Rethinking water and crop management in the irrigated district of Diyar-Al-Hujjej (Tunisia)

In Diyar-Al-Hujjej irrigated area, the aquifer’s over-exploitation, sea intrusion, and abandonment of irrigated areas and wells took place. A yield decrease for all crops was observed. Average aquifer water electrical conductivity (EC) jumps from 4 to 6.6 dS/m between 1969 and 2017. A fresh surface water transfer over more than 100 km was launched in 1998 to safeguard this irrigated area but this fresh water supply is not stable; it varies from 1 year to another (about 1,900,000 m3 in 2015 and only 60.000 m3 in 2018) while annual crop water requirement of the perimeter is about 2,500,000 m3. An adaptation by farmers to this new situation of saline and water stress was observed. The follow-up surveys of the farmer’s practices showed that (i) new crops with high added value grown during the rainy season were introduced in association with dry season crops (strawberry-pepper association), (ii) rainfed crops, fallow, and water blending are common practices; and (iii) growing of rainy season crops in the aim to reduce water supply. The instability of fresh surface water volume transferred constitutes the main threat for this perimeter. The use of aquifer salt water must be stopped; it is the cause of the large quantities of salts supplied (over 13,000 kg/ha) and also of the low annual net income achieved. Net revenue was less than 2000 US $/ha under salt water and reached even 8000 US $/ha when sufficient surface water is available. An agrarian reform policy must be applied for this perimeter; only crops whose water requirements are partially met by rain should be grown. The introduction of another more sustainable water source should be initiated (as desalination) even at the private farm level.


Introduction
Land degradation and the overexploitation of natural resources are unanimous around the world , especially in areas with high population density and intense economic activity (Kanzari et al. 2020a). Competition between the different users of these resources is often observed. Indeed, the Mediterranean region is vulnerable to the consequences of climate change. According to the Intergovernmental Panel on Climate Change (Eckstein et al. 2019), North Africa is considered a vulnerable area to climate risks. In Tunisia, good-quality water is reserved for drinking water and in case of water shortage; irrigated areas receive quantities of water lower than their actual needs (even no supply sometimes for vegetable crops). In our irrigated area, fresh surface water supply was about 1,900,000 m 3 and 60,000 m 3 during respectively 2015 and 2018 (Daghari et al. 2021).
According to Aragüés et al. (2011), 7.3 million hectare salty soils are located in Spain, Morocco, Tunisia, and Turkey for a total of 27.3 million hectares for the entire Mediterranean region. Coastal aquifers around the Mediterranean and more particularly in North Africa are increasingly degraded and over-exploited. In Tunisia, salty soils occupy 1.5 million hectares (25% of the total area of cultivable soils (Hamrouni and Daghari 2010). Agriculture will remain a key sector; it accounts for more than 11% of the gross domestic product and provides employment for more than 20% of the working population. The irrigated sector contributes to more than 35% of national agricultural production. Unfortunately, Tunisia is a country with very limited water resources. The quota of a Tunisian does not exceed 500 m 3 per inhabitant per year. A third of Tunisia's mobilizable water resources have salinity greater than 4.5 dS/m (Daghari and Gharbi 2014). Also, 50% of the sampled wells in Northern Tunisia have salinity higher than 3 dS/m (Hachicha et al. 2013). Furthermore, all regions of Tunisia suffer from high groundwater salinity; this is the case of our irrigated area and a lot of other regions such as all the south when average rain is about only 100 mm. Crop yield is very low due to salinity; in Tunisian oases, the driving force of development for all of southern Tunisia, the average yield of irrigated date palms is 4.6 tons/ha while it is 41 tons/ha (9 times) in Egypt (El-Juhany 2010; Zafar 2020). Salinization develops from the mobilization of salts stored in the soil profile due to human activities (Ferjani et al., 2013) and practices such as inappropriate irrigation practices and deforestation.
In the region of Diyar-Al-Hujjej, the site of our study, aquifer over-exploitation combined with the rain's irregularity led to a drop in the water table level and sea intrusion and yields decline. A low tomato yield (less than 20 tons/ha) has been observed. The electrical conductivity (EC) of the wells (dS/m) of more than 30 dS/m and depressions from sea level down to − 13 m were measured (Chekirbane et al. 2014).
