Assessment of a New Approach for Systematic Subsurface Drip Irrigation Management

This paper aimed to assess the reliability of a new approach that provides systematic irrigation management based on fixed water suction in the vadose zone. Trials were carried out in the experimental farm of IRAGabès on subsurface drip irrigated (SDI) tomato plot. The SDI system was designed so that the soil water content is to be maintained within prescribed interval ascertaining the best plant growth. Irrigation management was systematically monitored by water suction evolution in the vadose zone. Recorded results showed that all-over irrigation season lateral pressure head ranged within 93.3 ± 20.0; 119.95 ± 53.35 and 106.6 ± 40.0mb, respectively, at the upstream, middle, and downstream. The correspondent lateral pressure head distribution uniformity ranged within 97.1% and 99.6%. Soil water content varied within 0.2175 ± 0.0165; 0.206 ± 0.0195 and 0.284 ± 0.100 beneath the inlet, the behalf, and the lateral end tip. The correspondent soil water distribution uniformity was higher than 80.7% all-over irrigation season. Based on the recorded results, the proposed approach could be a helpful tool for accurate SDI systems design and best water supplies management. Nevertheless, further trials are needed to assess the approach reliability in different cropping conditions.


Introduction
Water scarcity is among the main problems to be faced by many societies and the world in the 21st century [1].The use of water-efficient irrigation is one of the most practical options to reduce global water scarcity [2].Subsurface drip irrigation (SDI) provides the opportunity to record consistently water use efficiency over traditional methods, including surface drip irrigation (DI) [3][4][5].Several field trials revealed relevant profits on managing SDI for crops' production.In fact, SDI system allows the direct application of water to the rhizosphere maintaining dry the nonrooted top soil.This pattern generates numerous advantages such as minimizing soil evaporation and then evapoconcentration phenomenon [6].Comparing evaporation from surface and subsurface drip irrigation systems, Evett et al. [7] reported that 51 and 81 mm were saved with drip laterals buried at 15 cm and 30 cm, respectively.Patel and Rajput [8] recorded maximum onion yield (25.7 t ha −1 ) with drip laterals buried at 10 cm, whereas Ombódi et al. [9] recorded an average yield ranging between 40.7 and 54.6 t ha −1 for onion in irrigated conditions.Also, with SDI systems more uniform moisture distribution, in the vadose zone (than with drip irrigation systems), was observed, and thus drainage and surface evaporation were less with SDI [10,11].Automation of irrigation systems has the potential to provide maximum water use efficiency by maintaining soil moisture within an optimal interval ascertaining the best plant growth [6].
This experimental study aimed to assess the reliability of a new approach for SDI laterals' design accounting for the soil water-retention characteristics and the roots water extraction.The proposed approach provides systematic irrigation management based on fixed water suctions in the vadose zone.

2
International Journal of Agronomy

Site.
Experiments were carried out in the experimental station of Arid Regions Institute of Chenchou (Gabès) whose geographical coordinates are latitude = 33.88∘ North, longitude = 9.79 ∘ East, and at an altitude of 59 m.Average monthly temperature ranges were between 10.4 ∘ C (January) and 28.6 ∘ C (August).Average annual rainfall is 162 mm while potential evapotranspiration (ETP) is 1430 mm/year.
Field trials were performed from May 26th up to September the 15th 2014 in tomato (Feranzi variety) plot (86.0 × 8.0 m 2 ).Seedlings' rows were 1.60 m distant while crop plants were 0.40 m apart.Each row crop was irrigated by a single SDI lateral buried at 15 cm depth.According to Najafi [12] and Zotarelli et al. [13], tomato crop irrigated with laterals buried at   = 15cm depth's leads to the better yields, whereas Machado and Oliveira [14] found that tomato roots were concentrated mainly within the [0-40 cm] top soil layer under DI and SDI irrigation systems.
For soil physical characterization, four representative profiles were randomly chosen (within the plot).In each profile, soil samples were collected on four layers: 0-20, 20-40, 40-60, and 60-80 cm.Analyses were focused on properties that account for soil moisture holding and water suction evolution, namely: texture, bulk density (  ), and water content-pressure head relationship.

