Effects of Deficit Irrigation on Yield and Quality of Onion Crop

The broad objective of this study was to test Deficit Irrigation (DI) as an appropriate irrigation management strategy to improve crop water productivity and give optimum onion crop yield. A field trial was conducted with drip irrigation system of six irrigation treatments replicated three times in a randomized complete block design. The crop was subjected to six water stress levels 100% ETc (T100), 90% ETc (T90), 80% ETc (T80), 70% ETc (T70), 60% ETc (T60) and 50% ETc (T50) at vegetative and late season growth stages. The onion yield and quality based on physical characteristics and irrigation water use efficiency were determined. The results indicated that the variation in yield ranged from 34.4 ton/ha to 18.9 ton/ha and the bulb size ranged from 64 mm to 35 mm in diameter for T100 and T50 respectively. Irrigation water use efficiency values decreased with increasing water application level with the highest of 16.2 kg/ha/mm at T50, and the lowest being13.1 kg/ha/mm at T100. It was concluded that DI at vegetative and late growth stages influence yields in a positive linear trend with increasing quantity of irrigation water and decreasing water stress reaching optimum yield of 32.0 ton/ha at 20% water stress (T80) thereby saving 10.7% irrigation water. Onion bulb production at this level optimizes water productivity without significantly affecting yields. DI influenced the size and size distribution of fresh onion bulbs, with low size variation of the fresh bulbs at T80.


Introduction
The ever increasing world population and the demand for additional water supply by industrial, municipal, and agricultural sectors exert a lot of pressure on renewable water resources forcing the agricultural sector to use the available irrigation water efficiently to produce more food to meet the increasing demand (Andarzian et al., 2011;Valipour, 2014a).To achieve sustainable use of limited water resources available for agriculture, it is necessary to develop guidelines for irrigation applications to be used by extension service providers and farmers (Geerts, Raes, & Garcia, 2010).
This study was conducted in Nandi County, in the western part of Kenya which experiences water scarcity during the dry season from October to March, often resulting in conflicts.During this season vegetable supply is low while its demand is high.Therefore it is desirable to utilize the scarce water resource during this period to produce vegetables under irrigation for the ready market using an appropriate water saving technology.Deficit irrigation (DI) strategy was chosen for use in the study as it maximizes irrigation water productivity and optimizes crop yields (English & Nakamura, 1989;Zhang & Oweis, 1999;Fereres & Soriano, 2007;Geerts & Raes, 2009).
A certain level of water stress is applied to the crops in DI strategy either during specific growth stages or throughout the growing season, without necessarily causing significant yield reduction compared with the benefits achieved by diverting saved water to irrigate other crops (Kipkorir, Raes, & Labadie, 2001;Leskovar, 2010).Olalla, Padilla, and Lopez (2004) reported that the sizes of onion bulbs under DI were in direct relationship to quantity of water applied.Geerts and Raes (2009) reported that DI could be used to raise the crop yield to crop water consumption ratio where crops have growth stages in their development where they are tolerant to water stress.This takes place through either suppressing the water loss caused by unproductive evaporation, or by increasing the proportion of marketable yield to the overall produced biomass, or by increasing the ratio of total biomass production to transpiration due to hardening of the crop but due to the conservative relation between biomass production and crop transpiration whose effect is considered to be limited (Steduto, Hsiao, & Fereres, 2007).In addition, crop yield could be improved through application of sufficient fertilizer (Steduto & Albrizio, 2005) or by avoiding bad agronomic practices during crop development (Pereira, Oweis, & Zairi, 2002).Leskovar (2010) reported that results of onion trials indicated that DI at the 50% ETc had a significant effect on yield, while the yield from DI at 75% was not much different from 100% ETc and produced a similar bulb size.It was further concluded from the study that it would be possible to produce onion by water-conservation practices to a 75% ETc rate, as a means to target high-price bulb sizes without reducing quality.However, he advised that DI should be avoided during the yield formation stage for high yield to be achieved.
By practicing DI during vegetative development and late season stages, which are considered to be water stress tolerant stages of onion, some water would be saved (Kipkorir, Raes, & Massawe, 2002;Tesfaye, 1997).The specific objectives of this study were: (i) to determine onion yield in response to various water deficit application levels, during stress tolerant stages, (ii) to determine the effect of deficit irrigation on the quality of onion crop produced based on colour, shape and bulb size and (iii) to determine irrigation water use efficiency of the onion; by setting up on-farm field trials.

