Variation in actual corn (Zea mays L.) evapotranspiration, single, and dual crop coefficient under different point source irrigation systems in a semiarid region

Close management of irrigation could have considerable impacts on water resources, especially for cropping systems dominated by corn. The experiment was carried out to compare the influence of porous capsule irrigation (PCI), surface drip irrigation (DI), and subsurface drip irrigation (SDI) systems, with or without mulching, on actual evapotranspiration (ETc act), crop coefficients (Kc single and Kc dual), biomass yield, and water use efficiency (WUE) of corn in a semiarid region of Iran. The experiment was arranged in a split-plot design with the three irrigation systems assigned to the main plots and two mulching (with or without foil type) treatments (M1 and M2) assigned to the sub-plots. The corn ETc act varied significantly (P < 0.05) with the different irrigation systems, being (mm) 389.8 for PCI, 377.0 for DI, and 372.8 for SDI. The highest Kc average and Kcb (0.82 and 1.22) and the lowest Ke (0.12) were observed under the PCI system. The dry and wet biological biomass yields were highest (29.98 and 107 t/ha) under the PCI + M1 treatment, and the lowest (23.19 and 58.54 t/ha, respectively) were under the DI + M2 treatment. The highest WUE (7.89 kg/m3) was also observed under the PCI + M1 treatment; PCI produced the best biological biomass yield, WUE and IWUE in comparison to DI and SDI systems. Accordingly, the PCI system could be a viable alternative to drip irrigation for areas with scarce water resources, particularly for smallholder farmers.


Introduction
Evapotranspiration (ETc) loss is one of the major challenges to irrigated agriculture in arid and semiarid regions, where the agriculture sector is the main consumer of water. Evaporation occurs predominantly from the exposed soil fraction in the field and free water surface (Allen et al. 1998). Mulching is a recommended water management practice for reducing water loss due to evaporation from soil surface and creating more favorable conditions for plant transpiration (Tiwari et al. 2014). Drip irrigation plus surface mulching is a new agricultural water-saving irrigation technology, which can prevent or reduce water loss from the soil and has been widely used in arid and semiarid regions (Qin et al. 2016;Reddy et al. 2018). Mulching increases soil temperature; on the other hand, drip irrigation in combination with surface mulch maintains soil moisture by reducing evaporation, improves soil fertility, prevents the growth of weeds, and also helps to evenly distribute moisture to reduce plant stress (Paul et al. 2013;Panigrahi et al. 2016). There are several reports of increases in crop yield and water use efficiency under combined drip irrigation and mulching in different agro-climatic regions and soil conditions (Biswas et al. 2015;Thentu et al. 2016;Liu et al. 2017;Zhang et al. 2017). Meanwhile, our knowledge of some traditional irrigation systems in relation to crop development is limited. For example, porous capsule irrigation is an effective adaptation of buried clay pot irrigation (Silva et al. 1981a, b;Bainbridge 2002;Bainbridge et al. 2008). Furthermore, the supply of water with small porous capsules is a cost-effective method that can be used in small-scale farms in arid and semiarid regions. This method was used for the planting of a wide range of annual and perennial plants including horticultural crops (Setiawan et al. 1998), corn (Bainbridge 2001), watermelons (Soomro 2002), tomatoes (Tesfaye et al. 2012), and turnip, okra, and eggplant (Siyal et al. 2011).
The crop coefficient (K c ), which measures the effect of crop development status on evapotranspiration, is slightly affected by climate, but significantly by crop type and irrigation method, soil surface conditions, and others (Zhang et al. 2013). The K c can be applied as a single crop coefficient (K c single ) which is influenced by the combined effect of evaporation (E) and plant transpiration (T p ) or as dual crop coefficient (K c dual ) which is expressed by soil evaporation coefficient (K e ) and basal crop coefficient (K cb ), separately (Allen et al. 1998). The K c dual can provide a more accurate assessment of the effects of soil moisture by rainfall or irrigation and also the effects of using mulches to control soil evaporation. Crop interception is not considered by the K e factor. Hence, to some extent, the K c dual can improve the accuracy of estimation of ETc when used with appropriate calibrated base data (Allen et al. 2011;Rosa et al. 2012;Zhang et al. 2013). Corn is one of the main cereal crops in the world, ranking third in cereal production in area cultivated and production (FAO, 2011). Close management of irrigation could have significant impacts on water resources, especially where corn dominates the cropping system. To further improve water management for corn under different localized irrigation systems in arid and semiarid conditions, this study aimed to investigate the variations in evaporation and transpiration, soil evaporation coefficient, and basal coefficient during growth stages for different point source irrigation systems under semiarid conditions.

