Assessment of surface water management institutions: a case of public irrigation schemes in northern Algeria

The pressure on water resources that Algeria encounters is due to the scarcity and the water demand increase by different sectors of activity. This induces managers to reconsider the traditional supply management approach. The new water policies aim to adopt water demand management by improving the efficiency of its use in the agricultural sector, which consumes more than 59% of freshwater withdrawals. Economic, regulatory, and voluntary instruments have been adopted for better governance of water resources. This study aims to (i) define institutional arrangements on the exploitation of surface water in large irrigation schemes in Algeria and (ii) assess the effectiveness of the current water management instruments. Findings from farmers’ and water distribution agencies’ surveys suggest that objectives expected by the application of the different economic instruments are far from being achieved. Moreover, the absence of effective executive agencies for the implementation of regulations suggests that an improvement in the performance of public administration is necessary to achieve better governance of irrigation water in Algeria. At the local level, and despite a lack of farmers’ will to formally engage in a collective of farmers, they organize themselves around informal arrangements to overcome the challenges associated with water scarcity.


INTRODUCTION
The water sector in North Africa has traditionally been and is still largely driven by investments in water infrastructures to increase water supply (Frija, 2009). In Algeria, the water sector is subject to a particular attention by public authorities, which allocate increasingly important financing resources. Then, to maintain water supply and increase resource availability, the government has implemented several projects since the beginning of 2000. Thus, many efforts have been made to build new storage capacities for surface water (Benblidia & Thivet, 2010;Drouiche et al., 2012). At the same time, a regional transfer program, which aims to ensure more equity and reduce geographical disparities for access to water, has been progressively implemented. These water transfers from the most watered areas are in line with the objectives of the country's food security strategy, which aims to promote regions with high agricultural potential, such as the high plateaus (Mozas & Ghosn, 2013).
It is true that the policy of improving supply and increasing water availability has served its historic function to face water needs and to develop the agricultural sector. 1 However, this supply strategy has reached its limits, resulting in increasing pressure on the water resource (Zeggagh et al., 2010). In addition, it is very likely that, in the near future, water will need to be diverted from irrigation to meet the needs of urban and industrial sectors. Indeed, the water code in Algeria stipulates that priority is given to urban freshwater supply (Benblidia & Thivet, 2010).
Given the importance of irrigation for food production, this must occur without compromising agricultural growth. In this sense, improvements in the irrigation sector are needed at the technical, managerial, and institutional levels to increase the efficiency of water use (Inocencio et al., 2002;Rosegrant et al., 2002;Wichelns, 2004;Boyer et al., 2011). Indeed, as stated by El Alaoui (2006), the mobilization of water, its exploitation, and its rational management require not only hydraulic infrastructures but also an overall legal and institutional framework that has equal importance. In response to this, the Algerian government has undertaken a reform process since 2005. Its objective is to restructure water institutions. These reforms include the establishment of economic and regulatory instruments to manage water as well as the decentralization of resource management and the integration of users in water management at the irrigated schemes level.
After a brief presentation of the physical situation and a description of the new institutional framework of the water sector in Algeria, this article aims to study the performance of water management institutions based on their ability to achieve the expected goals. To do so, four indicators are chosen: (1) impact on water availability and reduce pressure on the resource, (2) impact on water productivity, (3) acceptability, and (4) financial viability. Based on the revealed weaknesses and strengths, recommendations are derived to improve the efficiency of irrigation water institutions in Algeria.
The remainder of this paper is organized as follows. The next section presents a literature review of water management institutions. Then, we present the physical and organizational context of the water sector in Algeria. A methodological section presents our case study, and the analytical framework used to analyze water governance performance in Algeria. Finally, the results are presented and discussed in the 'Results and discussion' section.
In the literature, several studies have focused on the construction of analytical frameworks for the assessment of water governance, including the work carried out by the OECD (2011), Huntjens et al. (2011), and Pahl-Wostl & Lebel (2010. However, in this study, we will mainly rely on the framework developed by Wijnen et al. (2012), which itself is a hybrid framework based on the works of the authors previously cited. The study of Wijnen et al. (2012) decomposes water governance into three levels. The first is the national level (or policy level), in which objectives in relation to water are defined, integrated with land and environmental policies, and aligned them with other related policies affecting water (e.g. agricultural policies). The main objectives relate most often to sustainability in quantity and quality, efficiency in allocation and use, and equity by ensuring fair access and protection of water rights. The second is the strategic level, in which a nation sets up institutions and management instruments to align stakeholder behavior and expected outcomes with water policy objectives. Management instruments in this level can be divided into two groups: regulatory (including codes to access water and water rights) and economic instruments (including sanctions and financial incentives, especially pricing, water rights transferability, subsidies, and taxes) (Shah, 2005;Foster et al., 2010;Theesfeld, 2010). Finally, governance at the local level implies formal and informal organizations and institutions. This level includes voluntary policy instruments, also known as community participation instruments. They are instruments that motivate voluntary actions or behavioral changes without the application of direct incentive tools (Foster et al., 2010;Theesfeld, 2010). To achieve the goal of affecting the entire water governance structure, these three types of instruments are ideally combined (Stone, 2002). On the other hand, institutions must evolve at the same time as the context in which they evolve (Saleth & Dinar, 1999). The idea of the institutional change as a source of economic growth is motivated by the need for continuous adaptation to deal with the changing environment (Matthews, 1986). This adaptation is reflected in terms of legal, political, and administrative reforms (Saleth & Dinar, 2004).

