Research Article
Yield Response of Durum Wheat (Triticum durum Desf.) Cultivar Waha to Deficit Irrigation under Semi Arid Growth Conditions
Not Available
A. Aidaoui
Not Available
H. Bouzerzour
Not Available
A. Saci
Not Available
The high plateaus region of Algeria with a Mediterranean- type climate receives 70% of the 250-500 mm rainfall during the cold season from October to February. Cereal crops suffer from winter cold, terminal heat and drought stresses. Adoption of short cycle genotypes was intended for an effective utilization of the limited soil moisture and to reduce from the deleterious effects of terminal stresses, but still grain yield remains very low when compared to what is obtained in the neighboring countries (Annichiarico et al., 2002). Under such growing conditions the amount of rainfall received in the spring, once the soil water recharge has been exhausted, increased significantly crop yield (Chennafi et al., 2005). As a resource water is limited and it is used preferentially on horticultural cash crops which generate a high plus value. Cereals are rarely conducted under full irrigation, but it is a common practice to add small amounts of water, in the spring when rainfall fails to provide sufficient moisture to insure a normal plant growth and by doing so crop failure is generally avoided (Bouzerzour and Oudina, 1990).
Several researches demonstrated that substantial increase in yield of rainfed crops is obtained in response to limited irrigation applied at sensitive crop stages (Ehdaie, 1995; Li et al., 2001; Deng et al., 2002; Zhang et al., 2004). Kang et al. (2002) reported significant increases in wheat grain yield, varying from 20-45% after applications of 30-60 mm of reduced irrigation at jointing. The efficiency of the limited irrigation applied in the spring or early summer is highly related to the physiological state of the crop to which water is added (Perrier and Salkini, 1991; Singh et al., 1991; Ilbeyi et al., 2006). It is also dependant on climatic conditions such as air dryness, hot wind events and high temperature frequency, prevailing once the crop has been irrigated (Payero et al., 2006). A high evaporative demand around heading stage affects seriously grain yield and above ground biomass production through reduction of ear number, spike fertility and individual grain weight (Rasmussen et al., 2003; Zhang et al., 1998).
In order to optimize grain yield and water use efficiency of reduced irrigation, it is advisable to control the crop water deficit to sustain crop growth (Wichems, 2002; Kirnak et al., 2001). Regulated deficit irrigation consists in applying small amount of water at specific growth stages. The objective is to prevent severe water deficit which inhibits crop growth and subsequently reduces the expected yield gain from reduced irrigation applied at later growth stages (Kang et al., 2002). Asseng et al. (1998) reported that water deficit, at certain growth stages, increased grain yield and water use efficiency. The objectives of this study were to (1) quantify the grain yield response of durum wheat (Triticum durum Desf.) cultivar Waha to deficit irrigation, (2) determine climatic co-variables correlated to grain yield response under deficit irrigation and (3) evaluate the impact of regulate deficit irrigation on grain yield and water use efficiency under semi arid climate of the eastern high plateaus of Algeria.
Site and experiments description: The Trials were conducted at the ITGC- agricultural experimental site of Sétif (Lat 36°12'N, Long 5°24'E, alt 1081 m asl.), located in the eastern high plateaus of Algeria. The field experiment lasted 10 years (1987-88, 1989, 1990, 1991, 1994, 1995, 1996, 1998, 1999 and 2000), it was followed by a 3-year pots experiment (2001, 2002 and 2003). The soil of the experimental site is a deep brown calcareous earth classified as a steppic brown soil, with a pH of 8.2 and 1.35% organic matter. Soil composition is 13.94% sand, 41.32% silt and 44.74% clay. The mean water content at field capacity and at permanent wilting point was 25 and 12% per weight respectively. Soil bulk density is 1.35 and the soil infiltration rate is 8.27 mm h-1.
The field experiment was conducted in a randomized complete block design with four replications. Plot dimensions were 12x6 m with a 2 m border strip between adjacent plots. A short cycle durum wheat cultivar Waha was seeding at a rate of 250 viable seeds m-2, by mid- November of each cropping season.
