58, Winter 2005
Soil management for drylands
"Our carbon is gone; we have to bring it back!"
Soil fertility management and social learning in Senegal's drylands
by Petra Tschakert
"... knowledge partnerships and social learning processes with respect to soil carbon can engage land managers more actively in research programs, complement their environmental wisdom, and, ultimately, contribute to more sustainable soil management decisions. ...Farmers, for their part, are ready. "
(Back to top)
Gains in soil carbon, it is claimed, are best achieved through conservation agriculture and improved grassland management, including reduced tillage, cover crops, crop rotation, crop residue management, mulching, green manure, composting, and manuring (Lal 2002; FAO 2004; Vagen et al. 2005). One or several of these practices are typically proposed as optimal agronomic solutions for semi-arid and sub-humid regions, often without verification of presently most viable solutions under given household constraints. Such a postulation reflects the "received wisdom" (Leach and Mearns 1996) about African drylands, including the notion of a continuing "downward spiral" of land degradation, soil impoverishment, and rural poverty against which immediate action has to be taken (Scoones 2001). Carbon sequestration, then, is perceived as a promising option to counteract this negative trend. If implemented in time, it may prevent looming environmental disasters. A largely technology-centered plan of action stipulates that inappropriate management practices of uninformed land users must be rectified through awareness and capacity-building. The French equivalent, "sensibiliser les populations," is even more problematic as it portrays rural groups as one ignorant conglomerate of people dependent on guidance from external agencies. With notable exceptions from indigenous soil conservation research (Tabor 1990; Reij 1991; Reij et al. 1996; Mazzucato and Niemeijer 2000), these stakeholders are rarely perceived as key resources that ought to be tapped for their long-time experience and accumulated knowledge on soil fertility management. And yet, while researchers, scientists, and policy makers are likely the ones to elect "best" practices, there is no doubt that it will be the decision of individual land owners and managers whether or not to implement them. Ignoring their points of views and rationales, therefore, seems a luxury we can ill afford.
Farmers and other land users in drylands, as elsewhere, will be unlikely to participate in carbon sequestration projects unless the environmental, social, and economic benefits are convincing and clear (Tschakert 2004a). Knowledge and learning, in this context, become particularly crucial when the ultimate goal is to capture something as seemingly nebulous as CO2 that is neither visible nor tangible. Thus, framing explanations and expectations in a way that embeds this largely foreign concept of carbon into local perceptions, conceptual frameworks, and soil management decisions appears fundamental for the success of long-term programs.
This article does not deny the fact that land degradation and soil fertility losses in drylands exist and that local knowledge to resolve them can be incomplete or even erroneous. Rather, it adopts a model of knowledge partnerships that merges expert opinion and local environmental wisdom to facilitate sustainable soil management decisions. Sustainability, as argued by Moller et al. (2004), can best be achieved through complementary use of scientific and local knowledge. Omitting community inputs in knowledge construction, on the other hand, is likely to undermine processes of engaging and empowering local knowledge (Zanatell and Knuth 2002). The article demonstrates how Sahelian land managers can play an active role in a social learning process that puts carbon sequestration in the center of existing conceptual frameworks of soil fertility management in a way that is understandable and useful to agricultural decision-making. Drawing from Borrini-Feyerabend et al. (2000) and Schusler et al. (2003), social learning is defined here as learning to enhance common knowledge, awareness, and skills by engaging multiple participants, sharing diverse perspectives, and thinking and acting together. The lessons learned by the author and others from 14 months of carbon sequestration research with small-scale farmers in Senegal suggest that participatory approaches can be a valuable tool for collective learning and problem solving that serve the needs of community-based soil management in drylands.
(Back to top)
By following a highly participatory research approach based on social learning, the conventional order in carbon sequestration studies was reversed: instead of identifying best management practices for dryland farming systems and then encouraging individual adoption, this research put farmers' knowledge, strategies, and constraints at the center of analysis. This reflects a new mode of research collaboration in natural resource management that, as stressed by Wiber et al. (2004), focuses on appropriate knowledge for decision-making and processes that make this knowledge accessible to local managers. We began with local conceptual understandings of the various processes that govern soil fertility and degradation. These mental frameworks then, in addition to an assessment of household assets and environmental and agricultural policies, helped us to identify and evaluate management preferences and spatial-temporal patterns of land use. This phase of field work consumed eight months during which carbon sequestration was not once mentioned! It was only when we had gained sufficient insight into soil management decisions at the household and community level and built trust and confidence between our "expert" team and the research participants in the villages that we tackled the second part of the study. This included:
These three components of the social learning process for soil carbon sequestration assessments constitute the main focus of this article.
