Evaluation of irrigation scheduling to maximize tomato production using comparative assessment of soil moisture and evapotranspiration in restricted irrigated regions

Abstract

Evaluating and comparing two different irrigation scheduling methods like soil moisture (SM) and evapotranspiration (ET)-based irrigation scheduling methods were assessed scientifically to maximize tomato yield, especially in areas where restricted irrigation facilities are available. The experimental design was arranged in a Randomized Complete Block Design (RCBD) with 100%, 75%, and 50% water requirements with three replications. The water was applied to furrows using Randomized Block Design (RBC) flume with an inflow rate of 1.62 l/s. The total amount of water applied in SM- and ET-based methods were 229.1 mm and 280 mm, respectively. The collected data were analyzed using ANOVA (statistical software) at 5% significance level. Values on agronomic characteristics, yield attributes, and yield were determined for those two methods. T1–T6 are the total treatments used in this research. Out of which, T1, T2, and T3 are taken as SM-based represented for 100%, 75%, and 50% of requirements and T4, T5, and T6 as ET-based represented for the same percentage of requirements. The results did not illustrate a huge difference in yield potential. The highest net benefit on yield was obtained at T2 (SM-based), whereas the lowest in T6 (ET-based). It was also observed that SM-based method saved around 18.2% of irrigation water compared to ET-based approach. Therefore, to cultivate high water demanding crops such as tomato in water scarce areas, SM-based irrigation scheduling method is comparatively better than ET-based as far as yield is concerned.

Keywords:

• ET-based method
• irrigation scheduling methods
• RBC Flume
• Sm-based
• tomato

1. Introduction

Impeccable use of scientifically tested irrigation system reduces the food production crisis irrespective of any climatic change intervention. The methods in particular are too crucial for agricultural production in areas where either the water resources are scarce or irrigation facilities are limited (Gu et al., 2017). Agriculture is the backbone of the Ethiopian economy. The rural community is completely dependent on farming for their sustainability. In water scarce areas of the country, cultivation of high water demanding crops without irrigation is impossible. Limited irrigation facilities and non-perennial a supply of water through it in the country are practically inadequate (Awulachew & Ayana, 2011). The primary issue occurring in the irrigation schemes is improper irrigation water application, particularly over or under irrigations of crop fields that result in lowest water use efficiency (WUE). Test of relative irrigation supply (the ratio of irrigation supply to crop water demand) at scheme level for 10 irrigation schemes in Ethiopia ranged between 0.5 (insufficient) and 5.0 (excesses) for the period of January–May and between 0.8 (insufficient) and 7.0 (excesses) for the period of June–December (Agide et al., 2016). Non-scheduled and inconsistent water application regardless of crop type (Daniel et al., 2020) has also been observed in many cases. Currently, this has become a major worry among the scientists and scholars for enhancement of WUE (Daniel et al., 2020; Derib et al., 2011; Yohannes et al., 2019). In order to bring an effective agricultural water management and increase agricultural productivity, farmers, irrigation experts, water resources managers, and decision makers suggested proper irrigation scheduling techniques (Daniel et al., 2020; SAI Platform, 2010).

Scientific irrigation scheduling methods are required to determine when, where, and how irrigation needs to be facilitated (O’Shaughnessy & Rush, 2014). Especially, there are two distinctive methods: one by directly monitoring SM using SM sensors and the other using weather data to account for soil-water by soil water balance approach (Sharma, 2019). Others classify the methods into three “scientific” irrigation scheduling methods: plant, soil, and climate based (O’Shaughnessy & Rush, 2014). Ali’s (2010) classification is more explanatory than other classifications that put them in two main groups as Farmer and Advanced method. From these different classifications, soil- and climate-based methods are considered as “typical” methods as they represent SM and ET-based irrigation scheduling methods (Hanson et al., 2000; Sharma, 2019). In this study, these two commonly applicable typical methods are comparatively evaluated for their WUE.

The general approach for the ET-based method is maintaining balance of current available SM to the plant by tracking the ETc losses and the additions from irrigation and precipitation (Nouri et al., 2016). This method sometimes performed better with significantly less water for the same yield compared to some other irrigation systems (Bartlett et al., 2015; Migliaccio et al., 2016). The method normally uses climatological parameters to estimate ET for a “well watered” crop. If it is correctly applied, this scheduling practice will have conserving water and maintaining crop yield and quality (McCready & Dukes, 2011).