Outcomes of weak governance
It is well known that weak governance and poor water management will have a negative impact on different levels, including environmental and socioeconomic development of local communities. In a comparative study of irrigation water management institutions in several countries, Saleth & Dinar (1999) identified some similarities on the water sector and institutional order consequences resulting from weak governance. According to these authors, increase in water scarcity, degradation of water quality, conflicts between different sectors in terms of water allocation, and low-cost recovery are the most recurring problems. While these problems are rather too obvious, their origin and interdependence are of particular importance to formulate specific policies for each case. For example, water scarcity is not specific to countries that have depleted or are close to the depletion of their natural resources (cases of Algeria or Qatar with availability less than 500 m 3 /capita/year (FAOSTAT, 2015)). The shortage also affects countries with excess water potential. In this case, the shortage is caused either by acute spatial imbalances in supply (e.g. Australia, Brazil, and China), or because of poor water quality (Australia, Brazil, China, and South Africa) (Saleth & Dinar, 1999).

Criteria to evaluate the performance of water management instruments
In the literature, the assessment of government programs or natural resource institutions is defined as a review of their relative impacts on their specific objectives (Nadeau, 1988;Dore, 2007). Indeed, Pahl-Wostl & Lebel (2010) claims that a governance system should achieve its stated objectives. Failure of doing so is an indicator of nonsatisfactory performance without alluding to any normative claims. This definition suggests that the criteria for evaluating the water management instruments must be defined in relation to the objectives expected by their application. For example, to assess groundwater policies in France, Lenouvel & Montginoul (2010) selected five indicators: (1) abatement level of the aquifer, (2) impact of the instrument on aggregate income, (3) acceptability, (4) predictability, and (5) durability of the instruments. Other authors, such as Faysse et al. (2011), used three criteria to evaluate different types of groundwater instruments applied in selected North African countries: (1) increase of water resource availability, (2) reduction of aquifer withdrawals, and (3) increase of water productivity. Otherwise, in this study, Frija et al. (2015) used four criteria to evaluate the performance of groundwater management instruments in Tunisia: (1) the acceptability of the instrument, (2) the implementation degree and the implementation cost, (3) the impact on water productivity, and (4) the impact on the aquifer withdrawals.

