The soil in some of the peasant cotton fields I visited in Mali and Burkina Faso was so hard and light-colored that one would be amazed at how such a fragile a plant could ever have sprouted and grown in there. In genetic terms, these cotton plants are not fundamentally different; they came from either the same or very similar seeds.Whence therefore the sharp discrepancy in appearance and productivity? Commonsense would get the answer right: context. Even the same seed, when enmeshed in a different set of relations, will look different; roughly, this is actor-network theory’s basic claim about technology transfer. But why is this difference important? This question requires taking a step further, because the difference observed here is not just mere difference, but is perceived as a lag that must be closed. That is, at bottom, what the actors in the project were working so hard to achieve: to make the first plant look more like the second. In order to close this gap, they had to work simultaneously on both context and technology. The lag between peasant farms and the no-till experiments shown in the picture above was not however the only one that mattered. The project’s experiments, as the following chapter will detail, involved a controlled comparison between cotton as it was grown in the conventional system prevalent in the C-4 countries,nft growing system and cotton managed through the Brazilian no-till system. But the difference observed in the picture was true not only for the parcels managed through notill.

This is because there was a fundamental gap in the conventional system itself, between what local research and extension recommended to peasant farmers, and what they actually did in their fields. Technology adaptation and transfer therefore unfolded along at least three contextual levels, from Embrapa to the African research institutes, and from the institutes to peasant fields. It is in terms of this configuration that a first adaptation to no-till was introduced: the system made its way into the West African context not just as an alternative to the conventional system aimed at long-term soil conservation , but as a repertoire ofpotential fixes to the current system itself, which showed an “internal” lag between research institutes and peasant land. A second set of adaptations concerned more fundamental differences between the agricultural context in which no-till was assembled back in Brazil, and its expected ultimate destination in West African peasant land. These differences were evinced through a demonstration of how no-till emerged and worked back in Brazil, during the capacity-building trainings, study visits, and other technical project activities. Like the demonstration of Brazil’s success story of agricultural development in CECAT, these efforts involved stabilizing a version of the no-till system. This version was organized according to three “pillars”. As the Embrapa agronomists put it, the system’s main difference in relation to conventional agriculture is to avoid tilling; this is its first and indispensible pillar. No-till does not dispense entirely with soil preparation, but typically this happens only once , as a preliminary stage before the system is put in place.

In the Brazilian cerrado, this stage involves breaking down the structure of the soil in depth by specialized machinery, and applying massive amounts of lime and fertilizers. In Brazil, these procedures are usually guided by detailed soil analysis and mathematic equations estimating the optimal amount of each chemical element, to avoid both under and over-application. To get this process right is vital so that the soil is in a good starting condition; otherwise, the system may go astray over the long term. The Brazilian front liners quickly learned about the difficulties entailed in carrying out these preliminary preparations, as well as in keeping up with subsequent fertilization cycles, in the West African context. In none of the project countries was there availability of specialized no-till machinery, or sufficient or readily available supply of lime and even some fertilizers. Even soil analysis – a very basic step in conventional agriculture, necessary for estimating the dosage of fertilizers to be applied – was far from common practice among West African peasant farmers. For the project plot in Sotuba, the alternative was to apply a good amount of organic fertilizer – 6 tons per hectare of animal manure – in order to at least rebalance the soil’s low organic matter content. The project got the manure from the institute’s livestock program, but such amount far exceeded what was typically produced within most local farms, and even what was generally available for purchase by peasant farmers. Nonetheless, part of the trainings versed on technical parameters for using animal manure as organic fertilizer, since, in the absence of an open market of chemical inputs and credit like the one existing in Brazil, this was one of the few options potentially available to local farmers. The second pillar – soil cover – encountered even more “resistance” from the recipient context. In Embrapa’s version of the system, soil should not only be left undisturbed, as it must remain covered in-between cropping seasons – a period of four to five months in the cerrado, but which may linger for over half the year in the C-4 countries.

