A Benchmark to Analyze On-Field Performances: A Case From an Irrigation Scheme in Tunisia

Irrigation systems entail interacting processes that should be considered when analyzing the performances of irrigated areas. When analyzing the cause of performance gaps, one may consider the effect of one factor without taking into account the effect of other influential ones. This study addresses the performance of irrigated areas using a global approach. The analysis includes different factors of the production process and examines the relationship between irrigation and the economic performances of farmers’ irrigated fields. Results showed that technical efficiency (TE), considered as our economic indicator, evaluating the degree to which the inputs are used efficiently, was about 0.85 for fruit orchards, tomato, wheat and 0.66 for olive trees. The on-farm water distribution efficiency (Efarm) that evaluates water lost during its transport to plots was poor and can decrease to 36%. A great potential for improving water management exists. Efarm was not a significant factor for TE. So, Efarm causing substantial water waste has no significant impact on economic performance, the main concern of the farmer. However the easy access to water was a significant factor for TE (p-value < 0.032 in all the cases). The easy access to water is a possible lever for improvement. We pointed out that irrigation performance in the studied area has no significant impact on economic performance. There is a divergence between the farmer’s interest (the economic efficiency) and the community’s objective to save water through better irrigation performance. Government efforts to provide incentives for farmers for better water management seem to have born no fruit. This study argues for the use of a bechmarking in building global representation adapted to the actual local context. The analysis approch suggests that more attention should be paid to the “water saving program” of Tunisia, by subsidising irrigation investments in a better way and focusing on subsidies that create the most of economic growth.


Introduction
The evaluation of the irrigation performance is an essential activity to improve performance at plot, farm and scheme scale (Molden & Sakthivadivel, 1999). Irrigation engineers have usually focused on a set of performance indicators related to application efficiency, on-farm distribution efficiency and uniformity of water distribution (Pereira, 1999;Horst, Shamutalov, Pereira, & Gonçalves, 2005;Pereira, Dong, Mao, & Fang, 2007;Juana, Rodriguez-Sinobas, Sànchez, & Losada, 2007;Francis, Noble, Dennis, Rickerby, & Knight, 2017), and sometimes extended to equity, flexibility, sustainability and productivity (Lorite, Mateo, & Fereres, 2004;Gorantiwar & Smout, 2005;Vandersypen, Bengaly, Keita, Sidibe, Raes, & Jamin, 2006). Precise water use concepts and performance descriptors were proposed by Pereira, Cordery, and Iacovides (2011) in order to distinguish between water that can be reused and the actual lost water from a defined system by introducing the concepts of beneficial and non-beneficial water uses.These highly precise indicators take into account only variables and constraints related to irrigation activity, however, they do not consider economic and agronomic constraints affecting the performance of irrigated areas. Economists use the concept of efficiency to simply account for the criterion of maximal outputs and the possibility of minimal inputs. Agronomists have focused on the analysis of the soil-plant-atmosphere continuum. Several studies have focused on agronomic indicators based on Water Productivity (WP), which do not consider irrigation efficiency, or the water use efficiency (WUE) with more focus on irrigation efficiency (Pereira et al., 2011). These studies did not introduce in parallel the economic constraints for better irrigation performance. As presented above, irrigation engineers, economists and agronomists focused on performance indicators independantly. However, irrigation systems entail complex and interacting processes including physical, social, economic, political, technical and environmental dimensions (Le Grusse et al., 2009). Thus, there exists a gap that requires performance evaluation in-line with the inter-twined complex indicators. A split-up of domains is needed because a universal or single menu cannot be applied due to the diversity of issues (Malano, Burton, & Makin, 2004;Zoebl, 2006). Thus, benchmarking identifies critical success factors of an organization to achieve its objective (Goodstein, Nolan, & Pfeiffer, 1993). It is based on comparative indicators established when an irrigation system is more or less efficient than another and evaluating the gap between the current and the sought performances (Burt & Styles, 2004).
In the Mediterranean basin, agriculture consumes more than 80% of the available water supplies (Capone, Bilali, Elferchichi, Lamaddalena, & Lamberti, 2012). Despite the technical researches and technologies water losses stand over 50%, and several studies show that irrigation performance at plot and farm scale remains unsatisfactory (Hamdi, Ragab, & Scarascia-Mugnozza, 2003;Miao, Shi, Gonçalves, & Pereira, 2015). Several technical and management problems remain unsolved in systems in many areas, along with a real danger of simply directing large investments into traditional public irrigation schemes (Merrey & Sally, 2008). From this perspective, it seems that future research is still needed to analyse the low performance of the irrigation systems wich consists of a conveyance system, a distribution system and a field application system.
In Northern Tunisia, the irrigation schemes of the Medjerda Valley constitute a strategic lever. Studies on the performances of irrigation systems in these schemes were conducted by specialists who usually focused on single specific subjects (Zairi, El Amami, Slatni, Pereira, Rodrigues, & Machado, 2003;Mailhol, Zairi, Slatni, Ben Nouna, & El Amami, 2004;Yacoubi, Zayani, Zapata, Zairi, Slatni, Salvador, & Palayan, 2010;Slatni, Zayani, Zairi, Yacoubi, Salvador, & Playan, 2011). When analyzing the cause of some low performance, these studies may consider the effect of one factor without taking into account the effect of other influential ones. Moreover, the interaction of different factors amplifies the expected effect of a given constraint. Pereira, Oweis, and Zairi (2002) identified the need to develop appropriate methodologies to analyze social, economic and environmental benefits of improved irrigation management. Ghazouani, Marlet, Mekki, and Vidal (2009) have studied the farmers' discourse, practices and perceptions in the modernization program for a better management of an oasis in Southern Tunisia. Le Grusse, Mailhol, Bouaziz, Zairi, Raki, Chabaka, Djebara, and Ruelle (2009) have proposed an approach to simultaneously take into account the technical, agricultural, economic and environmental objectives at the farm level. Dirwai, Senzanje, and Mudhara (2019) proposed a method to assess causal factors of irrigation infrastructure without, however, remaining limited to hydraulic aspects. Rather, they introduced human, institutional and environmental factors.
The objective of the current study is to depict effective and pertinent factors explaining the field-level performances in view of a better management. It particularly aims to analyze the field level performances considering simultaniously irrigation and economic indicators, including different production factors, given our hypothesis that there is an interaction between these later indicators.