(Back to top)
While studies on local soil classifications in West Africa have provided valuable explanations for spatial land use and cropping patterns, farmers' theories about soil fertility and processes of soil formation and degradation offer additional insight into specific management decisions (Mazzucato and Niemeijer 2000). Here, the attempt was to gain a detailed understanding, first, of local knowledge regarding the origin and components of soil fertility and the factors that lead to fertility increase and loss; and second, of the management strategies farmers in the Old Peanut Basin pursue based on these conceptual frameworks. This information later served as the basis for social learning activities around carbon. Following Zanatell and Knuth (2002), I define local knowledge as "the information, experiences, and predictive insights of resource users and persons whose lives are closely linked to a resource of concern." As illustrated below, such local knowledge also embodies a "knowledge-practice-belief" complex that provides an additional dimension often missing in conventional science (Berkes 1999; Moller et al. 2004). This can be seen as the "soul" of local soil management. If omitted, carbon projects in the Sahel are likely to fail.
Doolé (a Wolof term, equivalent to dakan in Serer-Sine)—the richness, strength, or fertility of a soil— is the centerpiece of the ethnopedological framework in the Old Peanut Basin (Tschakert 2001). As explained by farmers, doolé depends largely on management decisions and practices and only to a lesser extent on the soil itself. Some farmers also call it la dose. There is a general consensus that fertility comes from saletés, translated as "dirt," a generic term for various types of soil organic matter inputs. Saletés include manure, droppings from stubble grazing, decomposing branches, leaf litter, agricultural residues, grass, ashes, and household waste, all ingredients (ndjame-ndjame) that make soils look dirty. Dirt, as further explained, contains vitamines (vitamins), the essential elements of good and healthy (soil) food. Farmers are generally unaware of the origin of these vitamines, they just seem to be in leaves and roots and, when grazed by animals, are transformed into feces and manure. Some trees, particularly kad (Faidherbia albida), seem to dispose of additional vitamines that are released as sap when their wood is burned. Also, farmers argue that richer soils have a much better taste (djafka) than poor soils. Once again, farmers draw analogies from their culinary sphere: "A good jëb u jënn (national dish based on rice and fish), prepared with a variety of appetizing ingredients (safsafan) is more flavorful and appreciated than a bowl of rice prepared without any ingredients." Only very few participants (three individuals in a total of 14 communities) emphasized the role of other elements (ferniente) that contribute to a soil's strength, in particular nitrogen, calcium, potassium, and iron. Surprisingly, these elements were not at all associated with chemical or mineral fertilizer. Nitrogen was believed to be part of leaves and roots of the kad tree while the origin of the other elements was unknown.
Perceptions of soil degradation were closely related to the loss of vitamines and doolé and to three distinct processes:
The role of insects for soil fertility and decline remained ambiguous, mainly because farmers were hesitant to discuss underground processes that escape their empirical knowledge. Half of all participants perceived insects as destructive, devouring roots and other plant parts, while the other half viewed them as rather beneficial, adding organic matter once they had died inside the soil. Despite this ambiguity, it was agreed that more fertile soils also carried more insects. As explained by one participating child: "If somebody offered you a bowl of rice with nothing in it, would you eat it? Of course not. You would want to have the one with yummy meat and sauce. Insects do the same—they go where they find lots of vitamins and a tasty soil."
