAROO 145 households and particularly women, who the area in general, alternative livelihood Anseeuw, W., Laurent, C. (2007): Occupational currently benefi t least from communal options need to be explored further for paths towards commercial agriculture: The key roles of farm pluriactivity and the commons. – farmland (Kleinbooi & Lahiff 2007, the region. Pluriactivity is a widespread Journal of Arid Environments 70: 659–671. Lebert & Rohde 2007, Subchapter IV.4.5). and successful strategy for farmers in Armstrong, A.J., Siegfried, W.R. (1990): Selec- This aim would also be in line with the ob- this relatively unpredictable natural and tive use of heuweltjies earth mounds by sheep. – South African Journal of Ecology 1: 77–80. jectives of the South African land reform social environment (Anseeuw & Laurent Aronson, J., Blignaut, J.N., Milton, S.J., Le Mai- programme (May & Lahiff 2007). 2007, Cousins et al. 2007). However, tre, D., Esler, K.J., Limouzin, A., Fontaine, C., In a medium-term perspective, access pluriactivity and alternative livelihoods Wit, M.P. de, Mugido, W., Prinsloo, P., Elst, L. van der, Lederer, N. (2010): Are socioeconomic to more farmland is needed to make farm- require broadening the fi eld of experi- benefi ts of restoration adequately quantifi ed? A ing economically viable for a larger group ences and skills in the rural communities. meta-analysis of recent papers (2000-2008) in of community members in Namaqualand, Therefore, in a long-term perspective, restoration ecology and 12 other scientifi c jour- nals. – Restoration Ecology 18: 143–154. for whom small-livestock farming appears secondary and tertiary education has to Bainbridge, D.A. (2007): A guide for desert and to be the most important, if not only, liveli- be mainstreamed and vocational training dryland restoration. New hope for arid lands. – Washington, D.C.: Society for Ecological hood option. All those working and living opportunities are needed in order to de- Restoration International. Succulent Karoo in Namaqualand need to appreciate the velop alternative livelihoods for the rural Baker, L.E., Hoffman, M.T. (2006): Managing historical legacy of apartheid in the area community members of Namaqualand. variability: herding strategies in communal rangelands of semiarid Namaqualand, South and the signifi cant social, cultural, and Africa. – Human Ecology 34: 765–784. ecological impact it has had on people’s Bazzaz, F.A. (2000): Plants in changing environ- lives as well as the current spatial arrange- ments. Linking physiological, population and Acknowledgements community ecology. – Cambridge: Cambridge ment of communal and private farms. The authors’ general acknowledgements to the University Press. However, communal farmland, currently organisations and institutions, which supported Belnap, J. (1995): Surface disturbances: their about 37% of Namaqualand area (Desmet this work are provided in Volume 1. role in accelerating desertifi cation. – Environ- mental Monitoring and Assessment 37: 39–57. 2007), has to be shared with other landuse References Belnap, J. (2002a): Biological soil crusts of Ara- options and related ecosystem services AG Boden (2005): Bodenkundliche Kartieranlei- bian Sabkhat. – In: Barth, H.J., Boer, B. (eds.): like farming on privately owned land (cur- tung. Ed. 5. – Hannover: Schweizerbart’sche Sabkha Ecosystems: 227–237. Dordrecht: Verlagsbuchhandlung. Kluwer Academic Publishers. rently about 50% of the area in Namaqua- Agriculture Research Council (undated): AGIS. Belnap, J. (2002b): Nitrogen fi xation in biologi- land, Desmet 2007) and nature conserva- Agricultural geo-referenced information sys- cal soil crusts from southeast Utah, USA. – tem. – www.agis.agric.za [acc. 10.02.2010]. Biology and Fertility of Soils 35: 128–135. tion and (eco-) tourism (only 3% of the Allsopp, N. (1999a): Consequences of vegeta- Belnap, J. (2006): The potential roles of biologi- land in 2007, Desmet 2007). The latter has tion change associated with livestock manage- cal soil crusts in dryland hydrologic cycles. – become increasingly important in the Suc- ment to soil patterns and processes in southern Hydrological Processes 20: 3159–3178. Africa. – In: D. Eldridge, D. Freudenberger Belnap, J., Eldridge, D. (2003): Disturbance and culent Karoo Biome over the last decade, (eds.): People and rangelands building the recovery of biological soil crusts. – In: Belnap, which has resulted is a perceived ‘com- future. Proceedings of the VIth International J., Lange, O.L. (eds.): Biological soil crusts: petition’ for land with other landuse op- Rangeland Congress 1: 96–101. Aithkenvale: structure, function and management. – Eco- International Rangeland Congress. logical Studies 150: 363–383. Berlin: Springer. tions like communal farming or farming Allsopp, N. (1999b): Effects of grazing and Belnap, J., Gardner, J.S. (1993): Soil microstruc- by emerging farmers. This has resulted cultivation on soil patterns