Conversely, early season water deficits are typically avoided due to the risk of poor fruit set , and irrigation is often ramped up at the end of the growing season to prevent root damage due to low soil moisture at dormancy in some regions . Based on our observations, the North Coast vineyards effectively use irrigation to regulate water supplies, yet in the vineyards over the Lodi and Madera regions irrigation closely satisfies atmospheric water demands. Different viticultural strategies across vineyards are also noticed by the trajectories observed in LAI . Unstressed water conditions led to significantly larger canopies in the Central Valley vineyards, which play as feedback in terms of water demands. Thus, larger canopies have a greater transpiration potential, which is well satisfied based on our observations. Different production goals in terms of yield and fruit quality underline the magnitude of targeted irrigation inputs across California’s viticultural regions. While irrigation management in viticulture aims to balance maximizing yield and achieving high-quality fruit, different regions are subject to different fruit quality expectations. Such expectations are related to reputation and the overall expected fruit quality potential recognized as terroir .In the Madera region, a focus on high yield is more prevalent since the area does not seem to have a special recognition for wine production. The Lodi area has some recognition as an emerging wine region,nursery pots therefore production goals are not as biased towards yield or quality as in the former regions. Vine water use throughout the season is largely supported by irrigation. ETa fluxes represent vine water use across different viticultural regions and illustrate the use of irrigation management as a tool to accomplish distinct production goals within a viticultural production program.

For instance, growing season water use in the North Coast is about half in comparison to vineyards in the Madera region, and so it is the expected resulting yield.Based on our results, for any given meteorological condition, we would expect a wide range of ETa depending on the different aspects of viticultural management. Therefore, further developing tools that can accurately estimate plant water use and stress has been addressed as a key aspect of further advancing ET modeling methods and irrigation tools . As vineyards’ water demands highly depend on production goals and viticultural management, it seems unlikely that any derivation of ETc or ETo could provide enough information to effectively fine-tune irrigation while controlling stress levels. A comparison of daily ETo and ETa for vineyards in the three viticultural areas analyzed in this study shows that there is not a reliable relationship between these two variables across sites and years . When assuming a linear response between ETo and ETa is not possible to distinguish a consistent pattern in the regression coefficients underlying these relationships. The intercept range is greater than 5 mm day−1 while the slopes could lead to ETa estimates of near half to twice ETo. Independent of a given viticultural production goal, our results highlight the need for accurate estimates of ETa. Estimates of ETa in combination with ETo can provide a clear depiction of the amount of water demanded by a vineyard in comparison to a reference well-irrigated crop. Then, irrigation management could aim at a consistent ratio of these two variables that would satisfy plant water demands and regulate stress when needed. For the studied sites within the GRAPEX project, water managers in the North Coast vineyards might have more water available than their counterparts in the Central Valley; they routinely require accurate ETa information to regulate vine water stress to optimize the relationship between fruit quality and yield. On the other hand, when yield is a priority, such as in many vineyards across California’s Central Valley, water demands can be much higher. In these cases, optimizing irrigation to satisfy atmospheric demands throughout an entire growing season becomes more relevant, especially during drought years.

A comparison of daily ETa, ETc, ET0, and vine water demands following RDI strategies shows that irrigation management can lead to a wide range of vine water use . Our results also show that in all study sites, ETa tracks closely or is usually above ETc, indicating that limited stress conditions are prescribed across these vineyards. Consequently, important water-saving opportunities could be possible if RDI strategies were effectively implemented. Considering the time–frequency and latency resulting from satellite remote sensing ETa estimates, ground-based sensors can offer a unique complement to timely inform irrigation decisions within an RDI program. Unfortunately, there is limited availability of commercial products able to measure ETa, and further advancing these technologies might be a critical component to develop as part of irrigation management tool kits. Further studies within the GRAPEX project will integrate the relationship between water use and yield as well as fruit and wine quality for these vineyards. Those studies will aim to advance the understanding of how much water is needed to achieve a given production goal, and also explore potential water-saving opportunities while not compromising yield and quality objectives.This is the case of Embrapa, to which we now turn.Embrapa is currently Brazil’s chief provider of technical cooperation in agriculture to Africa and elsewhere, and even of Brazilian South-South cooperation at large. It is a publiclyowned national research institute, created during Brazil’s military rule in the early 1970’s as part of a broader governmental effort to enhance domestic food supply and open up new agricultural frontiers in the country’s hinterlands. This historical background is itself key to understand Embrapa’s cooperation, and it will be discussed in Chapter 3. According to its own self-account, Embrapa engages in two main types of international cooperation: scientific cooperation with Northern research institutions and other emerging economies in the global South; and “technology transfer to developing countries”, or SouthSouth cooperation. The first type follows the so-called Labex model, where Embrapa researchers spend time doing cutting-edge research in the facilities of partner institutions abroad, in the United States, France, England, and more recently South Korea and China.