Experimental site
The study was conducted at the experimental farm of Agricultural Engineering Research Institute, Karaj (35°46′ N; 50°55′ E; 1260 m asl), Iran. The local climate is semiarid with cold and dry winters and hot and humid summers. The mean annual precipitation and annual temperature were 271 mm and 15.5 °C for the period 2009-2014. The daily climate data for the growing season (August to November 2014) were collected from a synoptic meteorology station located 5 km from the farm. Table 1 and Fig. 1 show the monthly average and daily variations of some meteorological variables during the corn growing seasons.
The mean daily maximum and minimum air temperature during the corn growing season ranged from 20.8 to 41.4 °C and 6.9 to 23.5 °C, respectively. The mean daily relative humidity varied from 8 to 57%, wind speed from 3 to 7.5 m/s, and solar radiation from 10.43 to 14.13 MJ/m 2 day ( Fig. 1). On average, wind speed and relative humidity were highest in August and November, while the average solar radiation was highest in August. The experimental soil was loamy with a mean bulk density (0-30 cm) of 1.42 g/cm 3 and slightly alkaline with a pH of 7.9. Soil characteristics and irrigation water quality for the experimental site are given in Tables 2 and 3.

Crop management
To monitor water consumption under the different point source irrigation systems, 21 volumetric mini-lysimeters filled with excavated soil were installed in the corn field of 18 ha. The mini-lysimeters were cylindrical with a diameter of 40 cm and a depth of 70 cm (area of 1256 cm 2 and volume of 87,920 cm 3 ) (Dehghanisanij et al., 2020). Inside each mini-lysimeter, three forage corn seeds (Single Cross 704) were sown to 6-cm depth and 13-cm spacing, on 6 August 2014 (Fig. 2).
Fertilizer was injected into the irrigation water, starting from the 3-4 leafed stage to 45 days before harvesting. The fertilizer was applied as 250 kg/ha ammonium phosphate (16-20-0) and 200 kg/ha nitrogen (urea-46-0-0). Hand weeding was used to eliminate weeds during the corn growing season.

Irrigation management
The corn experiment was irrigated by porous capsule irrigation (PCI), surface drip irrigation (DI), and subsurface drip irrigation (SDI) system. The DI and SDI systems were equipped with 40-cm emitter spacing and discharge of 4 L/h. The SDI and PCI systems were installed at 30-cm soil depth. Each porous capsule was 30 cm long and 6.5 cm wide as shown in Fig. 3 and was buried vertically in the soil.
Irrigation interval was twice a week for DI and SDI systems and daily for PCI system.
where ETc is crop evapotranspiration; ET o is reference evapotranspiration (mm/day); R n is net radiation (MJ/m 2 / day); G is soil heat flux density (MJ/m 2 /day); T is mean temperature (°C); U 2 is wind speed at 2-m height (m/s); ɣ is psychrometric constant (k Pa/°C); ∆ is slope vapor pressure curve (k Pa/°C); e a is actual vapor pressure (k Pa); and e s is saturation vapor pressure (k Pa). In our study, the K c coefficient for corn recommended by Farshi et al. (1997) was applied. The volume of water applied to each lysimeter was measured using a calibrated water meter.