and processes in ture in soils of the Colorado Plateau: The role in a somewhat irrational and unfortunate the Paulshoek area of Namaqualand. – Plant of the cyanobacterium Microcoleus vaginatus. Ecology 142:179–187. – Great Basin Naturalist 53: 40–47. dichotomy (see Benjaminsen et al. 2008) Allsopp, N., Gaika, L., Knight, R., Monakisi, C., Belnap, J., Gillette, D.A. (1998): Vulnerability of between “biodiversity conservation” on Hoffman, M.T. (2007a): The impact of heavy desert biological soil crusts to wind erosion: the one hand and “poverty alleviation” on grazing on an ephemeral riversystem in the The infl uence of crust development, soil tex- succulent karoo, South Africa. – Journal of ture, and disturbance. – Journal of Arid Envi- the other (Roe 2008), which is also known Arid Environments 71: 82–96. ronments 39: 133–142. from similar situations in other regions Allsopp, N., Laurent, C., Debeaudoin, L.M.C., Belnap, J., Rosentreter, R., Leonard, S., Kalte- (Shackleton et al. 2001). Again, a par- Samuels, M.I. (2007b): Environmental per- necker, J.H., Williams, J., Eldridge, D. (2001): ceptions and practices of livestock keepers on Biological soil crusts: ecology and Manage- ticipatory action research process should the Namaqualand Commons challenge con- ment. – Technical Reference 1730-2. Denver: tackle this perceived dichotomy and ex- ventional rangeland management. – Journal of U.S. Department of the Interior, Bureau of Arid Environments 70: 740–754. Land Management. plore the scope for synergies between the Anderson, P., Hoffman, M.T., Holmes, P.M. Belnap, J., Büdel, B., Lange, O.L. (2003a): Bio- different landuse options (see for instance (2004): The potential of Cephalophyllum logical soil crusts: characteristics and distri- a discussion of the fl ower display at fal- inaequale (L. Bolus) for the restoration on de- bution. – In: Belnap, J., Lange, O.L. (eds.): graded arid landscapes in Namaqualand, South Biological soil crusts: structure, function, and low ploughed lands as a potential tourist Africa. – Restoration Ecology 12: 343–351. management. - Ecological Studies 150: 3–30. attraction, Subchapter IV.4.3). This is par- Anderson, P.M.L., Hoffman, M.T. (2007): The Berlin: Springer. ticularly critical considering the threat- impacts of sustained heavy grazing on plant Belnap, J., Phillips, S., Duniway, M., Reynolds, diversity and composition in lowland and up- R. (2003b): Soil fertility in deserts: a review ened economic and ecological viability land habitats across the Kamiesberg mountain on the infl uence of biological soil crusts and of arid Namaqualand under the current range in the Succulent Karoo. – Journal of the effect of soil surface disturbance on nutri- climate change projections (Midgley & Arid Environments 70: 686–700. ent inputs and losses. – In: Alsharhan, A.S., Anonymous (1986): Agricultural development Wood, W.W., Goudie, A.S., Fowler, A., Abdel- Thuiller 2007, Article III.2.1). programme: Karoo region. – Unpublished re- latif, E.M. (eds.): Desertifi cation in the third Due to climate change projection and port. Pretoria: Department of Agriculture and millennium: 245–252. Lisse: Swets & Zeitlin- low economic potential of agriculture in Water Supply. ger Publishers.

146 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT Belnap, J., Prasse, R., Harper, K.T. (2003c): In- Carrick, P.J., Krüger, R. (2007): Restoring de- Dunkerley, D. L. (2008): Intra-storm evaporation fl uence of biological soil crusts on soil envi- graded landscapes in lowland Namaqualand: as a component of canopy interception loss in ronments and vascular plants. – In: Belnap, lessons from the mining experience and from dryland shrubs: observations from Fowlers J., Lange, O.L. (eds.): Biological soil crusts: regional ecological dynamics. – Journal of Gap, Australia. – Hydrological Processes 22: structure, function and management. – Eco- Arid Environments 70: 767–781. 1985–1995. logical Studies 150: 281–300. Berlin: Springer. Coetzee, K. (2005): Caring for natural rangelands. – Ellis, F. (2002): Contributions of termites to the Benjaminsen, T.A., Kepe, T., Bråthen, S. (2008): Scottsville: University of KwaZulu-Natal Press. formation of hardpans in soils of arid and semi- Between global interests and local needs: con- Connell, J.H. (1978): Diversity in tropical rain arid regions of South Africa. – Proceedings of servation and land reform in Namaqualand, forests and coral reefs – high diversity of trees the 17th WCSS, Paper No 572: 1–9. Bangkok: South Africa. – Africa 78: 223–244. and corals is maintained only in a non-equilib- Soil and Fertilizer Society of Thailand. Berg, J., Dickhaut, J., McCabe, K. (1995): Trust rium state. – Science 199: 1302–1310. Epps, G.A. van, McKell, C.M. (1983): Effect of reciprocity and social history. – Games and Cousins, B., Hoffman, M.T., Allsopp, N., Rohde, weedy annuals on the survival and growth of Economic Behavior 10: 142. R.F. (2007): A synthesis of