Based on these concepts, farmers pursue soil management strategies that can be grouped in four distinct categories:
Crop rotation, mainly with millet and groundnuts or cowpeas, was cited as most beneficial because of crop-specific consumption pattern, always leaving less appreciated vitamins for the crop to follow. What exactly these crop-specific preferences are, farmers don't know, although they are convinced of their existence. Fertilizer and stubble grazing were considered highly advantageous for soil fertility management, but usage had declined in rural communities due to lack of financial means. According to farmers' theories, anything that increases the amount of "dirt" in the soil is good, including manure, crop residues, leaf litter, and inputs from household waste. Enhanced nutrient uptake was perceived to occur through superficial tillage, increasingly practiced due to the lack of ploughs and other heavy equipment as a result of discontinuing agricultural subsidies under Structural Adjustment Programs (Gaye 2000). Finally, farmers emphasized the role of a healthy vegetative cover for minimizing the loss of vitamines from the soil. Clearly, the strategy perceived as most beneficial was a 3-5 year fallow in combination with manure application and reforestation to prevent wind erosion of fragile topsoil and to replenish doolé.
These four strategies were depicted in a composite drawing (Figure 1) that was then reviewed with community members for correctness. It became readily apparent that farmers had a good, albeit incomplete, empirical understanding of soil fertility and degradation; yet, they were entirely unaware of crucial soil processes occurring underneath the surface. Since the 1960s, agricultural extension work in the Peanut Basin had focused on correct applications of new techniques and anticipated gains (Gaye 2000) without ever explaining the actual processes that would yield the final results. This is certainly true for the use of fertilizer and compost. Despite years of subsidized and widely applied fertilizer, farmers did not know the meaning of NPK, the nitrogen-phosphorus-potassium ratio expressed on each bag of fertilizer as a three-number formula (such as 10-10-20, the most commonly used fertilizer for millet). Likewise, farmers could not offer any convincing reason for using a different fertilizer mix (6-20-10) on groundnuts, other than recommendations from extension agents. Thus, after this first phase of field work, what we felt was most needed was not a plan for timely and efficient delivery of "best" management practices but a social learning process around the basic elements of carbon and nitrogen, adapted to farmers' conceptual frameworks of soil fertility and management. This, we believed, was a fundamental prerequisite for long-term community-based carbon sequestration programs.
(Back to top)
Explaining the N cycle proved to be slightly more difficult. Farmers all agreed that millet usually grew faster underneath the canopy of the kad trees (Faidherbia albida); this was (incorrectly) associated with the presence of vitamines in falling leaves and healthy sap dropping from the stem. The nodules with N-fixing bacteria attached to the roots of this leguminous tree had gone unnoticed. Rather than uprooting a tree, we used a groundnut plant, also N-fixing, and examined and dissected the nodules for practical demonstration and discussion. Once groundnuts were recognized as suppliers of valuable ferniente (N), the rationale for rotating millet and groundnuts and using different fertilizer mixes all of a sudden became a "scientific" fact that every participant understood. The same was true for N in leucaena, often used as green manure. Consequently, emphasizing that C and N undergo cycles, including release back to the atmosphere, underlined the importance of healthy vegetation cover and tree densities. A hand-drawn image of these relationships (Figure 2) was used throughout the workshops as a highly effective visual aid to trace the two cycles. Smaller paper copies were handed out to all participants for further dissemination and discussion.
In retrospect, our explanations, deliberations, and demonstrations of the C and N cycle seem rather simplistic. Yet, they were more enlightening than the information farmers had obtained from extension agents in more than half a century! The reason we were successful was that participants saw their own knowledge represented on the poster, could directly relate to it, and viewed novel elements as valuable and practical additions to what they already knew. In fact, the President of the Rural Council in Dingiraye suggested that we take our poster to all villages in his Communauté Rurale where no such learning exercise had ever been undertaken. In Thilla Ounté, the chef de village declared it his personal mission to explain the C and N cycle to the influential marabouts (Moslem leaders) in his community (Figure 3). His message was clear: "Our carbon is gone; we have to bring it back!" Even today, five years later, he shares his knowledge about carbon with any visitor who expresses an interest in it (M. Sène, personal communication).
The final step in this C and N learning activity was to discuss and evaluate possible strategies to increase the amount of C in soil and biomass, as projected with CENTURY, a biogeochemical model (Parton et al. 1994; Tschakert et al. 2004). The model output was transformed from an Excel spreadsheet into a large-scale poster with an illustrative rather than descriptive legend. It depicted four scenarios of possible future soil management:
Participants were quick to understand the differential accumulation rates for C while scrutinizing the practical feasibility of each management scenario. Lively, sporadically even fierce discussions involving farmers, herders, extension agents, and policy makers revealed the various opportunities and constraints associated with each option (Figure 4). In a highly accessible and transparent manner, these deliberations brought the "best" practices, as proposed by the model, down to rural realities.
