Irrigation raises a number of important questions about the land tenure rights/water rights interface. First of all, the importance of irrigation and its contribution to global food security cannot be over-emphasized. Some forty percent of world food production is currently produced on around 250 million hectares of irrigated land an increase of some 200 million hectares over the course of the twentieth century. This increase is a result of huge investments in the sector that have the effect that on average some 73 percent of all water abstractions are for irrigation, with an even higher share in lower income countries: in India irrigation accounts for 93 percent of the gross amount of water used.
Furthermore, the level of demand for irrigation water is unlikely to decrease in the near future. At least 17 percent more freshwater than is currently available will be needed by 2025 to produce sufficient food for the 8.8 billion people who it is estimated will populate the planet, even if everything is done to make irrigated agriculture more water efficient. If this is not done, it is estimated that at least 55 percent more freshwater will be needed.
Like other water users, irrigators need secure water rights. In many countries irrigation is essential for crop production. Even in those countries where rainfall permits some levels of production, supplemental irrigation is very often necessary to render agriculture profitable. For irrigators a lack of water security negatively affects the utility and thus the value of any land tenure rights they hold and vice versa. It is therefore perhaps surprising that irrigators in many countries lack secure rights to water, secure land tenure rights or both.
The lack of secure water rights is frequently the result of a number of factors including the form and design of irrigation schemes and the manner in which they are funded, designed and developed. The bulk of investments in irrigation, in richer as well as poorer countries, has been provided by states (even though the beneficiaries are usually private farmers). Such state funded schemes often involve the damming of major rivers and the construction of concrete lined canals that may run for many scores of kilometres. At the same time, many small-scale farmer-built irrigation schemes exist around the world, often comprising little more than an earth channel from a river to convey water to the land to be watered. Whatever its scale or source of investment funding, the construction, operation and maintenance of an irrigation scheme is usually a relatively costly affair.
Although, as just noted, logic suggests that rights are as important for irrigators as for any other type of water user, water rights, of the type described in this paper, play only an indirect role as far as most individual irrigators are concerned.
This is because, as outlined in Part Two, water rights are concerned with the abstraction and use of water from natural sources. Apart from those (statistically relatively few) cases where water is abstracted and used on riparian land, the design of most irrigation schemes means that such rights are of secondary relevance. Typical irrigation schemes take water either from a reservoir on a dammed river or directly from such a river through some form of diversion structure (a so called run-of the river scheme). From the point of abstraction water may be conveyed for a large distance through so-called primary irrigation canals. To complicate matters further such canals may feed storage reservoirs or receive additional water from subsidiary canals. After passing through smaller secondary or tertiary distribution canals or pipes, water is conveyed to the land parcel to be irrigated. Final application of the water to the land may be through temporary earthen canals or channels. In some countries water is conveyed through piped networks and, if placed under pressure, may be delivered to the land through spray, sprinkler or drip equipment.
In these circumstances the operator of the scheme will, subject to the applicable water legislation, usually require a water right to impound and/or abstract water from the natural source. But as irrigation canals and pipes are not themselves natural water sources, those whose land is adjacent to them (and who are supplied with water from them) cannot hold water rights in respect of the water that they contain. Indeed the common law is quite clear that the owner of land adjacent to a canal or other artificial watercourse has no rights whatsoever to the water in the absence of some form of grant or arrangement. Under the common law tradition, to take water from such a canal would probably amount to theft. Indeed, as a matter of logic it is difficult to see how an ordinary statutory water right could be conferred on such a land holder as that person is not responsible for the abstraction of the water in the first place. A bare right to water would not be of much use without the ability to enforce it against the operator of the irrigation scheme. In the case of state funded schemes this is usually a state body such an irrigation agency.
As such, therefore, the right is not only to a volume of water but also to a service, namely the delivery of that water and the operation and maintenance of the relevant irrigation scheme that enables such deliveries to take place. Irrigators typically enter into annual agreements with the relevant irrigation agency for the delivery of water in return for the payment of an irrigation service fee. In some countries rotational irrigation rules exist that reflect a mixture of customary and statutory law. What happens, however, if an agreed volume of water is not delivered at the appropriate time is often left unspecified. Leaving aside the inequality of bargaining power between a state agency and an individual farmer it is clear that such arrangements fall far short of a water right in terms of the degree of security that they confer on the water user. In other words although irrigators in such circumstances need legal rights to water just as much as any other water right holder does, such rights to water are of quite a different nature to the type of water rights that are the focus of this paper.
The picture is often further complicated by ongoing programmes in many countries to transfer responsibility for the operation and maintenance of state irrigation schemes, in whole or in part, from irrigation agencies to farmer/irrigator operated water user associations, a process known as irrigation management transfer (IMT). While the original objectives of IMT were to provide a better service to irrigators, as well as to increase the collection rate of irrigation service fees, the process is increasingly propelled by the need to reduce government spending and thus the funding of irrigation agencies.
While water user associations go by a variety of different names and may undertake a range of water management activities besides irrigation, they invariably operate on a not-for-profit basis and are controlled by their beneficiaries in a democratic manner, usually through a general assembly of participants and an elected management board. The water user association is not a new concept. In some parts of the world ancient water user associations still operate on the basis of customary law. In many European and North American countries legislation regulating the establishment and operation of water user associations has been on the statute books for hundreds of years and IMT programmes in developing countries and transition states have increasingly focussed on the importance of developing specific legal frameworks for the establishment and operation of water user associations as a specific type of legal entity, a sui generis legal person. Such legislation, backed up through the water user association's internal governing document (variously described as the statute, constitution or by-laws) usually confers a number of rights on individual water user association participants including a right to a share of the water that the association receives.
Given the fact that participants in a water user association are often effectively in competition for the same valuable water resources, the importance of robust and clear legal frameworks that provide for the establishment of water user associations with fair and transparent governance structures and which confer substantive legal rights on participants cannot be over-emphasized. Of course law is but one factor in the successful operation of water user associations, but unless substantive rights are clearly conferred on all water user association participants there is a real risk that socio-economic factors will lead to control over individual associations being captured by local elites. In this connection the existence of clear legal rights will at the very least strengthen the bargaining position of the poor and disadvantaged.
The first question is, at the level of the individual water user association participant, is an entitlement to a fair share of the water received by the association equivalent to a formal water right? In some ways it is less secure because the water user association itself may fail for financial or other reasons. Furthermore an individual water user association participant has only an indirect role in ensuring that the association does not lose any water rights conferred on it. The issue is bypassed in much of the property law literature because the share is not, legally, a water right. Nevertheless, the practicalities of the matter are such that conferring what are essentially organization rights on individual irrigators is probably the most realistic solution of providing a degree of water security.
The next question concerns the rights of water user associations. There would appear to be three basic scenarios. First of all, if it has direct access to a water source a water user association will hold a water right in its own name on behalf of its members. Water user associations in Europe and North America have done this for many years.
In the second scenario the diversion structure and primary canal is operated by a federation of water user associations that also enjoys legal personality. In this case a formal water right is held by the federation with individual water user associations holding rights against the federation, including rights to a share of the water. In the third scenario responsibility for the operation of the primary canal and diversion structure remains with the relevant irrigation agency which then supplies water to water user associations on a contractual basis, again usually on the basis of annual agreements. Such agreements create a kind of water right, in the sense of an entitlement to a specific quantity of water, albeit one that as a result of its short duration confers very limited security on the water user association. Curiously, discussion of this topic has been largely absent within the IMT literature.
Given the continued and growing importance of irrigation, research is needed into the basis on which secure water rights could be conferred on irrigation water users, particularly in the context of IMT programmes. Experiences from the United States of America and Australia may provide a useful point of comparison. In the United States, where such arrangements have existed for many years, the relevant state agency, the Federal Bureau of Reclamation, enters into long term contracts with water user associations for the supply of irrigation water. Such contracts typically last for some 25-30 years. While on one hand such an arrangement appears to confer a greater degree of security on water user associations, no compensation is payable in the event that water is not supplied even if the agency is at fault.
The Australian approach is quite different. Within government irrigation areas, each land holding was initially granted a water right based on what the relevant irrigation agency thought it could supply in bad years. Each agency is obliged to supply farmers with domestic and stock water and irrigation water every year. Irrigation water rights are registered in relation to land within an irrigation district, which is usually supplied from irrigation canals operated by the local irrigation agency.
Quite often, however, a lack of secure rights to irrigation water is coupled by a lack of security, or at the very least a degree of ambiguity, regarding land tenure rights at the level either of individual irrigators and/or water user associations.