social and bio- transplants under arid conditions. – Journal of Berg, L. van den, Kellner, K. (2005): Restoring logical perspectives on sustainable land use in Range Management 36: 366–369. degraded patches in a semi-arid rangeland of Namaqualand. – Journal of Arid Environments Esler, K.J. (1999): Plant reproductive ecology. – South Africa. – Journal of Arid Environments 70: 834–846. In: Dean, W.R., Milton, S.J. (eds.): The Karoo: 61: 497–511. Cowling, R.M., Pierce, S.M. (1999): Namaqua- ecological patterns and processes: 123–144. Berzborn, S. (2007): The household economy of land – a succulent desert. – Cape Town: Fern- Cambridge: Cambridge University Press. pastoralists and wage-labourers in the Rich- wood Press. Esler, K.J., Cowling, R.M. (1995): The com- tersveld, South Africa. – Journal of Arid En- Cowling, R.M., Hilton-Taylor, C. (1999): Plant parison of selected life-history characteris- Succulent Karoo vironments 70: 672–685. biogeography, endemism and diversity. – In: tics of Mesembryanthema species occuring Beukes, P.C., Cowling, R.M. (2003): Evaluation of Dean, W.R.J., Milton, S.J. (eds.): The Karoo: on and off Mima-like mounds (heuweltjies) restoration techniques for the Succulent Karoo, ecological patterns and processes: 42–85. in semi-arid southern Africa. – Vegetatio South Africa. – Restoration Ecology 11: 308–316. Cambridge: Cambridge University Press. 116: 41–50. Beukes, P.C., Ellis, F. (2003): Soil and vegetation Cowling, R.M., Esler, K.J., Rundel, P.W. (1999): Etzold, S. (2006): Dynamics on quartz fi elds in changes across a Succulent Karoo grazing gra- Namaqualand, South Africa – an overview of the Knersvlakte/South Africa. With special dient. – African Journal of Range and Forage a unique winter-rainfall desert ecosystem. – references to disturbance and restoration. – Science 20: 11–19. Plant Ecology 142: 3–21. Diplom thesis in biology. Hamburg: Univer- Blignaut, A., Milton, S.J. (2005): Effects of mul- Cramer, V.A., Hobbs, R.J. (eds.) (2007): Old sity of Hamburg. tispecies clumping on survival of three succu- fi elds. Dynamics and restoration of abandoned Evans, R.D., Lange, O.L. (2001): Biological soil lent plant species translocated onto mine spoil farmland. – Washington, D.C.: Island Press. crusts and ecosystem nitrogen and carbon dy- in the Succulent Karoo Desert, South Africa. Cramer, V.A., Hobbs, R.J., Standish, R.J. (2007): namics. – In: Belnap, J., Lange, O.L. (eds.): – Restoration Ecology 13: 15–19. What’s new about old fi elds? Land abandon- Biological soil crusts: structure, function, and Bonet, A. (2004): Secondary succession of semi- ment and ecosystem assembly. – Trends in management. Ecological Studies 150: 263– arid Mediterranean old-fi elds in south-eastern Ecology and Evolution 23: 104–112. 279. Berlin: Springer. Spain: Insights for conservation and restora- Critchley, W., Reij, C., Willcocks, T.J. (1994): Falk, T. (2008): Communal farmers’ natural re- tion of degraded lands. – Journal of Arid Envi- Indigenous soil and water conservation: a resource use and biodiversity preservation. A ronments 56: 213–233. view of the state of knowledge and prospects new institutional economic analysis from case Botha, S.M, Carrick, P.J., Allsopp, N. (2008): for building on tradition. – Land Degradation studies in Namibia and South Africa. – Göt- Capturing lessons from land-users to aid the and Rehabilitation 5: 293–314. tingen: Cuvillier. development of ecological restoration guide- Debussche, M., José, E., Lepart, J., Houssard, C., Francis, M.L., Fey, M.V., Prinsloo, H.P., Ellis, lines for lowland Namaqualand. – Biological Lavorel, S. (1996): Changes in Mediterranean F., Mills, A.J., Medinski, T.V. (2007): Soils Conservation 141: 885–895. plant succession: old-fi elds revisited. – Journal of Namaqualand: compensations for aridity. Brunner, E., Domhof, S., Langer, F. (2002): Non- of Vegetation Science 7: 519–526. – Journal of Arid Environments 70: 588–603. parametric analysis of longitudinal data in fac- Desert Research Foundation of Namibia (2005): Gabriels, N. Allsopp, N., Hoffman, T. (2003): torial experiments. – New York: Wiley. Local level monitoring for enhanced decision- Seedling establishment under livestock pres- Büdel, B., Weber, B., Kühl, M., Pfanz, H., Sülte- making. Farmer’s Field Guide. – Windhoek: sure in Succulent Karoo systems: an impos- meyer, D., Wessels, D. (2004): Reshaping of Ministry of Agriculture, Water and Forestry, Na- sible scenario? – In: Allsopp, N., Palmer, A.R., sandstone surfaces by cryptoendolithic cy- tional Agriculture Support Services Programme Milton, S.J., Kirkman, K.P., Kerley, G.I.H., anobacteria: bioalkalization causes chemical (NASSP) and European Commission (EDF 8). Hurt, C.R., Brown, C.J. (eds.): Proceedings of weathering in arid landscapes. – Geobiology Desmet, P.G. (2007): Namaqualand – a brief over- the VIIth International Rangelands Congress. 2: 261–268 view of the physical and fl oristic environment. 