SM-based scheduling depends on soil water holding capacity between Field capacity (FC) and permanent wilting point (PWP). In this method, there are enormous options of tools to conduct irrigation scheduling, such as automatic irrigation controllers, neutron scattering, tension meters, capacitance sensors, electrical resistivity measurements, heat pulse sensors, watermark sensors, and fiber-optic-sensors (Hedley & Yule, 2009; Yadav et al., 2020). Comparatively, dielectric sensors like time domain reflectometry (TDR) require less field maintenance and have a greater potential for commercial adoption (Munoz-Carpena et al., 2005). During the research, volumetric water content was estimated by TDR, and the calibration was performed by gravimeter (Evett, 2008).

Under normal circumstances, either SM-based (Junwang et al., 2017; Ruixiu, 2018; Vellidis et al., 2016) or ET-based irrigation scheduling methods are adopted. However, the methods varied as per the data availability and performance based on non-scientific preconditions. The method that is more appropriate needs to be restricted to provide options to farmers for managing their irrigation systems according to their preferences and on-site conditions (Juan et al., 2009). Thus, technical knowledge, skills, and unbiased information about the performance of the methods are required to evaluate the extent of effect of the methods. In another way, comparing the effectiveness of the two irrigation scheduling methods for WUE to their climatic condition (Aguilar et al., 2015; Grabow et al., 2013; Juan et al., 2009; Miller et al., 2018; Sabine et al., 2016) needs to be well understood (BlonquistJr et al., 2006; Varble and Chávez, 2011).

Past research studies are still unclear about the extent of saving of water between irrigation scheduling methods. The results vary from experiment to experiment and place to place for different attributes and a number of variables such as climate (geographic and seasonal), crop (type and condition), irrigation method, irrigation manager’s level of expertise, and soil type (N. Cooley et al., 2010). This indicates proper planning and effective execution of similar types of comparative works for variations. Therefore, it is necessary to carry out a comparative evaluation of the two methods in terms of WUE for crops to provide options that could help the users to manage their irrigated farms effectively under prevailing conditions in the area. The major objectives of the present study were to evaluate SM, ET-based irrigation scheduling methods’ effects, and respective WUE of crop and to evaluate SM and effects on deficit irrigated crop on ET-based irrigation scheduling methods.

In the study area, many researchers have used either SM or ET-based irrigation scheduling method. Unfortunately, none of them tried to evaluate and compare the effectiveness of the methods. The farmers must be provided with options to manage their irrigation scheduling according to their preferences and on-site condition (Juan et al., 2009). The growing interest of farmers is to cultivate high water demanding crops such as tomato, which is a big challenge due to water shortage. Appropriate irrigation scheduling method needs to be identifed to sort out the problem for the sustainability of the farmers. This study has been focused on identifying a better irrigation scheduling approach to recommend a healthy irrigated tomato crop under deficit irrigation, especially during dry seasons.

2. Materials and methods

2.1. Study area description

The experiment was conducted near Arba Minch city, southern Ethiopia (Figure Figure 1). It is located around 450 km away from Addis Ababa, the capital city of Ethiopia toward South West direction, geographically at 6º04”N latitude and 37º34” E longitude and at an altitude of 1200 m amsl. Two Rift Valley Lakes, Abaya and Chamo, enclose the study area in the east and southeast, respectively. The experiment was conducted during the period of May–September in 2022. Thirty year’s (1987–2018) historical rainfall and temperature data (Figure Figure 2) were analyzed to understand the past climatic behavior of the area to select the dry season for conducting this research.

Figure 1. Location map of the study area.

Figure 2. Climate data of the study area.

Meteorological station is a source to collect daily weather data such as temperature, rainfall, sunshine hours, wind speed, and humidity that are ET computation parameters for ET-based scheduling located in Arba Minch University.

According to the data from the meteorological station, the area is characterized by bimodal climatic conditions. The average annual rainfall of the area is 921.8 mm.