Location
The study was conducted at Mosoriot Teachers College farm in Nandi County which is located in the western part of Kenya at 35 o 10′E longitude, 0 o 19′N latitude and an altitude of 2117 m above sea level.The soil texture at the site is sandy loam, deep red, well drained with good fertility.

Climate
The area experiences a bimodal type of rainfall with mean annual rainfall of 1365 mm.The mean annual minimum and maximum temperatures are 10 o C and 24 o C, respectively (Ralph & Helmut, 1983).Climatic data observed during the season were acquired from the nearby Eldoret International Airport (EIA) meteorological station.The station is approximately 5 km north of the field trial site.Rainfall received at the nearby EIA metrological station during the main growing season (March, 2013to July, 2013) was 939 mm, and 365mm during the dry season from October, 2013 to March, 2014.

Soil
Soil physical properties were determined according to Lambe (1951) and Valipour (2014d).Soil moisture content at field capacity (θ FC ) and permanent wilting point (θ WP ) were determined in the laboratory before the start of the trial based on texture analysis and pedo-transfer function (Saxton, 2003;Saxton, Rawls, Romberger, & Papendick, 1986).Soil moisture content was determined every 10 days during the field trial period by volumetric method.The soil chemical properties obtained from laboratory tests determined the type and quantity of fertilizer and soil amendments applied.The bulk density was also determined using the procedures outlined by Okalebo, Gathua, and Woomer (2002).Note.FC = field capacity, WP = wilting point, AWC = available water holding capacity (FC-WP).
www.ccsenThe crop was harvested 150 days after transplanting when bulb onions were mature and the leaves had collapsed or bent over and left to dry for 10-12 days.Mature bulb onions were manually uprooted from the soil and cured in the sun for 10-14 days before taking measurements of yield and quality parameters.Dried leaves were cut off at 3.5 cm from the bulb.During harvesting, two bulbs were left out in each row, one at each end.Off-types were removed together with small bulbs resulting from gap-ups.The remaining ranged from 35 to 45 bulbs per plot out of which 30 bulbs were randomly picked for determination of yield and quality.

Irrigation System
Water was pumped from the borehole and stored in a masonry tank situated next to the plots.The quality of water was obtained from the existing records and was found to be suitable for irrigation of onion crop (Ministry of Environment and Natural Resources [MENR], 2012).The water was tapped from the base of the tank and piped to the plots by gravity through a main line of 25 mm diameter.The main supply pipe branched into three sub-mains to supply 18 sets of laterals which separately served each of the 18 plots (Figure 1).
Flow from the sub-mains to the laterals (drip lines) was controlled using in-line control valves which regulated the supply of water to the crops during full and deficit water application to various plots.Drip pipes ran along the rows of onions which were spaced at 30 cm apart by 240 cm long.A filter was installed on the main line just before the sub-mains to guard the emitters against blockage by dirt and soil particles leading to non-uniformity of irrigation water application.Head losses in the pipeline distribution network were determined according to Valipour (2014bValipour ( , 2014dValipour ( & 2012)).Details of the setup are presented in Figures 1 and 2.
After transplanting on 1 st March 2013, specified irrigation schedules per plot were applied to each treatment.Irrigation water discharge from the drip system was regularly measured with transparent graduated cylinders to ascertain the accuracy of the drip system at regular intervals.

