The groundwater data and analytical issues highlighted in the preceding chapters place major limitations on the analysis of relationships between water and food security. To improve estimates of groundwater availability for irrigation beyond the calculations of Postel (1999), Seckler et al. (1999) and Shah et al. (2000) would require a major initiative to collect primary groundwater data and the associated information essential to interpret it correctly from widely dispersed locations. In addition to the relatively straightforward process of locating data sources and documents, this would require substantial effort to obtain approval from governments to access primary data. The research process undertaken to produce this paper investigated a wide array of material relating to the links between groundwater access, yields and crop production. However, little of this second-level material has been incorporated into this paper because of the core uncertainties relating to the status of groundwater resources.
Compilation of available case-study data on groundwater from country sources would improve substantially the understanding of groundwater-level trends in key agricultural areas. This would enable evaluation of probable changes in the economics of groundwater extraction and changes in access to groundwater for poor and marginal groups or those dependent on specific technologies such as dug wells. This type of information would be useful for evaluating probable distributional and economic impacts as development proceeds and water levels decline. However, it would not resolve the inherent problems of data quality and short periods of record that are often encountered. Similarly, it would not resolve questions relating to other key components of the water-balance equation, e.g. extraction rates, leakage between aquifers, and evapotranspiration by native vegetation and crops. As a result, while the data should enable improvements in overabstraction estimates, they would not resolve many of the major modelling issues and would probably not enable accurate estimates of groundwater overabstraction at a global or regional level. These types of uncertainties would be magnified if taken a step further and used as inputs for analysis of global food production. The parallels to debates over climate change are worth noting here. According to Rosenzweig and Hillel (1995): "The uncertainty inherent in predictions is a very important feature of climate change impact studies...Other uncertainties derive from the fast pace and unpredictable directions of future social, economic, political and technical changes. The world of the coming century will be different in many ways; unforeseeable developments in other sectors may change the way in which agriculture responds to climate change.... An even more challenging task is to estimate the probability of coincidental events that might happen in conjunction with global warming, spanning the range between low probability catastrophic events (called "surprises") and higher probability gradual changes in climate and associated environmental effects. A seemingly small change in one variable - for example, rainfall - may trigger a major unsuspected change in another; for example droughts or floods might possibly disrupt the transport of grain on rivers. Moreover, one "surprise" may then lead to another in a cascade, since biophysical and social systems are interconnected."
Given the status of groundwater information at national level and the inherent unknowns in the models for predicting the impacts of development, the above comments also apply to estimates of the impacts of groundwater-level change on global food production and security. Improvements in access to primary data on groundwater would improve models of food production and food security. However, the predictive value of such models will remain limited by data quality issues, incomplete understanding of systems and ongoing processes of climate, demographic, economic and agricultural change. Thus, the usefulness of this type of analysis is uncertain. In the climate change case, Rosenzweig and Hillel (1995) advocate courses of action that respond to this uncertainty and increase resilience: "Identifying potential surprises and communicating them to the public and policy makers may help to build the resilience that is needed to anticipate and mitigate harmful effects in a timely fashion." Similar courses of action appear appropriate in the debate on groundwater overabstraction and food security.
The above discussion suggests that, rather than attempting to analyse the macro implications of water availability or, more specifically, groundwater problems for food production or access on an aggregate level, it would be more productive to focus on a broad array of early-warning indicators that can trigger responses to food security concerns as they emerge at specific points of time in specific local contexts. Food security problems emerge because of a confluence of hydrological, climate, economic and social factors. Therefore, analysis could focus on developing indices of food security vulnerability that combine an array of long- and short-term physical, economic and social indicators. Groundwater conditions and availability would be among the more important water availability indicators that would need to go into this analysis. However, they would need to be combined with other indicators that reflect, for example, drought probabilities, general economic conditions (availability of alternative sources of work), global food availability and transport capacity. Such indices could be used to trigger proactive responses to emerging food security problems before they reach a critical level and thus reduce the need for post facto relief programmes. Thus, the role of analysis would move away from efforts to predict quantitatively the impact of groundwater depletion on aggregate food production and would focus instead on the development of more localized early-warning indicators.