These high ECs have led to the abandonment of wells and farms and the migration of farmers to other areas less affected by salinity, often in the form of tenants. The average sodium adsorption ratio (SAR) was 8.6 0.5 (meq L -1 ) and the average electrical conductivity (EC) was 6.6 dS/m, (Mekni 2017). The number of abandoned wells increased from 1.268 to 3.200 between 1980 and 2005 in Diyar-Al-Hujjej and surrounding areas (Cap Bon peninsula).
As emergency measures, there was recourse to the transfer of surface fresh water from another geographical region (north-west) of Tunisia, via the Cap Bon Medjerda canal over approximately 100 km with several pumping stations. This operation of water transfer is highly contested last years by the population living in the watershed from which the water transfer is made. On another side, the feasibility of using seawater desalination shows its limit , besides environmental problems, but renewable and clean energy usage can help sustain environmental conditions in the opposite of fossil fuel-based energy (Abbas et al. 2020). Desalinated water cost is very high (more than 0.5 US $/m 3 ) while the selling price of surface water is 0.05 US $/m 3 and the cost of pumped saline water is only 0.02 US $ /m 3 . The salt seawater desalination analysis showed clearly that the application of desalinated seawater in the current situation leads to a negative net marginal value except for crops with high added value (strawberry), . But no agroindustry exists for strawberries; the market is very limited and it is a crop that spoils in two or 3 days. As desalination is beyond the reach of farmers today, it was deemed more appropriate to analyze the farmer's practices.
In the irrigated area of Dyiar-Al-Hujjej, all farmers are aware of the salinity problem. Some new practices were adopted with the aim to manage salt, water, and crops over time: -Water blending and over-irrigation with freshwater in the aim to drop soil salinity -Introduction of new and more profitable crops such as strawberry following this supply of fresh water and the practice of associated crops on the same site (strawberrypepper combination) , a) -Growing of winter and spring crops with the aim to reduce water supply -Adoption of crop rotation by the application of successions between irrigated crops, rainfed crops, and fallow by all farmers in order to reduce soil salinity even if the farm size is very low; farms with an area of less than 4 ha represent 60%.
In several other fields, several authors have shown that the phenomenon of topsoil salinization can be cyclical and that the effect of salinity on crops can be mitigated by appropriate management which involves appropriate irrigation for leaching of the soil (Kanzari et al. 2020b). For this perimeter, it is only recently that the behaviour of farmers and their perception of salinity have started to be studied. There is no clear agreement on what constitutes best practices for managing salinity in this region especially since the fresh surface water volume is very irregular. Considering the heaviness of these experiments, only modelling was often used (Lahlou et al. 2000;Kaur et al. 2007; Arafat and Abdulrasoul 2019) but field studies are lacking. Some studies were carried out over short periods and only looked at salinity , a, (Bani et al. 2020. Total amounts of salt brought, farmer's net revenue, and salinity evolution for a long period (many years) were not studied before. The objective of this paper is to compare farmers' practices in an irrigated area whose irrigation water comes from a saline aquifer and very irregular transferred surface fresh water but certain prosperity is observed in this perimeter despite water and salt stress.
The average potential evapotranspiration and annual rainfall were respectively 1.166 mm and 441 mm for the period . Dyiar-Al-Hujjej's aquifer (EC) was 6.6 dS/m in 2017; well depth is about 20 m measured in 2011, 2017, and 2020; and it is the main region when strawberry is growing in Tunisia (more than 95%). Only localized irrigation is encountered and irrigation uniformity is about 90% .
Meetings with extension services, farmers, and farmers' association called "izdihar" were carried out. The crops predominantly grown are vegetable crops (pepper, potato, strawberry, squash, tomato, etc.). Soil profiles showed that the crust is often present from 20 cm depth sometimes; after granular analysis, percentages of clay, silt, and sand were respectively 9%, 36%, and 55%. Crop choice varies depending on available water sources. Thanks to the availability of surface water with a low (CE) less than 1 dS/m from the state network supply, the growth of a high added-value crop which is strawberry is increasing every year. Its agricultural production value is about 15.600 US $/ha while it is less than 3.000 US $/ha for all other crops. This explains the increase in the area occupied by strawberries from 0 ha in 1999, 55 ha in 2000 to 150 ha in 2011 and 121 ha in 2019. Cost production of strawberry is very high (0.2 USD/plant) and it is not within the reach of all farmers; it is about 6.000 US $/ha while it is less than 2.000 US $/ha for all other crops. The cultivated areas dedicated to tomato crop which is a summer crop with high water requirement decreased from 450 ha in 2000 to only 70 ha in 2019. The farm size is divided into two classes 0-4 ha and 4-10 ha, each occupying 435 ha with 185 farmers and 365 ha with 60 farmers, respectively.
