Author: naturehydro

Each of these has its own experience in domestic development and/or in cooperating with foreign agencies, and for the most part they have no contact with each other. Projects will therefore vary widely, sometimes even within the same institution, as is to some extent the case of Embrapa. Besides heterogeneity, another differential effect of South-South cooperation’s organizational outlook relates to flexibility and autonomy in implementation. Mawdsley suggested that one of the differences between emerging and traditional donors has been greater “dynamism”, “speed” and flexibility . In the case of Brazil, the absence of a pre-determined, standardized portfolio of expert solutions seems to allow not only for greater flexibility in translating principles and policy into practice, but also greater autonomy and some degree of experimentation at the lower end of the policy-front line spectrum. The limited amount of resources available especially for bilateral projects, for instance, may mean that their implementation takes the form less of an ambitious intervention over a broad slice of local reality guided by predefined technical prescriptions, than an initially circumscribed and somewhat openended enterprise that gradually expands along with its learning curve. Different than what was suggested by Mawdsley , however, this does not necessarily mean that project execution will proceed at a faster and smoother pace. In fact, the mediation apparatus that needs to be put in place for transferring resources abroad, normally through UNDP, adds an intricate layer of red taping that may weight heavily down on project activities on the ground,plastic nursery plant pot as will be further discussed in Chapter 4. Another set of effects resonates with the South-South principles of non-intervention, non–conditionality and demand-drivenness.

In some iterations of Brazil’s discourse, these principles appear as a negation of the country’s historical experience as a recipient of aid from the North, marked by impositions of various kinds, conditionalities in particular. South-South cooperation therefore makes a principled point of not imposing itself on recipient countries, but responding to their demands. The first cooperation model tried by Embrapa in Africa – a regional office located in Accra centralizing cooperation all over the continent – sought to apply the demand drivenness principle quite literally. It involved an “over-the-counter” mode of operation whereby African agents would come to them with their demands. This model was eventually changed with the centralization of cooperation in Brazil in 2010, for reasons that are both operational and political. Now most demands get to Embrapa through its international relations unit in Brasília, mediated by the Brazilian Cooperation Agency . In practice, projects may be, and occasionally are, offered to African counterparts. But one of the points to which the discussion in this chapter leads is that, from an organizational point of view, like most of its Southern counterparts Brazilian cooperation would not have the capability to impose itself on recipient countries anyway. The institutional presence of Brazilian cooperation in Africa is minimal, and it would not have the financial or organizational capacity to monitor, for instance, the implementation of conditionalities. This spares recipients from the bureaucratic burden that development aid usually adds to their already fragile institutional apparatuses and may make them even more dependent on foreign expertise and funds . But on the other hand, this non-interventionist approach brings other kinds of challenges, as it requires more extensive engagement by recipients in all stages of project design and implementation.

This requirement evokes a prominent issue in the development aid scene at large, that of ownership , or how to make local actors carry the projects forward after the donor leaves. This is an element of the Aid Effectiveness Agenda that has been addressed by Northern donors in bureaucratic terms – for instance, by incorporating aid projects into the recipient countries’ own policy systems. Even if ownership has not been explicitly incorporated in Brazilian cooperation’s official policy as such , it is something that individual actors, especially at the front line, care about deeply. Lack of ownership was among the reason why, for instance, the project between Embrapa and Ghana’s CSIR that I was following up in 2010 was eventually abandoned – the latter was not able, or willing, to invest in its last stage, which would demand a significant apportion of local funds for the trainings that were planned to take place next. But in other cases , the organizational configuration of Brazil’s cooperation could end up promoting ownership in the sense that, by requiring that recipient countries, institutions and individual actors invest their own, scarce resources in order to carry South-South projects forward, they would be encouraged to own them. Even if, as we shall see in later chapters, on the ground things are not as straight-forward, the point to retain from this chapter’s perspective is that to promote ownership by making African counterparts share some of the project costs is, again, as much a matter of principles as an effect of Brazil’s limited cooperation resources and particular organizational architecture. Finally, an analogous point could be made with respect to the provider vs. recipient dimension. In my experience with Brazilian initiatives, even if there was an asymmetry in terms of resources and capabilities between Embrapa and the African research institutes, the difference between providers and recipients did not appear as a “trustee-subject” boundary .In the C-4 Project, for instance, African partners were required to play a leading role in the production of diagnoses about the local situation and in the technology adaptation process.

If this concurs to fulfilling South-South principles of horizontality and mutual exchange, it seems to be, again, less a matter of applying principles through a clear policy path than an effect of the capabilities available to Brazilian cooperantes as they come to engage with local actors without a specialized apparatus for producing development-related knowledge. This may be the case even with more elusive assumptions about Brazilians’ supposedly higher socio-cultural capabilities for functioning in other Southern contexts . In Brazil’s cooperation for institution-building in East Timor, for instance, Brazilians’ presumed greater cultural openness for socializing with the locals was eventually made real by their poor fluency in central languages like English or French, which made socialization with other expatriates difficult.As I observed during fieldwork, the language barrier may also encourage the deployment of more practical and tacit idioms, such as joking relations, hands-on work,seedling starter pot and communicative mediation of technological artifacts and other non–humans. In this case and the others, what I wish to suggest is how assumptions about Brazilian South-South cooperation as being different than Northern aid may end up becoming true not because of an alternative, “Southern” bureaucratized path systematically linking principles to front line practice through policy, but due to characteristics at the level of organizations and resources that, from the point of view of established aid institutions, would be regarded as lacking. Finally, another effect of South-South cooperation’s organizational outlook is that those implementing it do not have the same mechanisms as their Northern counterparts for “recycling” failure back into the project pipeline .Especially now that smaller projects are being phased out in favor of structuring projects such as the Cotton- 4, Embrapa is likely to have fewer initiatives in Africa, and each of them will become more visible. Since, as remarked, domestic support for South-South cooperation is far from consensual, and implementing institutions are porous to pressures and influences by governmental sectors other than diplomacy, project failure could become a big issue for the institutions and individuals implementing them. As Leite also noted, many in Embrapa are indeed concerned about the potential for inefficiency that increased demand for South-South cooperation coming from Itamaraty could entail. Moreover, in contrast with scientific cooperation with Northern institutes, in technical cooperation with Southern countries there are no evident immediate returns from the point of view of the institution’s own interests – let us not forget, neither Embrapa nor most other implementing institutions are development agencies. The institute certainly benefited from the generous resources provided especially during the last years of the Lula administration.

CECAT’s infrastructure, for instance, has been useful to other ends like internal trainings, and it is hoped that some benefits may eventually accrue from technology transfer projects, such as access to new markets for Embrapa’s technologies. These and other dilemmas are being confronted as I write, and point to yet another effect of the emerging assemblage outlined here: its open-endedness. These aspects are not fully captured by the established approaches in the anthropology of development discussed here.This chapter argued that, like other emerging donors, Brazilian South-South cooperation is in a process of emergence, characterized by the formation of a unique assemblage made up of new interfaces that bring together, under the aegis of foreign policy, preexisting institutions, discourses, individuals, practices, and politics. What is certain is that it is changing and will continue to change, and if the forward drive unleashed during the Lula administration is not reversed , it could be that the current picture will eventually give way to a more stable and policy-oriented assemblage. However, it is unlikely that, even in this case, Brazilian cooperation will ever look like the picture described in ethnographies of Northern aid; but neither will it ever constitute an entirely alternative model to it. The historical genealogy proposed in this chapter indicates that South-South cooperation, even if discursively constituted in terms of an opposition to Northern aid, emerged from within a global apparatus built under Northern hegemony. As a result, Brazil’s process of emergence as a donor has been highly ambivalent and even contradictory, both internationally and domestically. While South-South cooperation involves a quest for recognition at a global scale, it has also been shaped by the domestic politics of foreign policy and its relations with other governmental and economic sectors such as agriculture. This double directionality is key for making sense of Brazilian cooperation at various levels, as the following chapters will continue to show. A similar point can be made regarding the organizational aspect, which was described here in light of a double claim commonly found in the anthropology of development literature: about development aid’s self-referential character, and its bureaucratization and de-politicization effects. A look at the organizational assemblage of Brazil’s South-South cooperation against this backdrop yielded a three-leveled architecture that is in a sense an inversion of Northern aid’s: instead of prevailing over discursive principles and implementation practice, the level of managerial policy is weak relatively to them. Rather than constituting a specialized, bureaucratized model alternative to its Northern counterpart, Brazilian cooperation has relied significantly on global bureaucracies such as UNDP on the one hand, and on the sector-specific experience of national institutions like Embrapa on the other. This state of affairs has an interesting effect. While South-South cooperation upholds principles that are largely crafted in opposition to Northern aid, it has no coherent bureaucratic apparatus to systematically translate them into practice. Yet, as I have argued in the last section, some of the effects of Brazilian cooperation do go in the direction of some of these principles, such as demand-drivenness, non-conditionality, mutual exchange, tailored projects, and even more elusive assumptions about Brazilians’ higher socio-cultural capabilities for operating in Third World contexts. Rather than being the outcome of planned policy, this seems to be an effect of the practical conditions under which cooperation operates – conditions that, when looked at from the point of view of established aid organizations, would be regarded as immature or lacking.By zooming in from the previous chapter’s hemispheric scale on relations between Brazil and Africa I am inverting my actual research path. The fieldwork on which this dissertation is based started in Africa, more specifically in Ghana, before it got to Brazil and its broader SouthSouth cooperation enterprise. Only then did I start to pay attention to the account Brazilian cooperation provided of itself, which was largely crafted by its diplomatic arm rather than by the front liners themselves. One of the things that stood out since then was a certain mismatch between the concerns shown by those pioneering Brazilian cooperation on African grounds, and what was said about it in official discourse back in Brazil.

The top layer served two purposes, one, it allowed the net infiltration rate to be calibrated to match measured average field infiltration rates of 0.17 cm/hr and two, it represented the expected increase in sediment uniformity expected in ploughed or tilled layers in agricultural settings. While, the impact of the top layer resulted in water being delivered more slowly to the heterogenous sediments below, varying rates of percolation occurred after reaching below the more homogenous layer allowing us to examine the effects of heterogeneity on nitrate transport and fate in the vadose zone. For each stratigraphy, we further varied the frequency and duration of water per application to investigate the impact of different AgMAR implementations that are similar to recent field trials conducted throughout the state . In addition, we tested the effect of antecedent moisture conditions on N biogeochemistry within the more complex stratigraphy by setting the model with a wetter initial moisture profile. Overall, a set of 18 simulation experiments were used to isolate and understand the contribution of different AgMAR strategies to enhance or decrease denitrification rates in deep vadose zone environments with homogeneous and banded configurations. A detailed model setup and numerical implementation is provided in Section 2.3. Although our reactive transport analysis was guided by a particular field site that is classified as a “Medium to Good” site for MAR , our aim was not to replicate site conditions in its entirety, but rather to enhance our understanding of how hereogeneity might impact nitrogen transport and fate under MAR.The study site is an almond orchard located in California’s Central Valley, southwest of Modesto, and north of the Tuolumne River . The surface soil is classified as a Dinuba fine sandy loam . The site is characterized by a Mediterranean climate,bucket flower with wet winters and hot, dry summers. Average annual temperature and total annual precipitation are 17.5° C and 335 mm, respectively.

As suggested above, the vadose zone typifies the valley with contrasting layered sequences of granitic alluvial sedimentary deposits consisting of predominantly silt loams and sandy loams. We therefore use these textures to design our modeled stratigraphic configurations with and without banded layers. The groundwater table in the study area typically occurs around 15 m below ground surface. Soil properties including percent sand, silt, clay, total N, total C, and pH are shown in Table 1. To specifically characterize the textural layers and subsurface heterogeneity at our site, we used electrical resistivity tomography . ERT profiles were generated along a 150 m transect to 20 m depth prior to flooding to quantify subsurface heterogeneity while the subsurface was relatively dry . Further, to validate the texture profiles generated by the ERT data, a set of six cores were taken along the transect of the ERT line down to nine meters with a Geoprobe push-drill system . The first meter of the core was sampled every 25 cm. Thereafter, cores were sampled based on stratigraphy as determined by changes in color or texture. The ERT profiles were used to develop the stratigraphic modeling scenarios and the coring guided the specification of the hydraulic parameters. Redoximorphic features were noted throughout the cores. Several scenarios were developed based on the soil textures identified in cores and the ERT profiles to provide insights into the effect of stratigraphic heterogeneity and AgMAR management strategies on NO3 – cycling in the deep subsurface, as described in section 2 above. The five stratigraphies modeled in this study are shown in Figure 1. The limiting layer in the ERT scenario spans 187 to 234 cm-bgs based on field core observations. For each lithologic profile, three AgMAR management strategies were imposed at the top boundary between 20 m and 150 m of each modeled profile . For each AgMAR management strategy, the same overall amount of water was applied, but the frequency, duration between flooding events, and amount of water applied in each flooding event varied : a total of 68 cm of water was applied either all at once , in increments of 17 cm once a week for four weeks , in increments of 17 cm twice a week for two weeks , and all three scenarios with an initially wetter moisture profile.