Physical context
The renewable water resource in the country is estimated to be around 12.2 billion m 3 /year (23% for groundwater, representing 2.8 billion m 3 , of which 800 million m 3 in the south of the country, and 77% for surface water, i.e. 9.4 billion m 3 ) (Drouiche et al., 2012;Hamiche et al., 2015). Furthermore, there are large quantities of groundwater in the south of the country (from the intercalary continental and the terminal complex) estimated at more than 50,000 billion m 3 . In recent years, the potential of water resources per capita and per year has declined considerably. Water availability per capita declined from an average of 1,500 m 3 /capita/year in 1962 to 630 m 3 /capita/year in 1998 and to 292 m 3 /capita/year in 2014 (FAOSTAT, 2015). This ratio is low compared to other Mediterranean countries such as Morocco (879 m 3 /capita/year) and Tunisia (420 m 3 /capita/year) in 2013 (FAOSTAT, 2015). This situation places Algeria among countries that live below water stress, with an average per capita resource much lower than the World Bank's scarcity threshold of 1,000 m 3 /capita/year (World Bank, 2007). To deal with this situation, the use of non-conventional water has recently received an interest from public authorities, as it can provide a significant water potential for Algeria. It involves the production of freshwater by desalination of seawater or by demineralization of brackish water, artificial recharge of groundwater, and wastewater treatment (Hamiche et al., 2015). These options offer the opportunity to reduce or, at least, conserve conventional water resources by increasing water availability for the agricultural sector. Indeed, with 11 large seawater desalination plants all over the Algerian coastline, the production in 2021 is 561 million m 3 /year, producing 17% of the amount of drinking water consumed at the national level (Titouche, 2021). However, this solution has its limitations, especially the high costs for its mobilization. In this logic, studies conducted in Algeria by Akli et al. (2019) have shown that, to enhance water availability, there are alternatives and more cost-efficient solutions than seawater desalinization, in particular, the subsidy of water-efficient technologies and the reuse of treated wastewater. Concerning this later, Algeria can derive a significant water potential from domestic wastewater. Its volume is estimated to be 1,100 million m 3 in 2014 (FAOSTAT, 2015).

Historical and organizational contexts
In the last decades, the government undertook a series of institutional processes and reforms. The resulting water policy is based on the principles of decentralization, integrated management and user involvement, and the search for the financial balance and the conservation of the existing structures.
Until 1984, large irrigated schemes 2 were managed by a central public administration, i.e. the local departments of hydraulic services (DHW 3 ). However, because of technical and financial disadvantages, the management of irrigated schemes was granted to 11 irrigation agencies (or OPI 4 ) and was financed by water sales and government subsidies. These OPIs marked the transition from sectoral management to concerted management at the regional and national levels. In 1996, the principle of cost recovery was introduced, and the price of water increased in 1998 and 2005. In 2005, these irrigation agencies restructured into a single irrigation management body, the National Office for Irrigation and Drainage (ONID), which was established as a financially independent entity. This institutional change was also designed with the principle of decentralization and local management to facilitate the consultation and participation of stakeholders (administrations and water users). In 2000, the other institutional reform was the creation of a specific water resources ministry, which is the central authority responsible for the development and implementation of the national water management policy. This management is currently part of the water law (Law 0512 of 4 August 2005). This law has established a National Advisory Council for Water Resources, in which relations of consultation and coordination with the various economic sectors and, more generally, all users must be organized and developed.

METHODOLOGY Study area: Jijel-Taher and Relizane-Mina irrigated schemes
Our case study concerns two irrigation schemes. The first one is located in the western part of Relizane Wilaya (Mina scheme), and the second is located in the north of Jijel Wilaya (Taher scheme). The Mina scheme is an old large public scheme created in 1936. Its irrigated area is about 17,235 ha. It occupies a particularly important place in the economic life of the western part of Algeria through its livestock farming vocation. The study area has a Mediterranean semi-arid climate, with an annual evapotranspiration of 1,506 mm and an average annual rainfall (very irregular from 1 year to the next) of 253 mm. It is characterized by a dry period of 7 months which spreads from March to October, during which the demand for water is higher. The surface irrigation water comes from the Sidi M'Hamed Ben Aouda dam, with a capacity of 225 million m 3 . This dam allows the irrigation of the two parts of the scheme, the right and the left banks, by gravity. The conveyance is carried out by two main open-air canals. They are then connected to secondary open canals of different diameters. Finally, farmers' fields are organized around secondary canals. The flow rate is constant in primary canals throughout the irrigation period to ensure that the water reaches all terminals. In total absence of water counters at the farm level, the terminals are the last place for water control. These are equipped with valves where flow is adjustable through regulatory boards called 'mask module'. Each volume indicated on the contract corresponds to a mask.
Concerning Taher scheme, it is a new scheme initiated in 2010 with an area of 4,885 ha. It is characterized by a large number of small farms which mainly produce water-intensive crops within greenhouses (mainly tomato, pepper, and strawberry) and open field cropping (mainly cauliflower, watermelon, and seasonal tomato). Unlike Relizane, Jijel is a humid region mainly in winter. Rainfall is abundant with 1,200 mm/year on average. The scheme is water supplied by the El-Agrem dam with a capacity of 34 million m 3 , intended exclusively for irrigation. This dam allows to irrigate the first half of the scheme (i.e. 2,498 ha) by closed canals and by gravity. In 2019, there is still no water conveyance system in the second part of the scheme.
As with all large irrigation schemes in Algeria, the ONID is responsible to allocate and distribute water to farmers. For the Mina scheme, public officers open valves according to the timing indicated on contracts (duration, flow rate). As the amount of water is not sufficient to irrigate the entire scheme simultaneously, the total water flow is divided into sequential turns. Moreover, in the case of the Taher scheme, water turns do not exist, and farmers can irrigate at the same time during water releases.