Lack of land tenure security for individual irrigators arises particularly in those cases where farmers and communities have been re-settled onto irrigation schemes that have been newly constructed or funded by the state. In undertaking such developments, the state, acting through an irrigation or development agency, often has its own objectives such as maximising food production in the interests of seeking national food security. Control over the land is often seen as the best way for the state to fulfil such objectives. Consequently access to land or to its exploitation is almost always defined as precarious and conditional: the state affirms itself as the owner of the irrigated land, and the irrigated parcels are distributed to users in a conditional manner for as along as they respect the rules of exploitation and fulfil the obligations defined by the framework (such as the payment of irrigation fees, contributions towards works of maintenance and the supply of information needed for hydraulic research etc). At the level of an individual scheme, irrigators are accorded only personal non-transferable use rights with the responsible state agency usually holding the necessary legal and administrative powers to expel those who do not conform with its objectives and rules.
Indeed particular problems frequently arise in cases where schemes have been built on lands that were formerly subject to customary land tenure rules. On the scale of a regional development project, the conditionality of use and the suppression, as a matter of principle, of customary land tenure rights means that it is possible to attribute the developed lands (or to develop them in the case of concessions granted to private actors) to those who are judged most able to effectively exploit them even if they come from outside the area (or from abroad) and thus do not have any traditional rights over such lands.
Even though in many developing countries land within irrigation schemes remains under state control this is often only really the case at a formal level with informal land transactions taking place in any event. Nevertheless the uncertainties caused by the divergence between the de facto and de jure situation and the ensuing lack of land tenure security may deter investment in the land by individual irrigators who cannot use the land to raise capital and who may be afraid to lease irrigated land. In itself this is not a remarkable observation: such are common consequences of land tenure insecurity.
However, in the context of irrigated land a key point to note is that in the context of IMT it may be unrealistic to expect that an IMT programme can be successful in the absence of secure rights both to water and land. Establishing and operating a sustainable water user association is a form of investment in itself and it is probably unrealistic to expect such investments to take place on the basis of insecure land tenure and water rights.
Another common area of legal ambiguity relates to the degree of land tenure security that water user associations enjoy in respect of the components of irrigation schemes that are transferred to them, as well as the extent of their rights to access land held by third parties for the purposes of operating and maintaining those components. Again this is an area that has been largely neglected in the literature and in practice. While some irrigation schemes are transferred into the ownership of water user associations, very often such transfers take place in use (with the association receiving only a use right). Transfers in ownership have a number of potential advantages: they demonstrate a clear and strong commitment to the process of IMT, they arguably confer a greater degree of security on water user associations and send a clear signal to associations that long term maintenance and ultimately the eventual renewal/rehabilitation of such schemes is their responsibility and not that of the state. On the other hand states are frequently reluctant to transfer ownership of schemes that often represent significant investments and furthermore a transfer into ownership generally makes it harder for the relevant state agency to make sure that a scheme is operated and maintained correctly. In any event, whatever form of transfer is undertaken it is likely to amount to a transfer of real or immovable property rights. Consequently in most jurisdictions such transfers must therefore be registered in accordance with the applicable land registration rules. Yet all too often such obligations of land tenure law are ignored meaning that rights held by water user associations are imperfect.
In designing irrigation projects both national agencies and foreign donors frequently fail to take the complexity of land tenure rights and water rights sufficiently into account, partly because their top-down approach in planning and construction of irrigation is considered as a mainly technocratic process, and partly because the socio-economic impact on families is insufficiently realised.
What seems to be quite clear is that in the context of irrigated land it is necessary to take simultaneous account of both land tenure rights and water rights security particularly in the context of interventions by development agencies. There is also the risk of conflicting land tenure reform and water policies being developed and advocated in mutual isolation one from the other. Furthermore it is necessary to take account of all relevant aspects of both land tenure rights and water rights, including gender issues, an area which as will be seen below is usually neglected in the literature on water rights. A failure to do so can lead to a partial or complete failure in meeting the objectives of such types of intervention. The countries of Eastern Europe and Central Asia provide a fine example in this context (see Box D).
In short, at the very least, a greater dialogue is necessary between land tenure professionals and water professionals working on the area of reforms to irrigated and drained lands. While in most jurisdictions, land tenure legislation makes no distinction between irrigable and non-irrigable land, at the practical level such a distinction is essential if the land is to be used to its full potential. At the same time it is clear that more research, including field work, is necessary into the relationship between land tenure rights and irrigation water rights so as to ensure increased land and water security and the development of consistent policies. In the context of developing countries, and particularly in the case of long-established irrigation schemes, local customary rules regarding the use of both land and water may exist. Whether and how such customary rules should be integrated into, or recognized by, formal law is an issue that is considered below. However, in order for formal law to have a viable relationship with customary rights relating to land tenure and water it is essential that it has a conceptually viable and effective approach to offer.
Box D - Land and irrigation sector reforms in Eastern Europe and Central Asia
During the period of communist rule, major investments were made in the former socialist countries of Eastern Europe and Central Asia in the construction of both irrigation and drainage schemes designed to supply water to, and drain water from, the lands of the collectivised farms. Such farms in turn generally had their own work brigades to operate their on-farm irrigation and drainage systems.
With the collapse of communism, reforms in the agricultural sector saw the break up of the collectivised farms and their land holdings with the aim of creating private farms that would function on market economy lines. Depending on the jurisdiction, land was either restituted to the former land owners or distributed, on an objectively equal basis by reference to parcel size and land quality, to the former collective farm employees and their families. New legislation was enacted to permit the establishment of ownership or long term use rights and the introduction of land markets coupled with investments in survey projects and the introduction of land registers. It is fair to say that this agenda was driven by notions of land tenure law by land lawyers and other land professionals. However, in the process irrigation and drainage were largely ignored.
This was no insignificant omission. In many transition economies, particularly in the countries of the former Soviet Union, irrigation and/or drainage is essential for agriculture. The collapse of the collectivised farms not only meant that there was no-one to run the on-farm irrigation systems but the fields themselves had been split up into tens or hundreds of small privately farmed parcels, each with different crops and thus watering requirements. The policy response has generally been to introduce water user associations - a process which is ongoing. However, as water sector reform has lagged behind land reform, few of these water user associations have any type of water security as they are still supplied with irrigation water on the basis of annual delivery contracts. While the ensuing disputes and delays in resolving disputes over the irrigation water delivery has not totally derailed land and agrarian reform, the piecemeal approach has certainly not helped achieve the original objectives of viable private farming communities.
What is really surprising, though, is the way in which land tenure issues relating to irrigation were frequently dealt with. In Moldova, for example, where around one third of the collectivised farms had irrigation systems installed at great expense and which had/have the potential to significantly increase yields, irrigation infrastructure was treated in the government's farm privatization programme as a non-land asset to be distributed among former farm employees alongside tractors and livestock! The same happened in the Kyrgyz Republic with the effect that in some places the ownership of valuable irrigation systems has been effectively lost to farming communities. Indeed it is not just a question of irrigation. In Estonia, where despite the fact that two thirds of the arable land area is covered with intricate sub-soil field drainage systems, designed to serve the collectivised farms, the provisions of the relevant land law are completely unhelpful in determining how such systems should be operated following the land restitution process: they simply repeat the classic civil law tradition principle cited above that the owner of lower lying land may not hinder the flow of drainage water from higher land.
The management and use of groundwater resources also raises a number of questions about the complexities of the land tenure/water rights interface. Groundwater is contained in aquifers beneath the surface of the land: an aquifer is a geological formation that has sufficient water-transmitting capacity to yield a useful water supply in wells and springs. With the exception of so-called fossil groundwater, which is de-coupled from contemporary discharge, it is part of the hydraulic cycle and thus in constant movement, following the hydraulic gradient of the aquifer under confined or un-confined conditions. The degree of confinement is determined by the properties and geometry of the geological strata through which groundwater flows (see Box A). There is arguably as much variation in the characteristics of aquifers as there is with surface water resources. While all have the characteristics of storage and flow, they vary widely in hydraulic properties (permeability and storability) and reservoir volume (effective thickness and geological extension).
The very nature of groundwater, hidden beneath the earth's surface, has implications for data collection: as a general rule in most countries the availability of data on groundwater quantity and quality is variable, certainly compared with surface water data. The main problem is that a single groundwater observation is only a sample and not an integrated measure of upstream hydrology, as with a flow measurement made in a (surface) watercourse. Many groundwater observations are therefore required to characterise an aquifer and all monitoring wells have to be regularly checked to maintain a meaningful time series of groundwater change. This data requirement imposes a massive strain on groundwater agencies responsible for collecting groundwater data and they rarely have the commensurate resources. As a result, groundwater data coverage tends to be patchy and of highly variable quality.
