SOIL DEGRADATION ASSESSMENT IN MIXED FARMING SYSTEMS OF SOUTHERN AFRICA: USE OF NUTRIENT BALANCE STUDIES FOR PARTICIPATORY DEGRADATION MONITORING
Andrew Dougill1, Chasca Twyman2, David Thomas2 and Deborah Sporton2
1 - School of the Environment, University of Leeds, LS2 9JT, UK:
2 - SCIDR, Department of Geography, University of Sheffield, Sheffield, S10 2TN: ;:
Soil degradation assessments for mixed farming systems of the Molopo Basin (NW Province, South Africa and Southern District, Botswana) are provided from farmer-based research, designed to quantify nutrient fluxes across the farming system and to analyse the social, economic, political and environmental factors affecting nutrient management practices. This paper provides case study material that supplements recent research from elsewhere in sub-Saharan Africa and discusses the practical difficulties of how to use participatory studies to assess land degradation extent and causes, and to identify broader policy implications and recommendations.
Nutrient balance studies show that land degradation is evident on arable fields as soil nutrient depletion (the main threat to poorer manure-reliant farmers) and soil acidification (the main threat to wealthier farmers who have become dependent on inorganic fertiliser additions). Integrated nutrient management strategies involving both compound fertiliser additions and regular manure inputs can mitigate most soil degradation even on the sandy infertile soils, but remain infrequently practised. The need to retain nutrient flows through the livestock sector from rangeland to arable land is vital to environmental sustainability and offers an applicable entry point for agricultural development initiatives and support. Factors identified as threatening the flow of nutrients from rangelands to arable lands include policy settings in terms of the different support programmes offered to communal farmers, village level extension advice, household poverty levels and labour constraints.
Keywords: Soil Degradation, Nutrient balance studies, Participatory assessment, mixed farming, South Africa, Botswana
Research Context
In sub-Saharan Africa it is increasingly recognised that optimising the nutrient balance on a farm is one of the most difficult agricultural management challenges faced in avoiding land degradation (e.g. Reij and Waters-Bayer, 2001; Scoones, 2001). As a result there are an increasing number of nutrient balance studies conducted on a variety of scales, ranging from continental (Stoorvogel and Smaling, 1990) and regional analyses (Smaling et al., 1993, Folmer et al., 1998) to farm and field level studies (e.g. Krogh, 1995; Harris, 1998; de Jager et al., 1998). These studies have proved vital in recognising the diversity of factors affecting agricultural sustainability in mixed farming systems (Scoones, 2001) and highlighting the need for detailed case study material to better inform the policy process (Scoones and Toulmin, 1999). The participatory nature of nutrient balance studies also offers the opportunity to enable action-orientated research with farming communities (Defoer and Budelman, 2000). This has proved important in enabling improved development planning operations in regions where such research has been conducted (e.g. Hilhorst and Muchena, 2000; Scoones 2001).
Participatory nutrient balance studies have not yet been widely used to improve local and regional assessments of soil degradation, though some recent agricultural development projects have started to focus on how to incorporate farmers perspectives in land degradation assessments (Stocking and Murnaghan, 2001). The importance of this research priority is particularly apparent as regional policy statements continue to be based on standardised classifications of soil degradation extent, based on aggregated subjective data collected under the auspices of the United Nations Convention to Combat Desertification (UNCCD, 1995). For example, the Global Assessment of Soil Degradation study (GLASOD - Oldeman et al., 1990) is still seen as having a major impact on agricultural policy debates throughout Africa (Swift, 1996). Furthermore, where national land degradation classifications are available, including that produced recently for South Africa (NBI, 2000; Hoffman and Ashwell, 2001), classifications based on regional expert and agricultural extension workers views remain the primary information source. These classifications generally fail to engage with local farmers and consequently concerns remain over the validity of their assessments of both degradation extent and the underlying root causes of land degradation processes.
This paper focuses on one cross-border location in southern Africa which both international (Oldeman et al., 1990) and national (NBI, 2000) soil degradation assessments have classified as moderately degraded, principally as a result of wind erosion processes. By using an interdisciplinary farmer-centred approach, it aims to assess the applicability of these classifications by using partial nutrient balance methods (Defoer et al., 1998) as a tool to enable information gathering on soil nutrient fluxes and farmer decision-making. In addition, environmental survey techniques examine the scale of erosional losses and remaining soil nutrient reserves. Research outputs are also integrated with the findings of affiliated livelihood analyses (Twyman et al., in prep), conducted in-parallel as part of a wider interdisciplinary project (see for full project details). The link to broader sustainable livelihood analyses enables the impacts of contrasting policy and economic settings in South Africa and Botswana to be examined. Discussion here focuses on how such integrated soil degradation analyses can be used to provide policy information that can better enable suitable institutional frameworks to be devised capable of supporting sustainable agricultural development in mixed farming regions.