The farmers were grouped according to three different scenarios: (i) farmers who use only fresh surface water from the state network, (ii) farmers who blend surface water and saline groundwater resulting in decreased water salinity, and (iii) farmers who use only saline groundwater. Each group has its own strategy regarding crops grown and crop rotation. ECs measured on December 2, 2020, were 2.4 and 4.4 dS/m respectively for fresh surface water and aquifer saline water. These values were respectively 2.82 and 4.36 dS/m on February 26, 2021.

Experiments
EC was measured using the method of saturated paste extract and Geonics conductive meter (EM 38); the electrical conductivity of the soil was measured using the EMC method (Electromagnetic Conductivity Geonics EM 38). A CEM calibration was performed on several crops in the irrigated area. At the same time, soil samples were taken. The readings provided by the horizontal CEM apparatus (CEH) were correlated with the measurements of electrical conductivity (ECe) of soil samples obtained in the laboratory by the saturated paste method. A calibration equation of the form CEe = 0.141CEH +1.199 was established. Irrigation water amounts supplied were tracked by water-meter monthly. Also, wells (EC) and depths were measured by using conductive meter, and piezometric probe respectively. For different crop rotations encountered, EC was measured during September 2017-February 2021. The last salinity measurements were carried out on December 2, 2020, and on February 2021 over an area after three important rainfall of about 30 mm each. During the years 2016, 2017, 2018, 2019, and 2020, the volume of surface fresh water supplied is very low and many areas were not occupied by irrigated crops.

Main equations
The study is based on a set of equations described below for an agricultural product analysis produced by local farmers. These equations depend on several parameters. For the relative yield, it will depend on the irrigation water salinity threshold which causes toxicity to the roots which are no longer able to absorb water. For the second equation, water EC and the water volume supplied to give the correct value for the amount of salt supplied by the irrigation. Finally, for the net income, it will depend on 3 parameters which are the value of agricultural products (US $/ha), the cost of production (US $/ha), and the cost of water (US $/ha).
-Relative yield Relative yield (Y r ) is calculated by using the following equation (Maas and Hoffman 1977): where b is the curve slope expressed in percent per dS/m (equal to 14, 12, 10.5, 33, and 9.9 respectively for pepper, potato, squash, strawberry, and tomato), a is the salinity threshold expressed in dS/m (equal to 1.5, 1.7, 4.9, 1, and 2.5 respectively for pepper, potato, squash, strawberry, and tomato). EC s is the mean electrical conductivity of a saturated paste, measured in the root zone (dS/m).
-Salt brought to the soil by irrigation The quantities of salt brought (Q s ) are calculated using this equation: with Q s (kg/ha), V (m 3 /ha), and S (g/l) = water electrical conductivity (EC)*0.64.
with production costs (US $/ha), (= product selling price (US $/kg) * yield (kg/ha) and water cost (US $/ ha), (= volume of water (m 3 ) * selling price of water cubic meter (US $/m 3 )). Selling prices at farm level of agricultural products, water prices or costs, production cost, and crop yield considered are taken according to the Tunisian Ministry of Agriculture, Water Resources and Fisheries database.