26th July – 1st August 2003, Durban, South Af- Büdel, B., Darienko, T., Deutschewitz, K., Dojani, – Journal of Arid Environments 70: 570–587. rica: 65–68. S., Friedel, T., Mohr, K.I., Salisch, M., Reisser, Desmet, P.G., Cowling, R.M. (1999): The cli- Gebremeskel, K., Pieterse, P.J. (2007): Impact of W., Weber, B. (2009): Southern African bio- mate of the Karoo – a functional approach. – grazing around a watering point on soil status logical soil crusts are ubiquitous and highly In: Dean, W.R., Milton, S.J. (eds.): The Karoo: of a semi-arid rangeland in Ethiopia. – African diverse in drylands, being restricted by rainfall ecological patterns and processes: 3–16. Cam- Journal of Ecology 45: 72–79. frequency. – Microbial Ecology 57: 229–247. bridge: Cambridge University Press. Germishuizen, G., Meyer, N.L. (2003): Plants Cadenasso, M., Pickett, S.T.A., Morin, P.J. (2002): Dojani, S., Lakatos, M., Rascher, U., Wanek, W., of southern Africa: an annotated checklist. – Experimental test of the role of mammalian Lüttge, U., Büdel, B. (2007): Nitrogen input Strelitzia 14. Pretoria: South African National herbivores on old fi eld succession: Community by cyanobacterial biofi lms of an inselberg into Botanical Institute. structure and seedling survival. – Journal of the a tropical rainforest in French Guiana. – Flora Grime, J.P. (1973): Competitive exclusion in her- Torrey Botanical Society 129: 228–237. 202: 521–529. baceous vegetation. – Nature 242: 344–347. Callaway, R.M., Walker, L.R. (1997): Competi- Domingo, F., Sánchez, G., Moro, M.J., Brenner, Grime, J.P. (1979): Plant strategies and vegetation tion and facilitation: a synthetic approach to A.J., Puigdefábregas, J. (1998): Measurements processes. – Chichester: John Wiley & Sons. interactions in plant communities. – Ecology and modelling of rainfall interception by three Haarmeyer, D.H. (2009): Effects of domestic 78: 1958–1965. semi-arid canopies. – Agricultural and Forest livestock on the vegetation of the Knersvlakte, Campbell, N.A., Reece, J.B. (2005): Biology. Meteorology 91: 275–292. South Africa. – Diplom thesis. Hamburg: Uni- Ed. 7. – San Francisco: Pearson Education/ Dougill, A.J., Thomas, A.D. (2004): Kalahari versity of Hamburg. Benjamin Cummings. sand soils: spatial heterogeneity, biological Haarmeyer, D.H., Schmiedel, U., Dengler, J., Carrick, P. J. (2003): Competitive and facilita- soil crusts and land degadation. – Land Degra- Bösing, B.M. (2010): How does grazing in- tive relationships among three shrub species, dation and Development 15: 233–242. tensity affect different vegetation types in arid and the role of browsing intensity and rooting Dregne, H.E. (1992): Degradation and rehabilita- Succulent Karoo, South Africa? Implications depth in the Succulent Karoo, South Africa. – tion of arid lands. – Texas: International Cen- for conservation management. – Biological Journal of Vegetation Science 14: 761–772. tre for Arid and Semiarid Land Studies. Conservation 143: 588–596.

SUCCULENT KAROO 147 Hartmann, H.E.K. (2002a): Illustrated handbook Kleinbooi, K., Lahiff, E. (2007): „Die man is Merwe, H. van der (2009): Patterns of plant of succulent plants: Aizoaceae A-E. – Berlin: die hoof en vat voor”: Women’s attitudes to diversity in the Hantam-Tanqua-Roggeveld Springer. land and farming in the communal areas of Subregion, Succulent Karoo, South Africa. – Hartmann, H.E.K. (2002b): Illustrated handbook Namaqualand. – Journal of Arid Environments PhD thesis. Pretoria: University of Pretoria. of succulent plants: Aizoaceae F-Z. – Berlin: 70: 799–817. Meyer, S. (2009): Analysis and evaluation of ecolog- Springer. Knight, R.S., Rebelo, A.G., Siegfried, W.R. ical restoration experiments in the Succulent Ka- Hennekens, S.M., Schaminée, J.H.J. (2001): (1989): Plant Assemblages on Mima-like earth roo (South Africa). – Diplom thesis in Geoecol- TURBOVEG, a comprehensive data base mounds in the Clanwillian district, South Af- ogy. Karlsruhe: University of Karlsruhe (TH). management system for vegetation data. – rica. – South African Journal of Botany 55: Midgley, G.F., Musil, C.F. (1990): Substrate ef- Journal of Vegetation Science 12: 589–591. 465–472. fects of zoogenic soil mounds on vegetation Herpel, N. (2008): The scale-dependent variability Labitzky, T. (2009): Natural features and landuse composition in the Worcester- Robertson val- of topsoil properties refl ecting ecosystem patch- impact on soils of the Soebatsfontein region ley, Cape Province. – South African Journal of iness in drylands of Southern Africa. – Ham- (Namaqualand, South Africa). – Diplom the- Botany 56: 158–166. burger Bodenkundliche Arbeiten 62: 1–299. sis. Hamburg: Institute of Soil Science, Uni- Midgley, G.F., Thuiller, W. (2007): Potential Hiernaux, P., Bielders, C.L., Valentin, C., Ba- versity of Hamburg. vulnerability of Namaqualand plant diversity tiono, A., Fernández-Rivera, S. (1999): Effects Lavorel, S., Garnier, E. (2002): Predicting to anthropogenic climate change. – Journal of of livestock grazing on physical and chemical changes in community composition and eco- Arid Environments 70: 615–628. properties of sandy soils in Sahelian range- system functioning from plant traits: revisit- Midgley, G.F., Hannah, L., Millar, D., Ruther- lands. – Journal of Arid Environments 41: ing the Holy Grail. – Functional Ecology 16: ford, M.C., Powrie, L.W. (2002): Assessing Succulent Karoo 231–245. 