Study Area: Molopo / Barolongs Region
Research was undertaken around three main study villages (Logageng, Mathatheng and Mokatako) in the cross-border setting of the Molopo Basin, encompassing Molopo District, North West Province, South Africa and Barolong Farms, Southern District, Botswana (Figure 1). Villages were chosen to represent the range of policy and land-ownership issues identified in consultation with local NGOs, researchers and farmer groups. These villages are all characterised by mixed (pastoral and arable) farming systems, typical of the smallholder systems found throughout southern Africa (Whiteside, 1998). Environmentally, the region is typified by low fertility Kalahari sand soils and a typical semi-arid rainfall regime, with a mean annual rainfall of c. 450 mm associated with high interannual variability that makes drought an endemic feature and cause of livelihood vulnerability. This low fertility dryland setting has led to an opportunistic approach to farming, which has broad similarities to those discussed for other integrated soil nutrient studies recently published for sub-Saharan Africa, notably for Chivi District, Zimbabwe (Scoones et al., 1996; Chibudu et al., 2001), and for dryland zones of Mali (Dembélé et al., 2001).
Village residents belong principally to the Rolong group of the Tswana and retain strong ties across the international border (Drummond and Manson, 1991), as well as to the neighbouring white commercial farming sector in South Africa for whom many have worked as farm labourers (Francis, 1999). Land ownership around all villages comprises a mix of communally-owned rangeland where all village members have grazing rights, and privately-owned arable lands for which land ownership patterns are highly skewed both within and between villages, including some well developed share-cropping agreements.
Broad differences across the international border are a consequence of recent policies of land reform and agricultural development that have emerged since independence in Botswana and the end of apartheid in South Africa. Outside some slow-moving reforms towards land redistribution, it has been argued that in South Africa rural development policy lacks co-ordination and efficacy, something that has led to many small-scale farmers diversifying away from farming as their primary means of livelihood support. In Botswana, recent policies have shifted from a drive for food self-sufficiency to food security, based on imports from South Africa. This has led to a decline in financial and physical infrastructure support for farmers and a significant decline in the land under cultivation and overall yields (Carugi and Gareebine, 1996). Rural dwellers on both sides of the border are consequently changing or diversifying their livelihoods.
While moves away from farming partly result from policy changes, they may also be caused by declines in agricultural productivity due to degradation processes. This study was designed to examine the extent and causes of soil degradation affecting three typical village situations. At Logageng, as at many surrounding villages, there has been a recent shift to the increased growth of groundnuts, maize and sunflowers as cash crops sold through local co-operatives and dependent on increased fertiliser additions. Mathatheng was chosen due to its status as the poorest settlement in the North West Province (Drummond, pers. comm.) and the additional livelihood vulnerability expected due to the reported shortage of land and financial capital. In Botswana, studies focused around Mokatako due to the wide mix of farming households, the active involvement of the local Government extension officer and the problems faced by many farmers since the recall of agricultural loans by the Governments National Development Bank. At each village, integrated farmer interviews, workshops and environmental surveys were undertaken following the research methodology outlined below.
Methods
The research starting points are the yield concerns of farmers themselves and the need to involve rural communities throughout the research process. In this way, the approach followed is participatory or 'bottom-up', matching the focus formalised in the UNCCD, enabling findings to be gained from the farmers and then reported to development practitioners and policy makers.
The methodological framework was split into four distinct phases over a period of 18 months. The framework followed offers a template suitable for participatory soil degradation assessments in other regions where degradation is evident from an array of both visible and hidden indicators and is caused by a variety of social and institutional factors. Research phases are detailed separately below. However, analysis involves the integration of results from all phases, essential for the extent and causes of soil degradation to be fully understood.
- Initial environmental surveys and livelihoods analysis
Initial site visits involved independent environmental and social survey analyses. Environmental surveys involved quantification of soil degradation as used in GLASOD and national surveys. Studies focused on evidence of wind erosion and deposition due to its classification as the main cause of soil degradation regionally (Oldeman et al., 1990; NBI, 2000). Key indicators included fenceline deposits, nebkha dunes, rills and sheetwash deposits. Nutrient analysis (for both total-nitrogen and extractable-phosphate) of windblown deposits lost from agricultural fields assessed the impact of erosion on the natural balance between nutrient inputs and outputs to agricultural soils (Dougill and Thomas, 2002). In addition, social surveys were conducted in the initial research phase with background information collected through a broad-ranging livelihoods analysis involving over 70 semi-structured interviews in six settlements across the region (Twyman et al., in prep).
- Participatory nutrient balance studies and field visits
Farmer interviews aimed at enabling quantification of the major nutrient fluxes at both field- and farm-scales were then conducted for 15 farmers (five from each study village). Interviews involved discussions of existing soil nutrient management practices and the constraints, threats and opportunities identified by farmers as affecting their farm management and decision-making. Chemical analysis of the key organic constituents, notably dung and kraal manure, was also conducted at this stage, together with recording of fertiliser and cattle feed nutrient concentrations. This enabled quantification of all managed nutrient inputs for individual fields. Table 1 lists all the assumptions used to produce farm-scale nutrient flux models, estimations that must be considered when assessing the caveats of the nutrient balance information provided.