Four treatments were imposed, a rainfed check and three irrigated treatments: irrigation at jointing, at heading and at both growth stages. The amount of irrigation water applied averaged 50 mm. This amount represents 18.5% of the soil water holding capacity in the 0-80 cm soil profile and 35.6% of the plant-available soil water. Plots were irrigated using a solid-set sprinkler system which was arranged in a grid with a sprinkler head installed in the middle of the plot. Water for the system was pumped from the Boussalam River. Trials were fertilized with 100 kg ha-1 of super phosphate 46% at sowing and 100 kg ha-1 of urea 34% at the tillering stage. Weeds were controlled by application of 1l ha-1 of 2,4 -D [Dichloro Phenoxyacytic Acid] or 12 g ha-1 of Granstar [Methyl Triberunon] herbicide mixed with 250 L of water.
Pots, used in the second experiment, have 30 cm diameter and were 35 cm deep. They were poured with a plastic bag containing 20 kg of dry soil taken from the experimental site. Fifteen pre-germinated seeds of the genotype Waha were transplanted to each pot, during the second half of November, which represents a stand density of 212 plants m-2. The experiment was laid out in a randomized complete block design with four replicates. Five treatments were studied: irrigation to field capacity (FC), irrigation to ¾ field capacity (¾ FC), irrigation to ½ field capacity (½ FC) throughout the crop cycle, irrigation to ¾ field capacity up to heading (¾ FCH) and a rainfed check. A mobile plastic rain shelter was installed above the pot experiment to control soil water deficit. Pots were weighted twice a week and the evapotranspired water was added to bring them to their predetermined water content.
Measurements and data analysis: The beginning of stem elongation was recorded when the first node was detectable, this corresponds to the stage 3.1 of Zadoks scale (Zadok et al., 1974). Heading date was recorded when 50% of the number of spike in a segment of row 1 m long emerged half away above flag leaf ligules.
The vegetation from 3x1 m row-segment was sampled from each plot of the field experiment to get above ground biomass produced at maturity and the number of spike m-2. Grain yield and 1000 kernel weight were derived from the combine harvested trials. Only grain yield is reported and discussed hereafter. The same traits were measured in the pots experiment in addition of plant height. Straw yield and harvest index were derived from measured data. Water Use Efficiency (WUE) was calculated after Ehdaie and Waines (1993): WUE = GY/ TWE, with GY = grain yield and TWE = total water evapotranspired.
Weather data, monthly precipitations and monthly mean temperature, were obtained from an automatic weather station located 2 km North-East from the experimental site. Data analyses were performed using the statistical software Statitcf version 3.0. Grain yield from the field experiment and traits measured in the pot experiment were analyzed according to a factorial experiment conducted in a completely randomized block design with four replications. Least significant difference, at 5% level, was used for means separation. The degree of association between different traits and climatic co-variables was estimated through simple correlation and linear regression.
Field experiment: Grain yield analysis of variance showed significant main effects and irrigationxseason interaction (Table 1). Contrast analysis indicated that the supply of a limited amount of water in the spring has a significant effect on grain yield increase.
Table 1: | Mean squares of the grain yield analysis of the field experiment |
** Significant effect at 1% |
Table 2: | Grain yield means (g m¯2) obtained under rainfed and limited irrigation applied at jointing, heading and at both growth stages during 10 cropping seasons, mean values of water use efficiency (kg ha-1 mm-1) and seasonal precipitations (mm) |
GY = Grain Yield, J = Jointing, H = Heading stage, = I = Irrigation, S = Season, LSD5% GY = 26.3 g m-2 |
Fig. 1: | Percent grain yield increase over the rainfed treatment due to limited irrigation applied at jointing (J), at heading (H) and at both growth stages (J+H) |
Averaged over the 10 cropping seasons, grain yield obtained under limited irrigation showed a 93.4% yield increase over the rainfed treatment whose mean grain yield was 194.5 g m-2 (Table 2). Supplying a limited amount of water at heading was more beneficial, with an advantage of 19.14% yield increase than irrigating at jointing. Grain yield mean for both treatment was 364.1 and 305.6 g m-2, respectively. Irrigating at both growth stages increased grain yield by 36.4% compared to grain yield obtained under irrigation done at either growth stages (Table 2).
Grain yield of the different seasons showed high variation and ranged from 248.3 to 434.1 g m-2 as the accumulated precipitation over the cropping cycle varied from 206.5 to 411.6 mm. Variation in the distribution pattern of rainfall among seasons induced a variation in the crop response to the amount of water added through irrigation. In fact yield increase ranged from 0 to 219% when irrigation was applied at jointing, from 29 to 316% for heading irrigation and from 70 to 381% when the added water was applied at both growth stages (Table 2 and Fig. 1).