Note, that for all scenarios, the same reactions were considered, the water table was maintained at 15 m, and temperature was fixed across depths at 18°C, the mean air temperature for January to February in Modesto. For all scenarios, the modeling domain consists of a two-dimensional 20-meter deep vertical cross-section extending laterally 2,190 m and including a 190 m wide zone of interest located at its center, thus distant from lateral boundaries on each side by 1,000 m to avoid boundary effects. The zone of interest was discretized using a total of 532 grid blocks with a uniform grid spacing of 1 m along the horizontal axis, and a vertical grid spacing of 0.02 m in the unsaturated zone increasing with depth to 1 m in the saturated zone. A maximum time step of 1 day was specified for all simulated scenarios, although the actual time step was limited by specifying a Courant Number of 0.5, typically resulting in much smaller time steps during early stages of flooding. Before each flooding simulation, the model was run first to hydrologic steady state conditions including the effect of average rainfall . The water table was set at a depth of 15 m by specifying a constant pressure at the bottom model boundary , and the model side boundaries were set to no-flow conditions. Under these hydrologic conditions, the model was then run for a 100-yr time period including biogeochemical reactions and fixed atmospheric conditions of O2 and CO2 partial pressures at the top boundary, a period after which essentially steady biogeochemical conditions were achieved, including the development of progressively reducing conditions with depth representative of field conditions. For these simulations, the concentrations of dissolved species in background precipitation and in groundwater at the bottom model boundary were fixed, with compositions described in Table 2 to yield similar vertically distributed NO3 – concentrations as were measured in the soil cores. Flooding scenarios were then started from the initially steady flow and biogeochemical conditions developed as described above and run for 60 days. For these simulations, a free surface boundary was implemented for scenario S1 where 68 cm of water was applied all at once. In contrast, a specified flux boundary condition was imposed for the scenarios S2-S3, where floodwater applications were broken up over a week. The flood water composition is discussed in Section 2.3.5.

The groundwater composition was taken from analyses reported by Landon and Belitz for a groundwater well located near our study site. For simplicity,cut flower bucket the background recharge from rainfall was assumed to have the same composition as groundwater except that it was re-equilibrated under atmospheric O2 and CO2 conditions prior to infiltration. In addition, the concentrations of N species in the background recharge were set to values determined from our own analyses of N at the top of soil cores. The composition of the flood water was set to that of the background precipitation diluted by a factor of 100 for most constituents except for Cl-1 . Ratios of NO3 – to Cl-1 were used to trace the difference between dilution and denitrification effects on NO3 – . Denitrification and N2O production were simulated as aqueous kinetic reactions coupled to the fate of pH, CO2, Fe, S, NO3 – , and NH4 + based on the Spearman correlation analyses discussed above . Apart from pH and nitrate species, Fe and S have been linked to denitrification through chemolithoautotrophic pathways in addition to heterotrophic denitrification , and are therefore included in our reaction network. Heterotrophic denitrification of NO3 – to N2 was represented via a two-step reduction process of NO3 – to nitrite and NO2 – to dinitrogen . Additionally, chemolithoautotrophic reduction of NO3 – to N2 with Fe and bisulfide as electron donors were implemented. Further, dissolved organic carbon was observed throughout the nine-meter profile at our field site, and CO2 and N2O profiles showed strong correlation . Therefore, DOC degradation was simulated using Monod kinetics, although individual DOC components were not simulated consistent with other modeling studies . In particular, we considered a single solid phase of cellulose in equilibrium with acetate as the source of DOC. Parameters for cellulose dissolution were calibrated using the total organic carbon concentrations obtained for each cluster. Biodegradation of acetate was coupled to multiple terminal electron acceptors, including NO3 – , Fe and SO4 2- which follow the hierarchical sequence of reduction potential of each constituent implemented by using inhibition terms that impede lower energy-yielding reactions when the higher energy yielding electron acceptors are present. These microbially mediated reactions and their kinetic rate parameters are shown in Table 5. Rates for denitrification were calibrated using the results from the acetylene inhibition assays as described above. Enzymes involved in denitrification include nitrate reductase, nitrite reductase and nitrous oxide reductase. To remain conservative in our estimates, we chose values typical for oxygen inhibition of nitrous oxide reductase L -1, the most sensitive to oxygen of the enzymes . Spearman rank correlation indicated that pH, DOC, S, NO3 – , and Fe exhibit significant correlation with N2O and therefore, these geochemical species were included in the reaction network. Cluster analysis was used to further detect natural groupings in the soil data based on physio-chemical characteristics, textural classes and the total dataset. Cluster analysis revealed three clusters representing distinct depth associated textural classes with varying levels of substrates and biogeochemical activity. Table 5 shows the median and range for N2O, CO2, NO3 – -N, Fe, S and total organic C for each of the clusters. The first cluster is dominated by sandy loams within the top meter with highest median values of total N2O, total CO2, NO3 – -N, Fe, and total organic C concentrations, indicative of greatest microbial activity and denitrification potential. The second cluster is dominated by silt loams below one meter and had average values of total N2O, total CO2, NO3 – -N, Fe, and total organic C concentrations when compared to the other groups. The third group is dominated by sands and sandy loams below 1 meter and had the lowest median values of total N2O, total CO2, NO3 – -N, Fe, and total organic C concentrations amongst all groups. While most concentrations followed a decreasing concentration trend from cluster 1 to 3, the highest median values of S were associated with cluster 2. Liquid saturation profiles and concentration of key aqueous species predicted at different times for the homogeneous sandy loam column are shown in Figure A1. The sandy loam vadose zone is computed to be 32% saturated with near atmospheric concentrations of O2. As a result of oxic conditions, model results demonstrate significant residual NO3 – concentration within the vadose zone . Evolving from these conditions, Figure A1d shows that with flooding scenario S1, water reaches depths of 490 cm-bgs and saturation levels reach 40% in the sandy loam column. Deeper in the column, lower saturation and only small decreases in O2 concentration are predicted . Calculated concentration profiles show that O2 introduced with the infiltrating water is persistent at shallow depths down to 100 cm-bgs, below which O2 declines slightly as floodwater moves below this zone. Model results further indicate higher NO3 – reduction in the shallow vadose zone including the root zone with 35% of NO3 – being denitrified . Overall, this scenario results in NO3 – concentration persisting at depth. While other redox reactions, such as iron reduction and HS reduction of NO3 – to N2, may be important, conditions needed to induce these reactions were not realized in the sandy loam vadose zone due to the high pore gas velocities of the homogenous sandy loam allowing for large amounts of O2 to penetrate the profile from the incoming oxygenated water. In comparison to the homogenous sandy loam column, the predicted water content is higher and O2 concentration is 53% lower in the vadose zone of the homogenous silt loam column at steady state . This result is expected because of the difference in porosity, with silt loams having higher water holding capacity and lower pore gas velocities compared to sandy loams.

The last chapter will conclude by foregrounding certain trends in the project’s emerging technology adaptation and transfer strategy, as I came to see it: relatively hands-off, fluid, and open to inputs by actors on the recipient side. This has meant greater potential for horizontality, but also vulnerability to all sorts of noises in the translation chain . It could even be argued that this mode of engagement marks many of the relational interfaces in the Brazil-Africa cooperation assemblage: between nation-states, between research institutes, and between researchers, technicians, and farmers. Whether this is better than what happens in Northern aid, depends on the perspective; but it is certainly more open-ended. It comes however at a risk: that of the denying of, or disinterest in, the relation by the African partners – in which case it will inevitably die out. To address this last issue, instead of the technical jargon of ownership or an emphasis on the self-referential character of development cooperation , I will deploy the more relational and open-ended idiom of robustness. In their account of the rise of Mode 2 science in the contemporary “age of uncertainty” , Helga Nowotny and colleagues argued that robust scientific knowledge is that which, rather than shielding itself from society in a position of expert authority , actively seeks to be strongly contextualized in it . When discussing an ethnographic example of such a configuration amidst the “rise of management audit”, Strathern complicated the abstract, preemptive notion of society that is being conjured up in some of these schemes. But by acknowledging the inherent open-endedness of social transactions, she points to the possibility of the same schemes activating new relations, producing “in real time, on [their] own scale, unlooked-for effects” . Similarly to Mode-2 science, technology transfer has always been about not just avoiding isolation from society,procona london container but finding ways to actively produce strong contextualization in it; if the travelling technology’s relations with its new environment are not robust enough, it will neither thrive nor disseminate.

In the case of Brazil-Africa cooperation, robustness seems to be an open question, both theoretically and practically. Theoretically, one may say that, precisely because the hold of policy and bureaucracy over front line practice is, as Chapter 1 will argue, less firm than in development aid organizations, the possibility should be acknowledged of its effects being different than the ones described in the ethnographic literature. And since, as Chapter 1 will also argue, one of the effects of this loose grip of policy is heterogeneity in the implementation of projects, it is likely that if robustness is achieved, it will be more so in some cases than in others. But robustness was a question that imposed itself less through my readings of the literature than through my relations with some of the cooperantes. Having had little experience with international development before the recent surge in Brazil’s South-South cooperation, many of these people – who are research scientists, rather than development workers – seemed to have a sense of possibility that I did not envisage in the ethnographies of aid. I suspect that experienced and professionalized development workers in a way already know, and probably expect, that particular development initiatives will be short-lived, even if the overall system will certainly keep on going. By and large, however, this was not the case of Brazilian front liners. I would not equate this with naiveté, though; this sense of possibility seemed indeed real, precisely because of the emerging character of Brazil-Africa cooperation. In other words, at this point, future directions are open-ended, and, given the gloomy picture provided by the anthropological literature on development aid, that is a good thing. Finally, this leads to another reason why I have chosen the analytics outlined here over all-encompassing notions common in the anthropology of development such as governmentality.

In a discussion with fellow anthropologists in Brazil , I pondered whether, at least if deployed in a totalizing manner, these kinds of approaches would not run the risk of congealing virtualities that are today in full effervescence in this moment of emergence of Brazilian South-South cooperation – a moment marked by vitalities and internal tensions that seem to be no longer present in the cold landscape of Northern development aid. I therefore preferred, I affirmed, to instigate these vitalities and their multiple potentialities – to evoke a Deleuzian-Guatarrian idiom, the lines of flight flowing from their smooth spaces – than to incarcerate them in one single analytics that, albeit perhaps acceptable to, or even expected by, my relations at the desk, would not be productive, or even fair, in terms of my relations in the field. If, as may as well happen, Brazilian cooperation eventually goes on to be engulfed by development’s self-referential machine and these lines of flight fade away, this dissertation will nonetheless remain as the register of a moment – of a window that may be already closing – when things could have been different. This leads to the last cluster of general questions to be tackled in this Introduction, that of knowledge production and reflexivity in anthropology.How and when to pull the breaks on this “fractal unfolding of complexity” – or, in Akrich’s terms, the “propagation of causal chains in all directions” – inevitably brought about by ethnography? Akrich was a student of technology transfer, a phenomenon that is particularly conducive to the evocation of scales of context beyond micro-practice. She once responded to this question much along the lines of how her supervisor Latour probably would have: “On what grounds would the analyst stop [to extend the network] – apart from the arbitrary one of lassitude? Quite apart from the indefinite amount of time such a study would take, there is also the question as to whether it would be interesting” . I find Akrich’s response highly unsatisfying, and the reasons why point precisely to a broader problem with Latour’s version of actor-network theory and its “myopic” method : the “flatness” of the network, which in my view stems from Latour’s poor investment in reflexivity, as opposed for instance to the work of Strathern or Haraway. His perspective assumes that the analyst is as an observer whose task is to witness and register immanent relations in a flat network until some “arbitrary point of lassitude”. She supposedly operates at the same scale of the actors in the network, but is not situated in it, in the Harawayan sense ; what she must do is to describe what she sees and hears.