Irrigation Scheduling
The crop season length (150 days) of Red Creole onion variety considered in the trial was divided into four growth stages (initial, development, mid-season and late season) each with specific duration in days.Reference evapotranspiration (ETo) was calculated from climatic data from Kapsoya Meteorological Station (located approximately 20 km north of the site) using the FAO Penman Monteith equation (Allen, Pereira, Raes, &Smith, 1998 andValipour, 2015).
The crop coefficient (Kc) was determined (Allen et al., 1998;Valipour, 2014c) for each of the 10 days during the growing season.The estimation was made from growth stages curve with an assumption that the humidity and wind speed were medium.With reference evapotranspiration (ETo) and crop coefficient known, crop evapotranspiration (ETc) was calculated using the expression The net and gross irrigation depths in mm/day were estimated (Gupta & Larson, 1979) and used to calculate the amount of water applied and application intervals.
The first to be determined was the Total Available Water (TAW), Equation 2.
Where, θ FC is the soil moisture content at field capacity and θ WP is the soil moisture content permanent wilting point.
The equivalent depth of total available water, d, was then computed from the obtained total available water, bulk density, D b of the soil and the root zone depth, Zr (FAO, 1986) Net irrigation application depth (Inet) was considered to be equivalent to readily available water which is the maximum allowable depletion based on the rooting characteristics of the crop and calculated thus.
Gross irrigation depth (Igross) was then computed from the obtained net irrigation and irrigation efficiency (Ea) of the system.
The application efficiency for drip system range between 70% and 95% (Howell, 2003).For this study, 85% was used.The irrigation interval (i), was also calculated from net irrigation application depth and the crop evapotranspiration (Valipour, 2014c) i = Inet/ETc (6) The volume (V) of water applied to meet the demand of the crop was calculated from the ground wetted area (A) and gross irrigation application, Igross (Valipour, 2012).
V=A × Igross (7) Irrigation time of application (T) was determined from volume (V) of water applied and the drip emitter discharge (Q) in litres per hour: The integrity of drip discharge was regularly checked by recording the time taken for the discharge to fill a vessel of known volume.The calculated irrigation schedule for zero effective rainfall for the field experiment during the entire growing period of the crop is given in Table 3. Inet (mm) 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 13.0 Note.A decade is 10 days.

Onion Yield and Quality Characteristics
Onion bulbs were harvested 150 days after transplanting and cured for 10-14 days before yield and quality characteristics data were collected.This was after the bulbs had attained moisture content for storage.

Yield
Fresh bulb yield was estimated by weighing 30 randomly picked onion bulbs from each treatment harvested and converted to fresh yield in ton/ha.
For dry bulb yield, ten randomly selected bulbs from each treatment were weighed, chopped and dried in an oven at 70 o C until a constant weight was achieved and converted to dry yield in ton/ha.

Quality Properties
Thirty onion bulbs were randomly selected from each treatment to determine some quality parameters composed of size represented by the equatorial diameter, colour and texture, moisture content; and shape index.The equatorial diameter (mm) of onion bulbs were measured using a digital vernier caliper.The diameter measured was the maximum width of the onion in a plane perpendicular to the pole.Bulb diameter was determined as one of the parameters of crop quality (Murthy, 2007).
The colour and texture of onion bulbs vary according to varieties which are normally available in three colours namely yellow, red and white.The colour and texture of the skin and inner flesh of harvested onion bulbs were established through CIE colour system (Smith & Guild, 1931).Red Creole onion crop variety was used in the field trial.
Ten randomly selected bulbs at storage moisture content from each treatment were weighed, chopped and dried in an oven at 70 o C until a constant weight was achieved.Moisture content was calculated based on Equation 9.
Where, MC = moisture content; W w = wet bulb weight and W d = dry bulb weight.
The shape index is used to evaluate the shape of onion bulbs and was determined using Equation 10 (AbdAlla, 1993).A shape index greater than 1.5 indicates that the bulbs are oval while index lower than 1.5 means a spherical shape, Where, D e : Equatorial diameter, D p : Polar diameter and T -Thickness.

Irrigation Water Use Efficiency
Irrigation water use efficiency (IWUE) was determined according to Jensen (1983) using Equation 11. (11) Where, DFY: Dry fibre yield, ETa: Actual reference evapotranspiration Irrigation water use efficiency (kg/ha/mm) values were used to evaluate the effectiveness of the irrigation treatment practices on maximum water utilization by onion crops.

Data Analysis
Statistical analysis methods were used to analyse the data obtained from the trial for effect of water stress on yield and quality components of onions.Analysis tools used comprised Analysis of Variance (ANOVA) mean and standard deviation computed using Microsoft Excel, 2007.Analysis of variance for the yield and quality components was carried out to determine the significance of the impact of water stress on yields and quality according to the RCBD principle.The probability level for determination of significance was 5%.