2.2. Pre-experimental activities

Five furrows per plot of 80 cm spacing, 20 cm deep, and 35 cm wide were provided. The furrow slope along each plot measured as the ratio of vertical distance (m) to horizontal distance (m) is about 0.2%. The plot size is 3.6 m wide and 4 m long with five furrows in each plot. The crop was planted at the side of ridge with 40 cm and 80 cm between plants and rows, respectively. The spacing between each plot and treatment block was 1 m and 1.5 m, respectively. The prepared land comprised of 18 plots each with 4 m × 3.6 m (14.4 m²) (Figure Figure 3). Total experimental field of length and width including work and block space was 28.6 m×17 m (0.048 ha). For establishment, seeds were sown in a nursery bed of 5 m long, 1 m wide, and 25 cm raised bed within 20 cm spaced line.

Figure 3. Layout of experimental field.

Note: T1, T2, T3, T4, T5 and T6 = Total treatments, T1, T2 and T3 = SM-based represented for 100%, 75% and 50% of requirement and T4, T5 and T6 = ET-based represented for 100%, 75% and 50% of requirement. R represents rows (Replications).

2.2.1. Soil sample collection and analysis

Laboratory analysis was conducted to determine the soil bulk density and soil texture from the depths of 0–20, 20–40, and 40–60 cm from nine representative locations. The texture was determined in the laboratory using Bouyoucos hydrometer method. Soil infiltration test was carried out using double ring infiltrometer with outer and inner diameters and height of 53 cm, 32 cm, and 25 cm, respectively, on bare field procedurally (Savva & Frenken, 2002) (Figures 4 and 5).

Figure 4. Soil infiltration test carried out in the field.

Figure 5. Infiltration test graph.

2.3. Treatments

Of the total treatments of T1, T2, T3, T4, T5, and T6, SM-based was represented by T1, T2, and T3 for 100%, 75%, and 50% of requirement and ET-based was represented by T4, T5, and T6 for 100%, 75%, and 50% of requirement, respectively. Comparative treatments from both sides are T1 = T4 = 100% requirement, T2 = T5 = 75% requirement, and T3 = T6 = 50% requirement. For both sides, six treatments were irrigated by three levels of application for total of 18 Randomized Complete Block Design (RCBD) plots (Table 1).

Table 1. Treatments and levels of application

Plots	Application level	Plots	Application level
ST1R1	100%	ET4R1	100%
ST1R2	100%	ET4R2	100%
ST1R3	100%	ET4R3	100%
ST2R1	75%	ET5R1	75%
ST2R2	75%	ET5R2	75%
ST2R3	75%	ET5R3	75%
ST3R1	50%	ET6R1	50%
ST3R2	50%	ET6R2	50%
ST3R3	50%	ET6R3	50%

S and E are for SM- and ET-based methods, respectively; T is treatment; R is replication; 100%, 75%, and 50% are deficit levels.

3. Irrigation scheduling methods

3.1. ET-based irrigation scheduling

FAO Penman-Monteith method was used to compute daily ET_o from meteorological data of the area (Allen et al., 1998; Edamo et al., 2022; Ukumo et al., 2022). Since the specific crop coefficient was not done in the area, the Kc values used in this work were obtained from the findings of Melkasa research center, with similar agro-climatic conditions (Dirirsa et al., 2017). Crop ET is calculated as indicated by Subedi and Chávez (2015) (Equation 1(1) $\mathrm{E}\mathrm{T}\mathrm{c}=\mathrm{K}\mathrm{c}\ast \mathrm{E}\mathrm{T}\mathrm{o}$(1) ).

(1) $\mathrm{E}\mathrm{T}\mathrm{c}=\mathrm{K}\mathrm{c}\ast \mathrm{E}\mathrm{T}\mathrm{o}$(1)

ET-based irrigation scheduling is carried out by managing soil water balance estimation based on the water lost by ET, and its replacement through irrigation or effective rainfall. The balance was performed using Van (2004) for row crops (Equation 2(2) $\mathrm{C}\mathrm{S}\mathrm{W}\mathrm{C}=\mathrm{P}\mathrm{S}\mathrm{W}\mathrm{C}+\mathrm{I}\mathrm{n}\mathrm{e}\mathrm{t}+\mathrm{P}\mathrm{e}-\mathrm{E}\mathrm{T}\mathrm{c}$(2) ).