Results and discussion
In our irrigated area, crops are grown and crop rotation depends mainly on available water sources; surface water is available in uncertain amounts. The volume of surface water transferred is irregular and varies from 1 year to another (1,929.322 and 60.503 m 3 respectively for the years 2015 and 2018) depending on the water volume stored in dams (Table 1). Average net crop water requirements calculated by the CROPWAT model (Allen et al. 1998) are about 2,500.000 m 3 for this irrigated area. Tunisia is going through a dry period since 2016; the water stored volume in dams in Tunisia is less than 50% of their capacity; priority is given to potable water. In the irrigated area of Diyar-Al-Hujjej, for the year 2019, irrigation was applied by 150 farmers with an area of 276 ha (about 39% of the total equipped area). The volume of water distributed in 2019 was 620.000 m 3 while the expected volume is 1,225.000 m 3 (about 200%). Even in 2020, fresh water supplied is still low (920.000 m 3 ) and represents about 50% of the volumes supplied before 2015.

Farmers having only fresh surface water
These farmers are tenants who do not invest in the drilling of wells and the building of tanks necessary for water blending. They grow only crops with high added value, no rainfed crops. Soil profiles (EC) measured were about 2 dS/m for the first year (1.6 dS/m under strawberry in May 2018 and 2.2 dS/m under pepper in August 2018) (Table 1). This relatively low salinity measured is the result of the use of good-quality surface water and the leaching of salts under the effect of rain, especially since it is a sandy soil. Exceptional winter rains of 260 mm allowed the evacuation of more than 60% of the initial salt stock. The combination strawberry-pepper is maintained during a second year; especially in September 2018, more than 200 mm have fallen, which further promotes salts leaching. During 2019, the volume of surface water pumped was 780.389 m 3 while the volume distributed was only 620.000, a loss rate of 21%. The number of hydrants used was 138 for a total of 266 with a rate of 52%.
The EC measured at the end of the second year was 3.2 dS/ m (Table 1) (Fig. 2, blue curve). For this conductivity, no strawberry agricultural production can be obtained whereas it is 100% for squash by applying Eq. (1). At the start of the third year, farmers no longer alternate crops between strawberry and pepper to fall back on other crops less sensitive to salinity such as cabbage, potatoes, and squash. This cultivation practice is understandable by seeing that the at the end of the third year, EC was only 2.1dS/m in August 2018 under squash (Table 1).
With respective average net irrigation water requirements of 3.600 m 3 /ha, 6.000 m 3 /ha, 1.345 m 3 /ha, and 7.174 m 3 /ha respectively for strawberries, pepper, potato, and squash and with surface water's salinity of 1.5 dS/m, the quantities of salts  (2) were 9.600 kg/ha (= 3.600 kg/ha + 6.000 kg/ha) under strawberry-pepper and 8.519 kg/ ha (= 1345 kg/ha + 7174 kg/ha) under potato-squash. Even if the quantities of salts added are almost identical for the 3 years, the crops grown during the third year are resistant to salinity. For squash, even with an EC of 5 dS/m, yield reduction is negligible. Considering the average EC between these measured at the beginning and at the end of each season (Table 1) and using Eq. (1), relative yield (Y r ) drops from 80 to 47% for the strawberries and from 90 to 76% for pepper between the first year and the second year; this is a significant drop in yields, especially for strawberries. Indeed, the yields really observed in the field during the second year are low compared to the first year. Irrigation water productivity (kg of agricultural product/m 3 of irrigation water) decreases from 80 to 47% and from 90 to 76% respectively for strawberry and pepper between the first year and the second year. In the largest state-run study carried out for the agricultural sector in the 1960s in Tunisia called CRUESI, a reduction yield of 50% and 40% was observed respectively for tomato and for pepper for a water salinity exceeding 3 dS/m, (CRUESI 1970). For fodder sorghum, watermelons, and beans, irrigation water with a salinity of 5 dS/m leads to a yield reduction of 30% (Van Hoorn et al. 1968). Irrigation with saline water may be tolerable for a crop and presents a low risk of soil salinization (Jahin et al. 2020). However, saline water with ECw = 7 dS m −1 reduces agricultural yield by almost 50% even for halophyte crops and increases soil salinity in the short term (Bueno et al. 2020) and could threaten the agricultural season by what follows.
In several irrigation cycles (Oster et al. 2012), the use of a drip irrigation system combined with appropriate management practices such as increasing the frequency of irrigation (Azad et al. 2018) can decrease the effect of salinity on the soil and cultures. Repeating this experiment over at least 3 years can provide strong statistical significance for the effects of brackish water on crop yields. With an EC of 1.95 dS/m at the end of the third year, transplanted strawberry yield reduction will be of 30% but the autumn rains will contribute to the soil desalinization. Farmers over-irrigate if there is a lack of rain.