Analytical framework for water governance in Algeria
The analytical framework used to describe and analyse water governance in our case study is based on the framework developed by Wijnen et al. (2012). Then, the institutional water management instruments in Algeria will be analyzed, and their level of application will be examined critically over the three levels of governance as stated by Wijnen et al. (2012).
As with Nadeau (1988), Dore (2007), and Pahl-Wostl & Lebel (2010), and in order to analyze the performance of these institutional instruments, the approach used in this work is to examine whether the various instruments put in place allow to achieve the stated objectives. For our case study, we will build on the objectives related to the Algerian water law of 2005. The latter stipulates that irrigation water management instruments must achieve the following objectives: reduce waste, water valuation by appropriate processes and equipment, and the consideration of the real costs of water supply services and the recovery of these costs. Then, for each objective, we selected an indicator that can assess the performance of the institutional instrument implemented (Table 1). Four indicators were developed. These indicators are (1) the impact of instruments on enhancing water availability to agriculture and reducing water pressure, (2) the impact on water productivity at the farm level, (3) cost recovery and financial viability, and (4) the implementation level and acceptability. The acceptability of the instruments was introduced in the analysis, because, similarly to Lenouvel & Montginoul (2010) and Frija et al. (2015), we thought that the latter represents an important criterion for good governance. Table 1 also presents a description on how these criteria are assessed.

Data sources
Information used in this research comes from several sources. First, the applied institutional water management instruments and those that are under discussion have been documented and defined. Information on these instruments was obtained from official documents and departmental reports, and supplemented by interviews with irrigation water managers, including the ONID and the National Agency for Hydraulic Resources (ANRH). These interviews focused on managers' perceptions of the results achieved by the application of these instruments as well as the main constraints of their implementation. Secondly, at the local level, data from farmers' surveys at the level of the two irrigated schemes are collected. The interviews aimed to collect data on farmers' perceptions about the institutional instruments, and farmers had to describe the effects of these instruments at the irrigated scheme and the effects on their farming activities. Surveys also aimed to collect information on farmers'

RESULTS AND DISCUSSION: STRENGTH, WEAKNESSES, AND ACTIONS TO IMPROVE WATER GOVERNANCE
At the national level

Alignment of irrigation water policies with water and agricultural sectors
In Algeria, several sectors compete for surface water. This later accounts for 57% of freshwater withdrawals, including all sectors of activity (drinking water, irrigation, and industry), of which one-third of it is devoted to irrigation. In practice, it is the institutional reforms, since 2000, that have eased the alignment of irrigation water policies with the other water policies. Indeed, prior to the 2000 reform, policies were characterized by a predominantly vertical sectoral approach, characterized by shortcomings in horizontal coordination between the water sector and other sectors, including the irrigated sector. In fact, it was noted that the water management and irrigation water was under the responsibility of two different authorities. Water was under the responsibility of the Ministry of Water Resources, while irrigation water was under the jurisdiction of the Ministry of Agriculture. However, since 2000, the water sector (including irrigation) is under the responsibility of the same ministry (i.e. Ministry of Water Resources). Furthermore, there is an opposition between the water policy and the agricultural sector policy. Indeed, when the first seeks to conserve and reduce the pressure on the resource caused by the agricultural sector, the second encourages investment by providing financial support to produce water-intensive crops such as bananas, particularly in Jijel.