(2) $\mathrm{C}\mathrm{S}\mathrm{W}\mathrm{C}=\mathrm{P}\mathrm{S}\mathrm{W}\mathrm{C}+\mathrm{I}\mathrm{n}\mathrm{e}\mathrm{t}+\mathrm{P}\mathrm{e}-\mathrm{E}\mathrm{T}\mathrm{c}$(2)

Where CSWC is current soil water content (today, mm), PSWC is previous soil water content (yesterday, mm), Pe is effective rainfall since yesterday (mm), Inet is net irrigation since yesterday (mm), and ETc is crop ET (mm/day).

The first step was to determine the amount of water in the soil (PSWC) at the start of the water budget. Usually, PSW Cat one day earlier from first irrigation is returned to FC. The net irrigation water requirement is determined using Equation 3(3) $\mathrm{I}\mathrm{n}\mathrm{e}\mathrm{t}=\mathrm{E}\mathrm{T}\mathrm{c}-\mathrm{P}\mathrm{e}$(3) (Bennett et al., 2014).

(3) $\mathrm{I}\mathrm{n}\mathrm{e}\mathrm{t}=\mathrm{E}\mathrm{T}\mathrm{c}-\mathrm{P}\mathrm{e}$(3)

Where Inet = net irrigation requirement (mm), Pe = effective rainfall (mm), and ETc = crop ET (mm).

Meteorological data indicates that out of total 117.6 mm of rainfall recorded during the experimental season, 58.9 mm is effectively connected to the crop root zone. Assuming no input water except rainfall, the effective rainfall was determined by a comparison of the rainfall with the crop water used (ETc). If ETc of the crop for the day is greater than the rainfall of that day then Pe becomes equal to the rainfall. Otherwise, Pe equals to ETc.

3.2. SM-based scheduling

For SM-based irrigation scheduling, the amount of irrigation water applied was based on the moisture content at the time of sampling in the root zone. The calculation equations of this method are mentioned under Equations 4 and 5 (Haberle & Svoboda, 2015). Total available soil water was computed (Equation 4(4) $\mathrm{T}\mathrm{A}\mathrm{W}=\left(\mathrm{F}\mathrm{C}-\mathrm{P}\mathrm{W}\mathrm{P}\right)\ast \mathrm{R}\mathrm{D}$(4) ).

(4) $\mathrm{T}\mathrm{A}\mathrm{W}=\left(\mathrm{F}\mathrm{C}-\mathrm{P}\mathrm{W}\mathrm{P}\right)\ast \mathrm{R}\mathrm{D}$(4)

Where TAW is total available soil water (mm), FC is Field capacity (%), PWP is permanent wilting point (%), and RD is effective root depth of the crop (mm). Readily available soil water was computed by Equation 5.(5) $\mathrm{R}\mathrm{A}\mathrm{W}=\mathrm{M}\mathrm{A}\mathrm{D}\ast \mathrm{T}\mathrm{A}\mathrm{W}$(5)

(5) $\mathrm{R}\mathrm{A}\mathrm{W}=\mathrm{M}\mathrm{A}\mathrm{D}\ast \mathrm{T}\mathrm{A}\mathrm{W}$(5)

Where RAW= readily available water (mm), MAD = maximum allowable depletion (mm), and TAW= total available soil water (mm).

3.3. Discharge measurement (RBC flume)

Irrigation water is allowed to pass through submerged circular pipe and then to Replogle Bos-Clemmens (RBC) flume (Figure 6a).

Figure 6. RBC flume and b) discharge measurement.

Procedurally, the flow rate was determined by measuring the water level (SH1) in the flume. The sill-referenced water level (SH1) was converted to flow rate by using Equation 6.(6) $\mathrm{Q}=7\mathrm{E}-07\mathrm{x}\left(\mathrm{S}\mathrm{H}1\right)+0.000626\mathrm{x}\left(\mathrm{S}\mathrm{H}1\right)+0.01569\mathrm{x}\left(\mathrm{S}\mathrm{H}1\right)-0.0665$(6) From the three alternatives of discharge measurement (Figure 7),

(6) $\mathrm{Q}=7\mathrm{E}-07\mathrm{x}\left(\mathrm{S}\mathrm{H}1\right)+0.000626\mathrm{x}\left(\mathrm{S}\mathrm{H}1\right)+0.01569\mathrm{x}\left(\mathrm{S}\mathrm{H}1\right)-0.0665$(6)

Figure 7. The relationship between RBC flume discharge (l/sec) and head (mm).

where Q is discharge (l/sec) and SH1 is sill-referenced water level (mm).