During the second year, costs production for strawberry and pepper are less than the first year since the drip Irrigation installation already exists (about 50% compared to the cost of a new facility); costs of tillage, seeds, ad transplanting were not considered; crops were kept during 2 years.
For the first and the second year, net revenues were high (7.770 and 5.482 US $/ha respectively) compared to the third year (2.534 US $/ha) (Table 1). It is for this reason that the strawberry-pepper combination is very coveted. Water cost is low by comparison to production cost (only 3% for strawberry).

Farmers using only saline aquifer water
These farmers do not have access to surface water. Their plots are adjacent to the irrigated perimeter or they have not paid their irrigation water bill; agriculture is not their main activity. They have very small irrigated areas, coming from the heritage. They grow rainfed crops throughout rainy period (September-April). Thus, the soil observes desalination throughout the rains, which allows them to cultivate mainly tomatoes intended for processing and which are grown early between February and June. These tomatoes resist the salinity of the soil compared to the tomato intended for sale in the markets. The tomato sold in markets to customers is cultivated between May and August (dry season). The main Tunisian's tomato and pepper processing factories are located in this region and in neighboring areas.
Due to the high salinity of the water, the growth of the tomato is incomplete giving a small fruit that is not unsaleable and therefore mainly intended for processing. The selling price is fixed in agreement with the factories, and it is lower than the selling price of the tomato intended for fresh consumption but they have a guarantee for the sale of their production. The factories advance all operating costs (the price of seeds, fertilizers, phytosanitary products). It is the same thing for pepper which is not a crop cultivated by farmers since the resulting fruit is very small. ECs measured during the rainy season under rainfed crops and fallow are less than 2 dS/m while under tomato grown during the dry season, EC reached 5.37 dS/m even at the end of the first year (line 4, Table 2). This increase in EC forces farmers to choose rainfed crops or follow their plots to avoid irrigated with saltwater from the groundwater (Fig. 2, red curve). The average measured EC of Diyar-Al-Hujjej's aquifer was 6.6 dS/m, (Mekni 2017). Tomato irrigation water requirements are 6.467 m 3 ha. The amounts of salts brought are 13.658 kg/ha (6.6 dS/m * 0.64 * (6.467/2) m 3 /ha) under tomato assuming that only half of the water requirements are supplied in the form of salt water (50% of the water needs are met by rain) while the amount of the salts supplied is only 9.600 kg/ha for a whole year of irrigation under the strawberry-pepper combination when only surface water has been used. If all tomato water requirements were satisfied by saline water when tomato was grown during the dry season, the amount of salts brought was 27.317 kg/ha, which explains the farmers' choice for industrial tomatoes, the water requirements of which are largely satisfied by the rain. The areas intended for tomato growth fell from 450 ha to less than 100 ha between 2000 and 2019. Some farmers without access to surface water and observing low yields abandon their farms for one or 2 years to witness desalinization of soil profiles under the effect of the rain; it is the case of all farmers located near the sea. In Gambia, faster germinating varieties of rice and peanuts have been developed to mature in a shorter rainy season and avoid the need for saltwater pumping as part of a challenge of adaptation to climate change (Yanl 2011). Pickson et al. (2020) indicated that it is necessary to develop new cereal crop varieties and suggest that the prices of agricultural products must be subsidized to improve cereal variety production under climate change threat.
Annual net revenue was very low (about 1.000 US $/ha) ( Table 2) while it reached more than 7.000 US $/ha when only surface water was used (Table 1). Average annual returns simulated showed that fallow-wheat rotation was the most beneficial choice, compared to pearl millet-based sequences and pearl millet-wheat rotations (Kaur et al. 2007). Ahmed et al. (2019) indicated that fallow increases water stored for the next year and found that planting lucerne in rotation with canola, wheat, and triticale crops used more water, as did native vegetation. But here, it is clear that in the irrigated areas, rainfed crops and fallow decrease soil salinity also because of no salt water supply.