545–556. the vulnerability of species richness to an- Hilton-Taylor, C. (1994): Western Cape Domain Lebert, T., Rohde, R. (2007): Land reform and thropogenic climate change in a biodiversity (Succulent Karoo). – In: Davis, S.D., Hey- the new elite: exclusion of the poor from com- hotspot. – Global Ecology and Biogeography wood, V.H., Hamilton, A.C. (eds.): Centres of munal land in Namaqualand, South Africa. – 11: 445–451. Plant Diversity 1: Europe Africa South West Journal of Arid Environments 70: 818–833. Mills, A., Fey, M.V. (2003): Declining soil qual- Asia and the Middle East: a guide and strategy Li, X.Y., Liu, L.Y., Gao, S.Y., Shi, P.J., Zou, X.Y., ity in South Africa: effects of land use on soil for their conservation: 204–224. Cambridge: Zhang, C.L. (2005): Microcatchment water organic matter and surface crusting. – South IUCN Publications Unit. harvesting for growing Tamarix ramosissima African Journal of Science 99: 429–436. Hoffman M.T., Rohde R.F. (2007): From pas- in the semiarid loess region of China. – Forest Milton, S.J. (1995): Spatial and temporal patterns toralism to tourism: The historical impact of Ecology and Management 214: 111–117. in the emergence and survival of seedlings in changing land use practices in Namaqualand. Linke, T. (2009): Ethnographische Rahmenbe- arid Karoo shrubland. – Journal of Applied – Journal of Arid Environments 70: 641–658. dingungen und rezente Entwicklungen der pas- Ecology 32: 145–156. Hoffman, M.T., Allsopp, N., Rohde, R.F. (2007): toralen Landnutzung in Soebatsfontein, Südaf- Milton, S.J., Dean, W.R.J. (1990): Mima-like Sustainable land use in Namaqualand, South rika. – Master’s thesis in Social Anthropology. mounds in the southern and western Cape: Africa: Key issues in an interdisciplinary Hamburg: University of Hamburg. are the origins so mysterious? – South African debate. – Journal of Arid Environments 70: Loik, M.E., Breshears, D.D., Lauenroth, W.K., Journal of Science 86: 207–208. 561–569. Belnap, J. (2004): A multi-scale perspective of Mokolobate, M.S., Haynes, R.J. (2002): Com- Hook, P.B., Burke, I.C., Lauenroth, W.K. (1991): water pulses in dryland ecosystems: climatol- parative liming effect of four organic residues Heterogeneity of soil and plant N and C asso- ogy and ecohydrology of the western USA. – applied to an acid soil. – Biology and Fertility ciated with individual plants and openings in Oecologia 141: 269–281. of Soils 35: 79–85. North American shortgrass steppe. – Plant and Lovegrove, B.G. (1991): Mima-like mounds Moore, J. M., Picker, M.D. (1991): Heuweltjies Soil 138: 247–256. (heuweltjies) of South Africa: the topological, (earth mounds) in the Clanwilliam District, Jafari, M., Tavili, A., Zargham, N., Heshmati, ecological and economic impact of burrowing Cape Province, South Africa: 4000 year old G.A., Zare Chahouki, M.A., Shirzadian, S., animals. – Symposium of the Zoological Soci- termite nests. – Oecologia 86: 424–432. Azarnivand, H., Zehtabian, G.R., Sohrabi, M. ety of London 63: 183–196. Mucina, L., Jürgens, N., Le Roux, A., Rutherford, (2004): Comparing some properties of crusted Ludwig J.A., Tongway, D.J. (1996): Rehabilita- M.C., Schmiedel, U., Esler, K.J., Powrie, L.W., and uncrusted soils in Alagol region of Iran. – tion of semiarid landscapes in Australia. II. Desmet, P.G., Milton, S.J. (2006): Succulent Ka- Pakistan Journal of Nutrition 3: 273–277. Restoring vegetation patches. – Restoration roo Biome. – In: Mucina, L., Rutherford, M.C. James, I., Hoffman, T., Munro, A., O’Farrell, Ecology 4: 398–406. (eds.): The vegetation of South Africa, Lesotho P.J., Smart, R. (2007): The economic value Luther-Mosebach, J. (2009): Hierarchical clas- and Swaziland. Strelitzia 19: 221–229. Pretoria: of fl ower tourism at the Namaqua National sifi cation of the vegetation on the communal South African National Biodiversity Institute. Park, South Africa. – South African Journal farmland of Soebatsfontein in Namaqualand, Myers, N. (2003): Hotspots revisited. – Bio- of Economic and Management Sciences 10: South Africa. The role of grazing history and science 53: 916–917. 442–456. habitat conditions. Diplom thesis in Botany. Myers, N., Mittermeier, R.A., Mittermeier, C.G., Jones, C.G., Lawton, J.H., Shachak, M. (1994): Hamburg: Biocentre Klein Flottbek and Bo- Fonseca, G.A.B. da, Kent, J. (2000): Biodiver- Organisms as ecosystem engineers. – Oikos tanical Garden, University of Hamburg. sity hotspots for conservation priorities. – Na- 69: 373–386. Mauchamp, A., Janeau, J.L. (1993): Water fun- ture 403: 853–858. Jürgens, N. (1986): Untersuchungen zur Ökolo- nelling by the crown of Flourensia cernua, a Myster, R.W., Pickett, S.T.A. (1990): Initial con- gie sukkulenter Pfl anzen des südlichen Afrika. Chihuahuan Desert shrub. – Journal of Arid ditions, history and successional pathways in – Mitteilungen aus dem Institut für Allge- Environments. 