Farm walks or drives were also conducted to enable farmers to discuss their specific concerns in a field setting, and to collect soil samples from their chosen fields for subsequent laboratory analysis. Soil samples were analysed for total-N, extractable-P, exchangeable-K, pH and organic matter (OM) content using standard tropical soil methods (Anderson and Ingram, 1993). This information enabled identification of the limiting factors to crop growth and investigation of the soil chemical processes associated with different agricultural management strategies, in addition to enabling subsequent extension advice. Field visits also facilitated broader discussion of farmers' concerns and identification of potential indicators of degradation that are often only evident in the field setting (Stocking and Murnaghan, 2001).
- Soil fertility feedback discussions
The next research phase provided extension style advice to farmers by reporting the measured soil nutrient and pH values for their study fields. The feedback process allowed discussion of views on perceived environmental threats and constraints on agricultural management. Additional discussions on the amounts of organic inputs possible for each farmer, acidification problems and the impacts of the above average rainfall in the 1999/2000 wet season added to the information collected in the previous growing season. These less rigidly structured follow-on discussions provide an ideal forum for ensuring that participatory research is of value to farmers involved in the research process, in guiding their future decision-making on integrated nutrient management options.
4.Farmer Group Workshops and Policy-Makers Discussion Workshop
As the final stage of the PANRUSA project, community workshops were held in both Logageng and Mokatako for farmers from these villages and surrounding communities. In these workshops, researchers detailed the main study findings, including those on assessments of environmental sustainability, for discussion and verification by local residents. These workshops were followed by a Policy Workshop aimed at providing summary information to representatives from policy institutions, international organisations and local collaborators. This forum was designed to enable the key messages raised by rural dwellers on their livelihood constraints and opportunities to be reported to those with an influence on the institutional frameworks and policies recognised as controlling farmers' decision-making.
The sequential interdisciplinary methodological framework followed is one that offers the potential for use in other settings. The ability to integrate methods, combine simple indicators of degradation from different information sources, and the ability to maintain farmer involvement throughout are all vital to facilitating participatory action-orientated research.
Results and Discussion
Analyses of all environmental sustainability findings are provided here to enable integrated assessment of land degradation extent and the underlying socio-economic causes. The focus is principally on the extra information available from combining information from the different sources. In this manner, it is hoped to demonstrate how the simultaneous collection of several sources of soil degradation assessment information can provide a more reliable and practical (to the farmer) classification of land degradation extent and cause.
1a.Erosion losses and sediment redistribution
Sedimentological evidence suggests that windblown nebkha dune sediments found across communal rangeland areas are predominantly locally derived (Dougill and Thomas, 2002). In addition, nutrient (total-N, extractable-P and exchangeable-K) analyses of wind blown sediments suggest that nutrient losses from arable fields are limited by the low flux of aeolian erosion. These losses are therefore not predicted to be a major cause of reductions in the yield potential of arable land, and thus the presence of nebkha dunes should not immediately imply soil degradation, as previously hypothesised (e.g. Tengberg, 1995). This view is largely confirmed in farmer interviews that detail the longevity of nebkhas in this landscape. In addition, water erosion processes are confined to the immediate vicinity of homesteads. Therefore, no soil chemical evidence exists to support the classifications of erosional processes causing widespread soil degradation. Vitally for nutrient balance studies, this information suggests that erosion losses from arable fields remain relatively small. It is therefore assumed that the natural balance between the deposition and fixation nutrient inputs and leaching, denitrification and erosion outputs is not significantly affected by agricultural practices and the increased incidence of wind erosion. Subsequent modelling therefore focuses solely on the managed inputs and outputs that are directly affected by the farmers' nutrient management strategies.
1b.Livelihoods analysis
Semi-structured livelihoods interviews, together with regional policy analyses, investigated the flexible adaptations and livelihood diversifications that rural residents are making in response to environmental changes, changing land and agricultural policies and reforms to social institutions (Twyman et al., in prep.). These studies demonstrated many changes in cropping strategies and nutrient management practices in response to perceived environmental threats, and to changes in farmers' social and financial capital assets. It was clear that non-farm and non-natural resource livelihoods have a central role in abating poverty for many households and also for maintaining the viability of farming for wealthier large-scale commercial farmers. However, given the lack of alternative livelihood pathways open to many rural residents, dissatisfaction with farming (as expressed especially by the young) remains a major concern to the co-ordination of future rural land and agricultural policy. These studies suggested that to aid the adoption of farming livelihoods, improvements in micro-finance credit schemes, extension service advice and alternative livelihood options should all be considered (Twyman et al., in prep.). However, for the immediate needs of the poor to be met and supported by local institutional structures it is essential that environmentally appropriate measures and advice are provided to maintain the sustainability of farming practices which remain the main livelihood option for the rural poor.
2.Nutrient flux modelling and quantification
The central importance of arable farming to livelihoods in mixed farming regions implies that sustainability assessments (and thus land degradation classification) must focus on the processes of soil degradation capable of reducing arable crop yields. Concerns raised over the perceived threat of chemical soil degradation led to nutrient budget studies being undertaken. These focused on both the field-scale (at which farmers' decision-making is focused), and the farm-scale, due to the importance of nutrient flows across the farming system.