Method.
According to Hammami et al. [6], the minimum pressure required at the upstream end of nontapered flat SDI lateral is whereas the maximum pressure head (ℎ  ) required at the upstream end of the lateral is with ℎ  and ℎ  being minimum and maximum required pressure heads (m) at the beginning of the lateral.  is laterals depth of burial (m).  are total pressure head losses (m) along the lateral.ℎ op is optimal soil water suction (m) for crop's growth.Δℎ op is interval of variation of the optimal soil suction (m).Δℎ min is minimum differential pressure head for emitters operating.According to Hammami et al. [6], the soil capillary capacity () is the highest if the second derivative of the soil moisture content with respect to the suction head is zero.Thus, using van Genuchten [15] model, The optimal suction is straight fully derived as follows: Nonlinear adjustment of discrete data (, ℎ) allows deducing   , , and  values from the fitted expression (ℎ) (14).
Gärdenäs et al. [16] reported that tomato crop tolerates (without noticeable yield decrease) a soil water pressure variation in the interval [−800, −2 cm].Then, Δℎ op = ± 400 cm was considered.Therefore, for an optimal tomato crop's growth, the soil water pressure (ℎ) should be maintained within the interval as follows: To avoid any soil saturation risk, we retained 00.00 > ℎ (cm) ≥ −401.47.
(5  ) Consequently, the correspondent optimum water content ( op ) should be maintained within the interval as follows: A minimum value Δℎ min for the emitter operation is required.This threshold Δℎ min is dependent on the structural form, dimension, and material of the emitter pathway.For any emitter model, Δℎ min may be inferred from the emitter discharge-pressure head relationship provided by the manufacturer.Then, the minimum pressure ℎ * min into emitter should respect the following condition: A trapezoidal labyrinth long-path emitter with a minimal differential operating pressure head of Δℎ min = 500 cm was used; then So, the required pressure in the emitters should be between ℎ min req and ℎ max req , with Since the pressure head, in the soil around the laterals, should vary between −401.47 and 00.00 cm, emitters discharge  (l/h) should be maintained between Each lateral is equipped with  = 86/0.4= 215 emitters; therefore its flow rate  should comply with 160.5 ≤  (l/h) ≤ 462.5. ( The proper laterals' diameter used to ensure the maximum discharge ( max = 462.5 l/h) was Ø = 16 mm.Thus using Watters and Keller [17] formula, the total lateral's pressure head loss is equal to Finally, the maximum inlet lateral pressure head was determined using (2) as follows: In order to maintain the lateral inlet pressure head (ℎ  ) constant (equal to or less than 641.53 cm), two interconnected reservoirs were used.Water was pumped to the first reservoir (capacity = 120.0m 3 ) that supplies the second one (capacity = 1.00 m 3 ) which diverts water to the irrigation network.The water level inside the second reservoir was maintained constant thanks to a mechanical float.The pump was controlled by an electric float (Figure 1).
To record lateral's pressure head (ℎ  ), suction, and the correspondents soil water content (ℎ) spatial-temporal evolutions, three measurement sites were set along the lateral: at the inlet  = 0.0 m, at the behalf  = /2, and at the end tip  = .In each measurement site, the installed pieces of equipment were a  piezometer (connected on the lateral), three TDR access tubes, and 9 Watermark probes (three probes per layer buried at the distances  = 0.0 cm;  = 16.0 cm; and  = 32.0 cm perpendicular to the lateral (Figure 2)).Soil water content values were recorded for the following depths:  = 10; 15; 30; 50; and 70 cm.A water meter device has been installed at the laterals' inlet in order to record the delivered water volume.Simultaneously lateral flow rate was measured several times a day.Such measurements allow determining the average daily flow rate variation (from crop transplantation to harvest season).In sum, the following variables were recorded: (i) The spatial-temporal soil water content (, , ) variation within the root zone around the lateral.(ii) The spatial-temporal soil water suction ℎ(, , ) variation.(iii) The spatial-temporal pressure head ℎ  (, ) variation inside the lateral.(iv) The temporal lateral's flow rate () variation.