Irrespective of data quantity and quality, groundwater is a major source of water in many countries being used for urban water supply as well as to meet the needs of industry and agriculture. South Asia has in particular seen a huge increase in the number and rate of groundwater abstractions. In Punjab Province, Pakistan, the number of mechanized wells and tube wells has increased from barely a few thousand in 1960 to 0.5 million today. In Bangladesh the number of deep and shallow tube wells increased from 93 000 in 1982-1983 to almost 0.8 million in 1999-2000. India saw an increase in the number of mechanized water extraction mechanisms from less than one million in 1960 to almost 26-28 million in 2002. The benefits of groundwater for farmers and irrigators are that the water is produced at or near the point of use, needs little transport, can be supplied on demand and just-in-time. In addition, because it entails significant incremental costs for lift, farmers tend to economize on its use, and therefore maximise application efficiency. The effect is that as much as 70-80 percent of India's agricultural output may be groundwater dependent. Interestingly, most South Asian countries have active markets in pumped irrigation water in which tube-well owners sell groundwater to their neighbours at a price that exceeds their marginal cost of pumping.
Notwithstanding the continued abundance of groundwater in many parts of the world, as a resource it is quite fragile and one that is increasingly coming under threat. One of the main threats to groundwater resources is that of over-abstraction leading to a lowering of the water table. If the rate of withdrawal is greater than the rate at which an aquifer is re-charged from surface water intrusion, there is a risk of eventual exhaustion. Over-abstraction from aquifers in coastal areas has an additional risk, that of salt-water intrusion. The other main threat to groundwater resources concerns quality: threats to aquifers include anthropogenic pollution, pollution caused by excessive abstraction, well head contamination and naturally occurring contamination as a result of a change in relative acidity/alkalinity caused by minerals in the aquifer which may be exacerbated by excessive abstraction. Once polluted, the costs and the timescale for remediation of an aquifer vastly exceed those of surface water resources. Aquifers in many parts of South Asia, for example, are under serious threat of depletion and degradation as indeed are some of those in North America particularly the Ogallala Aquifer, which crosses the boundaries of a number of Western States, and others in parts of California.
Groundwater use and over use can raise important equity issues, particularly in rural areas when, following increased consumption, the level of water tables has been lowered. Only the richer farmers or land owners can afford to sink deep wells and boreholes and to purchase the more powerful pumps necessary to abstract groundwater. Those with access to such water have a captive market comprising those who do not. But it is not only among farming communities that equity considerations arise: cities and factories and other industrial sites also compete for groundwater against users in poorer rural areas.
The threats to groundwater resources are compounded by several different factors. First of all, it is only relatively recently that its nature has been understood. As already mentioned in many parts of the world data is either unavailable or of poor quality. Finally, and largely as a result of the two previous factors, the manner in which it has been regulated, its legal treatment, has lagged behind that of surface water resources.
At a theoretical level, rights over groundwater should serve the same basic purposes as rights over surface water and, for that matter, land. They should enable societies to provide for the orderly allocation of valuable resources while at the same time conferring the necessary degree of security on rights holders. The water rights/land tenure rights interface is one of the key reasons why this has not happened.
As already noted in this paper, historically the principal focus of water law, and thus of water rights, has been on surface water resources. It is only relatively recently, over the last hundred or so years, that specific legal responses have been formulated in water legislation to the questions of groundwater management and use.
Traditionally, within the civil law tradition, in accordance with the basic principles of Roman law, groundwater was seen as part of the property of the owner of the land above it. This approach is reflected in article 552 of the French Civil Code which states as follows:
Ownership of the ground involves ownership of what is above and below it.
An owner may make above all the plantings and constructions which he deems proper, unless otherwise provided for in the Title Of Servitudes or Land Services.
He may make below all constructions and excavations which he deems proper and draw from these excavations all the products which they can give, subject to the limitations resulting from statutes and regulations relating to mines and from police statutes and regulations.
While the approach of the common law tradition was slightly different, the effect was largely the same. Under the common law there is no property in water percolating through the sub-soil until it has been the object of an appropriation. The effect is that a land owner is entitled to sink a borehole or well on his or her land to intercept water percolating underneath his property even though the effect is to interfere with the supply of underground water to nearby springs. Yet at the same time, the owner of land through which groundwater flows has no right or interest in it which enables him or her to maintain an action against another landowner whose actions interfere with the supply of water.
In practice, once modern well drilling techniques and pumps were developed, neither the civil law tradition nor the common law could offer a viable means of effectively regulating the use of groundwater, although in some jurisdictions, such as the State of Texas, the common law rules described above, sometimes described as the doctrine of capture, still apply.
The most common legislative response has been first to vest groundwater in the public domain of the state, or to bring it under state control through a statutory assertion of superior use rights or a declaration of public trust on behalf of the people. In other words, groundwater is deemed to have the same legal status as surface water.
Next, various restrictions are placed on the use of groundwater. These include requiring well drillers to hold licences, that are issued in accordance with technical criteria relating to competence and experience, licensing the drilling of individual wells and boreholes and specifying minimum distance requirements between wells and boreholes. Finally, the abstraction and use of groundwater is typically brought into the same, or a similar, administrative regime to that for the use of surface water, through the creation of water rights based on administrative permits. As with rights to surface water, such rights are generally subject to a variety of conditions relating to their duration, monitoring, the quantity of water that may be abstracted and so forth. And as with surface water rights a groundwater right may be lost through non-compliance with applicable conditions.
Furthermore, as with surface water rights, legislation typically provides that a formal right to abstract and use groundwater is not necessary in connection with certain specified purposes provided relatively small volumes of water are used: in Australia, for example, a formal water right is not necessary for the abstraction and use of groundwater for stock and domestic purposes (including household garden irrigation). Exemptions for small wells are usually justified on the basis that their use will have little impact on the total available water supply as well as the administrative burden of seeking to regulate them. Nevertheless, the sheer number of individual wells can ultimately have a major impact on the quantity and quality of groundwater and related surface water resources.
From a hydrological, indeed hydro-geological, perspective it makes perfect sense to regulate groundwater and surface water under the same regime, after all the hydraulic cycle means that they are closely interlinked. Indeed increased pressure on surface water resources may well see greater use of conjunctive management whereby surplus water from precipitation and surface water bodies is captured and stored in aquifers for release during drier periods.
In practice using the principles of surface water rights to regulate the use of groundwater is not always effective. The difficulties lie not in establishing the necessary regulatory systems but in monitoring and enforcement. While it is by no means cheap or easy to monitor abstractions from surface water bodies, it is at least relatively easy to identify the possible abstraction points - the banks and shores of streams, rivers, lakes etc. In contrast, unauthorised and illegal wells and boreholes may be located literally anywhere above a given aquifer and the sheer numbers involved will often make the monitoring of abstraction rates from authorised and thus legal wells virtually impossible. Legislative techniques to indirectly address the issue, such as imposing a duty on suppliers only to provide electricity to well owners who hold valid water rights, can themselves be avoided by illegal connections (power theft) or the use of diesel pumps. And a further important problem is institutional. The regulation of groundwater rights is usually the responsibility of a groundwater department of the relevant water administration. Given that the primary legal and institutional focus of water legislation has been on surface water, such a department often has a somewhat junior role. But what is more serious is that the powers of such a body are really extremely limited. What can be done about illegal groundwater abstractions even if they are detected? The usual regulatory response is to initiate criminal or administrative proceedings against the wrongdoer, a time consuming and expensive process which at the end of the day has no direct impact on the resources in question, the land or water. A wrongdoer may be fined or even imprisoned, the well may be blocked up, but a crucial instrument of the offence, and indeed the basis on which access is given to groundwater, namely his land, remains intact.
As a result many states, including developing states and those in transition, struggle unsuccessfully to apply and enforce groundwater regulatory regimes. A number of North Indian states have debated groundwater laws for some 30 years with little progress as regards enactment. One exception is the State of Gujarat, the state with most severe groundwater overdraft problems. Shah et al. describe how the legislative assembly passed a bill as far back as in 1974, but the Chief Minister refused to sign it into a law. His reasons were convincing. First, he was unable to see how the law could be effectively enforced against a million small private well operators scattered throughout a huge countryside. Second, he was certain that it would become one more instrument of rent seeking for the local bureaucracy.
In any event does it make sense to try and regulate groundwater as if it is surface water when common sense suggests that it is part of, or at least closely integrated with, the land above? If not, what is the solution? Placing the operation of wells and boreholes under state control is clearly unrealistic. Indeed experience in India with public wells has not been very encouraging as the cost of their operation has been relatively much more than private tube-wells, the maintenance poor, and the access of small and marginal farmers to the resource, disappointingly low.
Providing for the seizure or confiscation of land on which illegal wells and boreholes are sunk would be a dramatic and largely undesirable solution. But perhaps the solution is to re-examine the relationship between land tenure rights and rights to use the water beneath that land.
Rather than expecting the state to regulate groundwater resources, perhaps it might be a better solution to confer management responsibilities on those who primarily benefit from them. This could possibly be done through the introduction of localised management of groundwater resources based on the holding of land tenure rights given that it is those rights that confer access to groundwater. Re-connecting rights to groundwater with the land above that resource would enable those who hold rights to make decisions concerning the management and use of that groundwater or a fraction of it by reference to the size of their land parcels and the size of the aquifer. As already mentioned in the context of irrigation, there is after all a long tradition in the water sector of participatory co-management of natural resources.