Placing greater emphasis on indicators of food security vulnerability does not reduce the importance of groundwater management. While data limitations and other factors restrict the ability to quantify with any degree of confidence whether or not groundwater overabstraction and falling water levels have major implications for aggregate food production and access, it is known that they could. It is also known that water-level changes have major implications for poverty, environmental values, health and regional economies irrespective of whether or not global food security is at risk. Thus, the critical importance of responding to groundwater problems in locations where they are evident should be clear. While the global warnings about groundwater overdraft may be welcome, they cannot in themselves point to any meaningful intervention or solution. These types of natural resource problems cannot be treated globally. The issue is one of finding an appropriate physical and administrative level at which the common property can be addressed successfully. The aggregate impact of such interventions will yield an improvement, but it will be incremental.
A Malthusian perspective on groundwater overabstraction leading towards a global food security crisis suggests courses of action designed to respond in an equally dramatic fashion to perceived threats to the resource base. In the groundwater case, this has led to a preoccupation with regulation, extraction control, water allocation and aquifer protection. While such activities will remain important, a lens that recognizes both variability and uncertainty requires a wider focus.
The global water management literature tends to emphasize the importance of complete institutional restructuring in the water sector and the development of comprehensive and integrated water management strategies. There are widespread calls for the development of integrated water rights systems, markets or other frameworks for allocating limited supplies and for the establishment of basin or aquifer management authorities with substantial regulatory powers. The focus is on management of the water resource base per se and not on the range of related services derived from groundwater use. This paper argues that a more clumsy approach is essential to complement but not replace the standard management paradigm.
The standard integrated approaches imply that systems are understood and can be managed in a comprehensive manner. They presume an ability to identify and quantify the nature of interactions and to define the boundaries of systems clearly. They also presume that social institutions (rights systems, regulatory organizations, etc.) are present and contain sufficient capacity that management can be implemented in a planned and integrated manner. In the case of groundwater overdraft, assuming new supplies are not available, managing overdraft will require a reduction in use. In this situation, effective management requires:
Sufficient technical capacity to quantify required use reductions, identify aquifer boundaries, and monitor the impact of management on aquifer conditions.
Institutions for the equitable allocation of use reductions, which in turn generally depend on a rights or licensing system that is legally recognized, acceptable to users and enforceable.
Effective mechanisms for coordinating the variety of agencies and users whose actions affect groundwater extraction and for planning the overall approach to management. In many situations, there are multiple departments and multiple users each with their own interests, perspectives and agendas.
Effective mechanisms for educating users and for building social and political support for difficult management actions.
In many developed and developing countries, organizations and institutional frameworks capable of the above tasks are either weak or absent. Developing such frameworks can be a major task and must occur before any real management benefits will begin to flow. India has more than 20 million wells and a weak legislative framework for management. The task of well registration would require years of effort coupled with the introduction of a new legal framework. Therefore, effective implementation of the standard management paradigm is a long-term prospect at best.
Clumsy approaches that respond to current constraints while not attempting to manage entire systems (and are therefore not fully integrated) are as important as more standard integrated management paradigms. Such adaptive approaches often work through lateral interventions (such as encouraging non-agricultural activities rather than attempting to control agricultural extraction) and can be designed to respond to variations in conditions and uncertainty. Because adaptive approaches do not require full understanding of resource dynamics and build off coping strategies that populations are already engaged in, such approaches may be able to produce results more rapidly than can integrated management initiatives. In addition, they often do not require the introduction of new institutions (such as water rights) and may be able to minimize politically difficult decisions (such as extraction controls) in the short term. This is important because the institutional capacity and data essential for active integrated management of the resource base is absent in many countries and could take decades to develop. Adaptive strategies can provide a critical breathing space while the capacity for more direct integrated management is being built. Furthermore, adaptive strategies may prove more effective than integrated management in many cases because of their ability to respond rapidly to the process of social and environmental change that is ongoing in most parts of the world.
