An analysis of current practices might lead to the conclusion that there is no effective system of groundwater management. It is a rare exception when wells are closed down and capped off to prevent abstraction, or limits set on pumping durations or volumes. It is not such a rare exception to observe local, consensual enforcement of pumping limits, e.g. for irrigated agriculture in Eritrea and Yemen when pumped groundwater is rationed during the dry season. These arrangements are 'customary', but have only been occasioned by the advent of inexpensive motorized pumps.
The variable patterns of groundwater use and the varied services that aquifer systems provide do not form a clear aggregate picture or status of groundwater, nor do they present an opportunity for systematic management response. Despite the highly technical work that is carried out and presented in the hydrogeological literature, the status of knowledge of the aquifer systems is often limited at the level at which a management response is required. Highly detailed studies in contaminant transport are carried out in high-value settings (usually because regulatory systems are enforced). However, accurate and reliable monitoring and regulation in the crucial aquifers of northern India (FAO, in press) and Baluchistan, Pakistan, are not available. Even if they were, it is questionable as to whether the data would provide an effective tool for regulation or furnish a clear message for the education of users and the basis for behavioural change.
One major concern is the fact that the issues outlined above are a symptom of current water management as a whole. In general, these management practices continue to ignore the integrity of groundwater systems even in arid regions where groundwater is the 'lender of last resort'. This is particularly true in the case of large sedimentary aquifer systems that are decoupled from contemporary recharge and are effectively non-renewable. In addition, the varying scales at which groundwater systems occur and are developed or exploited pose particular management challenges beyond those of conventional surface water or river basin management. Therefore, a clear articulation of the specific guiding principles in groundwater development and criteria for evaluating policy responses to groundwater depletion and degradation is warranted. Such principles may have little to do with the more conventional principles of 'integrated water resource management', which are generally predicated on hydraulic control and regulation over river basins. This 'engineering hydrology' focus on water management continues to colour water resource management styles, which remain largely centralized and technocratic. One example is Namibia, where groundwater provides some 60 percent of bulk water, yet the institutional arrangements and investments have focused on the development of intermittent surface flows.
River basins and surface water irrigation schemes present 'neat' arrangements. The resource is naturally integrated at any point in the basin's watercourse and it is easy to monitor measurements, diversions, storage and abstractions. The impacts of irrigation abstraction and return flows occur in near real time and are immediately apparent upstream and downstream through hydraulic continuity. The system is neatly bounded, there are clear solutions of continuity, and systems of rights in use are generally clearly established. The same cannot be said for aquifer systems and groundwater development. Aquifer systems are known imperfectly, there are no clear solutions of continuity (Burke, 2000), responses are highly non-linear (geological heterogeneity and anisotropy) and can be lagged over centuries with none of the clear 'water year' rhythm observed in surface basins.
The largely technocratic, vertically integrated basin management models built around surface water schemes and the sets of incentives to surface water managers and users differ markedly from the management 'models' and incentives associated with the more imprecise character of aquifer systems and groundwater use. In addition, while the 'client base' for a basin manager would typically consist of a set number of well-identified user groups (irrigation schemes, water user associations, municipalities, etc.), the manager of an aquifer system may in practice have to engage with millions of individual users.
Therefore, the transaction costs of applying 'cross-sectoral integrated water resources management' in a classic sense can be expected to be several orders of magnitude higher. This does not appear to augur well for progressive conjunctive use management as a means to reconcile competing surface and groundwater demands. While conjunctive use is common in many irrigation schemes where individual irrigators have wells in the surface command area, the specific management of the combined sources is less common and mechanisms for doing this appear to be lacking (Kloezen, 2002; Wahaj, 2001). However, progress in applying conjunctive use and the use of extensive informal water markets in and around surface-water and groundwater irrigation schemes demonstrate the ability of irrigation end users to adapt. Such adaptation often occurs in the face of contradictory signals and incentive structures established by higher order 'managers'. Whether such de facto arrangements offer more or less equity and economic 'efficiency' is debatable. Where regulation is weak or absent, the opportunity for the richer members of a groundwater user group to capture through technology or access to land will always be there. However, it is also possible to observe enhanced equity and efficiency through the myriad of small water, energy and pumping transactions that occur among irrigation user groups (Shah, 1993).