25: 299–306. ten contrasting old fi elds. – American Midland meine Botanik Hamburg 21: 139–365. Max-Neef, M.A. (2005): Foundations of transdis- Naturalist 12: 231–238. Kaouthar, J., Chaieb, M. (2010): Changes in soil ciplinarity. – Ecological Economics 53: 5–16. Nel, E., Binns, T., Bek, D. (2007): ‘Alternative foods’ properties and vegetation followinglivestock May, H., Lahiff, E. (2007): Land reform in and community-based development: Rooibos tea grazing exclusion in degraded arid environ- Namaqualand, 1994–2005: a review. – Journal production in South Africa’s West Coast Moun- ments of South Tunisia. – Flora 205: 184–189. of Arid Environments 70: 782–798. tains. – Applied Geography 27: 112–129. Kemmis, S., McTaggart, R. (2000): Participatory Mayer, C., Soka, G., Picker, M. D. (2006): The Nobel, P.S. (1984): Extreme temperatures and action research. Handbook of qualitative re- importance of monkey beetle (Scarabaei- thermal tolerances for seedlings of desert suc- search. – London: Sage. dae: Hopliini) pollination for Aizoaceae and culents. – Oecologia 62: 310–317. King, E.G., Stanton, M.L. (2008): Facilitative Asteraceae in grazed and ungrazed areas at O’Riordan, T. (2002): Protecting beyond the pro- effects of Aloe shrubs on grass establishment, Paulshoek, Succulent Karroo, South Africa. tected. – In: O’Riordan, T., Stoll-Kleemann, S. growth, and reproduction in degraded Kenyan – Journal of Insect conservation 10: 323–333. (eds.): Biodiversity, sustainability and human rangelands: implications for restoration. – Mbagwu, J.S.C. (1992): Improving the produc- communities: 3–29. Cambridge: Cambridge Restoration Ecology 16: 464–474. tivity of a degraded ultisol in Nigeria using University Press. Klak, C., Reeves, G., Hedderson, T. (2004): Un- organic and inorganic amendments. Part 1: Oettlé, N.M. (ed.) (2005): Enhancing sustainable matched tempo of evolution in Southern Afri- Chemical properties and maize yield. – Biore- livelihoods in the Suid Bokkeveld. – Amster- can semi-desert ice plants. – Nature 427: 63–65. source Technology 42: 149–154. dam: Both Ends.

148 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT Oettlé, N., Goldberg, K., Koelle, B. (2009): The Roe, D. (2008): The origins and evolution of the Simons, L., Allsopp, N. (2007): Rehabilitation Heiveld Co-operative - a vehicle for sustainable conservation-poverty debate: a review of key of rangelands in Paulshoek, Namaqualand: local development. – Cape Town: Drynet Case literature, events and policy processes. – Oryx understanding vegetation change using bio- Study. http://www.heiveld.co.za/resources/ 42: 491-503. physical manipulations. – Journal of Arid En- Heiveld-Case-Study.pdf. Rohde, R.F., Hoffman, M.T. (2008): One hun- vironments 70: 755–766. Otto, R., Krüsi, B.O., Burga, C.A., Fernández- dred years of separation: the historical ecology Stock, W.D., Dlamini, T.S., Cowling, R.M. Palacios, J.M. (2006): Old-fi eld succession of a South African ‘Coloured Reserve’. – Af- (1999): Plant induced fertile islands as pos- along a precipitation gradient in the semi-arid rica 78: 189–222. sible indicators of desertifi cation in a succu- coastal region of Tenerife. – Journal of Arid Rohde, R., Hoffman, M.T., Allsopp, N. (2003): lent desert ecosystem in northern Namaqua- Environments 65: 156–178. Hanging on a wire: A historical and socio-eco- land, South Africa. – Plant Ecology 142: Padilla, F.M., Pugnaire, F.I. (2006): The role of nomic study of Paulshoek village in the com- 161–167. nurse plants in the restoration of degraded munal area of Leliefontein, Namaqualand. – Re- Surplus Peoples Project (1995): Survey of Soe- environments. – Frontiers in Ecology and the search Report No. 17. Cape Town: Programme batsfontein Community. – Unpublished Re- Environment 4: 196–202. for Land and Agrarian Studies, School of Gov- port. Soebatsfontein: Surplus Peoples Project. Petersen, A. (2008): Pedodiversity of southern ernment, University of the Western Cape. Surplus Peoples Project (1997): Namaqualand African drylands. – Hamburger Bodenkundli- Rooyen, M.W. van (1999): Functional aspects of district planning and management project: che Arbeiten 61: 1–374 short-lived plants. – In: Dean, W.R., Milton, report on the current situation concerning Petersen, A., Young, E.M., Hoffman, M.T., S.J. (eds): The Karoo: ecological patterns and livestock farming in reservations, municipal Musil, C.F. (2004): The impact of livestock processes: 107–122. Cambridge: Cambridge areas and areas with communal grazing in grazing on landscape biophysical attributes University Press. Namaqualand. – Springbok: Surplus Peo- Succulent Karoo in privately and communally managed range- Rooyen, M.W. van (2002): Management of the ples Project. lands in Namaqualand. – South African Jour- old fi eld vegetation in the Namaqua National Thomas, A.D., Dougill, A.J. (2006): Distribu- nal of Botany 70: 777–783. Park, South Africa: confl icting demands of tion and characteristics of cyanobacterial soil Petersen, A., Gröngröft, A., Miehlich, G. (2010): conservation and tourism. – Geographical crusts in the Molopo Basin, South Africa. – Methods to quantify the pedodiversity of Journal 168: 211–223. Journal of Arid Environments 64: 270-283. 