From this point of view, indeed, to reach for scales beyond the immediate scope of micro-practice would seem not just “uninteresting”, but illegitimate. Scaling and context-making are a major part of what they do,cut flower transport bucket and they do it asymmetrically; as many have noted before me , actor-networks are never flat. Far from being arbitrary, these moves have a direction – at times, a very clear one – driven by the actors’ interests and the politics in which they are enmeshed. To access these moves and later on provide an account of them necessarily entail, therefore, that the analyst situates herself in her field relations. Ethnographers are never free of the risk of being called upon, by her field interlocutors, to make this explicit, especially as the products of research go back to circulate in the field – among the examples that come to mind are Latour’s involvement in the nineties’ “science wars”, and the controversy that prompted David Mosse’s reflections on relations between “field” and “desk” . Any ethnography is therefore a compromise between at least two relational networks: one in the field, and one in academia . At the “desk”, when to pull the breaks on the unfolding of complexity unleashed by fieldwork is hardly arbitrary either. Often, however, the ways by which this happens are eclipsed by recourse to the supposedly disembedded domain of theory. Take, for instance, the debates about technification-depoliticization in the literature on development. To argue that development discourse and practice are about rendering technical or depoliticizing problems that find their roots elsewhere assumes that the analyst knows what is the actual cause of poverty, underdevelopment, and so forth: namely, politics, usually conceived in terms of historical and political economic processes. This assumption that it is historical and political economic processes that matter – that are the real ones – may have been embraced by the analyst even before she went to the field, during seminars or in preparation for qualifying exams. In the writing stage, theory is brought back in to select and weave field data together – Gramsci, Foucault, or some other European philosopher of choice. This kind of operation follows knowledge practices that are prevalent in academia, rather than in the field. I do not have qualms with the depoliticization-technification claim as such; in fact, it is an assumption that I do share with the literature. What I try to do differently is to make explicit how my account is anchored both on debates in academia and on my field relations. This kind of movement can be found, for instance, in De Laet and Mol’s explicitation of their “love” for the Zimbabwe bush pump and its creator as “what moves [their] writing” . As a Brazilian citizen who has been personally interested in the politics of agriculture since long before this PhD, I have my own, non-academic stakes on the topic this dissertation is addressing; these have shaped my relationship with informants as well as the way I looked at their claims about South-South cooperation.

This oriented, for instance, my choice of what elements eclipsed by the actors to bring to the fore here – not an abstract “political economy” or “history”, but concrete political processes that have been central to Brazil-Africa relations, or to Brazilian domestic politics. Moreover, the fact that the empirical processes that I describe here are unfolding, as Maurer put it, “coincidently” with writing, led me to provide feedback to my field interlocutors not separately, but in this dissertation. Parts of my account, especially in Chapters 3 and 4, are responding to demands and debates coming less from academia than from the field. In a similar vein, assumptions of flatness and transparency that underlie actor-network theory’s formulations are rendered problematic when one acknowledges that the ethnographer’s view on the network is itself also directed by the actors’ interests. In other words, rarely is she given access to any and all actors and interactions in the field, and when and where she does, this happens through an often implicit “contract” with interlocutors that further constraints what can or cannot be made explicit in writing. Indeed, questions of access have been common in the anthropology of development, especially for those wishing to study up: that is, the developers rather than the poor communities or peasant farmers to be developed – normally more accessible to, because more vulnerable than, the ethnographer . Quite often, ethnographers of developers have had to “pa[y] their way [into the policy world] with knowledge products” by working for these institutions as consultants or volunteers, and that is how they get direct access to their internal operations: what Rottenburg described as “the main rule of access – ‘No admittance except on business!’”Alternatively, access can be granted through connections of a non-academic type, and this has been the case of some of the Brazilian literature on which I drew here. Many students of institutions such as Itamaraty, Embrapa or the WTO have either personal ties to their officials,or worked at them.Indeed, in occasions when I recounted my fieldwork experience to Brazilians from academia or from Embrapa who were not directly within my research scope, their first assumption normally would be that I worked at Embrapa, or was providing consultancy for the Brazilian Cooperation Agency. This was not the case, however. I had neither personal nor professional networks whatsoever in any of these institutions before I started this project. This often made things slow and sometimes oscillating for me during fieldwork, and, coupled with the rapidly changing character of Brazilian cooperation in its early moments, resulted in a fragmented fieldwork experience in institutional, geographic, and temporal terms. Thus, my field ended up including multiple institutions ; multiple countries ; multiple cities within countries ; three fieldwork languages ; and multiple fieldwork periods – not to mention long-distance interactions through email, skype, or facebook. This fieldwork experience has shaped writing in at least two ways. On the one hand, this dissertation does not reproduce the development version of Malinowskian fieldwork: it is not an in-depth experience of one particular cooperation initiative as it was observed locally.

Eared Dove, Southern Lapwing, Chimango Caracara, Rufous-collared Sparrow are species commonly found in agroecosystems and urban areas . Interestingly, abundance of Eared Dove and Chimango caracara were negatively related to the proportion of fragments but positively associated to the presence of fragments, indicating that even small fragments could be favor these species. Plot and landscape scale were important for conservation within this agroecosystem . This result indicated that both core areas of native vegetation as well as fragments of native vegetation within vineyards are necessary to enhance birds in agroecosystem. My results agree in part with others findings that the plot scale are more relevant for predicting distribution of birds in agroecosystem . Landscape scale was important for some communities as for granivores, indicating that multiple scales should be considered for this study area. My findings also coincide others highlighting the significance of small structural features such as single trees or hedgerows for increase birds within agroecosystems . This could be related to agriculture is a recent driver of land use change in comparison with other regions such as Europe where agriculture has a long history of habitat modification . These results should be interpreted in light of study limitations. Firstly, I did not measure micro-scale variables, as inter row cover crops or bare soil, that could shape bird communities in vineyards . However, previous research showed that these variables could be more significant during winter and not for the reproductive season in central Chile . Secondly, I did not assess the influence of the agricultural management as a covariate with bird abundance. Our study design cannot disentangle the interactions between agricultural management and landscape complexity.

Indeed,30 litre plant pots some of the vineyards that present high landscape complexity were managed conventionally while other vineyards that were managed organically presented low landscape complexity or were surrounded by other crops . Recent literature reveals that in highly mobile species such as birds, the proportion of natural vegetation within the landscape appears to be more relevant than conventional/organic management . Conservation of Forest-Matorral will play a fundamental role for the dynamics of whole bird communities. Given the potential for habitat fragments in and around the vineyard to enhance bird composition across the vineyard landscape and the fact that these fragments can be easily maintained or restored within vineyards, at low economic cost for vineyard production, I propose that maintaining and restoring fragments of native habitat should be widely adopted by grape growers . Restoration approaches should vary depending on the conservation goal. For example, in order to increase the potential ecosystem services provided by insectivore birds, fragments and large areas or Forest-Matorral are essential for a landscape planning in this area. In contrast, in order to conserve endemic rhynocriptids, preserving and restoring large areas of native vegetation is crucial. Most Scleorphyllous forest and Chilean Matorral is located in private sector, and the majority of the vineyards participating in this research own adjacent hills with native vegetation. A lack of policies and incentives that promote conservation of these ecosystems is still a challenge. Recently, laws approved by the Chilean parliament favor private parks, but incentives for conservation are still poor, with a lack of practical legal recognition and depend on individual species preferences by landowners . Simultaneously, land use change by vineyards is a continuous threat for wild areas due to continuous growth trends in the last 30 years .

Vineyard industry sales are valued at US $2,200 million for the year 2011 and projected to reach US $4,500 million in 2020 . The main driver of this industry is international market demand . Lack of land planning and restrictions to agricultural expansion are not currently part of the Chilean economic growth model . However, consumers demand could play a significant role in driving wine industry innovation toward a more sustainable arena . Strategies that preserve and restore fragments of native vegetation and core areas within vineyards can provide a successful approach for bird conservation in Chilean agroecosystems. Agricultural water districts are perhaps the most important players in efforts to reform water-resource management in California. According to several observers, a key impediment to the evolution of California water markets is the requirement in state law that water districts must approve any transfer of water rights outside of their borders . Agricultural irrigation districts have been particularly reluctant to participate in sales that would apparently transfer water from low-valued agricultural uses to higher-valued urban and industrial consumption. How these districts might distribute the costs and benefits associated with these trades has been the focal point of removing this particular barrier to developing viable water markets . In addition, the 1992 Central Valley Project Improvement Act focused on water districts as the agents for implementing water conservation and efficiency measures . On the other hand, recent attempts to establish water market protocols in California that bypass district control have met stiff resistance to date from agricultural interests. 1 Proposals by economists to reform water-resource management and to develop water markets generally have not considered the institutional context in which the targeted agricultural districts operate. Most analyses of water rights markets assume that the participants are attempting to gain the maximum net profits or monetary benefits.

However, this presumption may be off target, particularly if public-enterprise agencies dominate the water management structure as is the case in California. Given that most future water transfers in California are likely to occur among public agencies, looking beyond typical neo-classical assumptions about the “theory of the firm” may be important to understanding how water markets might develop . Previous political economy studies of irrigation districts have looked at some of aspects of how district decision-making processes work , but none has examined California districts across political structures in an economic framework. The emergence of two recent issues adds to the importance of better understanding the incentives embodied in various water-district forms. The first is that use of any electoral system other than universally-enfranchised, popular-vote was challenged successfully in part in federal court . The Association of California Water Agencies intervened with an amicus curiae brief to defend the voting system now in use in California water districts . The second is the recent passage of Proposition 218 in November 1996. This new law requires in many instances that certain types of special-purpose taxes must be approved by a majority vote of the assessed-benefit, and fees and charges by a majority of “property owners” within the relevant jurisdiction . Many of the dynamics that now affect water districts using”assessed-value voting will come to bear in a larger context among many local governments.The one example of a political-science study examined how the various electoral rules affected voter participation . The authors, in a report done for the California Department of Water Resources, attempted to explain why property-based rules led to less “democratic” processes than the popular-based methods. Unfortunately,25 liter pot plastic the predictive theory was unclear in the analysis, and the statistical analysis did not strongly support the thesis. Two comparative studies used sociological methods. Coontz examined the historical development of the Kings River Water Association and maintains that districts single-rnindedly pursued physical acquisition and control of water rights either through construction of diversion facilities or by appealing to outside government agencies for assistance in funding of upriver storage structures . Eventually, a strong contractual arrangement was structured, and the previously strife-torn parties successfully stood in concert against the U.S. Bureau of Reclamation during contract negotiations. In the Grasslands Area, Coontz found that the legacy of Miller-Lux had left control of the region’s water rights and leadership role to the Central California Irrigation District . As part of the Miller-Lux operations, neighboring farmers were allocated water portions greater than they might have achieved in fighting MillerLux and losing. This cooperative arrangement among districts and farmers continues today. Both of these situations represent bargaining solutions driven by the perceived disagreement outcomes by each party. In the first case, the upstream districts could physically control the flow of the Kings River, while the downstream Tulare Lake farmers could appeal to outside political power in the USBR and the city of Los Angeles.