Experimental design
The experiment was laid out in a split-plot design with the three irrigation methods-(i) PCI, (ii) DI, and (iii) SDIassigned to the main plots and two soil coverage options ((i) mulch cover (M 1 ) with 100% density (aluminum foil) and (ii) no mulch (M 2 ) cover) assigned to the sub-plots. All the minilysimeters were located in an 18-ha corn farm within the maize canopy (six mini-lysimeters: porous capsule irrigation (PCI), six mini-lysimeters: surface drip irrigation (DI), and six mini-lysimeters: subsurface drip irrigation (SDI) system) ( Fig. 4). Each treatment was replicated three times in a randomized complete block design. (1)

Actual corn evapotranspiration (ETc act )
Daily actual corn evapotranspiration (ETc act ) of each minilysimeter was calculated by applying the water balance method using Eqs. (3) and (4) (Allen et al. 1998): where P is the rainfall (mm); I is the irrigation depth (mm); D is the water loss through drainage from the mini-lysimeter (mm); R is the runoff (mm); ΔS is the change in soil water storage in the mini-lysimeter (mm), and S t and S t-1 are the amounts of water in the root zone at the beginning and end of the period (mm). (3)

Soil surface evaporation (E) and plant transpiration (T p )
To measure evaporation (E) from the surface of bare soil and soil with mulch cover, three other mini-lysimeters were placed in the experimental field with the same soil but without plants along the other mini-lysimeters and within the maize canopy. The E was measured from the difference between the amount of intake and drainage water in mini-lysimeters for each irrigation interval. Plant transpiration (T p ) was estimated from the difference between ETc act and E according to Moran et al. (2009) and Dehghanisanij et al. (2020) thus:

Crop coefficient (K c )
Two forms of K c are presented for calculating crop coefficient including single and dual crop coefficient forms. In this study, the single crop coefficient (K c single ) and dual (Allen et al. 1998) were modified for use based on the climatic conditions of the study area.

Water use efficiencies
Water use efficiency (WUE, kg/m 3 ) and irrigation water use efficiency (IWUE, kg/m 3 ) were calculated by using Eqs. (6) and (7) (Sakthivadivel et al. 1999;Ati et al. 2012). IWUE would increase with increasing irrigation water application efficiency but also with increasing precipitation (because then, less irrigation is needed). Accordingly, IWUE can be used only for comparing different systems at the same site, but not for comparison between different sites or climates with different amounts of precipitation.

Plant and biomass sampling
Leaf area index (LAI), plant height (H), and stem diameter (D) as well as dry and wet biological biomass yields were measured across growth stages. Leaf area index (LAI) was measured with the electronic leaf area-meter, CI -202, seven times during the growing season.

Statistical analyses
Analysis of variance was carried out to evaluate the main and interaction effects of the three irrigations systems and mulch treatments on evapotranspiration, yield, and water use efficiency of corn by using the SAS package. Treatment means were separated using the least significant difference (LSD) tests at P ≤ 0.05. Table 4 shows the total ETc act during the corn growing season in plots without surface mulch. The total ETc act (mm) was 379.8 for PCI, 377.0 for DI, and 371.9 for SDI. The lowest ETc act (mm/day) occurred at the initial stage, being 3.31 for PCI, 2.34 for DI, and 2.34 for SDI. The daily ETc act increased rapidly to peak values (11.61 mm for PCI, 9.33 mm for DI, and 9.24 mm for SDI) at 44 days after sowing. Previous studies showed that seasonal variations in ETc act for corn ranged from 200 to 663 mm for different climatic and regional conditions (Dehghanisanij et al. 2006(Dehghanisanij et al. , 2020Chuanyan and Zhongren 2007;Liu et al. 2017;Zhang et al. 2017;Akhavan et al. 2018). Variations in ETc act during the growing season are shown in Fig. 5.