Farmers having access to 2 water sources (surface and saline waters)
In our study area, farmers with both irrigation water sources prefer to irrigate with surface water only if the quantities are sufficient. If this is not the case, the freshwater saltwater blending is carried out. We noted the massive presence of reinforced concrete tanks or ground tanks for mixing water

Case of farmers having a fair volume of surface water
These farmers are often homeowners and no farmer is interested in plots that do not contain large amounts of surface water. The strawberry-pepper combination is absent because surface water is present in small amounts. Unlike farmers who have only saline water from the aquifer and only grow rainfed crops during the rains, these farmers grow salinity-resistant irrigated crops during this same period. The small amounts of freshwater are blended directly in the wells, resulting in additional pumping costs. Three crop rotation are encountered: -Rainfed wheat or barley-tomato or pepper -Beans used as a green manure-potato-tomato or pepper -Cabbage-potato-squash-fallow The crops grown till April are rainfed crops or irrigated by blended water and rain contributes to their water needs and to salt leaching.
In general, an increase in salinity is observed during the dry season following irrigation with blended water and/or groundwater; subsequently, during the rainy season, desalinization took place. During the rainy periods (Sept 2017-Apr 2018 and Sept 2018-Apr 2019), (EC) varied between 1.26 and 2.5 dS/m under rainfed beans, cabbage, and irrigated potato  (Table 3). Thus, EC increased and reached even 5.36 dS/m in May 2017-August 2018, (Table 3). We noticed that the field yield observed is low. Bani et al. (2020) found an EC of 3.8 dS/m and 5.5 dS/m under pepper irrigated with blended water between the beginning (first May) and the end (August) of the dry irrigation season. Saidi et al. (2010) reported that an increase of salt salinity from 2 to 8 dS/m was observed under tomato between the beginning (first May) and the end (August) of the dry irrigation season too. Salt accumulation occurs mainly throughout the irrigation season. In the second year, the ECs measured are the same (Table 3). We have noticed that the measured means of EC are the same in both cases, when only saline water from the aquifer was present and the irrigation is only carried out during the dry season (Table 3) and when surface water is available in small quantities but only irrigation is carried out during dry seasons (Table 4).
By using Eq. (3), the net revenues were 826, 1.549,and 1.386 US $/m 3 respectively for potato, squash, and tomato. For pepper, the net revenue is negligible and it is rarely grown except in areas far from the sea. A reduced farmers' net income was compared to those who grow strawberries when net revenue reached 7.770 US $/ha.
The amounts of salt added were 15.600, 1.345, 18.652, and 16.814 kg/ha respectively for pepper, potato, squash, and tomato, considered as very large quantities, except for the potato cultivated during the rainy season and for which a large part of the water needs is satisfied by rain. Studies suggest that the global annual cost of salt-induced land degradation in irrigated areas could be US$ 27.3 billion because of lost crop production (Manzoor et al. 2014) Case of farmers having medium amounts of surface water For the strawberry-pepper combination, EC measured under strawberry in May 2018 is 1.6 dS/m ( Table 4). The strawberry plant being very sensitive to salinity is transplanted only on soils left fallow or occupied by rainfed crops the previous year. In addition, a large part of its water needs is met by rain and only good-quality surface water is used for its irrigation. Another advantage is that the soil is sandy, which allows the salt leaching. The pepper was planted on the same lines as the strawberry to take advantage of the mineral elements and fertilizers of the soil produced by fertigation for the cultivation of the strawberry therefore an intensification rate of 200%. Its flowering and fruiting period takes place during the dry period and it is often irrigated with surface water or mixed water. EC increases and reaches 3.5 dS/m in August 2018 (Table 4), which is considered important to consider a second agricultural season for the strawberry-pepper combination, so this rotation is abandoned. With an EC of 3.5 dS/m measured at the end of the first year, the relative yield (Y r ) of the strawberry calculated by using Eq.
(1) will be under 10%; farmers grow strawberry-pepper on other different lands due to its profitability. Under the less salt-sensitive crops grown in the second year, the measured EC decreased and they were 1.15, 1.25, and 2.12 dS/m under respectively cabbage, potato, and fallow (Table 4). For these crops grown during the rainy season, only surface water is used if irrigation is necessary; rain compensates a good part of their water needs. The highest EC was observed under pepper grown during the dry season.