Clumsy approaches assume that tasks such as coordination are difficult and expensive. As a result, instead of attempting to develop a comprehensive integrated approach, the technique focuses on windows of opportunity that move groundwater conditions in the general direction required to address the problem. For example, when groundwater levels decline rapidly, some users may already be migrating in search of alternative sources of livelihood, others may be trying to focus on less water-intensive crops, and yet others may be trying to harvest more water. In this situation, interventions that help people to do what they are already trying to do on their own are likely to reduce pressure on the groundwater resource base. While better information and coordination could be helpful, the general direction in which change needs to go is clear and many practical things can be done that fit with that direction. Furthermore, many of the interventions do not require any coordination at high levels in order to be of some benefit at local levels (e.g. construction of local water harvesting structures or encouragement of non-agricultural activities). However, this may be the best case. Farmers pumping groundwater may not be water sensitive if well adjusted to a crop with high water consumption. The example of sugar cane is instructive as it is one of the few crops for which companies will offer credit to smallholders from the day of planting. These types of inelastic behaviour inhabit the mosaic of water use in rural settings and need to be taken along with the more positive opportunities for enhancing adaptive behaviour with regard to groundwater.
As Moench (1999) argues, four elements appear central to the development of adaptive responses to groundwater problems under conditions of variability and uncertainty:
identification of an appropriate physical and administrative level at which the common property can be addressed successfully (i.e. the response needs to be context reflective);
social auditors (the engine of change).
Each of the four elements functions more effectively in a situation where whatever data exist are accessible and can be used by different groups within civil society. Thus, although this does not remove uncertainty, data access is also an important condition.
There is a core distinction between using a systemic perspective in developing water management responses and attempting to develop comprehensive integrated responses. A systemic perspective recognizes both the importance of interactions between systems and the limitations of knowledge regarding those interactions. It also emphasizes scale issues. Aquifers or watersheds are not discrete units but operate rather as systems within systems. The scale at which management needs to occur and the type of management both depend on the scale of system processes and the nature of system interactions. Because these interactions are complex, a systemic perspective leads to a focus on those factors that appear at any given time to be the drivers of underlying problems. Unlike the comprehensive integrated terminology, a systemic perspective does not convey the impression that all driving factors are known and incorporated in the approach proposed. The distinction is subtle but important. Describing an approach as being comprehensive and integrated gives readers (including those involved in critical policy and decision-making roles) the impression that all relevant factors have been reviewed comprehensively and integrated effectively into the plan of action. In most situations, this is not possible. Gaps in data, basic scientific knowledge, social dynamics and many other factors make comprehensive integration impossible. In contrast, a systemic perspective draws attention to the variety of factors influencing water management needs and options in a given situation but without claiming an unrealistic ability to be either integrated or comprehensive. As a result, the perspective should focus attention on the factors that analysts view as most important while leaving the door open to changing management approaches as knowledge improves and conditions evolve.
As social dynamics change, so do the factors driving resource conditions. A systemic perspective is intended to encourage recognition of dynamic change processes and to encourage adaptation to them. Because a systemic perspective does not generate the impression that all factors are being addressed, it encourages rapid responses to changes and new information. A systemic perspective is also intended to situate responses within the wider context while de-emphasizing the importance of comprehensive integrated management planning. In contrast, integrated approaches are based on the assumption that all the important interactions can be identified and incorporated into the management approach. Because it is generally impossible to achieve this in practice, approaches based on comprehensive integrated planning generate a false sense of accuracy and security regarding the directions management should take and the results it should have. In addition, such management planning often absorbs considerable resources and produces products that are obsolete before they are implemented.
Constraint analysis is the servant of a systemic perspective. Instead of attempting to manage a system as a whole, the approach is to focus on the constraints that affect key values. In the groundwater case, falling or rising water levels in regional aquifers are often the primary constraint affecting agriculture, the environment, regional economies and other values. Constraints also take the form of surprises (low-probability catastrophic events) such as the occurrence of regional droughts in areas affected by groundwater overabstraction. At a local level, these surprises may have more important implications for food security than the more gradual, long-term process of groundwater overabstraction or water-level decline. Compared with integrated analysis, focusing on constraints and surprises may provide more insights into the problems people face and potential courses of action to address them.