SDI
Daily ETc act (mm/day) varied from 3.31 to 11.61 for PCI, 2.34 to 9.33 for DI, and 2.34 to 9.24 for SDI. The total actual evapotranspiration under PCI was almost solely due to crop transpiration (T p ), because for this system a non-wet soil surface occurred during the growing season and the total irrigation water applied was the T p . There was no E, drainage and deep percolation in the PCI system, resulting to zero loss in E from the soil surface (Fig. 6). Under rainfall event, a small part of E from the soil surface and crop interception may occur.
Under PCI, the soil surface usually remained dry, in contrast to DI and SDI, leading to a reduction in E and an increase in transpiration, WUE, and IWUE (Romero et al. 2004;Badr et al. 2010). The next lowest ETc act values were obtained under SDI because the rate of E was lower (soil surface drier) than that under DI. As shown in Table 5, the effects of irrigation systems differed significantly (P ≤ 0.01) for ETc act and K c . However, there were little differences between mulching treatments in their interaction effects with irrigation systems on ETc act and K c ( Table 5).
The PCI system had a higher (P < 0.01) K c value (0.82) than DI or SDI over the whole crop season. The lowest K c values were obtained with mulching because the mulch reduced E compared with the no mulch treatments (Table 5). These results are in agreement with previous reports (Vickers 2001; Mata et al. 2002;Yaghi et al. 2013). Since the crop coefficient changes with location and planting dates, a location-specific estimation of K c for calculating corn ETc act is essential. The changes in K c and LAI during the corn growing season are plotted as a function of days after planting (DAP) (Fig. 7).
Analysis of changes in K c was done using the relationship between LAI and K c observed on the same days when LAI was measured. Based on these data, a new K c curve was plotted for 7 days of LAI sampling as illustrated (Fig. 6). This was done to divest days with wet plants and soil because there was no irrigation or precipitation when LAI was sampled in the middle of the growing season. As shown in Fig. 6, an increase in LAI increased the K c markedly and the LAI generally peaked similarly with K c .

Variations in single and dual crop coefficient (K c single and K c dual )
The single (K c single ) and dual (K c dual ) crop coefficients for corn growth stages are shown in Table 6; their variations are also plotted as a function of days after planting (DAP) (Fig. 8).
The K c single values calculated by FAO-56 were 0.3 for the initial stage, 0.3-0.9 for the crop development stage, and 1.2 for the mid-season stages. In this study, the recommended K c single values were modified (based on the climatic conditions of the study area) to 0.3, 0.88, and 1.35 for PCI; 0.3, 0.86, and 1.34 for DI; and 0.3, 0.87, and 1.35 for SDI, respectively ( Table 6).
The K c dual includes the basal crop coefficient (K cb ) and the soil surface evaporation coefficient (K e ). The K cb suggested by FAO-56 for the initial stage, crop development, and mid-season stages (K cb ini , K cb dev , and K cb mid ) were   (Table 6).
The highest K cb was 1.22 for the PCI system at the midseason stage. K e , based on the mini-lysimeter measurements, varied temporally during the corn growing season. The average K e value was higher at the initial stage and decreased gradually to a minimum value of only 0.12 under PCI at the mid-season stage (Table 6). These results indicated that the soil evaporation coefficient decreased during the  corn growing season because of the increase in soil surface shading by the crop canopy. The K c dual was higher than the K c single at the initial, development, and mid-season stages. Figure 8 shows that the E was higher than the transpiration at the initial-season stage, and with the increase in plant shading, E became lower than the transpiration from the crop at the development and mid-season stages.

Canopy characteristics of corn
Data for leaf area index (LAI), plant height (H), and stem diameter (D) during the growing season are summarized in Table 5. The irrigation systems had a significant effect (P < 0.05) on LAI, H, and D, but mulching treatments had a significant effect (P < 0.05) only on LAI and D and not on H. The highest LAI, H, and D occurred under PCI, being 4, 182 cm, and 2.9 cm, respectively. These increased with mulching, reaching 4.1, 189, and 2.9 cm, respectively (Table 5). Although the interaction of irrigation methods and mulching treatments had little effect on LAI, H, and D, the PCI in combination with mulch (PCI M 1 ) treatment had the highest values of 4.23, 191.6 cm, and 3.0 cm, respectively.
The lowest values were obtained under DI without mulch (DI M 2 ), being 3.4, 169.2, and 2.9 cm, respectively. For all the irrigation and mulch treatments, the lowest LAI, H, and D occurred at the initial stage, then increased rapidly to peak values at the mid-season stage (Fig. 9).
The largest values of LAI, H, and D were found during the mid-season stage, 70 days after sowing under mulch condition. The positive effect of mulching on LAI was previously reported for wheat (Xie et al. 2005). The temporal variations in the H and D of corn are shown in Fig. 9.