The net revenues were 7.532 and only 2.926 US $/ha during the first and the second year (Table 4); the profitability of the combination strawberry-pepper is clear. For farmers having only surface fresh water, the net revenues were 7.770 and 5.482 US $/ha respectively for the first and the second year (Table 1). Salt brought under strawberry is 3.600 kg/ha while it is 6.000 kg/ha for pepper (three times) while the net income of strawberry is more than fourteen times that of pepper.

Case of farmers having an important volume of surface water and large irrigated areas
They are full-time farmers and they are often large tenants of land. They over-irrigate to leach as much salt as possible. The strawberry-pepper combination can often be kept for 2 years; later, more resistant crops are grown. EC measured under strawberry was only 1.53 dS/m in May 2018 (Table 5) because only surface water is used for irrigation. Thus, the relative yield obtained is about 80% calculated using Eq. (1). A slight increase in EC (2.2 dS/m) under the pepper in August 2018 was measured (Table 5). This is because sometimes the pepper is irrigated with surface water or blended water if it lacks surface water. Farmers are aware that soil salinity must be kept low in order to maintain the strawberry-pepper combination for another year. For these farmers, since the measured medium (CE) and strawberry plants are expensive and imported (the cost of the plant is about 0.2 US $ and the need for one hectare is about 30.000 plants), this combination was kept for a second year. However, the EC measured at the end of the second year was high compared to the first year (4.7 dS/ m under pepper in August) ( Table 5) (Fig. 2, green curve) and a reduction in field yield is observed. Using Eq. (1), the relative yield is only 40% for the strawberry. Already, in recent years, if a cycle of drought is observed and in the absence of heavy rains in September, the strawberry-pepper association has been removed at the end of the first year.
Given the high salinity, crops less sensitive to salinity such as cabbage, potato, squash, or rainfed crops are grown in the third year; a decrease in salinity was observed. The average annual (EC) was 2.36 dS/m for the third year while it is 4 dS/m for the second year (Table 5). These farmers are all tenants and they have large irrigated areas; they do not stop practicing the strawberry-pepper association by dividing their plots into several unit plots so as to always have the strawberry-pepper association on part of their land. One of the tenants, with whom we conducted the experiments, rented a total area of 33 ha and had 12 hydrants and 5 wells. He has access even to water from a nearby little dam (Lebna), which has very good water quality. The profitability of the combination strawberry-pepper is clear; the net revenue for the second year is low compared to the first year but it is two times more than this of the third year; the net revenues were 8.331 and 4.150 US $/ha for the first and the second year when strawberry is present while it was 2.094 US $/ha during the third year when only other crops were grown (last line, Table 5). This advantage in income under strawberry-pepper encourages farmers to keep this combination for a second year. Some farmers rent large land just to have several surface water hydrants at their disposal, and they can practice the strawberry-pepper combination annually.

Synthesis analysis
When only surface water is available or it is available on medium or important amounts, the average EC for a complete crop cycle is less than 3 dS/m (row 5, Tables 1, 4, and 5) while it is more than this value when only saline water or little amounts of surface water are available (row 5, Tables 2 and  3). Even if EC measured under crops reached almost 6 dS/m (case of tomato), we see that for total crop cycle, the EC was less than 4 dS/m in all cases; hence, the interest of the introduction of crop rotation. Keeping soil under rainfed crops reduces soil salinity, besides other advantages. Crop rotation breaks the cycle of harmful organisms affecting crops by restricting pathogens and weeds (Leteinturier et al. 2007). In our irrigated area, potato grown after a cabbage is of good quality compared to the potato grown after the strawberrypepper combination (it has blackish spots) after the Diyar-Al-Hujjej framers and some samples were noted by ourselves. As farmers fear the salinity, they often over-irrigate, further increasing the salinity in the soil. According to Saidi et al. (2018), EC increases from 3 to more than 8 dS/m within the root area between the start and the end of the tomato irrigation season in the Kalaat-Al-Andalous irrigated area. They measured a yield of 40 ton/ha while the average tomato yield in Tunisia is about 80 tons/ha. Farmers need to be convinced that rational irrigation water management is important both to the welfare of farmers and the environment (Lazaridou et al. 2019). Growing strawberry-pepper in the second year leads to high salinity (4 dS/m, (row 5, Table 5) even if surface water is available on important amounts but farmers who opt for this solution saw the net income achieved when strawberry is grown; net revenue obtained with 50% of strawberry yield is higher than any other net revenue realized with other crops even under low salinity. When only saline water or surface water is available in little amounts, not only net revenue is low but also quantities of salt brought were high. Extension services should recommend stopping this type of irrigations; over-exploitation of the aquifer leads to sea intrusion. In San Joaquin Valley, only 2.000 kg salt/ha is added to irrigate summer crops (cotton, alfalfa) when high-quality irrigation water is used (Kenneth et al. 1986).