Consequently one option might be to explore the extent to which the model provided by ground-water or resource management districts found in a number of American states could be replicated elsewhere. The participants in such districts, which are a form of water user association, are the owners (and sometimes users) of land above and dependent on stressed aquifers. A key advantage of such an entity is that is self-policing with operating rules being determined and enforced by its participants. As a result of reluctance on the part of legislators to interfere with existing vested interests and vested rights the first groundwater management districts tended to play more of a co-ordinating rather than an enforcement or managerial role. Recent experiences have, however, been more encouraging and as pressure increases on aquifers it may well be that land owners and users will take a greater interest in playing an active role in the management of the groundwater below that land. Regulation of all groundwater resources may well be both impossible and un-necessary, economic considerations may encourage land owners and users to take a short term view as regards aquifer over-abstraction and in any event no legal or regulatory response can in itself provide a complete solution to natural resource management problems. Nevertheless in certain situations a re-assertion of the land tenure/water rights link, possibly through the use of bodies like water user associations, may offer valuable legal means of contributing to improved groundwater management.
As described above, the other main threat to groundwater resources relates to their quality. As with surface water resources, however, land tenure rights are by their nature likely to play a relatively minor role in attempts to preserve and protect water quality. Instead it is necessary to consider the relationship between groundwater resources and land use planning regimes and legislation as well as with specific legislative measures, often contained in water legislation, that seek to protect the quality of groundwater resources. Examples include restrictions on development and agricultural activities within specified distances from wells and bore-holes or over groundwater re-charge zones. Integrating groundwater protection with land use planning processes can raise similar issues regarding jurisdictional priority to those that arise in connection with the management of surface waters, as outlined in Part Five. Indeed the situation is further complicated by the fact that the boundaries of aquifers do not necessarily follow the boundaries of river basins just as they seldom, if ever, reflect administrative boundaries.
A question that arises in many parts of the world is what should be the relationship between formal land tenure rights and water rights regimes and land tenure rights and water rights that exist under customary law. The issue can be put another way. In most jurisdictions the legal recognition of both land tenure rights and water rights depends ultimately on their inclusion in formal registers. Customary rights do not receive their normative effect on the basis of registration. Should customary rights be recognized by formal regimes and if so how can this best be achieved? And if customary rights over, say, land are recognized how should this affect the recognition of customary rights over water?
The influence of European concepts of land tenure rights and water rights was described in Part Two by reference to their reception into colonial jurisdictions. If the impression was created that this was a smooth and uniform process within colonial boundaries then such an impression was false. For while the precise process of reception varied from country to country, the uniform application of European law was not one of the outcomes. Generally speaking, the influence of European law was strongest in cities and urban areas. Beyond those areas its influence was largely dependent on the level of settler activity. For example, European law applied to land tenure rights and water rights in the latifundia or large estates of Latin America because that was the law of the European estate owners. Similarly, European law was used and applied by settlers in African colonies to defend their new interests. In India colonial land law was accommodated with traditional land tenure patterns, sometimes transforming social relationships.
Land not subject to such types of transformation was typically declared to be state land and while notionally under state management, in practice it was often not. European-influenced law had, and frequently continues to have, little formal impact on such land, either in those remote areas in which indigenous people live or, as in much of Africa, on state land where customary law applied. The situation is little changed in Africa where in many countries, the vast majority of land (more than 90 percent on average) remains under customary tenure.
A similar situation arises in many developing countries as regards water rights, either because water rights regimes have never really been systematically introduced or because those that do exist are ill-adapted to the need of water users. Research conducted in the Pangani River Basin in Tanzania in 1994 revealed that of 2 265 abstractions only 171 were subject to formal water rights. A typical lawyer's response is to assume that all such abstractions are simply illegal. In practice the majority are likely to be made in accordance with rights created under customary law. Indeed, sometimes the rights interface that is the subject of this paper can further complicate matters. To remain with the example of Tanzania, until relatively recently formal water rights could only be held by those who held formal land tenure rights, thus excluding the bulk of the population.
An important point to emphasize is that, in contrast to the land tenure sector, there appears to be much less published research into the relationship between water rights established under customary law and the requirements of formal law. Indeed it would be useful to establish what, if any, have been the effects of the provisions in modern water legislation that assert state control over water resources, a process commonly understood as one of nationalization, on customary water rights, an issue that has been researched in connection with land tenure rights. But first of all it is necessary to consider just what is meant by customary law.
The meaning, nature and scope of customary law raise a number of questions that do not find easy answers. One commentator has suggested that in the context of land tenure it would be better to talk in terms of socially determined land-use rules rather than customary systems since the latter could suggest something traditional or ancient with roots in the past. After all it is in some ways disingenuous to contrast traditional practices in Africa with official legislation as many communities have faced the latter for years since the dawn of the colonial era.
It is better, perhaps, to take a broader view and to clearly recognize the dynamic nature and adaptability of customary rule systems. Indeed research on legal pluralism has challenged simple static ideas about customary law. What may be recognized in formal legal terms as customary law captures only a small, and often distorted, part of the complexity of local law. For local rules are generally developed spontaneously by communities to allocate the use of important resources such as land and water, particularly in cases, where formal rules relating to their use are ill-adapted or simply not applied. The vocabulary and insights of legal pluralism, which recognizes the law as it is rather than as it should be, can shed important light on how such rules are applied and developed in practice and their relationship, if any, with formal law. Nevertheless, for the purpose of this paper the term customary law will be used but in a wide sense so as to include local law.
Another risk regarding customary law is that it is often taken to be inherently democratic, egalitarian, equitable and therefore to deserve support in contrast to formal law and regulations issued from distant capitals, which are not. This kind of romantic view is false. There is ample evidence that customary law frequently reflects unequal power relationships in local communities. Such relationships greatly affect the ways in which land and water are distributed and managed State law may often seek to confer more rights on the less advantaged members of a given community, on paper at least. This is one reason why it is important that legislation requires water user associations, which in the context of their internally adopted governing documents and operating rules develop and apply a form of local law, to have fair transparent governance structures and to confer clear and substantive rights on their participants. Indeed, recently concluded research over the use of land and water resources suggests that customary tenure acts neither as an obstacle to investment and increased productivity nor as an inalienable safety net for the poor.
The relationship between formal and customary rights is complicated by the fact that the form and substance of customary rights generally does not accord with those of their European-influenced equivalents which in the case of both land tenure and water are based on individual private property rights. A further complication is provided by the fact that customary law and customary rights are likely to vary quite substantially from community to community both as regards their substance and the extent to which they are still applied. Tanzania, for example, has over one hundred and twenty ethnic groups.
A full discussion of the scope and range of customary land tenure regimes is beyond the scope of this paper. Nevertheless to take the example of Africa, in broad terms customary land tenure regimes commonly confer a form of group right on a community with individual households holding long term rights to use specific land parcels that derive from that group right. Transactions regarding such household rights are often only recognized within the bounds of the community. Responsibility for decision making in connection with the allocation of household rights and the resolution of land tenure disputes typically lies with a community chief, perhaps guided by a council of elders.
As regards water rights, one commentator has claimed that there is a structural similarity manifested by pre-colonial water laws in parts of Africa inhabited by populations vastly different in culture and economic activity which create a pattern of stable core entitlements, rigidly protected from competition but circumscribed by rules enforcing a regime of sharing. In support of this claim the law and custom of the Tswana is cited. Water rights on pasture land belonged only to the tribal sub-group to whom the land had been allocated and those who had dug wells and made dams acquired private rights over them. Disregarding such entitlements constituted a punishable offence within the community and thus against other tribal sub-groups. Yet custom mitigated the exclusionary effects of these entitlements, as outsiders who drove their cattle through the district had the right to water their cattle there temporarily as they went along without having to obtain permission from the local right holders. However, while a well owner could refuse water for another man's cattle he was bound to give him water for his personal needs. In contrast, at the oases of El Kasr and Tozeur in Tunisia absolute and freely alienable private property rights existed but which were tempered by the Islamic right of Shafat which entitles strangers and their animals to slake their thirst with water belonging to another. Uses which expressed a claim to an entitlement as extensive as that of the original right holder were prohibited whereas functionally distinct claims were not.
Clearly these types of rights regimes and thus the rights that they create are quite different to modern land tenure rights and water rights regimes as described in this paper. While the concept of common property is certainly recognized under the common law tradition, the rights to use a common, this concept was not received into colonial law. Rights of common were essential feudal rights and indeed the enclosure of the commons and their transformation into privately owned land was a key driver of radical socio-economic change in England, in its way every bit as dramatic as the industrial revolution.