The sometimes unclear linkages between groundwater and food production have constrained the scope for management of the resource. Given the levels of uncertainty associated with groundwater information, the broader questions that remain are whether: (i) groundwater is amenable to the same types of management approaches associated with surface water irrigation (e.g. irrigation management transfer); and (ii) it sits well within the frame of so-called 'integrated water resource management'. It is important to resolve these questions because the aquifers that are being used intensively can be expected to be in arid and semi-arid zones. Moreover, in such areas, surface water alternatives are scarce or unavailable and drawdown and pollution externalities can be expected to affect users within and outside the zone of groundwater use. Therefore, solutions that require some sort of 'integrated' or consensual and expert management (such as conjunctive use) will become imperative. However, requests to individual users to sacrifice private opportunity for the sake of basin or aquifer-wide efficiencies and equity are likely to encounter resistance, particularly where groundwater access is the principal means for accumulating assets and escaping from poverty (Moench, 2001).
In irrigated agriculture, the scope for addressing such management problems is also conditioned by the need to serve several policy 'masters'. Irrigated agriculture is a key component of many national agriculture strategies and a factor in the political equations many national leaders face. However, it is also expected to conform to water and environmental policy initiatives. That management of groundwater use for irrigation should be part of a national and regional commitment to integrated water resource management is not in dispute, but precisely how and through which policy instruments it should be effected often remains unclear.
In terms of irrigation practice, efficiency gains are important at all scales if pressures are to be reduced on environmental flows and downstream/downgradient users. Irrigation is not in a position to foreclose on other users. It would be simplistic to assume that irrigation efficiency is not significant at basin level (Seckler, 1996). Irrespective of the impact across a particular basin or catchment system, it is the immediate deprivation of opportunity that may count. Similarly, it should also be appreciated that seeking efficiencies involves specific groups of water users and managers at the various levels. However, there may be no incentive for direct users to make efficiency gains if upstream managers cannot ensure conveyance efficiency. With groundwater, this may not apply as the incentive is generally internalized entirely by the user.
In some cases, it is already too late to talk about the sustainable development of groundwater because the aquifers are already depleted, polluted or salinized beyond the regenerative capacity of their natural hydrogeological regimes. Some industrialized countries (e.g. the United Kingdom) are moving toward a re-examination of groundwater management in a broader political and social context (Grey et al., 1995). Others (e.g. France) are maintaining a more technical perspective (Martin, 1996). Elsewhere, many developing countries that rely on pumped groundwater to sustain agricultural output and supply municipalities continue to permit the intensive levels of abstraction with little evidence of pro-active groundwater management being deployed. Technical regulation, economic incentives and participatory management approaches may offer the means to address groundwater management in the common interest. However, the character of initiatives will be determined necessarily by the local realities of the groundwater occurrences and the associated groundwater economy. Dealing with such diversity involves a different order of adaptability and flexibility than that normally associated with surface water or river basin management.
By the time groundwater arrives at the well head and enters irrigation ditches or raw-water pumping mains, it is perceived that groundwater management per se ceases and conventional water management takes over. Indeed, groundwater tends to be treated as the ultimate source of relatively high-quality water and the ultimate sink of used water. This occurs without any real appreciation of groundwater's regenerative capacity and its buffering role in the hydrological cascade. The management of surface water has fundamental implications for both groundwater quantity and quality at all stages and points of the hydrological cycle.