(Back to top)
Three teams of four designated community members (all male) in Thilla Ounté, Ngodjilème, and Thiaytou, respectively, participated in one half-day of theoretical and one half-day of practical training. This involved learning about the different tools for the soil and biomass measurements (including how to read a compass), the stratified random sampling strategy for selecting individual fields, and the detailed methodology for taking the desired samples from each field. Figure 5 depicts the complete set of tools as memorized and put on paper at the end of the training day by one of the participants. The only requirement for becoming member of the C teams was literacy.
In the first two communities, village elders selected one senior and three juniors to participate in the training and the measurements. In contrast, in Thiaytou the most influential and highly motivated elders were too eager to let this opportunity slip and selected themselves. As a result, the average age of our C "experts" there was close to 60 years old! From a learning perspective, this meant that in the first two cases a group leader was identified from the onset. In the third case, the learning experience was clearly more competitive as each senior team member experimented with every step of the measurement process in order to further refine his skills.
Under the supervision of the author, measurements were taken in seven fields per community, following the steps below, described in detail in Tschakert (2004b) and Tschakert et al. (2004):
Involving community members in the C measurements was not only an obvious next step in the social learning process but also an empowering experience. The three teams clearly took responsibility and ownership of the process. Slightly flawed measures were corrected or retaken. This was especially the case with bulk density samples, which require skill, precision, and patience. Since different group members took turns collecting the samples, certain minor inconsistencies in pressure and cleaning may have occurred. All in all, though, the teams proved convincingly that local "experts" can carry out such measurements with great care, enthusiasm, and minimal training. Also, the teams considered it their task to inform the individual field owners about the importance and the procedure of the C sampling (Figures 6a and 6b). The latter were invited to join and provide cropping and management histories of their plots. Our approach seemed in stark contrast to outside expert teams who had previously performed scientific work in the communities: on all of those occasions, local villagers had been excluded to avoid potentially disturbing interference.
Upon the teams' request, a French-Wolof C measurement manual was put together at the end of the research and distributed in each of the three communities. It contains a list of necessary tools, a detailed illustration of the various steps of the measuring process, the random number and distance tables, a tree sheet, and the lab results and final calculations of total system carbon for the 21 fields measured during this phase of the study. As shown in Figure 7, final numbers ranged from 20 to 48 tons of carbon per hectare (t C ha-1), with one outlier at 78 t C ha-1 in a high clay content field. It is significant to note that nearly 3/4 (72%) of total C was found in the soil; this confirms results from Manlay et al. (2002) in southern-central Senegal. This further stresses the importance of actively involving land managers in learning activities that complement their predominantly above-ground empirical knowledge, if carbon sequestration programs are to be sustainable in the long term.
Lastly, team members and other villagers debated how the local C teams could be employed in the future to practice their knowledge and conduct regular monitoring of C stocks under anticipated carbon sequestration programs. Two key requirements were put forward: first, such monitoring teams should measure at predefined dates (for instance every 3 or 5 years) in randomly selected fields. The calculation of average C numbers for individual communities would allow local land managers to flexibly use management strategies in response to household opportunities and constraints and pursue a reasonable—not optimum—maximum for the pre-set measurement dates. Second, to avoid free riders, C teams should sample only in fields outside their own communities and initially work with one external technician to build trust and refine their skills. In a rural context where employment opportunities are scarce, such local C teams would provide an additional source of income, empower project participants, and significantly reduce the costs of baseline and follow-up measurements. In combination with effective and equity-oriented land use plans, these local teams could represent viable institutional structures based on power sharing, well-established rules, and their enforcement. Wiber et al. (2004) identify such structures as crucial ingredients for participatory research protocols and community-based resource management.