The key questions regarding the relationship between customary land tenure rights and water rights are first whether or not they should be recognized under formal regimes and if so how? As regards the first question, attempts by formal legal systems to ignore customary land tenure law are often largely ineffective as outlined above. Notwithstanding the enactment of formal legislation, customary law frequently remains the only type of law that is applied at the local level. There is therefore, increasingly, recognition of the important role that customary land tenure rights continue to play as a well as a realisation that in place of costly and frequently ineffective attempts to impose European-influenced concepts of land tenure rights on peoples and communities some form of accommodation between formal and informal rights regimes is necessary. As regards customary water rights the situation is a little different. As already noted there is less published research into the relationship between formal and customary water rights and indeed there are fewer moves to give them recognition under formal law. In part this may be because the volumes of water that are subject to such rights are subject de minimis provisions on water entitlements, an issue that is returned to below.
The relationship between customary land tenure rights and water rights and formal rights is, however, somewhat complex. First of all, as noted in the previous section the type of rights recognized by customary law are often quite different to those recognized by formal law thus making it more difficult to accommodate them within formal rights regimes. At the same time formal law may in some circumstances be the best means of supporting the rights and interests of individuals who suffer from unequal power relationships within a given community.
What does seem quite likely is that the relationship between customary rights and formal rights will become more important over the coming years as a result of increased pressure on land and water resources.
Customary land tenure rights may, for example, operate quite effectively within a given community. As an accepted means of equitably allocating land resources they may confer the necessary security on community members within the ambit of that community. However, unless formally recognized by, or accommodated by, formal law such rights, whether held by that community or its members, may not survive confrontation with formal land tenure rights held by outsiders: persons from outside that community. After all, such formal rights, which may be held by rich investors from urban areas or from overseas, are backed up by the state and its enforcement processes. Such types of dispute over land rights can and do happen.
Similarly customary rights to use water in a given watercourse may continue to work satisfactorily among the members of a given community. But they too are not likely to survive intervention by formal rights holding outsiders such as, for example, the construction of an upstream hydro-power dam constructed and operated in accordance with a formal water right. A conflict between customary water rights and formal rights may to some extent be hidden from view given the inter-connection between water rights throughout a river basin. But conflicts over land tenure rights may be equally hidden in cases where land is not permanently occupied or settled, the rights of pastoralists being a case in point.
In examining this issue further it is convenient to distinguish between the rights of pastoralists and those of settled populations.
It is as regards the common property rights over land and water customarily enjoyed by nomadic pastoralist peoples that formal land tenure rights and water rights regimes really struggle to have relevance. The fragility of the arid grasslands of sub-Saharan Africa is such that it is simply unrealistic to imagine (technologically as well as economically) moving from common property resource management to formally recognized individual private land ownership.
The migratory patterns of pastoralist peoples who range over vast areas with their livestock offer the most sustainable and secure means of securing fodder. Such patterns are recognized by customary land tenure regimes: nomadic pastoralists do not own the land over which their cattle range either individually or by group. But their use and grazing rights are recognized under customary law. In this process water plays a vital role as the location of water sources influences pastoralists' choices of grazing areas, opens up new pasture areas and thus improving animals' nutritional status. Consequently water rights are the key to control and utilisation of arid and semi-arid areas. Here pastoralists are at a double disadvantage: holding no formal rights over land the law does not permit them to hold formal rights over water.
Indeed very often states have further weakened the role of pastoralists, through, for example bringing land under state ownership to the detriment of the customary rights formerly enjoyed by pastoralists. The same is true of development interventions such as the provision of water on rangelands and the irrigation of river margins. Whereas before customary rights holders controlled access to water, today public water points are accessible to all. This has concentrated people and livestock in new locations and transformed traditional land use patterns that have led to environmental stress. Furthermore, declarations of state ownership over water resources have the effect of denying customary rights and permitting what is effectively open access to water sources.
In fact in many countries pastoralists may in theory have a formal legal entitlement to use water for stock purposes on the basis of de minimis water rights as described in Part Two. Such rights are not, however, without their limitations. The essential problem is that a person who seeks to benefit from such an entitlement cannot lawfully prevent anyone else from also using the resource even if that use affects his own prior use/entitlement. Indeed the question arises as to whether or not they really amount to legal rights at all. The answer to this question probably depends on the legal tradition in question. The civil law tradition would probably answer that it is a legal right, albeit one that is difficult to enforce. The common law tradition, which has historically been more concerned with remedies than rights, might answer that it is not. Indeed the courts have held that if ... a person has a legal entitlement to something that can be enforced and protected, he has a right. But if a person has a legal entitlement which the court will not enforce, then he does not, practically speaking have a right. Whatever the ultimate answer to this question may be, the simple fact is that water rights regimes almost invariably fail to recognize the specific needs of pastoralists.
Reforms to land tenure rights regimes appear to have achieved little more. Attempts to provide in law for the needs of pastoralists have not always been followed through in practice. For example in Nigeria, notwithstanding a 1965 grazing reserve law that gave regional governments the power to acquire land for reserves, a 1978 Land Use Decree that extended the scope of that law and a National Agricultural Policy of 1988 that specified that reserves should cover a minimum of ten percent of national territory (9.8 million hectares), by 1998 only 313 reserves covering 2.8 million hectares had been acquired by the regional governments, of which 52 had been gazetted. Conflicts are caused when farmers deliberately block access to pasture or water, cultivate next to water points, or deliberately leave crop residues unprotected. The basic problem with the approach of creating reserves is that it fails to confer effective rights on pastoralists.
However, conferring individual rights is not without its problems either. In some jurisdictions, such as Botswana, private rights over rangeland have been conferred on individuals, elsewhere, as in the case of Kenya so-called group ranches have been introduced. Such reforms permit rights holders to exclude others, including other pastoralists, from their land parcels. In neither case, however, does it appear that the full impacts on access to water have been fully thought through. Indeed there is a tendency towards increased pressure on the remaining commons, the open access rangelands, as the holders of private rights keep their own land in reserve until those lands are denuded. The Rural Code of Niger is an exception to this trend. It creates a principle of priority rights to guarantee herders the right to control the use of the land that they had traditionally used. Under the Rural Code, while grazing land is considered to be common property, herding groups are allowed to use the same mechanisms to establish priority rights that farmers may use to establish exclusive rights, in other words private property. The effective content of priority rights has not yet been established by implementing regulations or observed experience, however, so that the adequacy of the legal guarantee is as yet unknown.
In short as far as the interests of pastoralists are concerned the interface between formal land tenure rights and water rights is at the same time both vital to their future survival - and largely irrelevant.
Generally speaking far greater attempts have been made to recognize customary land tenure rights of non-pastoralist peoples through a variety of different legal mechanisms. These including the grant of individual titles, the establishment of land boards, devolution to village and local government units and the recognition of customary chiefs. Controversy rages as to the desirability and impacts of programmes to confer formal title, relating to such questions as the impacts on security and investment and the risk of elite capture, in a debate that is beyond the scope of this paper.
While water legislation has sometimes sought to recognize customary water rights held by members of settled populations and to bring them within the formal rights regime more commonly it has ignored them. In part this too may be because customary uses fall within the de minimis rights provisions already described. In general, though, further research is needed not just about the content of customary water rights but as to how they inter-act with formal water law, and how they could best be recognized if this is deemed appropriate. Indeed it should be noted that customary water rights are not an issue only for developing countries. They have been described as the sleeping giant of water law in western and northern Canada as many of the older land claim treaties with native people did not address water rights.
What of the rights interface? Both regimes share common questions such as how best to recognize customary rights and in particular community or group rights. As regards community rights this may be an area where the long legislative experience in the water sector of formally constituted community based water user associations that hold water rights on behalf of their members can offer solutions as regards both water rights and land tenure rights. Otherwise as regards the interface an important issue would appear to be as follows. What is the impact, if any, on customary water rights of the recognition of customary land rights? Indeed does customary law distinguish between land law and water law in the way that formal European influenced law does? In this connection, given that customary water rights are generally based on the existence of customary land rights it would seem important to ensure that steps to recognize customary land tenure rights are not undermined by measures in water legislation that deny the validity of customary water rights.
Finally, there is one aspect of customary land tenure that can frequently conflict, not with water rights as such but with water legislation, and that concerns the use of river-side gardens, depressions (dambos) and wetlands.
The widespread introduction of tradable water rights would appear to have a number of implications not just for water rights regimes but also for land tenure rights. Is this likely and what would those implications be?