The flume was calibrated to a discharge of 1.62 l/sec for SH1 of 40.0 mm (Figure 6b).

The time (min) of water application to each furrow is calculated using Equation 7(7) $\mathrm{T}=\mathrm{G}\mathrm{I}\mathrm{R}\ast \mathrm{L}\ast \mathrm{W}/60\mathrm{Q}$(7) (Kifle, 2018) as follows:

(7) $\mathrm{T}=\mathrm{G}\mathrm{I}\mathrm{R}\ast \mathrm{L}\ast \mathrm{W}/60\mathrm{Q}$(7)

where T is application time (min), GIR is gross depth of water applied (mm), L is furrow length in (m), W is furrow spacing in (m), and Q is flow rate (l/s).

3.4. Application efficiency (Ea)

Application efficiency was determined from the relationship between time of water recession (tr) in minute, distance of water advance (x) in meter, time of water advance (tx) in minute, time of opportunity (To) in minute, and the infiltration data of the soil. Having the depth of irrigation application, application efficiency (Ea) is determined by using Equation 8.(8) $\mathrm{E}\mathrm{a}=\mathrm{I}/\mathrm{D}\mathrm{a}\mathrm{p}\mathrm{p}$(8)

(8) $\mathrm{E}\mathrm{a}=\mathrm{I}/\mathrm{D}\mathrm{a}\mathrm{p}\mathrm{p}$(8)

Where I is the average depth of infiltration stored in root zone, Dapp is the depth of application. The depth of irrigation application was determined using Equation 9.(9) $\mathrm{D}\mathrm{a}\mathrm{p}\mathrm{p}=1000\ast \mathrm{Q}\ast \mathrm{t}\mathrm{c}\mathrm{o}/\mathrm{A}$(9)

(9) $\mathrm{D}\mathrm{a}\mathrm{p}\mathrm{p}=1000\ast \mathrm{Q}\ast \mathrm{t}\mathrm{c}\mathrm{o}/\mathrm{A}$(9)

Where Dapp is the depth of application (mm), Q is the discharge applied (m³/min), tco is the time of cut off (min), and A is the furrow area (m²).

3.5. Crop data collection and harvest

Agronomic and yield data were collected from five randomly selected and tagged plants during the growth stages. Agronomic parameters were determined to compare plant height (cm), number of branches, number of fruits per plant, and number of flowers per plant. Four harvests were performed from 8th to 29th September 2021. Yield attributes included fruit length (cm), fruit diameter (cm), and fruit weight (g). Total fruit yield comprises marketable yield (ton/ha) and non-marketable yield (ton/ha).

3.6. Water productivity

Crop water productivity was calculated by dividing harvested yield by kg per volume of water applied. The indicators are crop water use efficiency (CWUE), irrigation water use efficiency (IWUE), and field water use efficiency (FWUE). These indicators can be calculated using procedures as suggested by Michael (2003).

3.6.1. Crop water use efficiency

The CWUE is the yield harvested per ha-mm of total water used and determined using Equation 10(10) $\mathrm{C}\mathrm{W}\mathrm{U}\mathrm{E}=\mathrm{Y}/\mathrm{E}\mathrm{T}\mathrm{c}$(10) (Micheal, 1997).

(10) $\mathrm{C}\mathrm{W}\mathrm{U}\mathrm{E}=\mathrm{Y}/\mathrm{E}\mathrm{T}\mathrm{c}$(10)

Where CWUE is IWUE (kg/m³), Y is total marketable tomato fruit yield (kg/ha), and ETc is crop ET (m³/ha).

3.6.2. Irrigation water use efficiency

The irrigation water efficiency is the ratio of crop yield to the amount of net irrigation (Zhang et al., 1998). It is calculated using Equation 11(11) $\mathrm{I}\mathrm{W}\mathrm{U}\mathrm{E}=\mathrm{Y}/\mathrm{I}\mathrm{n}\mathrm{e}\mathrm{t}$(11) .