Case Study
The study area is an irrigation scheme located in the Medjerda Valley (36°45′ N, 9°45′ S) and called "Borj Toumi" (Figure 1). This scheme is made up of 3 large farms of 100 ha each and 82 family farms of around 7 ha. The area was created by the Tunisian government in 1966 and covers 785 ha divided in two hydraulic sectors. A first one covers 425 ha and is supplied by a Gravity distribution System (GS). This system is made up of open-air pipes. A water bailiff opens the intake according to farmers demand. Farmers using under pressure irrigation techniques use their own pumping station. A second sector covers 360 ha and is supplied by a Pressurized distribution System (PS). Although water is supplied under pressure in this sector, farmers use surface irrigation on some fields. The water distribution system, which is buried, is built to serve farms on-demand, but secondary canals are not permanently filled. Some farmers are under the constraint of difficult access to water; some neighboring farmers take water from only one intake, some others at the end of the pipeline have a low pressure. In the other hand, mainly on the GS sector, farmers have an easy access to water. The easy or difficult access to water was a data given by the farmer. jas.ccsenet.

Survey
This study a typology of the sche by this typ area of fam farms into Table 1 su class 2 and were more the smalle irrigation. 2016 to th practiced machinery

Irrigation Indicators
Results of monitoring the irrigation events showed that the application efficiency was of 100% in all cases; soil-moisture samples before and after irrigation events showed that farmers applied low water depths which were stored entirely in the root zone; there was no water waste by percolation. There was nearly no water waste during the conveyance section. So, the E appl and the Econv were not considered as pertinent irrigation indicators in the context of the studied area.
In contrast, there was variability in the E farm . Results of measures of E farm are summarized in Table 2. With surface irrigation technique, the E farm vary between 36%, on fields far from the farm gate and 100% on the fields near the farm gate with a mean value of 80%. With under pressure irrigation technique, the outdated equipment affected strongly the E farm which ranged between 22% and 82% with a mean value of 59%. Poor E farm led to a great loss in water volumes destined for the crops. The on farm monitoring showed that the irrigation performances of some farms were dictated by the constraints of the public network. Obviously, the E farm varied from one field to another depending on the availability of the flow and the pressure, the quality of the used equipment and on the farmer's skills. To these constraints relate to the section of the network between the farm gate and the edge of the field (Figure 2) are added the constraints relating to the conveyance section of the network, between the reservoir and the farm gate ( Figure 2). As cited in Dirwai, Senzanje & Mudhara (2019), the regulators in the conveyance canals showed signs of deterioration in the PS sector and also showed abuse and vandalism in the GS sector.