At the strategic level
In Algeria, and since 2000, a vast program of reforms in the irrigation water sector has been carried out by the public authorities, and significant institutional changes have taken place. However, from our case study, some notable failures of water policies and institutions in Algeria have been identified, including (1) weak enforcement of the instrument, (2) weak involvement of local authorities, and (3) high cost of implementation (Table 2).

Weak enforcement of regulatory instruments
Regarding regulatory instruments, water right is the most important. Water right in the study areas refers to the right to use a certain quantity of the available water flow in the canals during a certain time. Before the beginning of the irrigation season, farmers require authorization and must be contracted to ONID. Each contracted farmer must provide necessary information to plan water shifts (timing, cropped area, and crop nature). The second regulatory instrument is compliance monitoring. Water managers have to curb water transgressions and ensure the protection of all water resources. They would ensure compliance with water legislation and enforcement where non-compliance is identified. However, in the study area, there is no effective and efficient executive agency for the enforcement of the rules, even though a water police (whose role is to ensure the enforcement of water laws) exists and is required by the water law of 2005. Through surveys at the Mina scheme, we note that 30% of farmers make false declarations about their cropped areas and on the type of crops to obtain more water, and no verifications of the veracity of farmers' declarations are made. In the same scheme, terminal operators ensure that water turns are respected, but their role in monitoring against opportunistic behavior (illegal pumping directly from distribution canals or from the wadi) is not effective. Indeed, it is informally recognized by irrigators that they can behave in this way without incurring any sanction. In addition, the absence of conflict resolution institutions further encourages this type of behavior. In the Taher scheme, water distribution is made through closed channels, and as a result, problems linked to water diversion from the distribution canal are avoided. However, pumping water from nearby wadi is a common practice. Within this situation, further efforts to address this problem must be made. The solution is either to invest more in promoting self-regulation or to invest in strengthening regional and local government capacity (strengthening control and sanctions) (Shah, 2005;Foster et al., 2010). Of course, the best is the situation where both options are combined (Frija et al., 2014).