Therefore, it is essential to examine the scope for groundwater management not only in the strictest sense, but also as a prerequisite for integrated water resources management. This involves appealing to individual groundwater users in ways that have to do with advocacy and demonstration. It is significant that some political scientists sense a continuing tension between the 'eminence' of the State and the customary user rights of the beneficiaries (Barraque, 1998). How this tension is resolved is critical for groundwater because the management of diffuse abstraction is highly dependent upon the approach local communities take to negotiating the use of a common property. Historically, it has appeared easier, though not necessarily beneficial or cost-effective, for the State to control surface water abstractions and disposal within a vertically 'integrated' river basin plan where a central authority undertakes all the operational and regulatory functions related to the water cycle. Here the purpose of the State's intervention is the protection of the broader public interest in the basin's resources. However, this tendency in river basin management risks ignoring the important but highly distributed physico-chemical and socio-economic buffering roles of groundwater. The approach may also rely heavily on regulatory measures as opposed to economic incentives to achieve desired results. More significantly, the array of users with which basin agencies tend to engage (large sectoral user groups and local government representatives) may differ markedly from the groundwater 'stakeholders'. Millions of individual farmers are not necessarily amenable to the same degree of association that can be recognized among urban utilities, industries and command area authorities. Thus, it can be argued that integrating groundwater use within both a physical and a socio-economic framework becomes not only an environmental necessity but also a political imperative where policies of decentralization and subsidiarity are adopted.
Under these circumstances, the choice of water resource management instruments to address specific groundwater issues may still fall within the bounds defined by technical regulation, economic incentives and participation. However, they can be expected to take on a markedly different character from those associated with conventional surface water management. Equally, the quality of the information about aquifers, groundwater and user behaviour will also need to assume a specific character.
Groundwater acts as the primary buffer against the impact of climate variability and spatial variability in drought. However, as human development has become more susceptible to such variability, three major gaps in groundwater management have emerged, each with significant implications for sustainable development:
The inability to cope with the acceleration of degradation of groundwater systems by overabstraction, and effective resource depletion through quality changes (pollution, salinity).
In general, a lack of professional and public awareness about the sustainable use of groundwater resources. In particular, a lack of coherent planning frameworks to guide all scales of groundwater development and the consequent lack of appropriate policy responses and institutional development to prevent and attenuate degradation to groundwater systems.
The failure to resolve competition for groundwater and aquifer services between sectoral uses and environmental externalities.
These specific concerns hinge upon the central issue of awareness. This relates as much to the groundwater related environmental concerns in industrialized countries as it does to the peri-urban communities in developing countries who depend on locally available groundwater sources.
In this sense, groundwater management regimes may be expected to encompass a set of economic, regulatory and ethical levers that are operated by markets, regulators/state institutions and user associations. Effective institutional approaches need to be aware of these socio-economic realities surrounding groundwater use. They also need to appreciate the inherent risks associated with development, the level of uncertainty (plus limitations in data quality) and the range of social pressures.
In contrast to technical views of groundwater management, an overview of the Indian experience (Roy and Shah, 2002) makes the point that although indirect demand management strategies, such as rice export controls in the Punjab, do not form part of the water management debate, they may ultimately have more impact on slowing the rates of depletion than equity-based institutional solutions. This begs the question as to whether there are practical approaches for responding to groundwater problems and their socio-economic impacts that are absent in current management styles. The strategies outlined below are starting points for expanding conventional management thinking by adapting to real points of socio-economic leverage and to aquifer and user-group scales.
In general, it is possible to observe the following characteristics of current groundwater 'governance':
Lack of data and scientific understanding limit the ability of society to predict aquifer functioning and to develop realistic rights systems.
Rights systems are difficult to design and implement in most situations for a variety of technical and economic reasons.
In most cases, social acceptance of private rights may be problematic.
Aquifer management is politically complex because it would require active modification of established use patterns.
The dynamic nature of both socio-economic globalization and global climate change makes management complex. People are increasingly mobile and often have little incentive to participate in long-term management initiatives.
From among this set of characteristics, two broad types of management approaches for groundwater emerge: (i) 'thin and wide'; and (ii) 'thick and deep'. 'Thin and wide' approaches may encompass blunt tools such as power pricing, subsidies for efficient technologies, economic policies that discourage water intensive crops, etc. They can be applied over whole countries or regions. 'Thick and deep' approaches deal with specific aquifers on the basis of command and control management whereby aquifer management targets are set and enforced through a resource regulator.