The result was a hard-driven bargain that required strictly defined behavior. In the second case, Miller-Lux, and later CCID, controlled the lion’s share of local water rights. As a result, its neighbors were quite willing to accept a cooperative rather than confrontational solution since they could face substantial losses if they defected. The classical paradigm in which the district maximizes the total net benefits of all members is the most frequently seen in the economic literature. In each case, the water district is entirely transparent to the motives of the farmers themselves. In other words, these models simply assume that district managers use maximizing aggregate net income as their objective function. The district managers have no individual motives themselves nor do they consider any other objectives than resource-use efficiency. The three more recent political-economy models approach differently the question of how districts’ policies are chosen . The first two models treat the institutional management-selection rules as the focal point of policy decisions, while the latter one examines the importance of informal political influence. The first and third models put the districts’ managers at the center of the decision making process, while the second one implies that decisions directly reflect the wishes of the districts’ members. The latter two models rely on information about individual members within each district, either about farming activities or relative political influence. None of the models assume that a district manager maximizes the total net benefits to member, but rather coalitions are built by targeting benefits to certain groups within a district. In the first model, district managers attempt to maximize district profits while maintaining a sufficient level of voting support in a median-voter or “isoprofit/isovote” model . This model focuses on managers as the decision-making unit. Unfortunately, McDowell, did not adequately specify the empirical model to give meaningful empirical results. In the second model, management policies are chosen based on which policy draws the greatest political support among the district’s members, which is done by comparing the relative economic benefits that each would receive . This approach views the members’ operations as the units of analysis and aggregates to the district level. The model sees the managers as simply transparent to the decision-making process. In the third model, the district managers attempt to maximize the benefits of the members subject to a distribution based on the relative political strengths of each member . This model examines the motives of both the managers and members and creates a two-stage optimization model, McDowell, examine whether government-enterprise managers respond to the sometimes divergent interests of “voter-consumers” in a manner different from those of private enterprises . Public managers must balance maintaining political support that ensures their tenure with maximizing net benefits to consumers of the districts’ services. The analytic framework uses the median-voter paradigm . M&U hypothesize that if political support is not proportional to revenue responsibility, i.e., the districthas many voters of whom few pay related fees or taxes, then interests diverge between the disparate groups within the district, They further ask whether cross-subsidies through pricing are more likely in the case of government enterprises. M&U build on Peltzman’s model in which the district manager attempts to maximize voter support subject to the constraint that total district benefits exceed a certain level . The dual of this problem is to minimize the economic benefits forgone to achieve a majority vote. The result is finding the tangency of the isovote and isoprofit curves in the multiple-group/price space. The isovote curve represents the combination of prices to the relevant groups within the district that maintain the same level of political support. The isoprofit curve represents the combination of prices to the relevant groups within the district that maintain the same level of total net benefits to the district. If the political process transmits voter support in proportion to the revenues generated by the consumers in each group, then the tangency should lie along the 45 degree line from the origin, Le., the relative prices for each group should be the same.

The land sharing and “wildlife friendly farming” approaches stress the importance of incorporating biologically diversity into farming practices as a way to reduce the reliance on agrochemicals, and to increase the value of the agricultural matrix for wildlife habitat within the agroecosystems . The main critique of the land sharing approach assumes lower yields from non-intensive agriculture coupled with the need for more agricultural land conversion . More recently a combined approach of land sharing/sparing was proposed to disentangle the debate and co-op favorable practices from both strategies . The importance of the surrounding natural ecosystem is equally important in both land sparing and land sharing strategies for biodiversity conservation purposes. Natural ecosystems provide irreplaceable habitat for wildlife, in particular for highly specialized forest-dependent specialist, migratory species, and local endemics . Core, well-preserved, extensive areas as well as fragments play a role in the landscape connectivity and meta-population dynamics . One proposed strategy to augment the quality of the agricultural matrix is to increase landscape complexity through conservation of native vegetation fragments within simplified agroecosystems . These remnants of vegetation and restored some of the benefit of these vineyards as wildlife habitat that allow increased levels of biodiversity compared to agricultural fields without structural diversification . Forest patches and fragments can contribute to higher species richness within the agricultural matrix and increase ecosystem services provide by biodiversity . Non-cropped areas increase the landscape heterogeneity in time and space, increasing sources of food, shelter, roosting, and nest sites which are important to support wildlife in agroecosystems .

Most current literature on conservation in agroecosystems comes from tropical systems, 30 plant pot whose climatic regimes and seasonality differ greatly from temperate ecosystems. These results provide new evidence of strategies for harmonizing food production with biodiversity conservation. extending these data to advance knowledge of the impact of agroecosystems on wildlife in mediterranean type ecosystems with not only provides regional insights but adds to the global land sparing- land sharing debate. In mediterranean type ecosystems, vineyards are a dominant land cover type and important for the economy of these regions . The remaining natural areas in these regions correspond to some of the most unique and threatened ecosystems worldwide in terms of number of local endemic species that are at risk from continued land use change . The immediate and future impacts of vineyards on wildlife are poorly understood and research in these agroecosystems can help long-term protection of these biodiversity hotspots. Birds are one of the most visible vertebrates in agroecosystems and can be used as indicators of habitat quality or levels of disturbance . Birds provide significant ecosystem services such as pest control within agroecosystems , although some species could be considered agricultural pests when they decrease yields . In these studies, I aimed to determine the influence of vineyards on bird communities in Mediterranean type ecosystems located in Chile and California, respectively. I utilized a natural experimental framework comprising a gradient of increasing vineyard land cover in both study areas. In Chapter one, I review the current literature to evaluate and summarize the impact of agroecological practices on bird conservation. Several practices, including increasing the structural complexity of habitat, organic management, and the supply of artificial habitats such as the addition of bird nest boxes, have been demonstrated to promote bird conservation.

In Chapter two, I examine the how vineyard and several site variables explain bird species detection rates and explore species co-occurrence patterns corrected by vineyard and shrub influence. Results showed that vineyards promote a subset of species, referred to as agricultural adapter birds that interact with non-agricultural adapter birds and thereby can change community composition. In Chapter three, I evaluate the influence of within-vineyard fragments of native vegetation and surrounding native vegetation on bird communities of central Chile. Results showed a positive influence on not only surrounding native ecosystems, but also remnant fragments within vineyards, on bird functional groups, including local endemics birds. In Chapter four, the influence of vineyard at the landscape scale in compared across both study regions by evaluating the influence of vineyards on species and co-occurrence patterns when corrected by vineyards. Results were similar in both study regions, with increasing vineyard proportion favoring agricultural adapters and increasing negative interactions between agricultural adapters and the rest of the bird community. Indirect effects in ecology are difficult to assess but potential consequences for ecosystems could affect resilience and inter specific interactions leading to trophic cascades . Similar findings in both California and Chile suggest indirect impacts from agriculturally adapted species on surrounding natural ecosystems may be important globally. Summarizing detailed information on birds and agriculture highlight the practical strategies that can be implemented across the vineyard landscape for the benefit of biodiversity conservation. The importance of not only the surrounding natural landscape but also fragments within vineyards highlights farm-scale management that can be implemented by growers seeking to conserve birds will have an impact on bird conservation. This result can inform practical environmental policy and management and promoted to growers or non-governmental organizations to facilitate restoration strategies within vineyards.

The landscape and farm-scale nature of these results are applicable not only for conservationists but also for growers implementing sustainable environmental programs, and may be part of a win-win strategy for conservation and sustainable wine production. Approximately ten thousands bird species are distributed across terrestrial and marine ecosystems . A significant number of these species cannot adequately survive in anthropogenic ecosystems . Today, about 13% of bird species worldwide are threatened , including more than 25% of omnivorous and frugivorous birds, and approximately 30% of herbivorous, scavenger, and piscivorous birds . Expected changes in climate could affect the distribution of birds species and their ecosystem functionality and ability to adapt to anthropogenic environments . Conservation beyond protected areas is needed to prevent mass extinctions , and improvement of suitability of agroecosystems for birds, will play an important role in conservation. Birds are one of the most visible vertebrates in agroecosystems. Many bird species have adapted to new anthropogenic environments through thousands of years of co-existence with humans, and although some can become pests of crops, multiple species provide multiple ecosystem services through their roles as scavengers and pollinators and through contributions to pest control, seed dispersal, nutrient cycling, etc. . However, land use change to agriculture and agricultural intensification have been reported as major threats to biodiversity conservation globally, as these activities have increased the extent of which natural habitat have become homogenized and simplified . Agricultural intensification have been documented to drive significant change in bird phylogenetic diversity . Loss of bird species is occurring at fast rates and therefore severely impacting ecosystem service provisioning in agricultural landscapes . These trends can be somewhat reversed by adopting agroecological practices and designs that create favorable habitats for birds in agricultural systems while maintaining acceptable crop yields . I review how agriculture drives land use change and its impacts on birds,grow raspberries in a pot and identify key threats to birds in agricultural systems. From an agroecological perspective I review field studies and meta-analyses to detail the role of the birds in agroecosystems, the associations of birds with specific crops, and the impacts of agricultural management on birds, including both generalizable trends and case studies from temperate and tropical agro-environments. I also explore opportunities to conserve birds through more biodiversity friendly farming systems and practices and how certain agroecological management strategies can enhance the beneficial roles birds play in agroecosystems. Literature was identified using search engines , first with general topics then with refined phrases . Different possible combinations of these key words were used in order to perform an extensive search on each specific topic. I compiled 205 documents used to conduct the review, consisting of 198 journal articles, four online resources, and three books. Studied used where selected and characterized based on a) experimental/influence design, meta-analysis; b) region , temperate areas , tropical areas , and tropical and temperate areas, and publication date .The studies used are publicly available and represent a non-random representation of this area of research. The data may be geographically biased by three main factors: firstly, temperate European countries and the USA have a longer history of research in agroecosystem biodiversity, likely resulting in more studies in this field; secondly, tropical countries have more research conducted in agroforestry systems , especially from long term studies in Central America; thirdly, developing countries from the Global South were not well represented in this study due to gaps in information.

However, this review compiles the most recent advances in this field, highlighting their significance for agroecological approaches for conservation. Biological control of insects by birds in agroecosystems is considered to be one of the main ecosystem services provided. It is estimated that reduction of 20-70% of the arthropod biomass in crop fields can result from consumption by birds, although this depends on season and arthropod population size . This ecosystem service provided by birds was recognized by the US Department of Agriculture in 1885, when the division of economic ornithology was created, but phased out in the second half of the 20th century when the use of chemical pesticides became prevalent . Recent research has revitalized interest in biological control provided by birds in agroecosystems. In Mexico, an exclusion experiment in shaded coffee reported that birds and bats significantly reduced insect pest populations, particularly during the wet season, suggesting that both birds and bats play a role in the provision of ecosystem services . In Jamaica, an exclusion experiment in coffee demonstrated that bird predation on the coffee berry borer reduced production costs by US $310/ha, equivalent to ~12% of the crop value. In the case of California alfalfa fields, an exclusion experiment reported that bird biological control of pests reduced insect abundance by 33% . The introduction of the threatened New Zealand Falcon in vineyards reduced grapes eaten by introduced song birds by 95% and by native birds by 55%, providing an ecosystem service that could reduce pest management costs by US $234-$326 per hectare . Western burrowing owls consumed more prey in irrigated fields in comparison with the semi natural shrub–steppe of sagebrush and perennial bush grasses in Idaho with no differences of prey abundance between the sites . Woodpeckers can act as biological control agents of insect larvae in apple orchards. For example a study in Nova Scotia between 52-90% of the insect predation was attributed to woodpeckers over seven years across 47 orchards . In Chinese flooded rice fields Huang et al. reported that field leaf hopper abundance was reduced by 63-77% and weed biomass by 50-94% in rice-duck systems, increasing the yield by 295 kg/ha while reducing the cost in agrochemicals. Further, the rice-duck system reported better quality rice of grains . In a cereal crop, biological control by birds was higher in an organic field within a diverse landscape than in conventional simplified monocultures . Interestingly, Maas et al. reported that birds and bats provide similar effects on biological control in the tropics as compared with temperate and boreal ecosystems. These noteworthy results provide novel quantitative insight into the role of birds in agroecosystems and their economic benefits. Birds form part of this interactions include intra-guild predation , functional redundancy , and facilitation or niche partitioning . In agroecosystems where few organisms are top down predators, like birds, the presence of these predator organisms could drive shifts in communities . Consequently, the effect of birds as biological control in agroecosystems could be considered essential for sustainable agriculture, and be based on direct predation decreasing abundance of arthropods or/and other vertebrates, or indirect by affecting the behavior of the prey . Additionally, one of the most profitable activities related to birds, with potential to be implemented in agroecosystems, is bird watching. Bird watching is a popular activity in some countries with great potential to help meet conservation goals . Biggs et al. found that empowering local communities in South Africa as guides for bird tourism permitted significantly increases in their monthly earnings by ~$US 250. Further such activities also increased the guides sense self-worth and self-determination, and instilled a sense of pride in their local environment, with in turn led to direct conservation benefits. Another economic argument tied to birds in agriculture is that many consumers are willing to pay more for produce derived from wildlife friendly farming systems.