1 km² areas — results from southern African Röwer, I.U. (2009): Small-scale vegetation and Thurow, T.L. (2000): Hydrologic effects on range- drylands. – Geoderma 155: 140–146. biodiversity patterns of termitaria (heuweltjies) land degradation and restoration processes. – Picker, M.D., Hoffman, M.T., Leverton, B. under different grazing pressures in the commu- In: Arnalds, O., Archer, S. (eds.): Rangeland (2007): Density of Microhodotermes viator nal area of Soebatsfontein, South Africa. – Dip- desertifi cation: 53–66. Dordrecht: Kluwer (Hodotermitidae) mounds in southern Africa lom thesis. Hamburg: University of Hamburg. Academic Publishers. in relation to rainfall and vegetative produc- RSA (2002): Government of South Africa; Ka- Todd, S.W. (2000): Patterns of seed production tivity gradients. – Journal of Zoology 271: miesberg Munisipaliteit: Weidingsregulasies. and shrub association in two palatable Karoo 37–44. – Provincial Gazette of the Province of the shrub species under contrasting land use inten- Pickett, S.T.A., Cadenasso, M.L., Meiners, S.J. Northern Cape No. 676, 04/2002. sities. – African Journal of Range and Forage (2008): Ever since Clements: from succession Rutherford, M.C., Powrie, L.W. (2010): Severely Science 17: 22–26. to vegetation dynamics and understanding to degraded rangeland: Implications for plant di- Todd, S.W., Hoffman, M.T. (1999): A fence-line intervention. – Applied Vegetation Science versity from a case study in Succulent Karoo, contrast reveals effects of heavy grazing on 12: 9–21. South Africa. – Journal of Arid Environments plant species diversity and community com- Pineiro, G., Paruelo, J.M., Jobbagy, E.G., Jack- 74: 692–701. position in Namaqualand, South Africa. – son, R.B., Oesterheld, M. (2009): Grazing ef- Samuels, M.I., Allsopp, N., Hoffman, M.T. Plant Ecology 142: 169–178. fects on belowground C and N stocks along a (2008): Mobility patterns of livestock keep- Todd, S.W., Hoffman, M.T. (2009): A fence line network of cattle exclosures in temperate and ers in semi-arid communal rangelands of in time demonstrates grazing-induced vegeta- subtropical grasslands of South America. – Namaqualand, South Africa. – Nomadic Peo- tion shifts and dynamics in the semiarid Suc- Global Biogeochemical Cycles 23: GB2003. ples 12(1): 123-148. culent Karoo. – Ecological Applications 19: DOI: 10.1029/2007GB003168. Schmiedel, U. (2002): The quartz fi elds of south- 1897–1908. Podani, J. (2001): SYN-TAX (2000): Computer ern Africa. Flora, phytogeography, vegetation, Tongway, D.J., Ludwig J.A. (1996): Rehabilita- programs for data analysis in ecology and sys- and habitat ecology. – PhD thesis. Cologne: tion of semiarid landscapes in Australia. I. Re- tematics. – Budapest: Scientia Publishing. University of Cologne. storing productive soil patches. – Restoration Pufal, G., Mayer, C., Porembski, S., Jürgens, N. Schmiedel, U. (2004): The phytogeography of Ecology 4: 388–397. (2008): Factors affecting fruit set in Aizoaceae the obligate quartz fi eld fl ora of southern Af- Turner, J.S. (2003): Trace fossils and extended species of the Succulent Karoo. – Basic and rica. – Schumannia 4: 181–205. organisms: a physiological perspective. – Applied Ecology 9: 401–409. Schmiedel, U., Jürgens, N. (1999): Community Palaeogeography, Palaeoclimatology, Palae- Querejeta, J.I., Roldán, A., Albaladejo, J., Cas- structure on unusual habitat island: quartz- oecology 192: 15–31. tillo, V. (2000): Soil physical properties and fi elds in the Succulent Karoo, South Africa. Valentyn, F. (1971): Description of the Cape of moisture content affected by site preparation – Plant Ecology 142: 57–69. Good Hope with the matters concerning it. – in the afforestation of a semiarid rangeland. – Schmiedel, U., Jürgens, N. (2004): Habitat Van Riebeeck Society Pulications, 2nd ser., No. Soil Science Society of America Journal 64: ecology of southern African quartz fi elds: 4. Cape Town: Van Riebeeck Society. 