Inefficient economic instruments
With the new 2005 water code, pricing is recognized as an important tool for conserving water, improving the efficiency of its use, and ensuring the sustainability of irrigation infrastructures. Since 2005, water price is binomial for the two schemes with a fixed fee of 400 DZD 5 in liters/s/ha (or 2.96 US$) and a variable charge of 2.5 DZD/m 3 (or 0.018 US$) reflecting the consumed quantity. However, water fees due by farmers do not even cover the variable costs of mobilizing water and maintaining the delivery system. Indeed, by comparing the cost of the resource (estimated at 21.97 and 4.94 DA/m 3 , respectively, for Taher and Mina schemes (ONID 2016) with the amount of fees due by farmers (2.5 DZD/m 3 ), the numbers show that the latter only covers 11 and 50% of water cost, respectively, in Taher and Mina schemes. This allows us to claim that tariff levels fixed by decree and applied are far from meeting these requirements of charges balance in the studied schemes. On the other hand, this low price has repercussions on farmers' consumption level because it does not encourage users to use water rationally (Oulmane et al., 2019). The enforcement of such a low price can be explained by the desire to improve farmers' incomes and to create jobs to limit the rural exodus, but also by the lack of information on the potential impact of a more rigorous pricing policy on farmers. Then, further studies must be carried out to help policymakers to establish appropriate pricing policies, especially since our studies also revealed that 41% of farmers are willing to pay more for water. In this context, Azzi et al. (2018) showed that farmers in the Mitidja-irrigated scheme are ready to pay 64% greater than the current price while the current water supply service being maintained, as it is now in terms of quantity, quality, and reliability. With higher willingness to pay, there is more scope for policymakers to raise the price of irrigation water and to achieve better rates of cost recovery. In addition to water pricing, other economic instruments have been implemented to reduce agricultural water consumption. The most important measure is water quotas. It consists of defining, at the beginning of the season, a total volume that can be withdrawn from the dams and distributed it to farmers. However, this policy is applied within the Mina scheme, while it is not in the Taher scheme. This difference is due to the difference in water availability in both regions. Indeed, according to our interviews with water managers, this quota policy is not applied in the Jijel region because the water availability in the dam is greater than the amount consumed by farmers annually. Based on survey data, we recorded that consumed water quantities (3,460 m 3 /ha on average) exceed the quota on the contract in the Taher scheme (3,000 m 3 /ha on average). Farmers in the Taher scheme declare that they do not accept the application of more stringent economic instruments. In fact, such measures will limit their water consumption, which could have repercussions on their production systems and consequently on their income (Oulmane et al., 2019). Furthermore, 35% of the farmers surveyed in the Taher scheme say that if the state applies the water quota policy, these will switch to groundwater use and will deepen their wells for a better flow. Indeed, in Taher's perimeter, 77% of farmers have access to groundwater by individual or collective wells where water is at 6-15 m depth.
Moreover, in the Mina area, and in order to better plan the water turns, quotas are not set at the beginning of the irrigation companion. Farmers can make seven to eight applications (spaced by at least 21 days) throughout the year. According to our estimates, the amount of water released by application is calculated on the basis of 900 m 3 /ha on average. However, this amount of water is insufficient to meet the needs of farmers. On top of that, deliveries do not occur on time in comparison to the dates ordered by farmers, which results in off-peak periods. For this reason, 52% of farmers use groundwater (which is at a depth of 50-60 m) to supplement and meet the needs at the farm level. This resource, despite representing only 8% for these farmers, remains essential for crop diversification.
Another economic instrument put in place by the government is the water-saving technologies subsidy. Since 2000, the Ministry of Agriculture, within the National Program for Agricultural and Rural Development (PNDAR 6 ), has been conducting a water conservation policy to reduce water use and enhance water productivity, through aids and subsidies actions to adopt water-saving technologies (sprinkler and drip irrigation systems). For example, the government has taken the lead in developing irrigation systems with an investment subsidy of 100% initially (in 2000). It decreased to 30% since 2005. Regarding the subsidy policy for water-saving technologies, the results of this policy in terms of equipped areas are very different from one region to another. In the Mina scheme, for example, only 37% of the irrigated area is equipped with these techniques, on the other hand, it is 69% for the Taher scheme. In the latter, farmers are adopting this technology despite the high cost of investment. According to Taher farmers' statements, drip irrigation is mainly used to intensify existing cropping systems, reduce labor need, and facilitate fertigation and improve yields rather than to reduce the need for water.
In the Mina scheme, farmers declare that the low equipment rate is mainly due to lack of technical know-how and training about the use of this technology, water quality (flow of water through open channels carries impurities that clog the drip irrigation tape), and the high cost of investment (the subsidy is conditioned by the payment of part of the investment cost, which is an obstacle for the poorest farmers), also because of the difficulty of accessing credit (cumbersome administrative procedure). In addition, tenants 7 are not recognized as eligible (by the public administration) for these grants and therefore do not have access to subsidized inputs. From the surveys, we found that only 2% of farmers have benefited from subsidies for this equipment. However, it is uncertain how much water the introduction of drip irrigation will save. In other countries with similar situations, water-saving techniques have led to an increase in water abstractions (Foster et al., 2010).
In recent years, there is a debate about the use of treated wastewater to meet the growing water needs of the agricultural sector (Hannachi et al., 2013;Elmeddahi et al., 2016). The latter has long been hampered due to the obsolescence of the country's treatment plants. However, in the new water policy, it has become a priority and investments have been made for the rehabilitation of old stations and in the construction of new ones. Despite the existence of a wastewater treatment plant in the Jijel region with a capacity of 30.000 m 3 /day and three in Relizane with a capacity of 8,900 m 3 /day, treated water produced is not used by irrigation. The government must, therefore, pay more attention to this issue especially during this period of water stress that Algeria is encountered. Concretely, this policy could be beneficial to reduce the pressure on traditional water resources by producing additional quantities of water for agriculture while increasing agricultural production by providing the fertilizer contained in these waters (Al-Shammari et al., 2013).