In recent years, donors and governments have increasingly targeted development aid to conflict-affected areas, often in the hope that aid will reduce conflict by “winning the hearts and minds” of the population. The idea is that by implementing development projects, for example by building roads, schools and hospitals, or by extending technology, cash transfers or insurance to poor people, we can increase popular support for the government and reduce support for insurgent movements. Facing a more hostile population, insurgents will find it harder to recruit fighters, acquire supplies and carry out attacks, leading to an overall reduction in violence. This idea—that development aid can be used to win hearts and minds—is widespread and forms the basis of much of the U.S. Armed Forces’ counterinsurgency strategy. Yet, there is no conclusive empirical evidence that development projects reduce violence. In fact, there is anecdotal evidence for the opposite.A recent report on civil counterinsurgency strategies by the RAND Corporation warns that “insurgents strategically target government efforts to win over the population. Indeed, the frequency with which insurgents attack schools, government offices, courthouses, pipelines, electric grids, and the like is evidence that civil [counterinsurgency] threatens them.”In a recent working paper, my coauthor Patrick Johnston and I offer a simple but frequently overlooked explanation for why the strategy of winning hearts and minds often backfires: if insurgents know that successful development projects will weaken their position,container raspberries they will try to derail them, which may exacerbate conflict. To help us think more clearly about this mechanism, we developed a simple theoretical model of bargaining and conflict around development projects.

The model’s premise is that the government tries to implement a development project while the insurgents threaten to use force to derail it—perhaps by attacking government staff or infrastructure, or by intimidating the population into not participating in the project. The model assumes that the government and insurgents engage in negotiations, during which the government can pay off the insurgents in return for allowing the project’s peaceful implementation. However, the insurgents know that a successful project will win the hearts and minds of the population and will make it harder for insurgents to achieve their political aims in the future. Thus, if the government wants to convince the insurgents to leave the project in peace, it has to compensate them for the shift in power that a successful project will bring about.Previous theoretical work on the causes of conflict has shown that a large shift in power between two parties can cause bargaining to break down. Our model shows that if a project causes a shift in power that is large enough, the government may not be willing or able to compensate the insurgents and conflict will occur. While this theoretical modeling exercise may seem somewhat abstract, it allows us to predict the conditions under which development projects are most likely to cause conflict. First, conflict is more likely if a successful project causes a large shift in the balance of power between insurgents and the government. Second, conflict is more likely if insurgents have a strong military capacity that they can use to effectively derail the project . I will come back to these insights at the end of this article and discuss what they can tell us about the best way to implement development projects in areas affected by conflict.To test the predictions of our theoretical model, we estimate the causal effect of a large development program—the Philippines’ KALAHI-CIDSS program— on casualties in armed civil conflict. From 2003 until 2008, KALAHICIDSS was the Philippines’ flagship anti-poverty program with a budget of $180 million, financed through a loan from the World Bank.

The program distributed grants for small infrastructure projects to the poorest 25% of municipalities in the 40 poorest provinces of the Philippines. In doing so, it followed a community-driven development framework that allowed the population to propose projects and decide which projects to fund through a participatory democratic process. We estimate the effect of this program on the ongoing conflict between the government of the Philippines and the country’s two largest organizations: the communist New People’s Army and the Muslim-separatist Moro Islamic Liberation Front . The New People’s Army is the armed wing of the outlawed Communist Party of the Philippines, a class-based movement that seeks to replace the Philippine government with a communist system. Since taking up arms in 1969, the NPA has relied on guerilla tactics rather than conventional battlefield confrontations against government armed forces. Its current strength is estimated at 8,000 armed insurgents who operate in rural areas all over the Philippines. The Moro Islamic Liberation Front is a separatist movement fighting for an independent Muslim state in the Bangsamoro region of the southern Philippines. It was formed in 1981, when the group’s founders defected from the Moro National Liberation Front, another long-standing southern Philippines insurgent movement. The MILF’s core grievances stem from disputes over lands considered by the southern Muslim population to be part of their ancestral homeland. With an estimated 10,500 fighters under arms, the MILF is larger than the NPA. However, the MILF has a more narrow geographic focus and only operates in parts of the southern island of Mindanao. Overall, conflict with these two groups has been ongoing for over four decades, caused more than 120,000 deaths, and cost the country an estimated $2–3 billion. We had access to information on all conflict incidents that involved units of the Armed Forces of the Philippines between 2001 and 2008.

These data were originally collected for the AFP’s own intelligence purposes, but a declassified version has recently been made available to researchers.Estimating the causal effect of development projects is difficult under any circumstances, and particularly so in conflict-affected areas. To cleanly identify the causal effect of development aid on conflict, one would optimally like to compare two places that are exactly identical in all characteristics, except that one of them received aid while the other did not.Since this is not possible in the real world, researchers usually use regression analysis to “control” for differences in observed variables. By controlling for a variable in a regression, we can “hold its effect constant,” which allows us to compare places that differ in the variable as if they did not. If we were able to measure all the differences between places that receive aid and places that do not, we could control for them in a regression and filter out the pure effect of development aid on conflict. Unfortunately, this is virtually impossible in the real world since many important variables are hard or impossible to measure. We may, for example,draining pots be able to measure and control for differences in demographics, poverty and access to infrastructure, but crucial variables like the strength and militancy of local insurgents and their level of support in the population are nearly impossible to measure. If these unmeasured variables differ systematically between places that receive aid and places that do not, we run the risk of misinterpreting these differences as the causal effect of aid, which would lead us to the wrong conclusions. For example, suppose an aid agency is worried about the safety of its staff and therefore targets aid to places with little or no insurgent presence. In this case, we would most likely find that the places that receive aid from this agency experience less conflict than the places that do not receive aid. However, this does not mean that aid caused a reduction in conflict, but merely that the agency targeted aid towards places that had a low propensity for conflict to begin with. The key to estimating causal effects is therefore to ensure that one is comparing like with like—i.e., that the places one is comparing do not differ in unobserved variables. To overcome this challenge and cleanly identify the causal effect of the KALAHI-CIDSS program on violent conflict, we employ a statistical method called Regression Discontinuity Design . This approach ensures that one is comparing like with like by exploiting arbitrary thresholds in the targeting of interventions. In our case, eligibility for the KALAHI-CIDSS program was restricted to the poorest 25% of municipalities. Thus, municipalities just below the 25th percentile of poverty were eligible and municipalities just above the 25th percentile were not. The basic idea of the RDD approach is that—since the location of the threshold is basically arbitrary—municipalities just above and just below the threshold should not differ systematically in any unobserved variables that determine conflict. We can therefore estimate the program’s causal effect by comparing the intensity of conflict in municipalities just below and just above the eligibility threshold.

The main results of our econometric analysis are summed up in the graph in Figures 1 and 2. The graphs compare the intensity of conflict—measured respectively by the number of casualties and the probability of having at least one casualty in a given month—in municipalities that were barely eligible for the KALAHI-CIDSS program and municipalities that were barely ineligible. Monthly averages of conflict in barely eligible and barely ineligible municipalities are denoted by solid and hollow circles, respectively. Smoothed time trends are plotted as solid lines for eligible municipalities and dashed lines for ineligible ones. The dashed vertical lines mark important dates in the project’s timeline. The first line at t = 0 marks the beginning of preparations for the project in eligible municipalities; the second line marks the start of the project’s implementation six months later. The third vertical line marks the project’s scheduled end after three years. The graphs show that both eligible and ineligible municipalities experienced similar levels of conflict in the period before the project. However, at the start of the project preparations, conflict increased sharply in eligible municipalities but remained virtually unchanged in ineligible municipalities. The difference in the intensity of conflict then became smaller over time and virtually disappeared as the project ended. Overall, the graphs suggest that the KALAHI-CIDSS program caused a large increase in the intensity of conflict over the three years of its duration. The regression results that correspond to these graphs, which are presented in detail in our paper , suggest that the KALAHI-CIDSS program caused a 70– 90% increase in the number of conflict casualties in eligible municipalities. In aggregate, we estimate that the program caused approximately 500 excess casualties over the three years of its duration. Our regression analysis also shows that eligible and ineligible municipalities did not significantly differ in pre-program or post-program levels of conflict, which supports our claim that the observed differences are really due to a causal effect of the program and not due to systematic differences in unobserved variables. Additional results show that the majority of casualties were suffered by insurgents and government troops, while civilians appear to have suffered less. We further find that the program caused similar increases in insurgent-initiated and government-initiated violence, suggesting that the effect is not the result of a one-sided offensive by either party.Our research makes two contributions to the study of civil conflict in developing countries. First, it provides empirical evidence that development projects can cause violent conflict. This evidence is particularly strong because our method of analysis is able to overcome the central problem of causal inference: that places that do and places that do not receive aid differ systematically in important unobserved variables such as the strength of local insurgents. By exploiting a discontinuity in the targeting of aid—the fact that only the poorest 25% of municipalities were eligible—we were able to compare municipalities that were barely eligible for aid with municipalities that were barely ineligible. Since the threshold at the 25th percentile was chosen arbitrarily, barely eligible and barely ineligible municipalities should not differ in unobserved variables, so that the difference in conflict between them reflects the causal effect of the development project. Of course, even though our results show that development aid can cause conflict, they do not suggest that we should stop giving aid to conflict affected areas. Many of the world’s poorest and most vulnerable households live in areas affected by conflict and cutting them off from aid would be throwing out the baby with the bath water. However, we believe that our theoretical model allows us to draw some conclusions about how to implement development projects while avoiding conflict.

Adjusted associations between a 10-fold increase in the amount of fumigants applied within 8 km of the home and the highest lung function measurements are presented in Table 4. We did not observe any significant adverse relationships between prenatal or postnatal fumigant use within 8 km and lung function. A 10-fold increase in wind-adjusted prenatal methyl bromide use within 8 km was associated with higher FEV1 and FEF25–75 . Additionally, a 10-fold increase in wind-adjusted prenatal chloropicrin use within 8 km was positively associated with FEF25–75 . Associations between methyl bromide and chloropicrin use and lung function observed in the prenatal exposure period were not observed in the postnatal period. Results were similar, although no longer statistically significant, for prenatal methyl bromide and chloropicrin use within 5 km of residences . There were no associations between fumigant use within 3 km of residences and lung function . We did not observe associations between postnatal fumigant use at any distance and lung function measurements or between fumigant use during the year prior to the assessment and lung function measurements .In sensitivity analyses using multi-variable models including other pesticide exposures that have been previously related to respiratory symptoms and lung function including childhood urinary DAP metabolites , proximity to agricultural sulfur use during the year prior to lung function assessment and prenatal DDT/DDE blood concentrations , the results were very similar to those presented in Tables 3 and 4. For example,blueberry plant size the relationships between prenatal methyl bromide use within 8 km were very similar for FEV1 and FEF25–75 .

Prenatal fumigant use was generally not correlated with other pesticide exposures that we found to be associated with lung function in this cohort, except for weak correlations between agricultural sulfur use within 1 km during the year prior to spirometry and prenatal use of metam sodium and 1,3 – DCP with r = 0.14 and r=0.26 respectively. The results were very similar when we only included children with two acceptable reproducible maneuvers in the analyses . The results were also similar when we excluded those currently using asthma medication, excluded the one outlier for FEV1 models or used inverse probability weighting to adjust for participation bias . Risk ratios estimated for asthma symptoms and medication using Poisson regression were nearly identical to the ORs presented in Table 3 and Supplemental Table 2. We did not observe effect modification by asthma medication use. Maternal report of child allergies modified the relationship between FEV1 and prenatal proximity to methyl bromide use and we only observed higher FEV1 among children without allergies . After adjusting for multiple comparisons, none of the associations reached significance at the critical p-value 0.002 based on the Benjamini-Hochberg false discovery rate. This is the first study to examine lung function or respiratory symptoms in relation to residential proximity to agricultural fumigant use. We found no significant evidence of reductions in lung function or increased odds of respiratory symptoms or use of asthma medication in 7-year-old children with increased use of agricultural fumigants within 3 – 8 km of their prenatal or postnatal residences. We unexpectedly observed a slight improvement in lung function at 7 years of age with residential proximity to higher methyl bromide and chloropicrin use during the prenatal period and this improvement was limited to children without allergies.