Conventional Economic Indicators
The entire tomato production in the study area is destined for manufacturing where farmers sell their production for the same standardized price. Similarly, the selling prices of olives and wheat are fixed by the government, so they do not display any variation among the sample farmers. As the study area is close to Tunis (30 km), most of the farmers sell their crops to the wholesale market for almost the same price. This indicates that the main difference in the farmers' income is due to their respective yields given that all the farmers buy their input from the same suppliers.
The results of descriptive statistics of economic data in relation to variable costs, gross margins and yields are summarized in Table 3.

Technical Efficiency Scores
Calculations using the DEAP program provided technical TE scores for each field. Basic statistics for these efficiency scores are reported in Table 4. The mean values of TE for fruit trees, tomato and wheat showed that the average efficiency level of considered fields is high (around 0.85). Fields of olive trees were still producing at a lower efficiency level (average of 0.66), indicating that the observed olives output could be 34% higher with the same inputs levels.

Variability of Efficiency Scores
The results of the variance analysis of the TE scores are summarized in Table 5. As mentioned in Section 2, the quantitative factor E farm was transformed into 2 classes (E farm  80% and E farm  80%) where differences between the means were significant.
Results showed that the easy or difficult access to water is the only significant factor explaining differences of TE for olive trees, fruit trees and tomato, among farmers. Table 5 shows that the water supply to olive trees (easy or difficult) was a significant factor for TE (p = 0.032). In case of an easy access, TE was 0.74 but it was only 0.49 when there was a difficult access.  Neither the distribution sectors -gravity or under pressure-nor the irrigation technique was a significant factor for TE. This raised some questions about the utility of the gravity to pressure network conversion program and the incentives for installing trickle irrigation to save water. The efficiency of on-farm distribution to the field (E farm ) was a significant factor only for olive trees (Table 5). Although most of the wheat and winter crop fields were rain-fed, some received only one complement irrigation in which case the irrigation indicators were insignificant.

Discussion
In a context marked by the necessity of saving water and by questioning the performance of irrigation systems which have not attained their objectives, this research intends to develop a bechmark which can provide elements of response to these challenges, by analyzing the performance of irrigated fields in a transdisciplinary dimension. The bechmark used to analyse the performance at the field level allowed to define the real and relevant indicators for improvement. In addition to the state of the science, this study will enable decision-makers to know what should be acted on to improve performance, especially since the Tunisian government is preparing for a modernization project.

Gravity Systems vs Pressurised Ones
In Tunisia, a modernization plan was planned for the old irrigation schemes of the Medjerda Valley. The effective implementation of this plan started in 2010, and aimed at upgrading the gravity water distribution systems to the pressurised ones. Within this framework, results suggest that the type of distribution sector (gravity or pressurized) is not a significant factor affecting the TE of farmers. This later efficiency indicator is, however, significantly affected by the difficult or easy access to water, which is in line with other research findings stating that farmers become inefficient when their access to water is uncertain (Varghse, Veettil, Speelman, Buysse, & Van Huylenbroeck, 2013). In this irrigation scheme, due to the low pressure at some points of the network, it will be easier to have a higher outflow on the Gravity distribution System (GS) compared to the Pressurized distribution System (PS). This constraint is an obstacle to the farmers located in the PS to reach better efficiencies. Although the GS suffers from other types of problems of upkeep and illegal water carrying, the PS also has his own shortcomings.It is indeed necessary to prioritize the modernization investments of the distribution networks based on relevant studies in order to maximize the social return and their impact on farmers. Paying attention to improving the access to water and controlling the illicit installation of water gates seems to be good and non-expensive options to enhance water management and irrigation performances in this study area.

The Compromise Between Economic and Irrigation Performances
Referring to the economic aspects, the majority of farms in our sample are actually producing at an efficient level despite the poor E farm reflecting considerable water waste (Table 4). In a context characterized by low levels jas.ccsenet.org Journal of Agricultural Science Vol. 13, No. 6; of intensification, water waste is not of great importance for farmers. There is a divergence between the farmer's interest (the economic efficiency) and the community's objective, which is to save water (better irrigation performance). The efforts of the government in providing incitation to farmers for better water use seem to be without results. In the context of this irrigation scheme, the water access and the reliability of water supply is a higher farming constraint compared to the irrigation efficiency.