Under these circumstances, groundwater management may be most realistic when applied in a limited manner to strategic aquifers for which a social consensus supporting management exists. This can be said to have happened in the case of the Qa'Disi aquifer in southern Jordan and the sister aquifers in Saudi Arabia. However, such examples are rare. Concerning recharge initiatives in India (Shah, 2000), here a 'thin and wide' approach to resource management may ultimately prove more successful than a 'thick and deep' approach to aquifer management, such as the technocratic initiatives to aquifer recharge along the Batiah coast in Oman and in the Quetta Basin in Baluchistan, Pakistan. In any event, the impacts of groundwater management approaches require specific monitoring and evaluation periods in order to make an accurate assessment of success in terms of aquifer response alone. It is doubtful whether enough time has elapsed to allow such an assessment of the various initiatives and to identify clear bright spots. These considerations apart, focusing management on strategic aquifers would also allow society to concentrate the required scientific, monitoring and enforcement tools on relatively small areas. In addition, any current users displaced by management could be absorbed more easily than if management were attempted over larger areas. The design of implementation strategies to initiate and then propagate across the area of concern will be key. The recharge movement observed by Shah (2000) will deserve attention where it is clear that such viability may have more to do with social structures than the technical feasibility of conservation and regulation.
In order to start addressing gaps in management, it is important to recognize that institutional innovation and adaptation will need to be more sensitive to the range of influences and management instruments. A diagnostic to develop such adaptations will need to cover:
macroeconomic policies;
sector policies;
rights systems;
institutions and capacities;
regulatory frameworks;
public involvement.
Against the 'soft' institutional strategies, it is possible to define sets of technical options that relate directly to groundwater. Although these options may present expanded opportunities to manage groundwater, they would have to be applied strategically in circumstances that are amenable (where uptake of technical strategies will succeed). Such technical options include:
conjunctive management (conjunctive use and aquifer storage and recovery);
conservation enhancement and protection;
water harvesting and supply enhancement;
irrigation efficiency improvement and demand management.
The design of a suite of institutional and technical strategies and their implementation at the scale required to make an impact (to conserve or reallocate groundwater resources) is unlikely to be achieved in the short term. Countries reliant upon groundwater irrigation, such as Mexico, are now considering periods of decades (H. Garduno, personal communication). It is up to agriculture to adapt and improve its water productivity and so release the pressure before the resource base is exhausted. In part the agriculture sector can achieve this through improvements in crop yields, irrigation efficiency and post-harvest processing. However, it may be non-technical interventions that play a greater role. Resource substitution through food imports and transfers to higher value uses are already making their mark. Setting up the socio-economic tools for more flexible allocation of scarce resources will be pivotal to addressing equity and efficiency concerns. For example, improved systems of land tenure and water use rights can help considerably in spreading local water risk and improving food security by optimizing local food production patterns. Equally, the economic signals to food producers need to be clear and stable. Indeed, sound agriculture policy can set a strategic balance between rainfed and irrigated agriculture where there is effective demand and real comparative advantage in domestic resource costs. Furthermore, agriculture policies that encourage different forms of agriculture in different areas could be critical. For example, India has nationwide policies governing the support price for rice and wheat. Such national-level policies could be tailored to reflect regional water availability differences. Thus, they could encourage production of water intensive and less water intensive crops in areas with different comparative advantages. Here again, 'thin and wide' approaches can complement 'thick and deep' approaches.
This is where there is real scope for savings in agricultural water demand. However, it will involve a change in the balance between rainfed and irrigated agriculture and a number of related socio-economic changes. People are making social transitions as a result of many other factors that are not related to water resource availability or management. Where people are making such a transition and the groundwater resource base is not being used in a sustainable manner, policies can be tailored in ways that encourage transitions towards low water-use livelihoods (often non-agricultural). This may not be part of conventional approaches to groundwater management, but it could be key where direct enforcement of water management targets is deemed impossible in either technical or political terms.