As outlined above, few jurisdictions currently permit unrestricted trade in water rights. Indeed comparatively few permit any trade in water rights separate to the land to which they pertain. A number of commentators consider that marketing offers a way to increase the efficiency of water use and allocation and to allow resources move from lower to higher value uses. The approach is also increasingly supported by a number of large donor agencies and indeed it can be seen as complementary to the increased role envisaged for markets in land tenure transactions. On the other hand, not everyone is convinced that markets offer a real solution to issues of water allocation and the likely need for some form of ongoing state approval, to prevent harm to the environment and third party rights holders, means that the relatively high transaction costs of individual trades will tend to negate some of the more optimistic claims for the power of markets. Yet to reduce the level of regulation would simply confirm other fears regarding speculation and hoarding. A full discussion of the claimed benefits and alleged disadvantages of tradable water rights is beyond the scope of this paper.
Nevertheless, at the present time, it seems quite likely that trade in water rights will increasingly be permitted by water legislation so as to provide existing right holders with an incentive to conserve water and to make more water available for cities in a way that does not involve the construction of expensive and often environmentally controversial dams, canals and pipelines. Ultimately, from a political perspective the introduction of tradable water rights will probably offer a relatively less painful and probably more efficient way of re-allocating water from less productive to more productive uses. So what will the implications be?
Trades in water rights will inevitably have impacts on land tenure rights, in respect of those land parcels from which water rights are lost, or traded away, as well as those from which they are not.
As regards land tenure rights of the first category, those that relate to a parcel of land from which water rights have been sold, the primary impact will most likely be a loss of amenity or value. In some senses this is not dissimilar to the situation when part of a land parcel is sold. Nevertheless, a number of additional questions will arise that also have implications for the land tenure rights that relate to that parcel. Take for example a parcel of irrigated land in respect of which all rights to water have been sold. How should the value of such land be assessed? It would not simply be a question of removing the value of the water as the effects of the loss of water on any irrigation infrastructure will also need to be considered. Without water, such infrastructure will arguably have no value. Similarly how would vineyards and fruit trees be valued without water? The assessed value of the land will depend on the answers to these questions. Similarly what would be the position as regards any taxes payable on the sale or purchase of the land? Should they take into account the loss of the water right? And how would the position of lenders be protected? The value to the mortgagee of a mortgage over irrigated land might be worth a lot more than a mortgage over land in respect of which no water rights exist. How would the mortgagee's rights be protected? Normally, a land tenure rights system contains sufficient safeguards to a mortgagor regarding subsequent transactions that have the effect of reducing the value of the land parcel, such as the sale of a portion of the land, by, for example, requiring the consent of the mortgagor. But how could appropriate notice be given in respect of a trade of a water right?
The question of land valuation could potentially have major impacts on the land tax receipts of local and municipal governments. When the advocates of tradable water rights talk in terms of them enabling water to be transferred from low value to high value uses, there is no doubt which shift in use they have in mind: from irrigation to municipal/commercial/industrial use. In one sense this is not surprising, given the fact that agriculture is by far the largest water use sector. There is, however, concern that towns and cities, being richer, will be in a position to buy up water rights for their own ends to the detriment of rural communities: without water agricultural production will fall, as will receipts from local land taxes meaning that communities would be unable to pay for essential social services.
Many rural communities have opposed transfers of water to outsiders in part on the assertion that such transfers can impose significant economic and social costs on the community. For example, where farming land is fallowed as a result of an external transfer, jobs can be lost, and tax revenue can decline (leading in turn to a reduction in governmental support services). In a community's worst nightmare, the social fabric of the community itself might begin to un-weave as local residents sensing economic trouble, leave the area.
The experience of Owens Valley in California, where the water rights were surreptitiously bought out by the growing and thirsty city of Los Angeles can only feed such fears. Apparently, the once fertile valley now has the appearance of the Bonneville Salt Flats and even hardy desert plants can no longer survive within it. This is no doubt a rather extreme example, but if the overall effect on the quantity of water used in agriculture would be small, it is conceded that local impacts may be significant. In short trades in water rights could have a major impact on the value of land tenure rights within a community. At present in those jurisdictions where trades are evaluated, this is generally by reference to the impact on other water rights and water resources rather than on other land tenure rights.
On the other hand, third party effects exist whenever significant resources are allocated or re-allocated or removed from a community, for example when a mine or factory closes down. Should water be treated any differently? Arguably it should, as it is a fundamental resource that once lost cannot be replaced. So should communities have a say in water rights trades, perhaps on the basis of a change of land use, from irrigated agricultural land to rain-fed agricultural land (or simply desert)? Should such approval be provided under land tenure legislation or land use planning legislation?
The widespread introduction of tradable water rights will also have a potentially significant impact on water rights themselves, particularly as regards their administration. As noted in the introduction, in many arid areas the value of land depends less on land tenure rights than on water rights. If tradable water rights are severed from the land the possibility arises of mortgaging them to raise credit. This would, however, raise a number of questions as to just how the mortgagee's interest should be protected. In short it might have a significant impact in the manner in which water rights registers are operated, either by including mortgage entries on the main register, thus effectively converting water rights registers into title registers or through the introduction of a Torrens type system of registered transactions rather than registered titles. Alternatively, would it be preferable to include such transactions on land tenure registers?
In conclusion, the issue of tradable water rights will raise a number of profound questions about land tenure rights, water rights and their inter-connection for which answers will be urgently needed. Further research in this area is called for, particularly as to the practical, as opposed to theoretical, aspects of water rights trades.
For the poor, securing access to land and water is often a key element of both survival and livelihood strategies. What impact does the interface between land tenure rights and water rights have on such strategies? Before considering the interface it is appropriate to consider the two resources separately.
Current livelihoods-based approaches to understanding poverty and the actions that can be taken to promote its reduction focus on improving the assets and capabilities of the poor. Since land is the primary means of both subsistence and income generation in rural economies, access to land and security of land rights are of primary concern to the eradication of poverty. In rural areas, land is a basic livelihood asset, the principal form of natural capital from which people produce food and earn a living. It is with this understanding that reforms seek to strengthen the land tenure rights of the poor, supported by a vast literature.
Generally speaking, the relationship between water rights, poverty and livelihoods has received less research. Many of the references in the literature regarding poverty and water relate to access to safe drinking water, which is not usually subject to individual water rights but more of a human right to water. Indeed, many uses of water by the very poor will frequently fall within the de minimis exceptions to the need to hold a formal water right. A first step in seeking to remedy this research gap might be to use a livelihoods approach to identify the extent to which water rights, or their absence, constrain poverty alleviation.
And while, as described above, poverty alleviation is frequently an objective of land tenure reform it is less often a priority of water rights reform. Indeed the lack of co-ordination between land reforms and water rights reforms may frequently lead to missed opportunities to ensure pro-poor outcomes, recent experiences of reform in Zimbabwe and Tanzania being cases in point. Nevertheless, co-ordinated reform of land tenure rights and water rights is possible if complex, one example being the Chilean reforms of the 1960s. More recently, the post-apartheid South African government has sought to co-ordinate water rights reform through the implementation of the recently enacted Water Act with the ongoing process of land reform as part of a programme to redress the inequitable apartheid era allocations of land and water.
Recognising that many of the poorer water users will fall within the de minimis exceptions to the need to hold a formal right, the South African Water Act creates a reserve for each surface watercourse that seeks to ensure that sufficient water is available for such uses. Another approach taken by the act is to provide for the establishment of suitable participatory mechanisms to ensure that the poor, along with other stakeholders, can participate in decision and policy making in connection with water resources management. Unfortunately, recent research suggests that notwithstanding the government's efforts, it is proving more difficult to include black communities in the former homelands who operate in the informal water sector in the reform process. Indeed the relatively complex institutional arrangements for water resources management, which frequently must take account of a state's obligations under international law, coupled with trends in the water sector such as the introduction of charging schemes seem almost by their very nature to militate against the interests of the poor.
Further research on the implementation of ongoing land and water reforms in Southern Africa is necessary to evaluate the extent to which they are capable of successfully taking an integrated approach to land and water reform. This is not an easy task and it is reported that both programmes are facing a rather uncertain future as a result of major in-competencies within government and community structures to make those ideas work and to get it implemented.
Finally, on the basis of the research undertaken in the preparation of this paper, apart from the relative lack of research into the relationship between water rights and poverty and livelihoods, it is striking that while gender issues have received considerable focus as regards land tenure rights research and reforms they appear to have received very little consideration as far as water rights are concerned. This too would appear to be an area calling out for more research.
 Gleick cited by
Salman, S. in Bogdanovic, S. (Ed) 2002. Legal Aspects of Sustainable Water
Resources Management YuAWL, Novi Sad.|
 Since 1940 irrigation has absorbed over half of the total amount invested in the agricultural sector in Pakistan and Indonesia and for China the figure is over 80 percent. In India 30 percent of all public investment has been in the irrigation sector.
 The World Bank, 1992. Development and the Environment, 1992 World Development Report, The World Bank, Washington D.C., p. 100.
 The World Bank, Water Resources Sector Strategy: Strategic Directions for World Bank Engagement, op cit., p. 20.