(11) $\mathrm{I}\mathrm{W}\mathrm{U}\mathrm{E}=\mathrm{Y}/\mathrm{I}\mathrm{n}\mathrm{e}\mathrm{t}$(11)

Where IWUE is the irrigation water use efficiency (kg/m³), Y is yield (kg/ha), and Inet net irrigation in (m³/ha).

3.6.3. Field water use efficiency

FWUE is the yield harvested per ha-mm of gross irrigated depth and determined by Equation 12(12) $\mathrm{F}\mathrm{W}\mathrm{U}\mathrm{E}=\mathrm{Y}/\mathrm{G}\mathrm{W}\mathrm{R}$(12) (Zhang et al., 1998).

(12) $\mathrm{F}\mathrm{W}\mathrm{U}\mathrm{E}=\mathrm{Y}/\mathrm{G}\mathrm{W}\mathrm{R}$(12)

Where FWUE is the field water use efficiency (kg/m³), Y is yield in (kg/ha), and $\mathrm{G}\mathrm{W}\mathrm{R}$ is gross irrigation in (m³/ha).

3.7. Data analysis

Significant difference among the measured parameters was analyzed by considering the above growth parameters to ANOVA (statistics trail 10) software after arranging and taking the average of the field data using Microsoft Excel. The mean square separation was done using the least significant difference adopting the normal procedures (Gomez & Gomez, 1984) for RCBD.

4. Result and discussion

The results of preliminary works such as soil properties are presented in Table 2. The effects of both irrigation scheduling methods on WUE and net benefit of the methods regarding water use is comparatively presented.

Table 2. Soil properties of the area

Depth cm	Sand,, %	Silt, %	Clay, %	Texture	Bulk density, g/cm^3	FC, %	PWP, %	pH	EC, dS/m
0–20	28	31	41	Clay	1.11	38.71	20.13	7.29	2.46
20–40	30	27	43	Clay	1.13	34.30	19.50	7.31	2.03
40–60	22	34	44	Clay	1.15	31.52	17.20	7.30	1.93
Average	26.67	31.33	42	Clay	1.13	34.84	18.94	7.30	2.14

The physical properties of soil were determined by Boyouncus method, clay soil is dominant in the area (Table 2).

The water ranges between FC and PWP available to the crop was 15.9% in average, which is within the acceptable range (Allen et al., 1998). The average hydrogen ion concentration in soil solution is represented by pH for the soil is 7.30 which appear to the neutral indicating suitability for crop production (Oshunsanya, 2019; Teklu, 2017). The electrical conductivity (EC) of the experimental site soil was 2.14 dS/m in average. Soils having EC less than 4dS/mas non-saline are suitable for the crop.

The basic infiltration rate at which the infiltration rate gets almost constant is 5.7 mm/hr. Sometimes, the infiltration rate for the soil is specified in a range of 1–5 mm/hr for the soil (Brouwer et al., 1988). However, Doorenbos and Pruitt (1977) argued the range to be 1-15 mm/hr for clay-type soils, which is consistent to the above result.

4.1. Irrigation scheduling

The net irrigation requirement was determined and presented in Table 3.

Table 3. Irrigation schedules

		SM-based			ET-based
No. of Ir	Date of Ir	Inet (mm)	Ig(mm)	Date of Ir	Inet (mm)	Ig (mm)
Ir 1	21-Jun	12.4	19.0	21-Jun	12.4	19.0
Ir 2	27-Jun	12.1	18.7	27-Jun	12.3	18.9
Ir 3	03-Jul	13.7	21.2	03-Jul	12.4	19.0
Ir 4	13-Jul	18.0	27.6	12-Jul	15.4	23.7
Ir 5	20-Jul	23.7	36.4	23-Jul	19.2	29.5
Ir 6	02-Aug	25.6	39.3	06-Aug	30.3	46.6
Ir 7	09-Aug	27.6	42.4	18-Aug	32.4	49.8
Ir 8	20-Aug	28.5	43.9	24-Aug	34.2	52.6
Ir 9	27-Aug	30.5	47.0	03-Sep	36.1	55.5
Ir 10	09-Sep	37.1	57.0	11-Sep	37.2	57.2
Ir 11	-	-	-	19-Sep	38.4	59.1
	Sum	229.1	352	Sum	280	430.9

Ir is irrigation, Inet is net irrigation, Ig is gross irrigation, and Dates are on Gregorian calendar.