Fig. 2: | Relationship between grain yield, water efficiency and total water evapotranspired of durum wheat cultivar Waha grown under limited irrigation |
Yield increase was explained by a better conjunctive use of rainfall and added water (WUE). The mean values of water use efficiency generally increased as the amount of water applied increased and ranged from 3.92 to 15.99 kg ha-1 mm-1 (Table 2). Some high WUE values were obtained with lower amount of water added indicating that a good rainfall distribution pattern within the season contributed to a better WUE. This supports the findings of Rasmussen et al. (2003) which showed that when rainfall is well distributed, it allows little stress to develop and usually matches the crop water needs. The relationship between grain yield, WUE and total water evapotranspired indicated that a higher proportion of grain yield variation between cropping seasons and water treatments was explained by the variation observed in the WUE, but only a small part of this variation was explained by the observed variation in the total water evapotranspired (Fig. 2).
The quadratic form of the relationship between grain yield and TWE suggested that high grain yield could be obtained with a relatively lower total water evapotranspired, mainly though improved WUE. The results of this study corroborated those of Ehdaie (1995) but were in contradiction with those of Condon et al. (1987) which indicated high mean WUE in droughty than in well watered conditions. Kang et al. (2002) reported a quadratic relationship between grain yield and seasonal evapotranspiration (ET), showing that grain yield did not increase when ET exceeded a critical value of 456 mm. Zhang et al. (2006) reported mean values of WUE ranging from 14.6 to 16.2 kg ha-1 mm-1 under regulated deficit irrigation and a quadratic relationship between WUE and GY.
The yield reduction from maximum yield observed under rainfed conditions was positively correlated with the winter precipitations accumulated in October- February (r = 0.6562, p< 0.05, n = 10). Yield increase under limited irrigation was positively correlated with May precipitations (r = 0.658, p<0.05, n = 10).This suggested that rainfed crop yield is generally dependent on the adequate storage of winter precipitations in the soil profile to delay the deleterious effect of drought stress and that the effectiveness of the limited irrigation applied in the spring is improved by late rainfall events. These results were in agreement with the findings of Rasmussen et al. (2003) and supported previous result's studies indicating that limited irrigation increased significantly grain yield (Deng et al., 2002; Kang et al., 2002; Perrier and Salkini, 1991).
Pot experiment: The pot experiment was undergone to answer the question what will be the yield difference between full and limited irrigation alternatives. The analysis of variation showed significant irrigation and season main effects but the interaction was non significant in most cases. Contrasts analysis indicated significant differences between controlled irrigation treatments (Table 3). The three cropping seasons accumulated 264.0, 168.7 and 517.5 mm rainfall. The distribution pattern of the precipitations is given in Fig. 3. Most of the rain is recorded during the winter months, with a high inter-season variation. The full irrigation treatment used 390.5, 392.5 and 380.0 mm, respectively in 2001-2003, Compared to the yield of the rainfed check, grain yield obtained under full, ¾FC, ¾FCH and ½FC irrigation treatments increased, on average over the three- season period, by 133, 75.5, 57.8 and 44.8%, respectively (Table 4).
This indicated that irrigation had a drastic effect on grain yield increase which was positively correlated with dry matter (r = 0.831, p<0.05, n = 15), straw yield (r = 0.699, p<0.05), spike number (r = 0.701, p<0.05), water used (r = 0.600, p<0.05), harvest index (r = 0.450, p<0.05), plant height ( r = 0. 507, p<0.05) and water used efficiency (r = 0.547, p<0.05). These relationships suggested that grain yield as well as above ground dry matter; plant height, straw yield and WUE of durum wheat were all affected by the controlled ranges of soil water content during the growing seasons (Fig. 4).
The yield increase/decrease under the different irrigation treatments was the result of the contribution from the increased/decrease increments in dry matter, spike number, plant height (Table 4). These findings corroborated the results of Giunta et al. (1993) which indicated that grain yield was linearly related with total dry matter yield and that reduced irrigation, compared to full irrigation, decreased all yield components. Hang and Miller (1983) reported a sharp decrease of total above ground dry matter, crop growth rate, head number and plant height under reduced irrigation compared to the well watered treatment.