In Sotuba, the project was experimenting with locally developed varieties of maize and sorghum, two of the most common cereal crops grown in West Africa. Ideally, however, coverage should be maximized by the introduction of a secondary, specialized plant called cover crop , which is sowed simultaneously with the main crop, in-between its lines . When the main crop is harvested , its residues as well as those of the cover crop are supposed to remain in the fields during fallow. The soil coverage that results reduces the loss of humidity through evaporation during the dry season, and erosion and runoff during the rains. Moreover, as it decays, it restitutes important nutrients to the soil that can be recycled to the next season’s crop,nft hdyroponic system which is planted directly on top of it without tilling the soil. Cover crops are an element of the system that concurs significantly to its overall fluidity. There are basic criteria for choosing a good cover crop: rapid growth and substantial production of aerial biomass; the more biomass, the better and longer covered will be the soil. This characteristic may also be combined with other benefits. Some cover crops are chosen for the capacity of their roots to infiltrate, and therefore aerate and capture nutrients from, deeper soil layers, bringing them up as they decay; others may be chosen for their capacity to fixate nitrogen from the air and restitute it to the soil after decay. Sometimes, these characteristics are combined in a same plant, or more than one cover crop is used. Much of the research on no-till carried out at Embrapa is about figuring out the best combination in each case. In the project, the choice of cover crops reproduced Embrapa’s extensive experience with certain grass and non-edible leguminous plant species , but researchers from the four project countries were encouraged to devise their own mix of cover crops to be experimented with, and to prospect for local alternatives or complements to the plants brought from Brazil.Indeed, some African partners proposed adaptations to this element of the system, such as experimenting with cover crops that could increase no-till’s potential for dissemination among local farmers. In the view for instance of the head Malian agronomist, edible leguminous crops were likely to be more appealing to peasants because of their more visible, short-term benefits. This immediacy was in fact a recurrent constraint met by the project as it eyed the transfer of Brazilian technologies to West African farmers – especially in the case of no-till, which is, at heart, a long-term conservation technique. But rather than constraints defined negatively as preventing effective transfer, project front liners sought to regard these and other pieces of context as potential enablers of technology dissemination already available locally – as long as the appropriate adaptations were figured out. In this sense, one of the system’s elements to be encouraged in its new context was the root system.

Indeed, much of the work during Phase I was focused on it. In the Brazilian version of no-till, a powerful root system was already a relevant experimental parameter, for its capacity to move and aerate the soil. In West Africa, it became also a way of performing the preliminary soil-revolving function that in Brazil was normally given a jump-start with the assistance of specialized no-till machinery. Thus, crotalaria has an aggressive pivotal root that pierces the soil vertically, while brachiaria has a very fine and abundant root system that spreads out horizontally . Besides revolving the soil in breadth and depth, their roots restitute organic matter and nutrients to it as they decay. In Brazil, the two were regarded as a good complement to each other, and Embrapa front liners reproduced this configuration in some of the project plots . Here, certain components of the system were tapped for their capacity to make up for elements present in the technology’s original assemblage that were missing in the new context: in this case, as a substitute for artifacts – machinery and inorganic fertilizers – that were not readily available in the C-4 countries.But just as, in this work of context-making, some of the technology’s components came to the fore as especially enabling at the arrival point, others receded to the background as their original purpose was rendered less significant or altogether reconfigured. This was the case of another challenge to no-till related to its second pillar, and which also came to be coped with by tapping on the root system: the difficulty, or even impossibility, of maintaining the soil covered during fallow in peasant land. In some areas, this is due to slash-and-burn, but most often it stems from the fact that cotton is normally grown as part of a highly diversified system that, in all four countries, almost always includes cattle and other livestock, which are left to feed on crop residues after harvest. Project plots were fenced off from animals and guarded full time; outside of the institutes, however, this was a variable neither Brazilians nor Africans could control. As one of the Embrapa agronomists put it when I asked him about this key issue early on, “there’s no way these fields can be all fenced off. They have, one, three hectares each. There isn’t even enough barbed wire in the world to fence all these”. His half-joking comment had some literal truth in it; as some of the African partners explained to me, unilateral attempts to fence off peasant fields generally failed. Even if a peasant does not own cattle, someone else’s animals will come and feed on his crop residues. Fencing or whatever alternative to it such as hedgerows or route management would have to be adopted by all or most peasants in a same region – something that does not seem to lie within the current horizon of possibilities. Ultimately, as local researchers were very much aware, this imbroglio is deeply tied to the broader land question , itself a legacy of colonial land regulations that are somewhat unique to Sub-Saharan Africa .In West Africa, the prevalence of customary tenure and peasant farming is even higher than in the rest of the continent, since settler plantation schemes were not widely deployed by colonial administrators. Customary tenure left however an ambivalent legacy: while it did prevent land concentration in the hands of a white elite during colonial times, it did not necessarily guarantee “land to the tiller” either. Lack of security of tenure is repeatedly mentioned by those in policy-making and development circles, including many of my African interlocutors, as a major obstacle to attempts at implementing agricultural development policies in general. But while development schemes such as the ones championed by the World Bank in the African continent have often called for private land titling and other market-based schemes as a way of encouraging investment and credit in agriculture, local scholars such as Sam Moyo have cautioned about the danger of assuming that in West Africa peasants would have ample and equitable access to land just because ownership has been customary.