Yield Response of Onion to Water Stress
The crop in this field trial was subjected to water stress at vegetative and late season stages with six different treatment levels, five of which were water stressed to different degrees (T90, T80, T70, T60, and T50) while one (T100) acted as control and was not stressed as given in Table 4. Water application depth in each irrigation event was 13 mm.Yield per unit area obtained from fresh onion in the experiment was found to increase with increasing irrigation water levels across the various treatments (Figure 3).
Yield from non-stressed treatments (T100) which acted as control was highest at 34.4 ton/ha while the most stressed treatment (T50) had the lowest yield of 18.9 ton/ha.The intermediate treatments T90, T80, T70 and T60 gave yields of 32.6, 31.9, 25.2, and 22.6 ton/ha respectively.The standard deviation varied between 0.6 and 1.7 ton/ha while the coefficient of variance ranged from 2% to 7.5% within the replications.The standard deviation within the treatments was low suggesting that yield was more clustered around the mean.
Analysis of variance across the treatments (at 5% probability) indicated that DI significantly affects yield, Table 4.  proportionally with the quantity of irrigation water applied.There is therefore a linear relationship between bulb size and quantity of irrigation water applied.The coefficient of determination analysis between diameter and irrigation water applied was (R 2 = 0.927), indicating that the increase in bulb diameter in different treatments was attributed to increase in the quantity of water hence quantity of water applied influences onion size.
The distribution of bulb sizes showed that large (> 60 mm) formed 27 % of the total production, medium (45-60 mm) made 40% and the remaining 33% were small (< 45 mm) as shown in Table 6.Large size was largely produced under treatments T100, T90, and T80 which received water amounts of 494 mm, 468 mm and 441 mm respectively.On the other hand standard deviation for the treatments varied with the highest being from T50 and the lowest T80.The low size variation of the bulbs as indicated by low standard deviation under T80 was an indication that the onion bulb diameters were more clustered closely around the mean under T80 than in other treatments.A similar effect of various irrigation water levels on size of onion bulb was observed by Olalla et al., (2004) under drip irrigation.Leskovar (2010) reported that it would be possible to adjust water conservation practices to a 75 percent ETc rate, as a means to targeting high-price bulb sizes without reducing quality.These results emphasize that adequate soil moisture content along the growing period encouraged the vegetative growth of the plant and enhanced the development of large and medium bulb size which is considered to be marketable.

Mass of Onion Bulbs
Fresh onion bulb mass across treatments was influenced significantly (at 5% probability) by DI (Table 7) with a coefficient of determination of 0.943 which suggests a direct relationship between DI and mass.The highest mean weight of bulbs (103 g) was obtained from treatment with the highest supply of water while the treatment with the lowest quantity produced the least mean bulb weight (57 g).There is a positive linear relationship between water stress and bulb mass.This means that water stress affects negatively the weight of individual bulbs.The moisture content of onion bulbs as depicted in Table 8, do not vary with treatments and range from 84% to 89%.This means that the influence of water stress on onion moisture content was not significant.

Comment
Insignificant Insignificant Insignificant Insignificant Significant

Shape Index
The shape index data is presented in Table 9.The results indicate that water stress at vegetative and late stages of growth of onion do not significantly affect the shape of onion bulbs.All bulbs were oval as shape index is greater than 1.5 (Figure 4).

Conclusions
From the findings of the study it is concluded that deficit irrigation at vegetative and late growth stages of onions influence yields in a positive linear trend with increasing quantity of irrigation water and decreasing water stress reaching optimum crop yield of 32.0 ton/ha at 20% water stress thereby saving 10.7% irrigation water.It was further concluded that production at this level optimizes water productivity without significantly affecting crop yields.
Further, it was established that deficit irrigation influenced the size and size distribution of fresh onion bulbs, with low size variation of the fresh bulbs at 20% water stress (T80) as attested by low standard deviation of 1.0 as compared to other treatments.DI therefore can be used in deciding onion sizes to produce for a particular prevailing market.Deficit irrigation does not affect the shape of onion bulbs as depicted by the shape index of more than 1.5.The colour of bulbs was also not affected by DI thereby maintaining attractiveness of the product to the users.
The irrigation water use efficiency for onion yield was affected significantly by DI treatments with the highest values obtained under the most stressed (T50) treatment while the lowest values were obtained under full water supply (T100) treatment.However, highest yields and quality were recorded for lowest efficiencies and vice versa.It is concluded that optimum production occurred at optimum water use efficiency (T80). Fig

Table 1 .
Soil physical properties of the study area

Table 2 .
So Top dressing with Calcium Ammonium Nitrate (CAN) fertilizer at recommended rate was done after three weeks of transplanting (Ministry of Agriculture [MoA], 2012).

Table 3 .
Irrigation schedule for the entire season

Table 5 .
Onion bulb size (in mm) as influenced by applied irrigation water stress levels

Table 6 .
Onion bulb size frequency distribution in response to applied irrigation water stress levels

Table 7 .
Mean mass (in g) of single fresh Onion bulbs

Table 8 .
Moisture content (in %) of fresh onion bulbs under different treatments

Table 10 .
Optimum production based on water saving and yield reduction