Adaptive or context-reflective responses follow from the combination of a systemic perspective and constraint analysis. Because the lens through which groundwater problems are viewed emphasizes variability, uncertainty and the lack of key systems information necessary for comprehensive management, adaptation needs to be a core component of any strategy. The argument behind this is not simply that the data needed to understand hydrological systems are lacking. It also reflects the role data play in developing social consensus around management strategies. Where overabstraction is present, management of the resource base would generally require large reductions in use and the allocation of available supplies through a rights system or other similar mechanism. For example, in the case of aquifers in northern Gujarat, India, 1976 estimates suggested that extraction would need to be reduced by 25 percent to reach sustainable levels (United Nations Development Programme, 1976). Percentage reductions would now need to be considerably higher. In most cases, building sufficient social consensus for this type of reduction to work would depend on a long-term process through which understanding of aquifer dynamics is generated and disseminated to the key actors, i.e. the well owners whose behaviour determines overabstraction conditions. In locations where this has occurred (e.g. the western United States of America), it has generally involved many decades of data collection and social debate among a limited set of highly educated players before action could be taken. Even so, the results in such cases have been mixed. Groundwater user associations in Mexico have reportedly agreed to substantial reductions in extraction much more rapidly, but the results are yet to appear. Expecting this approach to be effective on a short time scale under conditions in many developing countries appears unrealistic. While not suggesting that efforts to develop management capacity should be reduced, it is equally important to respond and adapt to constraints. In other words, building the capacity of society to adapt to water constraints is as important as the capacity to modify or manage water systems themselves.
The philosophy underlying an adaptive or context-reflective approach to groundwater problems emphasizes the development of responses based on the specific constraints emerging in a local situation. Rather than a uniform set of best management practices, it recognizes that management needs and capabilities vary considerably. It emphasizes adaptation in two ways: (i) adaptation of water management to the needs and opportunities present in local contexts; and (ii) adaptation of society to the constraints emerging from water-related limitations. The broad focus of the lens is evident in the latter. In most cases, governments attempt to respond to water problems through water-related interventions. Social dynamics (migration, changing regional economic systems, etc.) are not generally part of the formal management package. However, they are often the main response individuals are making as they face water-related constraints. In addition, while they affect water use patterns and management options, they are often driven by factors other than water. Adaptive approaches would attempt to identify these types of dynamics and build the types of governmental and other interventions that would support adaptation rather than controlling groundwater use. In this sense, adaptive responses constitute a form of strategic planning that is by nature dynamic, iterative and responsive to uncertainty. It suggests focusing management in the short term on priority aquifers where capacity and incentives are present at the grassroots level, possibly through networks of cooperative institutions, rather than relying on national institutions which often require long-term effort to develop. This does not diminish the importance of longer-term aquifer research and planning and the development of national institutions to undertake these tasks. These are still essential to framing management responses (Burke, 1996) and avoiding anarchic and potentially destructive development over the longer-term.
On one level, the proposed approach reflects a pragmatic response to power relations and dynamics within society as it is often not politically or socially viable to control extraction and, as a result, adaptation is essential. However, on another level, the approach recognizes that users are often adapting for reasons that reflect a larger set of aspirations and goals in addition to water constraints. The small farmers who decide to pump as much as possible in order to educate their children are probably behaving rationally. Forcing them to reduce extraction could lock them in a vicious cycle of marginal agriculture and poverty. In contrast, helping them to educate their children and move out of agriculture could create a more sustainable livelihood for their families and ultimately lead to reductions in groundwater extraction. Thus, adaptation may open new perspectives that support sustainable livelihoods and poverty alleviation rather than highlighting perceived conflicts between resource sustainability and current use patterns.
It is important to recognize that adaptive responses do not eliminate the importance of attempts to develop groundwater management systems. However, they do recognize that the development of management systems is generally a long-term process and often will not provide an effective avenue for addressing the impacts of groundwater overabstraction within the near future. Furthermore, adaptive strategies that move communities away from groundwater dependence could provide the breathing space and social conditions necessary for institutions capable of managing groundwater to evolve. For example, groundwater management in the United States of America is often effectively in the hands of a few tens to hundreds of actors (Blomquist, 1992). This enables negotiations to occur between small groups of stakeholders representing interest groups rather than the thousands of individual well owners often found in locations such as India. Adaptive strategies that move many individuals away from groundwater dependency could help to create the conditions under which direct use of groundwater becomes consolidated in a way that encourages long-term planning and management.