 Apart from the costs of operating a scheme, which may range from the operation of diversion structures, the costs of electricity when pumps are used, to the formation of temporary earth channels, a considerable amount of maintenance is invariably necessary to keep a scheme operational. Such maintenance may include the removal of silt and vegetation from canals through to the complete annual re-construction of earthen diversion structures in small schemes.
 Rameshur Pershaud Narain Singh v Koonj Behari Pattuk in 1878 cited in Howarth, W., op cit., p. 115.
 This is because when once appropriated, the existence of a property right in the water has the consequence that it is capable of being the subject of theft. Howarth, W., op cit., p. 14.
 Another complication for many irrigation systems in South Asia and parts of China is that they have absolutely no physical direct link with the types of water resources that are subject to water rights regimes. Instead monsoon rainwater is collected in reservoirs or tanks from which it is distributed through canals to irrigate crops during the dry season. Such irrigation systems can find themselves effectively beyond the scope of statutory water rights and the formal water management framework.
 The timing of a delivery of irrigation water is often as important as the volume: there is little purpose in irrigating crops that have withered due to lack of water.
 Such as the Subaks of the island of Bali in Indonesia.
 For example, the formal regulations for the operation and maintenance of the Benacher and Faitenar irrigation canals irrigators in Valencia, Spain were drawn up on 29 May 1435 (cited in Ostrom, E. 1990. Governing the Commons). These rules codified earlier customary rules that dated from many hundreds of years earlier. The statutory basis of the Dutch Waterschappe or Water Boards dates back to the twelfth century. See Hodgson, S. 2004. Legislation on Water User Organizations - a comparative analysis, Legislative Study No. 79, FAO, Rome.
 See further International Water Management Institute-Tata Water Policy Program Pro-Poor Irrigation Management Transfer, Water Policy Briefing No. 6, IWMI-Tata, Vallabh-Vidayanagar, 2003.
 Another issue that should be borne in mind is that, particularly in developing countries irrigation water may be used for a variety of purposes other than irrigation such as fish-farming. See Meinzen Dick, R. & van der Hoek, W. Multiple Uses of Water in Irrigation Areas draft of a paper to be published in a special issue of Irrigation and Drainage Systems IFPRI Washington D.C.
 Scott, A. & Coustalin, G., op cit., p. 940.
 Scott, A. & Coustalin, G., op cit., p. 941.
 Indeed in Spain the Water Law requires the establishment of a water user association if two or more people take water from the same source.
 Meinzen Dick R. & Bruns B.R. Negotiating Water Rights: Introduction in Bruns, B. & Meinzen Dick, R. 2000. Negotiating Water Rights ITDG Publishing, London, p. 37.
 In fact in accordance with the applicable legislation the Bureau may only supply water to water user associations as it is precluded from making supplies to individual water users.
 Clark, S., op cit., p. 34.
 Tiffen, M. 1992. Land Tenure Issues in Morris, J. & Thom (Eds) Irrigation in Africa West View Press, Boulder, p. 407.
 Mathieu, P., op cit., p. 66.
 Merrey, D.J. et al. 2002. Can irrigation management transfer revitalise African agriculture? in Hilmy, H. & Abernathy C., Private Irrigation in Sub-Saharan Africa IWMI, FAO & ACP-CTA, p. 98. and footnote.
 In other words a water user association that only holds a use right might rationally decide to defer maintenance on the basis that the ultimate owner, typically the state, will bear the eventual costs of renovation.
 Some states also oppose transfers in ownership on the basis that a transferred scheme could then be sold on to a third party, although generally speaking the market for second-hand irrigation schemes is rather thin.
 Huibers, F. 1999. Land tenure and irrigation development in the Senegal Valley African Studies Centre, Leiden.
 Barbara Van Koppen describes how Operation Riz projects in Burkina Faso took land and water rights from women, the primary rice producers, and re-allocated them to families, effectively to men. Van Koppen, B. Gender Water and Land Rights in Rice Valley Improvement, Burkina Faso in Bruns, B.R. & Meinzen Dick, R., op cit.
 This would also prevent the situation whereby the otherwise excellent paper on land tenure reforms in North India, referred to above, would, without any hydro-geological rationale, recommend increased use of groundwater for irrigation.
 Whether in the form of a collective farm (kolkhoz), state farm (sovkhoz) or cooperative etc.
 Article 163 Law on Real Estate.
 Foster in Salman, S. (Ed) 1999. Groundwater Legal and Policy Perspectives - Proceedings of a World Bank Seminar World Bank, Washington D.C. p. 15.
 FAO. 2003. Rethinking the approach to groundwater and food security. Water Report 24, Rome, p. 25.
 Mukherji, A. & Shah, T. Groundwater Governance in South Asia: Governing a Colossal Anarchy Water Policy Research Highlight No. 13 IWMI-TATA Vallabh Vidyanagar, p. 2.
 Muckjerji, A. & Shah, T., op cit., p. 2.
 Burke, J. & Moench, M.H. 2000. Groundwater and Society: Resources Tensions and Opportunities United Nations, New York, p. 22.
 While such trades are sometimes used, generally by economists, as evidence of the efficiency of markets for water, it should be noted that such transactions are quite different to markets in water rights, an issue which is considered in more detail below.
 Mukherji, A. & Shah, T., op cit., p. 5.
 Rema- Devi, P. 1991. Groundwater Development and Legal Regulation. 33 Journal of the Indian Law Institute 614, p. 619.
 See also the case mentioned in footnote 153 infra in which a large multi-national corporation is alleged to have deprived farmers of the groundwater on which they depend.
 Ballard v Tomlinson. 1885. 29 Ch. D. 115. Howarth, W., op cit., p. 14.
 An exception is made, under the common law, for underground water flowing in a defined channel in which case the riparian doctrine applies. Chasemore v Richards. 1859. 7 h.l. Cas. 349. See Howarth, W., op cit., p. 121.
 Howarth, W., op cit., p. 122.
 For example in Spain in 1985, Italy in 1994 and Morocco in 1995.
 For example Victoria, Australia in 1989.
 Burchi S. National Regulations for Groundwater: Options, Issues and Best Practices in Salman, S., op cit., p. 58.
 As a result of being subject to the same regime as rights over surface water resources, rights to groundwater are recorded in a register and are usually subject to the same charging regime.
 Additional provisions may also apply to the issue and operation of groundwater rights. For example, the legislation of the American state of Iowa restricts the term of the right to ten years if the aquifer capacity is uncertain.
 And as with rights relating to the use of surface water, undertaking an activity in circumstances where a licence or permit is required without such authorization or in breach of its conditions is punished in accordance with criminal or administrative law.
 See Drennan, J. 1997. Lassooing the Loophole: The Need to Rope in the Use of the Domestic Well Loophole by Subdividers in New Mexico, 37 Natural Resources Journal 923 and Caldwell, R.N. 1998. Six-Packs for Subdivisions: The Cumulative Effects of Washingtons Domestic Well Exemption, 28 Environmental Law 1056.
 Indeed it should be emphasized that groundwater resources and rivers are often intimately interlinked: the level of surface water abstraction can have a major impact on groundwater flows while at the same time excessive groundwater abstraction can impact on the flow of rivers.
 Blomquist, W., Heikkila, T. & Schlager, E. 2001. Institutions and Conjunctive Water Management among Three Western States 41 Natural Resources Journal 653, p. 654.
 By way of contrast a robber who uses a gun in the course of a robbery is likely to have that gun confiscated even if s/he lawfully owned it at the time of the offence. Similarly it is common for fisheries legislation to permit the confiscation of nets and gear as part of a punishment relating to the commission of a fisheries offence and similar provisions are found in other natural resource legislation.
 Shah, T., Makin, I. & Sakthivadivel, R., op cit., p. 11.
 Rema Devi, op cit., p. 619.
 Rema Devi, op cit., p. 634.
 The ability to sanction breaches of operating rules was shown by Ostrom to be one of the key design requirements of organizations like water user associations. See Ostrom, E., op cit., p. 94.
 Stephenson, K. 1996. Groundwater Management in Nebraska: Governing the Commons through Local Resource Districts 36 Natural Resources Journal 761. Ongoing research into community based groundwater management approaches in India funded by the United Kingdom of Great Britain and Northern Irelands Department for International Development may well shed valuable light on the merits and disadvantages of such an approach.
 This is not to suggest that all resource management decisions could be transferred down to the user level. Some form of aquifer-wide planning would be necessary for a start. Furthermore, as with surface water resources states have obligations under international law regarding their use of shared aquifers.
 Although, as will be seen below, attempts to bring customary rights within a formal rights regime may eventually entail this.
 Under the zamindari system in Northern and Eastern India revenue collectors (zamindars) received full rights to land subject to delivering a fixed amount of revenue to the colonial power. The users of the land became the tenants of the zamindars often paying their rent in the form of a share of the crop proceeds. On the other hand in peninsular (Southern) India the cultivator-owner (mahalwari) system was recognised, which by contrast, vested land rights in village bodies, essentially establishing individual land ownership by producers.