Lever for Improvement
Our results show that irrigation performances do not correlate with economic performances at the field level. In this case study, the E farm seems to be a very relevant irrigation indicator. The poor E farm denotes a serious source of water loss. This goes against the important efforts made by the Tunisian government to invest in water saving technologies in irrigated agriculture. Previous studies have been conducted in Tunisia to evaluate the irrigation water use efficiency (IWUE) at different locations and with different crops. Frija, Chebil, Speelman, Buysse, & Huylenbroeck (2009) found that IWUE was 42% in horticultural greenhouse, with pepper, tomato and melon, in the Eastern Central Tunisia. Dhehibi, Lachaal, Elloumi, and Messaoud (2007) evaluated IWUE in field scale for citrus production in Cap Bon (North East Tunisia) to 53%. Albouchi, Bachta, and Jacquet (2005) showed that it was possible to improve water efficiency by 53% in the Kairouan region (Central Tunisia). All of these studies show that a great potential to save water and to enhance the performance of its use still exists in Tunisia.
Improving the access to water can lead to better economic conditions of the farmers but not to the best irrigation performance. The low E farm is due to upgrated equipment in the majority of cases. However, the fact that the economic and irrigation performances are not related would make farmers unwilling to contribute to the on-farm improvement of the irrigatrion performances. The Tunisian National Irrigation Water Saving Program provides subsidies ranging from 40 to 60% to encourage farmers investing in new and modern irrigation equipment. Most farmers in this study area did not benefit from this subsidy program, due to many reasons (for example problems of land title). In this context, the improvement of irrigation efficiencies could be reached by further encouraging and assisting farmers' investments in adequate equipment and facilitating their access to the subsidy schemes offered for such type of investments. In line with this, the focus on irrigation pricing as a single (or dominant) irrigation management instrument can become a blocking factor and may lead to a decline of irrigated agriculture when it is associated with low levels of intensification (Pereira et al., 2011). Technical solutions combined within a set of appropriate intensification and water policy options might be a good alternative to stimulate sustainable use of irrigation water and creating positive socioeconomic dynamic at the local level. To this end, we recommend to further develop the capacities of local vulgarization services as well as farmer's knowledge by offering targeted trainings in relation to water productivity and intensification level.

The Bechmark
Our results justify the use of the bechmarking in building global representation adapted to the actual local context. This approach allowed understanding why irrigation performances are poor and determine the pertinent factors that may lead to their improvement. Such results might be useful for the development of the irrigation sector in Tunisia, especially in the current context of rehabilitation and modernization of the irrigation scheme. The approach outlined should be replicated in other irrigated areas to depict their context-specific levers of improvement. This study is a first step to define significant factors that contribute to enhancing the performance at field scale by bechmarking. Yet, in economic studies, the farm is seen as the relevant decision unit . Performance indicators should also be analyzed at farm level taken as a set of fields in interaction with each other and with other activities such as livestock.

Conclusion
The objective of this study was to depict effective and pertinent factors for a better management in the irrigated fields of a public irrigation scheme in Northern Tunisia. A benchmark was used to analyze the performance of irrigation systems at field scale. The relationship between irrigation and economic performance including other factors of the production process was investigated. For this, on-field irrigation measurements and monitoring events were processed to evaluate irrigation indicators on 31 individual fields. Then results of detailed techno-economic surveys on 91 fields were used to estimate the Technical Efficiencies (TE) for cultivated crops (olive trees, fruit trees, tomato and wheat). Finally, statistical variance analysis was used to evaluate a set of significant variables to depict effective and pertinent lever to better performance of irrigation systems. This study showed that irrigation performance, which has been widely studied with the aim of improving water management, does not always impact economic performance, which is the most important farmers' target. This is due to the irrigation water costs in comparison to the other input of the agricultural production in such low intensification context. In the context of the irrigation scheme support of this study, some farmers' strategies/practices are economically justified but may lead to a poor irrigation performance. Results showed that the on-farm distribution efficiency (E farm ) was the most pertinent irrigation indicator which was low ranging from 36% to 100%, and causing considerable waste of water. With regard to economic aspects, the majority of farms are operating at an efficient level; the TE was around 0.85 for fruit trees, tomato and wheat and only 0.66 for olive trees. The statistical analysis showed that the access to water is the way for a better TE in the context of the studied area. However, the on-farm distribution efficiency (E farm ) is indeed not always significant for the TE, which is the ultimate and most important goal of the farmer. This is due to the fact that irrigation water costs are usually low compared to the other inputs in such a low intensification context. In the irrigated scheme considered for this case study, some farmers' strategies/practices are economically justified but may also lead to a poor irrigation performance.
Based on these results, to overcome this problem of the divergence of the community and the farmer's concerns, higher attention should be paid to the "water saving program" of Tunisia, by better targeting subsidies for irrigation investments and focusing on subsidies which are creating the most of economic growth and have an impact at the farm levels. Another major recommendation for policy makers is to increase the water cost in order to consider water as a production factor. The second suggestion would be to watch closely the illicit private pumping extensions in the irrigated areas of Tunisia.