(Back to top)
Popular theater is a type of participatory performance that is used to explore social problems or other issues of concern through short skits or scenes followed by audience probing and discussion (Cornwall 1997). In the 1960s and 1970s, popular theater evolved out of the popular education movement, spearheaded by Augusto Boal in Brazil (Conrad 2004). It has been increasingly used as a participatory (research) method in health promotion interventions, including HIV/AIDS, domestic violence, and risk-taking among youth, in sub-Saharan Africa (Sliep et al. 2004; Bagamoyo College of Arts et al. 2002) and beyond (Conrad 2004). It derives its power from animating both actors and audience to critically analyze their collective problems, discuss possible solutions, and initiate behavioral changes. As such, this type of participatory engagement provided a valuable capstone experience in our social learning sequence.
Agatha Thiaw, one of our team members, worked with elementary school teachers and a group of young student volunteers to portray the story of an emigrating insect family that flees its native village due to deteriorating soil conditions, agricultural mismanagement, and famine, seeking refuge in another community (Figures 8a and 8b). In the skits, Mr. and Mrs. Carbon and their children told their tale of farmers offending and mistreating their soil insects, ending with a plea to their host family not to commit the same mistakes. While the skeleton of the storyline and some key elements were provided, teachers and students created the individual scenes and added their own twists and some humorous parts. Both girls and boys participated, which allowed the differential gender roles in soil fertility management to be addressed. In all three communities, the play attracted a large audience and people subsequently engaged in further discussion on the topic. Teachers emphasized the role of children as future environmental stewards; some parents also applauded the research team for having initiated this experience. This was particularly true for Thilla Ounté where no plays had ever been performed before. Last but not least, the children were proud and felt empowered for having made a contribution to their community research.
(Back to top)
This article has shown how knowledge partnerships and social learning processes with respect to soil carbon can engage land managers more actively in research programs, complement their environmental wisdom, and, ultimately, contribute to more sustainable soil management decisions. Such participatory learning processes require an intellectual interest and a substantial time commitment from researchers. who need to go beyond data collection and analysis. In our case, this process was neither planned nor neatly structured into the three distinct components that are described here. It evolved over time, driven by the curiosity and needs of participating land users. Although, so far, no community-based carbon sequestration project has been implemented in the Old Peanut Basin, mainly as a result of exceedingly complicated funding and emission trading guidelines, there is still a great need for collaborative, long-term soil fertility management programs. Farmers, for their part, are ready.
(Back to top)
(Back to top)
Bagamoyo College of Arts, Tanzania Theater Centre, R. Mabala, and K. B. Allen. 2002. Participatory action research on HIV/AIDS through a popular theater approach in Tanzania. Evaluation and Program Planning 25: 333-339.
Berkes, F. 1999. Sacred ecology, traditional ecological knowledge and resource management. Philadelphia: Taylor and Francis.
Borrini-Feyerabend, G., M.T. Farvar, J.C. Nguinguir, and V.A Ndangang. 2000. Comanagement of natural resources : Organising, negotiating and learning-by-doing. Heidelberg, Germany: Kasparek Verlag.
Centre de Suivi Ecologique. 2000. Annuaire sur l'environnement et les ressources naturelles du Sénégal. Dakar: The Centre.
Conrad, D. 2004. Exploring risky youth experiences: Popular theater as a participatory, performative research method. International Journal of Qualitative Methods, 3 (1): Article 2. Online: http://www.ualberta.ca/~iiqm/backissues/3_1/pdf/conrad.pdf
Cornwall, A. 1997. Performance and participation: An overview. PLA Notes 29: 23-28.
Food and Agriculture Organization (FAO). 2002. Carbon sequestration in dryland soils. World Soil Resources Report 102. Rome: the Organization.
Gaye, M. 2000. Région de Diourbel: Politiques nationales affectant l'investissement chez les petits exploitants. Drylands Research Working Paper 12. Crewkerne, UK: Drylands Research.
Lal, R. 2002. Carbon sequestration in dryland ecosystems of west Asia and north Africa. Land Degradation & Development 13: 45-59.
Leach, M. and R. Mearns, eds. 1996. The lie of the land: Challenging the received wisdom on the African environment. Oxford: James Currey.
Manlay, R., C.-L. Chotte, D. Masse, J.-Y. Laurent, and C. Feller 2002. Carbon, nitrogen and phosphorus allocation in agro-ecosystems in a West-African savanna. III. Plant and soil components under continuous cultivation. Agriculture, Ecosystems and Environment 88: 249-269.