The second type involves transferring technologies and building capacity in countries considered as being ata lower level of development than Brazil. The Secretariat of International Relations , a centralized units linked to Embrapa’s Presidency located at the institute’s headquarters in Brasília, is in charge of articulating and managing these activities, and many of Embrapa’s decentralized research units also have their own international relations personnel. When I did fieldwork, Embrapa offered four modalities of technical cooperation to African countries. The first were small-scale, short-term projects normally involving small teams of researchers from both sides . Embrapa researchers would eventually engage in this kind of cooperation even before the recent South-South cooperation wave. It typically involves study visits, technical training, and occasional transfer of small pieces of equipment and materials, and may also involve the production of new knowledge. I have followed up one such projects, with Ghana’s Center for Scientific and Industrial Research ,plastic planters which was not carried up to its conclusion. These smaller projects are supposed to be gradually phased-out in favor of more robust forms of engagement,the so-called structuring projects – this is a second modality, which will be described later on in this dissertation. A third modality of cooperation consisted in capacity-building trainings offered to trainees from Africa and elsewhere in a new center built for that purpose at Embrapa’s headquarters in Brasília, named Strategic Studies and Training . I participated in the first two trainings the center offered to Africans, in October 2010 and April 2011 . CECAT was literally built from scratch within a few months as part of the “PAC-Embrapa,” a generous supplement to the institute’s budget provided by Lula’s administration. Its small but hard-working team was recruited from other Embrapa units or hired anew, and a learning process was taking place as they accumulated experience with each new training cycle .Finally, there was the Africa-Brazil Agricultural Innovation Marketplace, a triangulation between Embrapa, the Forum for Agricultural Research in Africa , and several Northern agencies. This virtual platform provided relatively small research grants for bilateral teams made up of researchers from Embrapa and partner research institutes in Africa.Its format seems to be the closest to established models in the international development community, notably the World Bank’s Development Marketplace. This is certainly due to the involvement of one major donor and major multilateral institutions such as the World Bank, the UN’s International Fund for Agricultural Development , the FAO and the Inter-American Development Bank. I attended the forum of the first round of applications held in Brasília in October 2010, the only time when all 125 applicants gathered face-to-face. Due to the virtual and widely scattered nature of this platform, the Marketplace remained a secondary field site. To complement the ethnographic effort – and access to field sites was always negotiated ad hoc with the various actors, sometimes with success, sometimes not –, I conducted semistructured interviews with African and Brazilian participants in all these four modalities, as well as with ABC personnel working with Embrapa. Excluding the CECAT and Marketplace participants, to whom I had access mostly through informal conversations during the cooperation activities or after them through the internet, the remaining poll of more permanent actors in my research scope was no larger than twenty people.

While this reflects the circumscribed nature of Brazilian cooperation teams even in its largest initiatives, it has posed some challenges to writing and anonymizing research subjects. My narrative strategy is therefore a mixed one, combining omission of names and institutional or national affiliations, occasionally changing gender, expertise or other personal information, and, in some cases, slightly fictionalizing the narrative or making partial claims based on pieces of information that I assessed could be safely made available to other field interlocutors or the public at large. Here I will draw on my experience with all four modalities of cooperation, but the focus will be on the CECAT trainings and on a structuring project with four West African countries, the Cotton-4.In her ethnographic work on development as the “will to improve”, Tania Li described the aid apparatus in terms of a double process: what she called problematization and rendering technical . As was suggested earlier in this chapter, Brazilian South-South cooperation lacks a robust bureaucratic apparatus comparable to those of Northern agencies; neither does it have effectively centralized, standardized, across-the-board protocols for the design and execution of projects. Cabral and Shankland have recently characterized this as a “policy of no policy”. But as remarked earlier, here I am avoiding looking at it from a perspective of lack; indeed, more prominent during fieldwork were the ways in which what would be seen, from the perspective of “mature” development agencies, as lack may be turned into a positive asset for South-South cooperation. As both the historical and the organizational accounts above suggest, to look empirically at South-South cooperation in its own terms does not yield a coherent, alternative model to Northern development aid. It evinces, rather, a set of emerging interfaces loosely assembled around key nodes at the three organizational levels outlined here: principles, policy, and front line practice. Different from development aid as described by the literature, relations between these levels are not over-determined by a bureaucratized path that continuously recycles the messiness of practice back into the technicalities of policy; most often, these paths are loose, heterogeneous and intermittent. Principles are only partly translated into policy, and their deployment in front line practice has to do less with policy prescriptions than with the practical conditions under which cooperation is implemented. This section will bring some examples of the forms such relations may take, and suggest some of its differential effects vis-à-vis Northern aid. In their claims about South-South cooperation’s comparative advantages over Northern aid, which always implies a somewhat caricatured picture of it,emerging donors often argue that it would be better suited to other countries in the global South not only because of the supposedly similar experiences shared by provider and recipient, but because it would be tailored to each particular case. Indeed, due to its organizational architecture, Brazilian South-South cooperation is necessarily heterogeneous. Each initiative is to some extent tailored, because its technical content and implementation process are not over-codified by ABC’s policy and management apparatus, but are the task of the multiple institutions operating at the front line.