Although these results remained after adjustment for other pesticide exposure measures previously related to respiratory symptoms and lung function in our cohort, they do not remain significant after adjustment for multiple comparisons. There is a strong spatial pattern of methyl bromide and chloropicrin use during the pregnancy period for our study because of heavy use on strawberry fields near the coast at the northern portion of the Salinas Valley . There could be other unmeasured environmental or other factors that are confounding the relationship we observed between higher prenatal fumigant use and improved lung function. Previously published studies of prenatal exposure to air pollutants and lung function have generally observed links to alterations in lung development and function and to other negative respiratory conditions in childhood, and plausible mechanisms include changes in maternal physiology and DNA alterations in the fetus . Improved lung function was associated with higher estimates of recent ambient exposure to hydrogen sulfide in a study of adults living in a geothermal area of New Zealand . However, hydrogen sulfide has been shown to be an endogenously produced “gasotransmitter”, with anti-inflammatory and cytoprotective functions , and is being explored for its use for protection against ventilator-induced lung injury . In previous studies of this cohort, we found increased odds of respiratory symptoms and lower FEV1,  and FVC  per 10-fold increase of childhood average urinary concentrations of metabolites of organophosphate pesticides . Other studies of prenatal pesticide exposure and respiratory health in children have mostly evaluated exposure using cord blood concentrations of DDE, a breakdown product of DDT, and have observed an increased risk of respiratory symptoms and asthma with higher levels of DDE . Most studies of postnatal pesticide exposure and respiratory health in children have utilized self-reported information from mothers to assess pesticide exposure and have observed higher odds of respiratory disease and asthma with reported pesticide exposure . None of the previous studies of pesticide exposure and respiratory health have specifically evaluated fumigants.

Another strength of the study is that CHAMACOS is a prospective cohort followed since pregnancy with extensive data on potential confounders of respiratory health and other measures of pesticide exposure. Our study also had some limitations. We did not have information on maternal occupational exposure to fumigants or the geographic location of maternal workplaces during pregnancy, and we did not have the location of schools during childhood. These limitations likely resulted in some exposure misclassification during both the prenatal and postnatal periods. An important consideration in this study is that we estimated fumigant exposure using proximity to agricultural fumigant applications reported in the PUR data, which is not a direct measure of exposure. However, the PUR data explains a large amount of the variability of measured fumigant concentrations in outdoor air . In conclusion, we did not observe adverse associations between residential proximity to agricultural fumigant use during pregnancy or childhood and respiratory health in the children through 7 years of age. Although we did not observe adverse effects of fumigants on lung function or respiratory symptoms in this analysis,plant raspberry in container we have seen adverse associations in previous analyses of the CHAMACOS cohort between residential proximity to higher fumigant use and child development. We observed an association between higher methyl bromide use during the second trimester of pregnancy and lower birthweight and restricted fetal growth . We also observed decreases of ~2.5 points in Full-Scale intelligence quotient at 7 years of age for each 10-fold increase in methyl bromide or chloropicrin use within 8 km of the child’s residences from birth to 7 years of age . Future studies are needed in larger and more diverse populations with a greater range of agricultural fumigant use to further explore the relationship with respiratory function and health. It is typically assumed that output levels and prices in the U.S. food processing sector are directly linked to the availability and prices of the agricultural products or materials used for production. However, the traditional link between farm and food prices and production may be weakening. Adaptations in input costs and food consumption patterns are leading to changes in the production structure and technology of the food processing industries, that in turn affect demand patterns for primary agricultural materials. Such structural changes have been documented not only by anecdotal evidence, but in studies such as Goodwin and Brester, and Morrison and Siegel. In particular, Goodwin and Brester find that value-added by manufacture, both per worker hour and as a percentage of sales, increased in the 1980s in the U.S. food and kindred products industry overall, possibly implying an undermining of MA demand. Various economic and behavioral factors underlie these trends. As noted by Goodwin and Brester, relative prices of inputs important to food manufacturing, such as energy and labor prices relative to those for raw materials, shifted significantly in the past couple of decades.

The business environment also has experienced quite a transformation, including market structure and regulatory changes in the early 1980s. Tax changes have, for example, had a direct impact on relative input prices, by affecting the prices of capital inputs. Perhaps even more important than these alterations in the economic climate facing food processors are adaptations in food demand patterns. The fact that a greater proportion of adults are in the labor force today causes a higher demand for food products that require little home preparation time; they are at least in part prepared at the processing plant. These modifications in dietary preferences, combined with changes in food technology that allow processors to adapt foods to meet those preferences, could lead to more in plant processing of agricultural commodities. Other technical changes associated with capital equipment and the quality of agricultural materials, could also have an impact on the relative demand for agricultural products. These adaptations in food product costs, demand, and characteristics may mean that food processors are responding by altering their input composition. If they are using more capital, skilled labor, and non-agricultural materials to produce food products than in the past, these factors could become increasingly important elements in processors’ costs relative to agricultural commodities. The corresponding decline in agricultural materials input intensity is likely to result in weaker effects of changes in agricultural commodity prices on food prices, which has important impacts on both consumers of the final product and producers of the raw agricultural materials. To address these issues, this study assesses the role of changes in food product demand, input prices, and food processing technology on food processors’ costs and output prices, with a particular focus on the use of agricultural commodities as compared to other factor inputs. Our analysis of cost structure and input composition changes in the U.S. food processing industries is based on a cost-function representation of production processes in these industries. In our model we recognize a full range of substitution patterns among capital, labor, energy, agricultural materials, food materials and “other” materials inputs resulting from input price changes or technological factors. This allows us to explore modifications in input mix, costs and commodity prices resulting from changing agricultural commodity prices and output demand. It also facilitates consideration of technological factors affecting MA demand and production costs such as the quasi-fixed nature of capital , scale economies, technical change associated with either time trends or capital composition, and agricultural innovations or market power embodied in the MA input price. The model is estimated using data on 4-digit SIC level U.S. food processing industries, and the results summarized according to time period and 3-digit code . The base price and quantity data for output, capital, labor, and materials are from the National Bureau of Economic Research Productivity Database. The materials breakdown was drawn from data in the Census of Manufactures, which are only available at 5-year intervals – from 1972 to 1992. We therefore have a panel of data for 34 industries and 5 time periods, which are distinguished by fixed effects for estimation.Our empirical results suggest that agricultural materials demand has been affected by various technological and market characteristics of the food processing industry. Although own price effects have had the potential to limit MA demand, growth in the price of agricultural materials has fallen over time, and in the effective price has fallen even lower, so this effect was essentially erased – or even reversed direction – by the end of the 1980s. Substitution effects have also contributed to MA demand. Rising capital costs, especially in effective units, and its implied limitations on production flexibility, have particularly enhanced MA substitution. Scale effects have had a somewhat ambiguous effect, since MA use has increased slightly more proportionately than output increases in effective units, but less than the use of intermediate food products, so MA demand, especially in traditionally measured units, has weakened relative to these substitute inputs. We also, however, find a strong and increasing downward trend in MA demand over time.

The time interval between fire detections is not considered in this analysis, such that fires on consecutive and nonconsecutive days at the same ground location are treated equally. A 1-km radius is also consistent with fire spread rates of 200–5000 m h 1 for grass, grass/shrub, and deforestation fuel types , such that even slow-moving grassland fires would spread beyond the 1-km search limit on sequential days. Fires which burn on consecutive days at the same ground location can occur where fuel loads are very high, as is the case in deforestation fires when woody fuels that are piled together may smolder for several days.Specifically, the variety of days on which fires were detected was determined for each cell of the standard MODIS 250-m grid using a search radius of 1 km to interpret the center locations of all high-confidence fire detections for each year. This gridded product of fire days was then used to select those fire detections contributing to high-frequency fire activity and characterize fire frequency for recent deforestation events.To determine whether active fire detections associated with the conversion of forest to other land uses are unique in terms of fire frequency, we compared active fire detections from recently deforested areas with four additional types of fire management. In the following text, we describe the test datasets used to evaluate patterns in active fire detections for maintenance of cattle pastures, indigenous reserves in Cerrado savanna-woodland land cover, small properties associated with government settlement programs,blueberries in containers growing and sugarcane production regions. We used data on recent deforestation and land use following deforestation to identify and characterize active fire detections associated with forest conversion.Deforestation was mapped using highresolution Landsat Thematic Mapper or ChineseBrazilian Environmental Research Satellite data from approximately August of each year 2001–2005 .

We developed our approach for identifying deforestation fires with data for Mato Grosso state. For individual deforestation events 425 ha in size, we also evaluated differences in patterns of active fire detections for conversion of forest to pasture, forest to mechanized agriculture, and forest conversions not in agricultural production . The post clearing land use for each deforestation event was identified previously using phenological information from time series of MODIS data at 250 m resolution . Finally, we examined fire activity in the year before deforestation detection by PRODES, year of forest clearing, and for as many years post clearing as possible to characterize the nature of fire usage during the conversion process. These comparisons provide the timing, frequency, and degree of repeated burning detected by the MODIS sensors for forest conversion to different land uses. We selected annual deforestation from 2003– 2005 to utilize combined Terra and Aqua fire observations. Because few areas are deforested without the use of fire in Amazonia, deforestation events without any MODIS fire detections provide a measure of the extent of omission due to satellite observation and fire characteristics . We utilized data on historic deforestation and recent land use changes to identify maintenance fires on agricultural lands in Mato Grosso state. The dataset is derived from areas that were deforested before the initial year of PRODES digital data , buffered by 1 km from remaining forest edges to exclude fires from new deforestation. Next, we removed areas that underwent conversion from pasture to cropland during 2001–2004 and previously cleared areas that were identified as secondary forest . The resulting dataset isolates old deforestation not associated with forest edges, secondary forest, or recent conversion to cropland. To identify patterns of fire detections for extensive grassland fires in Cerrado regions, we selected 18 indigenous reserves in Mato Grosso and Tocantins states covering more than 42 000 km2 .

Fire is used during the dry season on some indigenous reserves to facilitate hunting, but extensive land cover change is rare . Small properties are an additional challenge for separating evidence of fire activity in the same location. To test the influence of property size on fire frequency, we considered a subset of the demarcated Instituto Nacional de Colonizac¸a˜o e Reforma Agra´ria land reform settlements in Mato Grosso without large deforestation events in either 2004 or 2005 . The typical lot size in these settlements is 100 ha, of which 20–50 ha may be cleared for agricultural use. Although some sugarcane is grown in the Amazon region, the majority of Brazil’s sugarcane industry is located in the southern and northeastern regions of the country. Sa˜o Paulo State had more than 3 million hectares planted in sugarcane in 2005. We evaluated active fire detections in 31 municipalities in Sa˜o Paulo state with 420 000 ha of sugarcane planted in 2005 to calculate the degree of high-frequency fire associated with sugarcane production.High-frequency fire activity is common in areas of recent deforestation but rare for other fire types in Amazonia . Deforestation in Mato Grosso state had more total fire detections than allother fire types in Table 1 combined and seven times the number of fires detected in the same location on 2 or more days during 1 year. High-frequency fire activity accounted for 27% of high-confidence MODIS detections associated with small producers in Mato Grosso 2004 and 2005, but the total number of detections was small , suggesting that property size is not the main component of the pattern of repeated fire usage associated with deforestation. Fires detected on 2 days at the same location are rare within indigenous reserves and agricultural areas of Mato Grosso state or sugarcane production municipalities in Sa˜o Paulo state; fires on 3 or more days are almost exclusively linked to deforestation. Mato Grosso had both the highest total fire activity and greatest fraction of high-frequency fire activity during 2003–2007 of any state in Brazilian Amazonia .

Combined with fires in neighboring Para´ and Rondoˆnia states, these three states contributed 83% of the fires that burn on 2 or more days and 74% of the total fire activity in the Brazilian portion of the Amazon Basin during this period. Inter annual variability in thetotal number of fires highlights drought conditions in Roraima state during 2003 and widespread drought in 2005 affecting Rondoˆnia, Acre, and Amazonas states. The fraction of total fire activity from burning on 2 or more days also increased during drought years in these states. Fire detections were highest in 2005 for Para´ and Amapa´ states, although these regions were less affected by drought conditions; the fraction of repeated fire activity did not increase in 2005 compared with other years. After a decrease in the fire activity in Brazilian Amazonia during 2006, fires in 2007 returned to a similar level as seen in 2004 and 2005, led by increased fire activity in southeastern Amazonia. Major contributions to this increase in 2007 were from low-frequency fires in Tocantins and Maranha˜o states and additional high-frequency fires in Mato Grosso and Para´. Overall, fires on 2 or more days during the same dry season accounted for 36–47% of the annual fire activity in Brazilian Amazonia during 2003–2007,planting blueberries in containers with greater contributions from repeated fires in years with highest fire activity. At the national scale, fire activity in Brazil and Bolivia accounted for 98% of all fire detections in the Amazon Basin during 2003–2007 . High frequency fires contribute a large fraction of MODIS detections in both countries, with peak repeated fire activity during 2004 in Brazil and 2007 in Bolivia. Small contributions to overall fire activity from other Amazon countries are primarily low-frequency fires, with the notable exceptions of 2004 and 2007 in Colombia, 2003 in Guyana and Suriname, and 2003 and 2007 in Venezuela. Spatial patterns of high-frequency fire activity in 2004 and 2005 highlight active deforestation frontiers in Mato Grosso, Rondoˆnia, and Para´ states in Brazil and in southeastern Bolivia . Isolated locations of high-frequency fire activity can also be seen across other portions of the Amazon Basin, but these areas have low total fire detections. Differences in the total fire activity and high-frequency fire detections between 2004 and 2005 highlight the influence of drought conditions in western Amazonia on fire frequency. Total fire detections in central Mato Grosso decreased slightly between 2004 and 2005, while fire detections in drought-stricken northern Rondoˆnia, southern Amazonas, and eastern Acre states in Brazil show higher total fire activity in 2005 than in 2004. The number of 0.251 cells with 450% of fire activity occurring on 2 or more days is similar during 2004 and 2005 , but the spatial distribution is broader in 2005 than in 2004, as fires associated with deforestation activity in Mato Grosso, Para´, and southern Rondoˆnia spread west into northern Rondoˆnia, Acre, and southern Amazonas states. In addition to deforestation-linked fires, slow-moving forest fires and contagion of other accidental burning events may also have contributed to the higher fraction of repeated fire activity in these regions.Among deforested areas in Mato Grosso, the intensity of fire usage varies according to post clearing land use . Forest conversion for cropland exhibits the most frequent fire usage; more than 50% of the 2004 cropland deforestation events had fire detections on 3 or more days during the 2004 dry season and 14% burned on 10 or more days.