Biomass yield
The average dry and wet biological biomass yields (28.41 and 97.70 t/ha) under PCI were significantly higher than those under DI (25.13 and 63.21 t/ha) and SDI (26.67 and 81.99 t/ha) ( Table 5). The better performance under the PCI can be explained by the maintenance of favorable moisture status in the root zone, enabling better utilization of nutrients from the wetted area (Zotarelli et al. 2008;Badr et al. 2010). The mulched (M 1 ) treatment had a higher (P < 0.05) yield (27.86 t/ha) than the un-mulched (M 2 ) treatment. Besides the reduction in E, there were previous reports of significant yield increases for chili and cantaloupe under mulching (Nijamudeen and Dharmasena 2002;Seyfi and Rashidi 2007). Mulching is known for reducing the weed dry matter due to physical hindrance and by reducing the amount of solar radiation reaching the weeds on the soil surface. Also, other studies have shown multiple positive effects of mulching on biological biomass yield in corn (Khurshid et al. 2006). The PCI system with mulch (M 1 ) prevented evaporation from the soil surface, leading to a general increase in measured T p from corn. The increase in biomass yield with mulch (M 1 ) is in agreement with earlier reports (Stein 1998;Chakraborty et al. 2010;Balwinder et al., 2011;Ram et al. 2013

Water use efficiency (WUE) and irrigation water use efficiency (IWUE)
The WUE and IWUE under PCI were superior to those under DI and SDI (Table 5). Under PCI, the soil surface usually remained drier than under DI and SDI and this led to an increase in T, WUE, and IWUE (Romero et al. 2004). The mulching (M 1 ) treatment also had a significant (p < 0.05) effect on WUE and IWUE which were 7.46 and 7.33 (kg/ m 3 ). The interaction of irrigation methods and mulching had negligible effects on IWUE and WUE during the growing season (Table 5). Across treatments, the IWUE ranged from 6.10 to 7.89 kg/m 3 , while the WUE ranged from 6.14 to 7.89 kg/m 3 . The PCI system with mulch (PCI M 1 ) treatment had higher IWUE and WUE than the other two irrigation methods due to its minimal E and absence of deep percolation. The lowest IWUE and WUE (6.10 and 6.14 kg/m 3 ) were under DI without mulch (DI M 2 ). Uniform mulching of the soil surface reduces E, increases soil moisture content, and leads to higher WUE and crop yield due to more plant available water for transpiration (Jin et al. 2009;Wang et al. 2015;Shen et al. 2012).

Conclusions
Water loss through evapotranspiration (ETc) is one of the major challenges to irrigated agriculture. A field experiment was carried out to study the influence of different point source irrigation systems, namely porous capsule irrigation (PCI), surface drip irrigation (DI), and subsurface drip irrigation (SDI), with or without mulching, on actual evapotranspiration (ETc act ), biological biomass yield, and water use efficiency (WUE) of corn in the semiarid climate of Iran. The irrigation systems had a significant impact on the ETc act . The highest biological biomass yield was recorded under porous capsule irrigation (PCI). The results confirmed that mulching significantly optimized ETc act , E, T P , biological biomass yield, and WUE, especially under PCI. The better performance under the PCI system can be explained by the maintenance of favorable moisture status (not saturated) in the root zone that helped the plants to utilize water and nutrients more efficiently from the wetted area. The dual crop coefficient (K c dual ) was recommended to estimate the ETc act . Water use efficiency was higher under PCI than other irrigation systems; we suggest that PCI can meet the water requirement of corn under the same applied irrigation water. Therefore, under water deficit conditions, PCI is recommended for semiarid regions. The daily actual corn evapotranspiration (ETc act ) in this study was calculated by applying the water balance method but using a weighing lysimeter could improve the quality of data collected. The PCI is not easy to install in large farms; we suggest improving irrigation management and scheduling under SDI close to the PCI performance with longer irrigation intervals and shorter irrigation time.