The net annual revenue is generally greater than 5.000 US $/ha/year when strawberry is present while it is less than 2.000 US $/ha/year when this crop is absent. For this crop, a large- scale processing industry must be set up to absorb the overproduction of a rapidly deteriorating crop. All the dry season crops must be discarded and replaced by crops whose water needs are partly met by rain such as artichoke, a crop very tolerant to salinity and highly sought after for export. Also, the sensitivity of crops to salinity depends on the vegetative stage; the effect of irrigation with saline water in development stages is low (Ashraf and Harris 2004). Tomato and pepper under greenhouse can constitute another alternative especially as these crops are profitable even if desalination is introduced, and this is a region where the greenhouse activity already exists. Other irrigation techniques must be introduced; subsurface drip irrigation (SDI) reduces evaporation losses during the pre-plant and early-season periods and improves water storage efficiency and crop yield even at low irrigation capacity (Bordovsky 2020). A significant impact in conserving groundwater can be obtained when improved irrigation at a farm level is applied (Ajaz et al. 2020). Additional effort is required from extension services; indeed, profitability and productivity of Boro rice, as well as water productivity, were comparatively higher for focal farmers than for control farmers (Uddin and Dhar 2020). The social benefits must be taken into consideration; Ryu et al. (2019) found that when only economic costs and benefits were considered, the benefit-cost ratio for the public system (0.02) was smaller than that for the private system (0.264) while the results of the two alternatives changed when we considered the social benefits.

Conclusion
This perimeter is still managed with old reflexes when only salt water was present; low-income crops (squash, tomato) tolerant to salinity are still grown even though they are the source of strong soil salinization, aquifer over-exploitation, and sea intrusion. Choice of crops must be reviewed. Crops grown during the rainy season (strawberry, potato) and for which a good part of their water needs is met by rain have led to a high net income and a low amount of brought salt (about 7.000 US $/ha and 3.600 kg/ha for strawberry) compared to the crops grown during the dry season when salt brought amounts were more than 13.000 kg/ha and net revenue was less than 2.000 US $/ha. In the presence of salty water, irrigation management is the key of success of any irrigation project.
The practice of rainfed crops and fallow during the rainy season allowed a salinity reduction within the soil profiles. The salinity measured under irrigated tomato is greater than 5 dS/m while that measured under rainfed wheat or barley is less than 2 dS/m in the same farms.
The economics of natural resources must answer these complex questions, what is better if surface water is missing? (i) Growing crops that coincide with the rainy season and/or faster germinating varieties but often with low net revenue (as processing tomato), (ii) accepting a low net income and protecting the soil against salinity by growing only rainfed crops and stopping sea intrusion but we must think about compensating these farmers, and (iii) are present crop rotations maintain or degrade soil proprieties; is an interdisciplinary research needed?
However, other durable solutions need to be examined on a larger scale; can desalinated water be an alternative if a water social price subsidized by the state will be applied? The case of this perimeter can constitute a roadmap for several other perimeters suffering from lack of water and excessive salinity.
Code availability Not applicable.
Author contribution ID prepared the initial document; he also collected data from the field and participated in meetings with farmers to investigate the problems of the irrigated area. FBA participated in meetings with farmers to investigate the problems of the irrigated area. HD offered us the methodology of the work and corrected the mistakes in the English language.
Funding This work is funded by the Eremology and Desertification Control Laboratory (Institute of Arid Regions of Medenine).
Data availability Data can be shared or used for any other analysis.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication Not applicable.
Competing interests The authors declare no competing interests.