2087–2096. studies on the thermal properties near the Vetter, S. (2005): Rangelands at equilibrium and Rahlao, S.J., Hoffman, M.T, Todd, S.W., McGrath, ground. – Plant Ecology 170:153–166. non-equilibrium: recent developments in the K. (2008): Long-term vegetation change in the Schmiedel, U., Mucina, L. (2006): Vegetation of debate. – Journal of Arid Environments 62: Succulent Karoo, South Africa following 67 quartz fi elds in the Little Karoo, Tanqua Karoo 321-341. years of rest from grazing. – Journal of Arid and eastern Overberg (Western Cape Province, Villiers, A.J. de, Rooyen, M.W. van, Theron, Environments 72: 808–819. South Africa). – Phytocoenologia 36: 1–44. G.K. (1992): Studies of the revegetation of Raper, P.E., Boucher, M. (eds.) (1988): Robert Schnegg, M. (2009): It’s the combination that saline mined soil at Namaqua sands. – Report Jacob Gordon. Cape travels, 1777-1786. – counts: diversifi cation of pastoral livelihoods in for Anglo American Corporation. Pretoria: Houghton: Brenthurst Press. northwestern Namibia. – In: Clemes, C., Kokot, University of Pretoria. Richardson, F.D., Hahn, B.D., Hoffman, M.T. W. (eds.): Networks, resources and economic Villiers, A.J. de, Rooyen, M.W. van, Theron, G.K. (2007): Modelling the productivity and sus- action. Ethnographic case studies in honor of (2003): Similarity between the soil seed bank tainability of pastoral systems in the commu- Hartmut Lang: 229–248. Berlin: Reimer. and the standing vegetation in the Strandveld nal areas of Namaqualand. – Journal of Arid Shackleton, B., Shackleton, S., Cousin, B. (2001): Succulent Karoo, South Africa. – Land Degra- Environments 70: 707–717. The role of land-based strategies in rural live- dation & Development 14: 527–540. Riginos, C., Hoffman, M.T. (2003): Changes in lihoods: the contribution of arable production, Vollan, B. (2008): Socio-ecological explanations population biology of two succulent shrubs animal husbandry and natural resource har- for crowding-out effects from economic fi eld along a grazing gradient. – Journal of Applied vesting in communal areas in South Africa. experiments in southern Africa. – Ecological Ecology 40: 615–625. – Development Southern Africa 18: 581–604. Economics 67: 560–573.

SUCCULENT KAROO 149 Vollan, B. (2009): Co-operation for common Webley, L. (2007): Archaeological evidence World Reference Base for Soil Resources (2006): pool resources: an experimental perspective. for pastoralist land-use and settlement in A framework for international classifi cation, – München: Dr. Hut. Namaqualand over the last 2000 years. – Jour- correlation and communication. – World Soil Wadsworth, Y. (1998): What is participatory ac- nal of Arid Environments 70: 629–640. Resources Report No. 103. Rome: Food and tion research? – Action Research International, Werger, M.J.A. (1974): On concepts and tech- Agriculture Organization of the United Na- Paper 2. http://www.scu.edu.au/schools/gcm/ niques applied in the Zürich-Montpellier tions. ar/ari/p-ywadsworth98.html [acc. 17.05.2010]. method of vegetation survey. – Bothalia 11: Yayneshet, T., Eik, L.O., Moe, S.R. (2009): The Ward, D. (2009): The biology of deserts. – Ox- 309–323. effects of exclosures in restoring degraded ford: Oxford University Press. Whalen, J.K., Chang, C., Clayton, W.C., Care- semi-arid vegetation in communal grazing Warren, S. D. (2001): Synopsis: infl uence of foot, P. (2000): Cattle manure amendments lands in northern Ethiopia. – Journal of Arid biological soil crusts on arid land hydrology can increase the pH of acid soils. – Soil Environments 73: 542–549. and soil stability. – In: Belnap, J., Lange, O.L. Science Society of America Journal 64: Zhang, Y.M., Wang, H.L., Wang, X.Q., Yang, (eds.): Biological soil crusts: structure, func- 962–966. W.K., Zhang, D.Y. (2006): The microstructure tion, and management. Ecological Studies Witbooi, B., Esler, K.J. (2003): Old fi eld resto- of microbiotic crust and its infl uence on wind 150: 349–360. Berlin & Heidelberg: Springer. ration: an assessment of the infl uence of res- erosion for a sandy soil surface in the Gurban- Warren, S.D., Eldridge, D.J. (2003): Biological toration techniques on the establishment of tunggut Desert of Northwestern China. – Geo- soil crusts and livestock in arid ecosystems: selected Succulent Karoo species. – Elsenburg derma 132: 441–449. Are they compatible? – In: Belnap, J., Lange, Journal 2004: 3–7. http://www.elsenburg.com/ Zink, A.T., Allen, M.F. (1998): The effects of O.L. (eds.): Biological soil crusts: structure, agriprobe/2004/ej2004-1.pdf organic amendments on the restoration of a Succulent Karoo function, and management. Ecological Studies Wong, M.T.F., Nortcliff, S., Swift, R.S. (1999): disturbed coastal sage scrub habitat. – Resto- 150: 401–415. Berlin & Heidelberg: Springer. Method for determining the acid ameliorat- ration Ecology 6: 52–58. Watkeys, M.K. (1999): Soils of the arid south- ing capacity of plant residue compost, urban western zone of Africa. – In: Dean, W.R.J., waste compost, farmyard manure, and peat Milton, S.J. (eds.): The Karoo: ecological pat- applied to tropical soils. – Communica- terns and processes: 17–25. Cambridge: Cam- tions in Soil Science and Plant Analysis 29: bridge University Press. 2927–2937.

150 BIODIVERSITY IN SOUTHERN AFRICA 3 – IMPLICATIONS FOR LANDUSE AND MANAGEMENT