The term social auditors represents a final component that is essential for the evolution of social capacity to respond to water constraints. Social auditors are: "the watch dog social activists as well as various organs of the state that are responsible for assuring appropriate justice. They are not users or managers, and their concerns often stem from different callings - those of equity, sustainability and fair play. Linear policy models that account for the users at the bottom and the managers at the top are often at a loss when these actors enter the fray - often in the event that contradictions emerge between the avowed objectives of management and its practice. Except for extreme cases of bureaucratic rigidity, social auditors from the activist mould and from within the government do often work together to assure proper functioning by the concerned water bureaucracy. Even as the managers of a department may advocate hierarchical administrative approaches to water management, significant sections of the state machinery, including the judiciary and units of local governance, often assert themselves in upholding points of equity, democratic process and social justice." (Moench et al., 1999).
Social auditors serve as catalysts that highlight emerging water problems and their social impact. They also focus social attention on potential solutions. In many cases, they represent a critical social element balancing the power of government bureaucracies or giving voice to socially marginal sections of society. Their ability to play this role is enhanced by access to information and technical-analytical capabilities.
The results of this research point to the need to rethink groundwater management for food security. The paper contains numerous contradictions. While acknowledging the fundamental importance of emerging groundwater problems, the research led in an unanticipated direction. Instead of facts and documentation accumulating towards an increasingly clear and well-founded perspective regarding the impact of groundwater overdraft on food security, variability and scientific unknowns have tended to accumulate. However, groundwater overdraft may be justified in economic terms where the immediate drawdown externalities are negligible and the long-term benefits derived from its use allow sustainable substitutes to groundwater as a factor of production to be developed (Schiffler, 1998). It is not always possible to state categorically that groundwater overdraft is inherently bad.
The demand for groundwater irrigation is diverse. It ranges from smallholder irrigation in remote locations to meet local needs (typical of Sahelian irrigation from basement complex aquifers) to large-scale commercial irrigation to feed national markets (wheat irrigation based on highly productive dolomite aquifers in Zambia). Each type of demand and style of irrigation has its own, usually highly individual, management strategies that are conditioned by specific groundwater dynamics. Some of these patterns of demand will change as populations migrate, subsidy structures change and alternative opportunities arise. These changes will need to be sensed before implementing new policy recommendations or advocacy.
Overabstraction, quality and other groundwater problems are severe in many parts of the world. However, they need to be set against the significance of reliable water services in maintaining global food security. At present, irrigated agriculture in developing countries accounts for only 20 percent of the arable lands but contributes nearly 60 percent of cereal production and 40 percent of total crop production (FAO, 2002a). However, groundwater generally produces substantially higher yields than areas irrigated from surface sources and, in locations such as Spain, can generate economic returns up to five times those found in surface irrigated areas (Hernandez-Mora et al., 1999). Because groundwater plays a major role in achieving high yields, particularly where demand for high-value crops is effective, emerging threats to the resource base do have important food security implications at local and possibly global levels.
However, the accessible groundwater data are insufficient to quantify the areas where food security implications may be severe. Furthermore, emerging groundwater problems are important regardless of their implications for global food security. Effective approaches to addressing these problems are essential. However, the same challenges that frustrate attempts to estimate the impact of groundwater overabstraction on food security also complicate the applicability of standard management models. The contradictions that are evident in this paper are also inherent in the groundwater story itself:
Problems are emerging but there is uncertainty regarding their nature, extent and implications.
Management needs are evident but the viability of management is unclear in many situations.
Groundwater problems create significant inequities in access to water for poor and marginal users but most standard management solutions would in practice exacerbate such inequities.
Solutions that ensure access to groundwater for all exist but they are likely to lead to the depletion of the resource base and ultimately increase inequity.
Solutions to overabstraction that build off inequity by helping those displaced by overabstraction to develop alternative livelihoods may ultimately be the most equitable if they ease pressure on the resource base and enable people to move out of poverty.
On closer examination, the above contradictions may prove more apparent than real. However, they point to a major need to rethink groundwater use for irrigated agriculture. Better understanding is needed regarding the implications of emerging problems and the best points of leverage for addressing them. Better understanding is also needed with respect to the coping strategies that people follow and how attempts to address groundwater problems can build off existing trends in society rather than attempt to work against them.