The yield comparisons within irrigated treatments showed that yield obtained under full irrigation was not significantly different from the one obtained under ¾FC reduced irrigation two out of three years. Saving 50% of wheat water requirement affected significantly yield level as grain yield under full irrigation was significantly higher than yield of ½FC reduced irrigation, which is however significant higher than the rainfed check grain yield (Table 4). When adopting the regulated deficit irrigation at ½FC level, yield losses relatively to full irrigation yield ranged from 47.0 to 88.8% of the rainfed yield. The yield gain from ½FC controlled deficit irrigation relatively to the rainfed yield varied from 27.8 to 64.8%. The amount of water used under the ½FC controlled deficit irrigation throughout the crop cycle averaged 198.5 mm, this value compares well with values reported in the literature. According to Musick et al. (1994) wheat grain yield required a minimum evapotranspiration of 206 mm while Zhang and Oweis (1999) reported a value of 156 mm for wheat in the Mediterranean region.
The results of the pot experiment agreed somewhat well with the conclusion drawn from results of the long term field experiment which indicated that limited irrigation increased significantly yield expectation but the yield gain was subjected to year-to-year variation. Yield increases were associated with increases in total dry matter, plant height and yield components mainly spikes produced per unit area of soil. The increase in dry matter and plant height generated an increase in straw yield too. Wheat straw is used for feeding livestock during the winter months and is an important commodity which has an economic value reaching 40% the grain price (Annichiarico et al., 2005).
Table 3: | Mean squares of the analysis of variance of the traits measured in the pot experiment |
æx10 3, GY = Grain Yield (g m-2), SN = Spikes Number (m-2), TKW = 1000-krenel weight (g), HI = Harvest Index (%), PHT = Plant height (cm), STR = Straw yield (g m-2), DM = Above ground dry matter (g m-2), TWE = Total Water Used (mm), WUE = Water Use Efficiency (kg ha-1 m-2). *Significant effect at 5% level |
Table 4: | Mean values of the measured variables in the pot experiment |
GY = Grain Yield (g m-2), SN = Spikes Number (m-2), TKW = 1000-krenel weight (g), HI = Harvest Index (%), PHT = Plant height (cm), STR = Straw yield (g m-2), DM = Above ground dry matter (g m-2), TWE = Total Water Used (mm), WUE = Water Use Efficiency (kg ha-1 m-2) |
Fig. 3: | Monthly rainfall distribution of the 2001-2003 cropping seasons |
The results corroborated also the findings of Oweis et al. (2001) and Ilbeyi et al. (2006) which showed that regulated deficit irrigation didnt reduced significantly wheat yield from yield observed under full irrigation.
Fig. 4: | Responses of Grain Yield (GY), Spike Number (SN), above ground Dry Matter (DM) and straw yield (STR) to regulated irrigation deficit |
They differed however from the findings Zhang et al. (2006) whom reported an increased yield under regulated deficit irrigation compared to full irrigation. Jamieson et al. (1995) mentioned that full irrigation was less efficient compared to deficit irrigation treatment because full irrigation generally has much water left in the soil profile at harvest. The discrepancy between the results of the present study and others arises from differences in soil, genotype and climate types under which studies were conducted. In order to cope with severe drought effects and erratic rainfall distribution patterns, it is important to combine genetic properties of the genotype (earliness to head, early vigor growth to minimize evaporation, the ability to avoid luxuriant growth, to adjust the growth pattern according to the rainfall events and to recover when soil moisture is restored) with proper management practices such as early sowing, weeding and adoption of deficit irrigation applied at sensitive crop growth stages around heading.
Available water appeared as the most important factor limiting wheat crop yields under the semi arid highland of eastern Algeria. The results of this study demonstrated that the amount of grain yield produced per water use increased with the increase in availability of soil water and consequently water used efficiency increased. Therefore the conjunctive use of rainfall and irrigation can be improved by limiting irrigation by restricting the applications late in the season to prevent crop failure or low grain yield. Water saved can be used to cope with grain yield variability at a larger scale by increase the irrigated area. The findings of this study support the idea that, under semi arid climate, the limited irrigation strategy offers the opportunity to substantially increase grain yield potential, even though the yield increase is highly variable in magnitude from one season to another. They suggested that it is advantageous to irrigate once at the heading stage than twice at jointing and heading. In the cases where water is more available, during the whole crop cycle, regulated deficit irrigation to meet 50% of the plant water requirement lead to more stable yield gain, but still significantly less than full irrigation.