Mazzucato, V. and D. Niemeijer. 2000. Rethinking soil and water conservation in a changing society: A case study in eastern Burkina Faso. Tropical Resource Management Papers 32. Wageningen, The Netherlands: Wageningen University.
Moller, H., F. Berkes, P. O'Brian Lyver, P., and M. Kislalioglu. 2004. Combining science and traditional ecological knowledge: Monitoring populations for co-management. Ecology and Society 9(3): 2. Online: http//www.ecologyandsociety.org/vol9/iss3/art2
Parton, W., D. Ojima, C. Cole, and D. Schimmel. 1994. A general model for soil organic matter dynamics: Sensitivity to litter chemistry, texture and management. In Quantitative modeling of soil forming processes, eds. R. B. Bryant and R. W. Arnold, 147-167. SSSA Special Publication 39. Madison, Wisconsin: Soil Science Society of America.
Reij, C. 1991. Indigenous soil and water conservation in Africa. London: International Institute for Environment and Development.
Reij, C., I. Scoones, and C. Toulmin, eds. 1996. Sustaining the soil: Indigenous soil and water conservation in Africa. London: Earthscan.
Schusler, T.M., D.J. Decker, and M.J. Pfeffer.2003. Social learning for collaborative natural resource management. Society and Natural Resources 15: 309-326.
Scoones, I. 2001. Transforming soils: The dynamics of soil-fertility management in Africa. In Dynamics and diversity: Soil fertility and farming livelihoods in Africa, ed I. Scoones, 1-44. London: Earthscan.
Sliep, Y., K. Weingarten, and A. Gilbert. 2004. Narrative theater as an interactive community approach to mobilizing collective action in Northern Uganda. Families, Systems, & Health 22 (3): 306-320.
Tabor, J.A. 1990. Ethnopedology: Using indigenous knowledge to classify soils. Arid Lands Newsletter 30: 28-29.
Tschakert, P. 2001. Human dimensions of carbon sequestration: A political ecology approach to soil fertility management and desertification control in the Old Peanut Basin of Senegal. Arid Lands Newsletter 49. Online: http://ag.arizona.edu/OALS/ALN/aln49/tschakert.html
Tschakert, P. 2004a. The costs of soil carbon sequestration: an economic analysis for small-scale farming systems in Senegal. Agricultural Systems 81: 227-253.
Tschakert, P. 2004b. Carbon for farmers: Assessing the potential for carbon sequestration in the Old Peanut Basin of Senegal. Climatic Change 67 (2-3): 273-290.
Tschakert, P., Tappan, G. 2004. The social context of carbon sequestration: Considerations from a multi-scale environmental history of the Old Peanut Basin in Senegal. Journal of Arid Environments, Special Series 59: 535-564.
Tschakert, P., Khouma, M., Sène, M. 2004. Biophysical potential for soil carbon sequestration in agricultural systems of the Old Peanut Basin in Senegal. Journal of Arid Environments, Special Series 59: 511-533.
Vågen, T.-G., Lal, R., Singh, B.R. 2005. Soil carbon sequestration in sub-Saharan Africa: A review. Land Degradation & Development 16:53-71.
Wiber, M., Berkes, F., Charles, A., Kearney, J. 2004. Participatory research supporting community-based fishery management. Marine Policy 28 (6): 459-468.
Zanatell, B.A., Knuth, B.A. 2002. Knowledge partnerships: rapid rural appraisal's role in catalyzing community-based management in Venezuela. Society and Natural Resources 15: 805-825.
(Back to top)
Petra Tschakert (firstname.lastname@example.org) is assistant professor of Geography and the Alliance of Earth Sciences, Engineering, and Development in Africa (AESEDA) at the Pennsylvania State University. She conducts training workshops on participatory research methods and currently leads a USAID-funded project on national carbon assessments and adaptation to climate change (Spatially Explicit Modeling of Soil Organic Carbon, or SEMSOC) in collaboration with national partners in Ghana, Mali, Niger, and Burkina Faso.
(Back to top)
SEMSOC Project web site
About the Arid Lands Newsletter