The balance method needs extra irrigation at the end of the season but moisture-based does not require so. Such conditions may indicate that sometimes the assumption-based balance may continue on its track of depletion regardless of actual SM. Figure 8 presents a comparative analysis of SM variation in SM-and ET-based scheduling methods.

Figure 8. SM variations in scheduling methods.

The net irrigation water applied in SM- and ET-based are 229.1 mm and 280 mm, respectively, as shown in the Table 3. Net irrigation comparison for the methods is presented in Figure Figure 9.

Figure 9. Net irrigation comparison for scheduling method.

A total of 117.6 mm of rainfall was recorded during the experimental season on ET-based. From this, 58.9 mm of rainfall is effectively stored in the root zone. The effective rainfall on SM-based is estimated from the change in SM reading before and just after rainfall by moisture sensor. In this case, 37.7 mm of water was supplied to the root zone.

Results in Table 4 indicate that SM-based saved 18.2% of water compared to the ET-based method. This indicates direct moisture monitoring could have a better result than the climate-based method (BlonquistJr et al., 2006) as 16% water was saved by time domain transmissometry (TDT) sensor-based method compared with ET-based irrigation recommended values. In addition, Miller et al. (2018) reported 31% water saved under SM-based compared with ET-based in tomato-growing fields. In the same way, this SM-based method applied the least amount of irrigation water in results obtained from Smajstrla and Locascio (1990). Although the findings are indicating the advantage of SM-based, all findings including the present findings are different on quantity of saving. The variation between the findings can be due to several factors such as geographic and seasonal, crop type, soil type, irrigation method, irrigation manager’s level of expertise, application technologies, and climatic conditions (N. Cooley et al., 2010).

Table 4. Net requirement of irrigation water

Treatments	NIR (mm)	GIR (mm)
T1	229.1	352.5
T2	171.8	264.4
T3	114.6	176.2
T4	280.0	430.9
T5	208.5	323.2
T6	139.0	215.5

*T1, T2, and T3 are for treatment one, two, and three, respectively.

4.2. Effects of irrigation scheduling methods on WUE

The crop water use, irrigation water use, and field water use for irrigation scheduling methods are presented in Table 5. In addition, the table compares the variation of the values for application levels.

Table 5. ANOVA for effects of irrigation scheduling methods on WUE

Treatment	CWUE (kg/m^3)	IWUE (kg/m^3)	FWUE (kg/m^3)
T1	15.4^c	17.9^b	11.6^b
T2	18.5^a	22.5^a	14.6^a
T3	16.6^b	22.0^a	14.3^a
T4	11.6^e	14.5^c	9.5^c
T5	13.5^d	18.2^b	11.8^b
T6	11.5^e	17.4^b	11.3^b
CV	2.61	2.6	2.5
LSD	0.312	0.399	0.249

* Similar spellings indicate the non-significant difference between the means

WUE values are significantly varied for irrigation scheduling methods and water application levels at P < 0.05. The values are higher for low water application. However, this may not be true under all conditions. Not only water applied but the yield obtained could also have effect on the efficiencies. In the same way, Table 5 indicates the lowest WUE from full application but vice versa is not true.

The highest CWUE (18.5 kg/m³) is in the SM-based approach compared to the corresponding highest value (13.5 Kg/m³) in ET-based. This is directly linked to the amount of water applied to the methods. SM-based consumed less water than ET-based, and the ratio of yield to water in T2 is higher than the ratios in T1 and T3. That is why the value is highest for T2. The lowest CWUE in SM-based (15.4 kg/m³) is greater than the highest in ET-based (13.5 kg/m³). This can indicate the dominance of the SM-based approach over ET-based for CWUE. This result agrees with Luke et al.’s (2018) report since SM-based irrigated plants have a greater WUE than the ET-based.

Table 5 shows the highest IWUE at T2 in SM-based and the lowest at T4 in ET-based. T4 is the highest water consumed but with the lowest yield to net consumed water ratio indicating its non-appropriateness. T2 is reasonably fair for both applied water and yield. In the case of application level, the highest deficit level may increase IWUE but it has negative impact on crop yield. This result is in agreement with Luke et al. (2018)’s report that revealed SM-based irrigated plants have the greatest irrigation water use efficiency than ET-based.