Results and Discussions
3.1.Physical Soil Characteristics.Mean values of particle size proportion, bulk density (  ), and soil water contents (at saturation   , field capacity   , and wilting point   ) for the four sampled soil layers are summarized in Table 1.These results showed that clay and silt proportions are relatively equiponderant all-over the soil profile while sand proportion decreases from the surface up to 60 cm depth.So the experimental plot is loamy sand textured soil all-over the profile but becomes as fine as it is deep.The bulk density and the soil holding capacity (roughly 100 mm/m) values confirm such texture tendency.
The   and their ℎ  correspondent values (measured in situ) were fitted to van Genuchten [15] formula, using RETC model (Figure 3).So the inferred analytical expression of the soil retention curve was  = 0.096 + 0.289 3.2.Soil Moisture Distribution.Temporal soil water content  evolution in the soil depth  = 10 cm, at the inlet ( = 0), at the behalf ( = /2), and at the lateral end tip ( = ), is depicted in Figure 4. All-over irrigation season, recorded  values ranged within 0.385 >  ≥ 0.184 cm 3 cm −3 for  = 0 and  = /2.Then, it was maintained within the predicted interval 0.385 >  op ≥ 0.184 cm 3 cm −3 (6) optimal for the tomato growth, while, underneath lateral end tip ( = ),  values were almost slightly lower than 0.184 cm 3 cm −3 .This difference could be attributed to the total pressure head losses occurring along the lateral that subsequently induces a slight emitter discharge decrease.Safi et al. [18] reported that an increase of SDI laterals' length leads to a decrease of all uniformity parameters.Also, such discrepancy could be due to measurement errors on  and/or ℎ values.Haverkamp et al. [19] reported that an error of only 2% of  value could cause a relative error of 24% of soil water pressure head.The same trends of the soil moisture distribution were recorded in the soil depth  = 15cm (Figure 5), where  remained higher than the minimum prescribed threshold (ℎ op − Δℎ op ) = 0, 184 cm 3 cm −3 , at  = 0 and  = /2 but still slightly lower than that threshold at the lateral end tip.Such trend confirms the above finding.
In the soil depth  = 30 cm, water content values remained roughly confused with the prescribed minimum threshold (at the inlet  = 0) at the lateral behalf ( = /2) but slightly lower (at the end tip  = ) than such threshold 0.184 cm 3 cm −3 (Figure 6).However, in the deeper soil layers  = 50 cm and  = 70 cm, water content values remained approximately invariant lower than 0.184 cm 3 cm −3 all-over irrigation season and for whole lateral length (Figures 7 and  8).These results could be explained by the fact that supplied water (by the lateral) was not so enough to reach such depths.So, deep water and then nutrients losses were negligible.Thus the used approach could be useful tool to improve SDI irrigation efficiency.The above results validate the systematic SDI irrigation management.Lazarovitch et al. [20] proved that soil hydraulic properties affect outlets flow rate in SDI irrigation system.To assess the water distribution uniformity along the laterals, we determined the coefficient of uniformity (CU) values throughout irrigation season.
where   is average soil water content for different depths in the three soil profiles ( = 0,  = /2, and  = ) and, at a given date,  (,) is soil water content in the coordinates (, ) and  is number of the sampled points.
The recorded CU values are always higher than 80.7% and the mean value was 84,3%.These results confirm those of Ben International Journal of Agronomy    Ali et al. [3] who reported that soil water content underneath SDI system was always higher and especially varied within narrower interval than under drip irrigation system.Gil et al. [21] recorded a lower variability of buried emitters' discharges compared to on surface ones.water losses were recorded.Yao et al. [22] reported that the back pressure risk (or over pressure) occurring underneath subsurface lateral could be addressed by rigorous network design.

Lateral Pressure Head and Flow Rate.
Because the pressure head  in the supplying reservoir was maintained constant equal to 641.53 cm, the pressure head at the lateral inlet ( = 0) remained also constant ( ≈ 640 cm).However,  values inside the behalf and at the lateral end tip were slightly lowered (ranged between 600 and 640 cm) (Figure 10).Such slight variation could be attributed to the linear and nonlinear head losses along the lateral.Though the lateral inlet pressure head was maintained constant, the correspondent flow rate   was noticeably variable within 236 ≥   (/ℎ) ≥ 184 but ranged within the fixed interval (11).Such variation could be explained by the soil (around the lateral) suction variation due to the soil water redistribution enhanced essentially by roots' water uptake.It should be International Journal of Agronomy  stressed that daily lateral discharge was always higher than nightly one (Figure 11).This slight difference (between daily and nightly discharges) highlighted the higher roots' water uptake enhanced by intensive physiologic activities by day times.

Conclusion
The objective of this work aimed to check the reliability of a new approach of SDI laterals design for a systematic irrigation management monitored by soil suction variation close to the outlets.Recorded results showed that, without human intervention for irrigation management, water content, in the soil layer 0-40 cm, remained within the interval ](ℎ opt +Δℎ opt ); (ℎ opt −Δℎ opt )], corresponding to the optimal humidity interval for tomato growth.Soil water content in the deep layers ( = 50 cm and  = 70 cm) remained roughly constant but lower than (ℎ opt − Δℎ opt ).So neither saturation risks nor water and nutrients losses by deep percolation were observed within the vadose zone.In addition, irrigation water uniformity along the lateral was almost higher than 80.7%.So the design procedure illustrated in this paper provides the appropriate emitters discharge and the inlet lateral pressure head that fit the plant roots water uptake.Even though, soil water content recorded at the lateral end tip remained lower than the minimum optimal threshold throughout the entire cropping cycle.Therefore, the proposed approach could be an efficient tool for rigorous SDI lateral design.But further field trials are needed to effectively confirm such finding.

Figure 2 :
Figure 2: Profile of a measurement site.