Over 70% of the forest clearings with fires on more than 5 days were subsequently used for cropland. Because of more frequent fire usage in preparation for mechanized agriculture, few areas deforested for cropland in 2004 had no high-confidence fire detections during 2004 . Deforestation for pasture averaged less than half as many fire days as deforestation for cropland, measured as either the maximum or mean days of fire detection per clearing. Even among very large clearings , fire usage was significantly higher for cropland deforestation than forest clearing for pasture . Only 13% of all deforestation events for pasture 4200 ha averaged 3 or more fire days in any year, suggesting that mechanized forest clearing and high-frequency burning are more related to post clearing land use than clearing size. For both pasture and cropland deforestation, polygons in which the conversion occurs within 1 year have a greater number of fire days in the year that the deforestation was detected than conversions occurring over 2 or more years , consistent with the expectation that higher fire frequency leads to higher combustion completeness. For those areas that showed no clear pasture or cropland phenology in the years following deforestation, fire activity was minimal. Nearly 50% of the areas described as NIP showed no high-confidence fires in 2004, and only 22% of these deforestation events exhibited fires on 2 or more days typical of other deforestation events. The timing of fire use during the dry season also differed for cropland and pasture deforestation . Deforestation fire activity may begin during the late dry season in the year before the deforestation is mapped and continue for several years post clearing as the initial forest biomass is gradually depleted to the desired conditions for cropland or pasture use. September was the most common month of fire activity for all types of deforestation in Mato Grosso in 2004. More than 70% of the fires associated with 2004 deforestation for pasture during 2003–2005 occurred during the late dry season . In contrast, fire activity for conversion to cropland was more evenly distributed through the dry season, with 45% of fire detections occurring in May–July. Burning activities initiated in the early dry season for both pasture and cropland deforestation continue to burn in subsequent months. The highest percentage of fires without detections on additional days occurred during the late dry season; approximately 30% of the fires for conversion to pasture during September and October were the first fire detection for those deforestation events compared with 11% of all fires for cropland conversion during this period. High-frequency fire activity may last for several years following initial forest clearing, further increasing the expected combustion completeness of the deforestation process . Forty percent of the areas deforested for cropland during 2003–2005 had 2 or more years during 2002–2006 with 3 1 fire days. The duration of clearing for pasture was more variable. Most areas cleared for pasture had 0–1 years of high-frequency fire usage, although a small portion had frequent fire detections over 2–3 years typical of mechanized forest clearing.

To isolate the effects of bio-solids and TCS amendments on microbial community composition, the data was analyzed using pCCA considering TCS and bio solid amendment as environmental variables, and incubation time as a covariable . This confirmed the results of the CCA indicating that the strongest determinant of microbial community composition was addition of bio-solids to soil. TCS concentration, on the second axis, described only 3.6% of the variation, showing TCS effects were overshadowed by the effects of bio-solids amendment. Bio-solid amendments caused an approximately two-fold increase in PLFA biomarkers for Gram-positive bacteria, actinomycetes and eukaryotes in SB compared to soil samples . Even larger increases were observed in biomarkers for fungi and Gram-negative bacteria, which were up to three times higher in SB than soil. Again, these changes were likely due to increased nutrient availability in the bio-solid amended samples and/or the biomass added along with the bio-solids, consistent with previous studies that found that the fatty acid 18:2 ω6, 9c and monounsaturates were increased by addition of these materials . The effect of TCS on microbial community composition was greater in soil than SB. Spiking with 10 or 50 mg/kg TCS decreased the abundance of Gram positive and Gram negative bacteria as well as fungi, with reductions ranging from 14 to 27% by day 30. Additionally, actinomycetes, which are Gram positive bacteria,plant pots with drainage were reduced in the 50 mg/kg TCS samples after 30 days of incubation . Eukaryotes were negatively affected after 7 and 30 days of incubation at both concentrations of TCS in soil but not SB samples.

Biomass results for all microbial groups were consistent in suggesting that the presence of bio-solids mitigated the potential toxicity of TCS. It is important to note that the spiking levels used here are similar to levels found in the upper half of U.S. bio-solids, but would be unlikely to be achieved in bio-solid amended soils even after continued long term application. Therefore, the effects observed at the 10 or 50 mg/kg spiking levels should be viewed as a conservative upper bound on potential effects expected in the field. In addition, since all of the results in this study are based on an observation period of 30 d, the extent to which the observed effects persist is not known. Future studies should, in particular, investigate longer term changes in community structure in response to addition of bio-solids both with and without specific contaminants. There have been many efforts across the world to mitigate wetland habitat lost over the past century. This movement is echoed in California’s Central Valley where stakeholders have established the goal of creating and protecting over 60,000 ha of new wetland habitat in the state . Many of these wetlands are, or will be, ephemeral, flow through wetlands receiving irrigation return flows during the growing season . Most wetlands in CA are restored with the primary objective of enhancing waterfowl habitat, however, these systems also have the potential to retain and remove nutrient loads that would otherwise be exported directly into major waterways . Therefore, wetland treatment of agricultural return flows is being considered as a beneficial management practice to reduce algal and nutrient loads that contribute to seasonally low dissolved oxygen in the lower San Joaquin River, California . Many studies have demonstrated that natural and constructed wetlands are generally effective at removing nitrogen from municipal and agricultural waste waters . Removal efficiencies as high as 98% have been reported, though other studies report significantly lower N removal rates typically between 35 and 55% .

A study of three wetlands used to treat subsurface tile drainage water in the Midwestern, USA demonstrated NO3 removal rates of 28% . Similarly, high but variable NO3 removal rates have been documented from water seeping through side berms of a constructed wetland in Illinois . Variation in nitrate removal is a result of many factors such as hydraulic residence time, soil properties, vegetation characteristics, variability in input loads, N loading, temperature, dissolved oxygen concentration, climate and nitrogen form in input waters . Using wetlands as a beneficial management practice to reduce non-point source pollution from agricultural drainage waters may introduce a problem as these wetlands could leach contaminants such as nitrate directly into the groundwater. This could compound an existing problem in California where groundwater NO3-N loading rates of 200 Gg per year have been reported in areas of intensive agriculture such as the Salinas Valley and Tulare Lake Basin . Several studies of dairy lagoons summarized in Harter et al. document high seepage rates , and elevated groundwater N concentrations beneath lagoons. Similarly, Huffman found NO3-N concentrations exceeding the EPA drinking water standard beneath two thirds of 34 swine lagoons in North Carolina. More studies of nitrogen fate and transport in wetlands receiving tail water from cropland are needed because the existing literature base for this topic encompasses a wide range of environmental characteristics that govern nitrogen transformations . The primary objectives of this study were to determine the fate of nitrogen in seepage waters of a restored surface-flow through wetland and to determine the importance of hydrologic- as well as soil- and bio-geochemical-factors that regulate nitrate removal. We addressed these objectives by: monitoring nitrogen concentration in nested piezometers throughout the wetland and comparing them to surface water; measuring spatial patterns in selected soil and hydrological characteristics; and, developing wetland hydrologic and nitrogen mass balances to evaluate the fate of nitrate. The results from this study provide information relevant to the optimization, design, and management of restored wetlands for nitrate removal. Moreover, these findings expand upon the limited number of published studies that document nitrate removal by constructed wetlands receiving nitrate runoff from irrigated agriculture .

The wetland received agricultural return flows during the irrigation season from April to September, with no rainfall occurring during this time. Surface water inflow and outflow volumes were measured at 30-min intervals using v-notch weirs and barometric pressure compensated water level loggers . A digital elevation model was created using a Trimble RTK GPS with 3 cm accuracy. The DEM was used to relate water depth measured at two locations with water depth throughout the wetland,plastic plant pots as well as to determine changes in the wetted surface area throughout the irrigation season. Vertical hydraulic gradients were calculated at 12 piezometric monitoring locations in the southern section of the wetland, using biweekly water height measurements at 10- and 100-cm depths . Surface water residence time was calculated using a plug-flow model . Temperature was measured at 15-min intervals near the output. Wetland evapotranspiration was estimated using meteorological data obtained from the California Irrigation Management Information System Patterson station, approximately 15 km from the study site. ET rates for vegetated upland areas were presumed to approximate the CIMIS values calculated for grass cover. Evaporation for the sparsely vegetated wetland area was assumed to be 1.28 times that of the grass ET value . ET volumes were calculated at 30-min intervals to account for fluctuations in the wetted surface area. A season-long seepage volume was calculated by subtracting total outflow volume from total inflow volume, accounting for water loss due to ET. An independent measurement of the seepage rate for the northern and southern sections of the wetland was determined on 6/4/2007 through 6/9/2007 by preventing all inflow and outflow, and measuring the rate of water level drop over a 120- h period. Seepage volumes were then calculated for each 30-min interval by multiplying the seepage rate by the wetland wetted surface area . Assuming similar seepage rates across the different hydrologic zones, we calculated the percentage of the water surface area covering each hydrologic zone at 30-min increments based on the high-resolution DEM and water height at the output location. The seepage volume was summed for each 30- min increment to obtain a total seepage volume for each hydrologic zone.Pore water was collected from piezometers at 12 locations on a biweekly basis at depths of 10, 50 and 100 cm below the soil surface. Screened sections of the piezometers were surrounded in a layer of pure silica sand and sealed above and below with bentonite clay to prevent water intrusion from adjacent horizons . Prior to sampling, piezometers were purged and allowed to recharge for 1–2 h. Water samples were maintained at 3 C between the time of collection and analysis . Aliquots of samples were filtered through a prerinsed 0.4mm polycarbonate membrane filter for quantification of NO3-N , NH4-N , and DOC . Determination of NO3 was made using the vanadium chloride method and NH4 using the Berthelot reaction with a salicylate analog of indophenol blue . DOC was measured using a Dohrmann UV enhanced-persulfate TOC analyzer .