FWUE was presented in Table 5 indicating 14.6 kg/m³and 11.8 kg/m³ for SM- and ET-based, respectively, as the highest value for 75% application level. All the WUE values in the above findings are higher in SM-based methods than ET-based, and the results are within a range suggested by Bisa et al. (2018), who suggested WUE of tomato from 5.9 kg/m³ to 30.79 kg/m³ as acceptable.

Comparative studies by Varble and Chávez (2011) and BlonquistJr et al. (2006) also indicated that the SM-based method is more efficient than the ET-based on water saving. Results regarding application levels and WUE in this study are more or less agreed with the previous findings (Ayas, 2019; Fereres & Soriano, 2007; Luke et al., 2018; Marazky El, 2018). Despite significant WUE for the higher deficit, it can negatively affect the crop development process by lowering ET rate (Zheng et al., 2013). Determining the appropriate level of irrigation to balance crop yield and water use can be complex and depends on biophysical and economic factors (Payero et al., 2008). However, the dependence of crop development process on the rate of water use is real. Thus, WUE is needed to be balanced not to cause significant yield reduction. Regarding this work, T2 in SM-based irrigation scheduling method could possibly be more appropriate.

4.3. Effect of irrigation scheduling methods on yield performance

The total yields determined from the analysis are shown in Table 6, which indicate that the two methods have not shown considerable differences for irrigation scheduling methods at P < 0.05) (Table 6).

Table 6. ANOVA for the effect of irrigation scheduling methods on yield performance

Treatment	Total yield (ton/ha)
T1	42.70^a
T2	41.87^a
T3	29.40^b
T4	42.50^a
T5	41.63^a
T6	28.90^b
CV	2.71
LSD	0.943

Similar spellings indicate the non-significant difference between means

The average total yields obtained from the SM-based experimental field (39.2t/ha) is similar to ET-based (39t/ha). In a similar way, Jose (2016) compared checkbook and sensor-based treatments located on two zonal experiments as he named it zone A and B. Result at zone A shows little advantage to ET-based over sensor-based. However, at zone B, the sensor-based achieved marginal advantage. These variations on yield are in agreement with the present finding. The total average yield from both methods in the present study is nearly similar to the predicted average yield (about 40 ton/ha) for the selected determinate open pollinated Roma VF variety. The total yield was decreased toward higher deficit levels, which is in agreement with other findings.

5. Conclusion

The results of the study revealed that the net water applied to perform this experiment under SM-based and ET-based scheduling methods was 229.1 and 280.0 mm, respectively. The comparison of agronomic properties, yield attributes and yield, has not shown a significant difference for irrigation scheduling methods (P < 0.05) used in the area during the investigation period. The cross comparison on results of WUE has shown a clear advantage of SM-based over ET-based scheduling methods in every treatment. This is resulted from 18.2% water saved by the SM-based method during irrigation. On the other hand, the total average yield harvested in SM- and ET-based scheduling methods were 39.2 t/ha and 39 t/ha, respectively, which are nearly similar. Thus, for higher WUE, the SM-based irrigation scheduling method with 75% level of application (T2) is more appropriate and profitable combination for tomato producers in water scarce areas. In water scarce areas like Arba Minch, application of irrigation water management practices is very important to improve WUE. This issue is highly linked with the quantity and timing of irrigation practices that also depend on the use of appropriate irrigation scheduling methods in a given area, since the use of irrigation scheduling with appropriate scheduling methods can alleviate inefficient use of available water during the irrigation seasons. This field experiment was specifically conducted in the area to evaluate the SM- and ET-based irrigation scheduling methods under deficit irrigated tomato crop field based on CWUE, FWUE, and IWUE.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Solomon Mathewos Boltana

Solomon Mathewos Boltana and Manuel Tejada Moral are working as Lecturers in Faculty of Water Resoures Engineering in Arba Minch University, Ethiopia.

Demelash Wondemeneh Bekele

Demelash Wondemeneh Bekele and Tigistu Yisihak Ukumo are employed as Assistant Professors in Faculty of Water Resoures Engineering in Arba Minch University, Ethiopia.

Tigistu Yisihak Ukumo

Demelash Wondemeneh Bekele and Tigistu Yisihak Ukumo are employed as Assistant Professors in Faculty of Water Resoures Engineering in Arba Minch University, Ethiopia.