Low integration levels of water users
Up to now, many professional organizations have been responsible for representing farmers with resource managers (producer associations, water users associations, etc.). However, we found that the principle of participatory management in relation to the allocation of irrigation water is weakly adopted and that farmers do not find any incentive to consider the management of schemes, because the management of large irrigation schemes is dependent on ONID. Indeed, this latter is responsible for the distribution of water and the maintenance of the distribution network. This puts farmers in a complex situation. On the first hand, farmers continue to depend on a water management body, most often state, on the other hand, they are called to participate in associations of water users and cooperatives by agricultural laws decreed by the government. In addition, according to interviews, farmers do not take personal initiatives and denounce the absence of leaders to preside associations. This low ability to establish associations undermines the negotiation power of farmers with other organizations. For example, in the Taher scheme, this situation makes it difficult to plan water releases. Indeed, because of this situation, bargaining power with the ONID is low for farmers, especially in high-demand periods.

Voluntary instruments: the emergence of informal institutions
One year after the institutional reform for the decentralization of water management initiated in 2005, Algeria has launched the principle of integrated irrigation management that encourages the involvement and the empowerment of users in resource management. 8 Farmer relations on water resource use can be formal, especially through the creation of agricultural water user associations as part of participatory irrigation management policies, whose main mission is to facilitate communication between farmers and irrigation authority. In addition to formal arrangements, and being faced with multiple constraints to access to irrigation water, farmers continue to operate around informal arrangements. Water-related arrangements deal with the informal sharing of dam water between farmers. In the field, we recorded significant differences from one scheme to another. In the Mina scheme, two or more farmers agree to exchange their water turns. The purpose of this operation is to accumulate several turns when the need arises. For example, during the planting season, when a farmer needs to irrigate successive days. He will have to lend another farmer's water turn to meet his needs. In the Taher scheme, arrangements on irrigation water occur around water counters. In fact, while each farm has a water counter, farm divisibility and informal land tenure lead to the situation where several farmers share the same water counter. Sharing water bill, besides the divisibility of farmlands, is most often according to the irrigated area and does not consider crops nature or irrigation technique.

CONCLUSION
This study explores and analyses the performance of water management instruments at the different governance levels in two public irrigation schemes in Algeria (Taher and Mina).
We showed that the government has made significant efforts to ensure the coherence between the objectives set for the development and the protection of water and the irrigation water policies. This has resulted in a decentralization of water management and the implementation of several institutional instruments. However, this study highlighted some shortcomings reflecting lower performances of these instruments. Indeed, it appears that, despite the enforcement of the economic instruments, these latter do not achieve the desired objectives of cost recovery and reducing the pressure on the resource. Irrigation water remains highly subsidized, and the current price covers only a small part of the operational costs. This price does not encourage farmers to use water in a rational way, affects the financial viability, and leads to poor maintenance resulting in infrastructure deterioration.
Moreover, the study shows that quota policy is leading to greater pressure on the groundwater resource (which is in free access in both irrigation schemes). A better consideration of groundwater and the application of appropriate policies for optimizing its conjunctive use with surface water are then necessary. Furthermore, farmers face multiple constraints on access to grants (especially for the acquisition of water-saving technologies) due to cumbersome administrative procedures and to the lack of recognition of tenants as eligible. The government must then interfere in order to secure them. In the case of treated wastewater, more attention by public policy must be paid to this process, since this technology not only allows allocating new quantities of water to the agricultural sector but also makes it possible to face environmental pollution problems. The study also shows that regulatory instruments exist but their enforcement is not effective. Local actors interviewed claimed that, except in rare cases, illegal practices are not severely punished by the administration, and the government remains tolerant, which further encourages this kind of behavior. This is mainly caused by the absence of an effective executive agency for the enforcement of the regulations.
Furthermore, the participatory approach is not sufficiently adopted by public managers and farmers. It has been affirmed that as long as water governance remains in the hands of the central administration, and as long as farmers find informal alternatives, farmers will not be devoted to the efficient water use and management of public schemes. In addition, there is no ambition on how to boost water users' association activity, or other ways of including farmers in decision-making for the management of large irrigation schemes. It is then clear that a more rigorous enforcement of regulatory instruments in the irrigation water sector and a better control at the distribution networks level are needed.
Finally, the lack of political determination to adopt a more rigorous enforcement of the institutional instruments is partly due to lack of political awareness. In our sense, this is due to failures in providing adequate information on the current situation to decision-makers. Then, raising awareness among decision-makers to adopt a sustainable water-use strategy is an important step in improving water policy and governance.

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.