 Deninger, K., op cit., p. 2.
 Huggins, C., op cit., p. 17.
 Delville, P.L. 2000. Harmonising Formal Law and Customary Law Rights in French-Speaking West Africa in Toulmin, C. & Quan, J. (Eds.) Evolving land rights, policy and tenure in Africa DFID/IIED/NRI London, p. 99.
 Delville, P.L., op cit., p. 102.
 Also in those countries in which Islamic law was introduced, for example those of the Maghreb and Sahel, this did not cause the pre-existing customary rules to disappear. Instead such rules co-existed and indeed continue to co-exist with Islamic law. Bouderbala, N. Problématiques de la Transformation des Régimes Fonciers en Afrique Francophone: Reflexions Autour de Quelques Experieinces (Burundi, Guinee, Maritanie, Rwanda, p. 9 in Bouderbala, N., Cavérivière, M. & Ouderaogo, H. 1996. Tendences dévolution des legislations agrofoncières en Afrique francophone, Legislative Study No. 56, FAO, Rome.
 Mathieu, for example describes how in a North African irrigation scheme local rules gave priority to the original users of a state built irrigation scheme over subsequent users. This was contrary to the intentions of the relevant irrigation agency and had no basis in formal law. But it was one of the rules of the original irrigators. Mathieu, P. 2001. Water rights, investments and meanings: conflict and change in a traditional irrigation system in northern Morocco 1, 3/4 International Journal of Water 271.
 The flexibility of customary law is such that it is quite capable of re-inventing itself, as necessary, as well as adapting to new situations. Thus in Senegal, for example, share cropping arrangements are increasingly replacing labour tributes from landless farmers of slave castes. Tzeutschler, G.G.A. 1999. Growing Security: Land Rights and Agricultural Development in Northern Senegal, 43 Journal of African Law 36, p. 51. Tzeutschler also describes how higher castes have acquired a disproportionately large share of newly irrigated lands, contrary to the responsible agencys stated intentions as regards inter-caste equity (pp. 55 and 56).
 Woodhouse, P. African Enclosures: A Default Mode of Development 31 World Development, p. 1717.
 Du Bois, F., op cit.
 In Parts Two and Three.
 Historically, English rights of common are usually attached (appurtenant) to the dominant tenement (or land parcel) which they benefit. Confusingly, however, grazing rights can also exist in gross. In this event they are not attached to a specific tenement and are a separate property right belonging to the rights holder. The most important rights, in practice, are rights to graze, to take timber for fencing or bracken for animal bedding and the right to take turf or peat for fuel. All rights are now subject to registration under the Commons Registration Act 1965. Some 1.4 million acres of land are common land - over four percent of the land area of England and Wales. Sydenham, A. 2002. Managing Common Land for Environmental Benefit: the Difficulties After Bettinson v Langton, 4 Environmental Law Review 1. Many of the commons in England and Wales were seized by rich landowners in the eighteenth and early nineteenth century in a legal process known as enclosure. The former right holders, the commoners lost their rights to use the former commons in the process and many fell into poverty thus creating a new rural working class.
 Herrera, A., Riddell, J. & Toselli, P., op cit.
 Huggins, C., op cit., p. 12.
 Lane, C., op cit., p. 9.
 In the language of the new institutionalism such water resources would probably be considered to be common pool resources as they are public goods which are used simultaneously or sequentially by different users because of difficulties in claiming or enforcing rights, or because they are so sparse and uncertain that it is not worth doing. When the rights and duties of groups of users in relation to these resources are defined and enforced, the resources become common property resources, as distinct from open access resources. In the latter, which Hardin famously confused with common property, rights and duties are not well-defined, and a tragedy of over-exploitation is a possible, and likely, outcome. According to both legal and social science approaches to property, a fundamental feature of all regimes, including common property, is the possibility of excluding those without property rights. Cousins, B., op cit., p. 158 As regards the case of entitlements created on the basis of de minimis water rights, the problem is that no-one has any right to exclude anyone else. Consequently it is hard to argue that they can even amount to common property rights.
 This statement, which may appear controversial, was recognized by the courts in the eighteenth century in the case of Ashby v White, 92 Eng. Rep. 126 (K.B. 1703). In that case, which concerned voting rights the court held that want of right and want of remedy are reciprocal or without remedy there is no right. Scott and Coustalin, op cit., p. 824.
 Maina, cited in Cousins, B. Tenure and Common Property Resources in Africa in Toulmin, C. and Quan, J., p. 173.
 Bloch, P. & Foltz, J. 1999. Recent Tenure Reforms in the Sahel: Assessments and Suggestions for Redirection, Land Tenure Center, Madison.
 The Japanese River Law of 1896 provided that customary water use rights were to be deemed as permitted water rights. Samnbogi, K. Formation of case law and principles of watershed management in Bogdanovic, op cit., p. 409. In Indonesia, the 1974 Water Act and subsequent irrigation regulations recognise traditional system of irrigation water allocation and rights on the island of Bali. But that is in many ways an exception as Bali is after all an island.
 In a sense the integration of customary water rights regimes into a formal water rights regime would appear to be more complex given the constant inter-action between all water rights in a given river basin.
 Percy, D.R., op cit., p. 29.
 Louks 201. See also Briscoe, J. 1997. Managing Water as an Economic Good: Rules for Reformers, The World Bank, Washington D.C.
 See for example Dellapena, J.W. 2000. The Importance of Getting Names Right: The Myth of Markets for Water 25 William and Mary Environmental Law and Policy Review 317; and DuBois, op cit.
 Bauer, C.J., Slippery Property Rights: Multiple Water Uses and the Neo-liberal Model in Chile 1981 - 1995.
 And indeed such fears are to some extent confirmed by the Chilean experience as regards non-consumptive water rights. See Bauer C.J., op cit.
 Bjornland, H. & OCallaghan, B. 2003. Property Implications of the Separation of Land and Water Rights, Paper presented at the North Annual Pacific-Rim Real Estate Society Conference, Brisbane, Queensland 19 - 22 January 2003.
 Thompson, B.H. 1998. Water Law as a Pragmatic Exercise: Professor Joseph Saxs Water Scholarship, 25 Ecology Law Quarterly 363, p. 380.
 Reisner, M. (2001) Cadillac Desert: the American West and its disappearing water Random House, London, p. 101.
 Briscoe, J., op cit., p. 11.
 Thompson, B.H., op cit, p. 381.
 Bjornland, H. & OCallaghan, B., op cit., p. 13.
 Clark, S.D. op cit, p. 38.
 Soussan, J. & Chadwick, M. Legal, Institutional and Policy Issues Related to Water Resources Stockholm Environment Institute at York University, p. 1.
 Soussan, J. & Chadwick, M., op cit., p. 4.
 See Deininger, K., op cit.
 In Uganda, for example, following the introduction of the 1995 Water Statute only some two hundred water users would require a formal water abstraction right. Garduno, H., op cit., p. 70.
 The Chilean agrarian reform process of the 1960s sought to change the old agricultural structures characterized by the existence of huge, partially exploited estates accounting for 80 percent of the total arable land in the country, and the smallholdings owned by 80 percent of the peasant population accounting for 15 percent of the total arable land area. The Agrarian Reform Law was supposed to make it possible to expropriate agricultural lands that were not being properly exploited, subject to certain conditions, for re-distribution among the poor farmers who were normally farm labourers, sharecroppers and smallholders. Land was not the only subject of the reform - the Agrarian Reform Act comprised 336 sections of which 85 were devoted to modifying existing legislation on water. Manríquez Lobos, G. Transferability of Water Use Rights and the Case of Transitional Legislation in Chile in FAO Issues in water law reform, op cit., p. 62.
 See page 29 above.
 The Reserve consists of two parts: (1) the basic human needs reserve, which includes water for drinking, food preparation and personal hygiene; and (2) the ecological reserve, which must be determined for all or part of a water resource such as rivers, streams, wetlands, lakes, estuaries and groundwater.
 Catchment Management Agency (CMA) Formation in Olifants, South Africa, Merrey (2000:9) writes: ... rural communities were unaware of the provisions of the new water law and the CMA process, despite the efforts to inform people and offer them opportunities to express their views. Small-scale farmers had not heard about the CMA.. [But] the Irrigation Boards providing water to large commercial farmers were participating actively in the process.. A small number of large stakeholders is easy to work with; the ballgame changes fundamentally once we have to deal with a huge number of tiny stakeholders. Shah, T. et al., op cit., p. 10.
 Kirsten et al., p. 23.
 The work of Van Koppen, op cit. was concerned primarily with rights to irrigation water rather than formal water rights as the term is used in this paper.