A non-filtered sample was used to determine total N following oxidation with 1% persulfate using the method described above for NO3-N. Surface water samples were collected adjacent to the piezometers and at input and output locations on a weekly basis and were analyzed as described above. Depth splines were used to model nitrate distribution over the 100-cm depth of the piezometer monitoring nests. The segmentation procedure involved fitting an equal-area or mass-preserving quadratic splineacross the discreteset ofporewaterNO3-N sampling depths , producing a continuous depth function segmentedat1-cmintervals . Mean values at each1-cmdepth increment were calculated across all sampling dates and sampling locations within each hydrologic zone. The segmenting algorithm was implemented using the ‘GSIF’ and ‘aqp’ packages for R .Inflow and outflow seasonal loads for total nitrogen, nitrate, and ammonium were calculated using the period-weighted approach from weekly constituent concentration and weekly water flux . Nitrate seepage loads for each hydrologic zone were also calculated with the period-weighted approach using average biweekly nitrate concentration at the 100- cm depth and weekly seepage flux.Linear mixed effects models were used to analyze data from water analysis and DNP incubations using S-Plus . As samples were taken at the same location several times throughout the season, location was treated as a random effect in the model to account for auto correlation between measurements at the same site. The NH4-N, NO3-N, DNP and DOC values were log transformed prior to statistical analysis to better approximate a normal distribution. For each analysis, the initial model accounted for main effects, as well as all possible two-way interactions between main effects. Interactions that were not significant were removed from subsequent models to gain sensitivity. Mean separation was determined using a conditional t-test. Raw are reported in Tables 4 and 5 to reflect measured field conditions. The water sampled from piezometers was termed seepage water. Nitrate concentration was markedly lower in seepage water than in surface waters . Concentrations of NO3-N were significantly lower at the 50-cm depth than the 10-cm depth, but there was not a significant difference in NO3-N concentrations between the 50- and 100-cm depths among the three hydrologic zones . Modeled nitrate removal rates from Fig. 4 in the top 10-cm soil depth relative to the water column were 932, 631 and 143 mg NO3- N m 2 d 1 in the flowpath, finger and upland zones, respectively. In the wettest hydrologic zones there was a significant increase in NH4-N concentrations from the surface water to the 10-cm depth . NH4-N concentrations decreased at the 50- and 100-cm depths and were not significantly different from those in the surface waters . DOC concentration in seepage water ranged from 3.2 to 6.0 mg L 1 . There were no significant differences in DOC between the surface water, 10-, and 50-cm depths; however, DOC concentration decreased significantly at the 100-cm depth of the upland sites. Among the hydrologic zones, DOC in seepage water was significantly higher in the uplands .Soil texture was generally similar among hydrologic zones and no abrupt changes in texture were observed with depth . Sedimentation was highest in the flowpath zone totaling over 35 kg m 2 yr 1 compared to sedimentation rates <5 kg m2 yr 1 in the fingers and uplands. Saturated hydraulic conductivities estimated for these textural classes were similar to measured seepage rates . Average soil organic carbon concentration was relatively low in all hydrologic zones . Organic carbon decreased with depth in all hydrologic zones.

They allowed for a 1% chance of mutation of each experiment and component to allow for global search. They also discovered that the response space was multi-modal and had interactions between components, which confirmed the need for global optimization of fermentation and bio-processing problems.Microbes perform key ecosystem functions , necessitating a better understanding of how microbes and microbial diversity respond to environmental stressors . Streams are examples of threatened ecosystems, where watershed modification decreases stream integrity and water quality , altering macro-invertebrate, fish, and microbial diversity . The use of macro-invertebrate and fish indices to assess stream conditions is fundamental to stream ecology and depends on known relationships between stream integrity and community structure . The Benthic Index of Biotic Integrity is one such index, using the abundance and diversity of stream benthic macro-invertebrates to accurately distinguish degraded streams classified based on stream chemical and physical criteria . Biotic indices are calibrated to specific regions, as the distribution of stream macro-invertebrates is controlled by a combination of dispersal limitation and local environmental conditions . Despite their abundance and ecological importance, natural microbial communities, unlike macro-invertebrates, are not used in stream monitoring programs to assess stream conditions. As with macro-invertebrates, dispersion and environmental selection control the spatial distribution of microbes along stream continuums . Dispersion,growing raspberries in pots or the advection of microbes from the surrounding landscape, impacts head water stream community composition, and with increasing stream order, environmental sorting becomes more important as stream residence times increase .

Several studies have demonstrated the influence of the surrounding landscape on stream microbes, showing that watershed urbanization leads to shifts in bacterial communities . While alpha diversity generally remains constant , the abundances of taxa associated with anthropogenic activity and high-nutrient conditions increase in urbanized streams . Similar to larger organisms, microbes respond to environmental disturbance and are strongly influenced by watershed land use ; therefore, their distribution may be used to further characterize stream conditions. Microbes mediate important stream ecosystem functions, controlling the movement of carbon and nitrogen through freshwater ecosystems . Previous studies demonstrate the effects of urbanization on stream nutrient transformations, such as nitrogen uptake , nitrogen retention , and carbon processing . Community respiration determines the fate of terrestrial carbon in head water streams, where carbon is either lost as carbon dioxide during respiration or transported farther downstream . Community respiration is often used to assess ecosystem function , as rates are influenced by watershed land use , correlated with stream chemistry , and sensitive to pollutants . The effects of urbanization on stream dissolved organic matter quality and respiration have previously been demonstrated, and stream microbial community structure can potentially be used to monitor these ecosystem functions. In addition to respiration, dissolved organic matter fuels stream denitrification and the microbial reduction of nitrate to nitrous oxide and dinitrogen gases . Denitrification removes nitrogen from streams and is credited as the major source of the greenhouse gas N2O . Watershed land use and anthropogenic nitrogen loading alter rates of stream denitrification , increasing the amount of nitrogen transported downstream and emissions of N2O to the atmosphere . Urbanization has been linked to changes in denitrifier community composition , and a previous study linked changes in denitrifier composition to changes in denitrification potential, and therefore nitrogen loss, in urban streams .

However, it is less clear how changes in microbial community composition in response to land use modification alter N2O production. The goal of this study was to identify stream microbes that respond to watershed urbanization and agricultural development. These anthropogenic factors alter microbial diversity and community structure, which can be used to assess stream conditions and ecosystem functioning. We measured microbial diversity using 16S rRNA gene amplicon sequencing across 82 head water streams within the Chesapeake Bay watershed in the state of Maryland in the spring and summer for 2 years. Measurements were collected in conjunction with stream physicochemical parameters and a macro-invertebrate indicator of stream health. Additionally, at a subset of streams, water column and sediment community respiration were measured using oxygen consumption methods, and N2O concentrations were measured using gas chromatography. We determined how stream bacteria and archaea are distributed across gradients of watershed land use and stream conditions, assessed how changes in microbial community composition relate to benthic macro-invertebrate diversity and traditional indices of stream conditions, and determined how these changes influence stream function by relating microbial community composition to rates of microbial respiration and concentrations of N2O.The aims of this study were to understand how stream bacteria and archaea are distributed across gradients of watershed land use and water quality, to assess how changes in microbial community composition relate to benthic macro-invertebrate diversity, and to discern how changes in stream conditions alter stream ecosystem processes, as reflected in community respiration and N2O concentrations. Bacterial and archaeal diversity significantly differed across the geographic regions of Maryland , demonstrating the influence of the surrounding landscape on headwater stream microbial communities.

Regional alluvium composition likely influenced stream alpha diversity, causing lower alpha diversity in Coastal Plain streams . Sediments on the Coastal Plain of the eastern United States are composed of gravel, sand, silt, and clay , making streams more embedded . Embeddedness was the environmental factor that most strongly negatively correlated with Shannon diversity , and homogeneous fine sediments have been shown to have lower diversity than that of sites with riffles, shallow turbulent sections . Similarly, community structure varied across the geographic regions and strongly correlated with DOC concentration, pH, and embeddedness , all of which significantly differentiate Coastal Plain streams from the other regions . This finding is in agreement with those from previous studies, demonstrating the strong influence of DOC concentration and pH on freshwater communities . Despite the strong influence of stream chemistry on microbial communities , in this study, geodesic distance explained more of the variation in community composition than environmental distance. Partial Mantel test results indicated that community structure was correlated with geographic distance rather than the measured environmental variables. Geographic distance is likely a strong controlling factor in structuring head water stream communities because there are regional differences in landscape, and stream microbes are locally seeded from the surrounding soil . Alpha diversity was greatest in spring, when water flow through the landscape is greatest and, therefore, when advection of microbes from the surrounding landscape is greatest to head water streams. While seasonal changes in microbial diversity during fall and winter are unknown, the higher diversity in spring than in summer was likely due to higher terrestrial inputs in spring, further demonstrating the influence of landscape on stream microbial communities. Distance-decay relationships were also observed between water column and sediment community similarity and geodesic distance , further highlighting the finding that head water stream microbes display geographic distribution patterns. Alternatively,plant pot with drainage the distance-decay relationship could be a result of spatial differences in unmeasured environmental variables. Microbial distance-decay relationships have been observed previously in streams . Z-values represent the rate at which species similarity decreases with increasing distance; in this study, Z-values are similar to microbial values from soil, salt marshes, and lakes but lower than regional differences observed in salt marshes , suggesting different dispersal limitations across regional scales. In contrast to highly urban and agricultural streams, community dissimilarity in highly forested streams did not increase with distance. Neither geographic distance nor environmental distance correlated with community structure, implying that highly forested streams have a similar terrestrial microbial source. Microbial diversity differed in streams in watersheds with high urban, agricultural, and forested land use. In contrast to previous studies, degraded streams had lower alpha diversity than that of forested streams , likely due to elevated pollution and habitat loss. Several abundant and pervasive taxa found in urban and agricultural streams are often associated with high-nutrient and low-oxygen environments. Members of the order Burkholderialeswere abundant in urban streams and correlated strongly with several anthropogenic nutrients .

Comamonadaceae are often associated with high nutrient conditions and are ubiquitous in many environments, including aquatic, soil, activated sludge, and wastewater . Comamonadaceae have previously been associated with urban streams and have been found to have the highest number of urban-tolerant taxa . Sulfurospirillum spp., in the order Campylobacterales, were abundant in highly agricultural streams and are often associated with microaerophilic polluted habits, commonly growing on arsenate or selenate using NO3 and sulfur compounds as electron acceptors . In contrast, an unclassified and potentially phototrophic member of Acidobacteria and Hyphomicrobiaceaewere more abundant in forested streams. These taxa are often associated with low nutrient conditions and were previously identified as indicators of forested streams and shown to decrease in abundance with increasing watershed urbanization . Only weak associations were detected between sediment and water microbial community composition and B-IBI scores. This is in contrast to findings of Simoninet al., who found that stream microbial community structure correlates with a macro-invertebrate biotic index in North Carolina . Simonin et al. identified concurrent changes in microbial taxa and environmental conditions associated with the biotic index, finding a higher number of negatively responding taxa than positive responding taxa . Here, only one taxon was more abundant and pervasive in streams in good condition, while several taxa were found to be abundant and pervasive in streams in very poor condition . Hydrogenophaga spp. and an unclassified member of Desulfobacterales, commonly associated with anaerobic, reducing, and contaminated environments , were both more abundant in streams in very poor condition . Forested water communities were more even than agricultural and urban communities, suggesting that certain taxa increase in abundance disproportionality in degraded streams, which is likely why the indicator analysis identified more taxa in streams in very poor condition. The findings here suggest that land use cover and stream chemistry are better predictors of head water stream microbial community composition than are macro-invertebrate indices of stream conditions. In agreement with the idea that structure determines function, in this study, water community respiration correlated with microbial community composition. Similarly, previous studies report that changes in community metabolism, specifically, the degradation of organic matter, are related to shifts in community composition and diversity . In contrast, other studies report that respiration depends on substrate availability rather than community composition due to functional redundancy , finding no connection between stream bacterial diversity and the activity of enzymes associated with carbon cycling . The weaker correlation between community respiration and community composition in sediments compared to water samples could be due to a high level of functional redundancy within sediment communities; if dominated by generalists, shifts in community composition would likely not significantly affect rates of respiration . Degraded streams had lower rates of community respiration than forested streams, as evidenced by the positive correlation between water respiration and forest cover and the negative correlation between respiration and urban cover . Rates of community respiration also negatively correlated with several physicochemical variables , including conductivity, Cl , Ca, Mg, pH, SO4 2 , and ANC. All signatures of anthropogenic influence, ANC, Cl and Zn concentrations, and pH, were found to previously correlate with benthic stream respiration across the Highlands, Piedmont, and Coastal Plain regions of the eastern United States , and Zn is a common urban pollutant . These results suggest that environmental conditions associated with land use drive differences in community respiration, altering carbon processing in head water streams. In addition to altering carbon transformations, watershed modification affected stream nitrogen processing. Agricultural streams had higher N2O concentrations than those of forested streams , with higher N2O concentrations being associated with elevated TN, NO2 , NO3 , and NH4  concentrations . The N2O concentrations measured in this study, at 0.22 to 4.41 g N2O liter 1 , were comparable to the values reported for agricultural streams in Illinois but lower than values published for agricultural drainage waters in Scotland, UK and higher than values reported for other forested and agricultural streams . N2O production is known to vary by land use, with higher production from denitrification in streams in agricultural and urban basins , and changes in community composition have been shown to influence denitrification rates in agricultural and urban streams . However, Audet et al. found no difference in N2O concentrations measured in forested and agricultural streams in Sweden .