Effects of Supplemental Irrigation and Nitrogen Applied on Yield and Yield Components of Bread Wheat at the Saïs Region of Morocco

BA designed the study, performed the experiment and statistical analysis, wrote the protocol, and wrote the first the manuscript. Authors MI and RD managed the analyses of the study. Authors RD, MI, AK and KD supervised the study and managed the literature searches. All authors read the final manuscript. ABSTRACT The objective of this study is to determinate the optimal growth stage for applying supplemental irrigation and nitrogen to enhance bread wheat ( Triticum aestivum L . ) productivity and water use efficiency under rainfed condition in Morocco. Field trial was conducted during two years (2007-2008 and 2008-2009). Three genotypes (G) of Moroccan bread wheat; Achtar, Arrehane and an Advanced Line II were combined with five nitrogen doses (N); 0, 40, 80, 120 and 160 kg N ha –1 and three water treatments (I); rainfed (I 0 ), irrigated (60 mm) at 21 according to the Zadoks scale (I 1 ) and irrigated (60 mm) at 59 according to the Zadoks 2 ). Results in the drought year (2007-2008) show that I 2 water use efficiency, grain yield and its components by 91 and 60% respectively, compared to I 0 and I 1 treatments. I 2 has limited the effects of the Chergui –hot and dry wind coming from the Sahara- which has a negative impact on growth mainly at the end of the cycle of the crop. Grain yield and other yield components increased simultaneously with the increase in nitrogen inputs during the second 2008-2009 crop year which was rainy. The 120 kg N ha the highest results, with no significant difference with the yield obtained after a 160 kg N ha –1 input. Genotype appeared to have no significant effect neither on grain yield nor on water use efficiency in both the years.


INTRODUCTION
Cereals are the most important crops in Morocco; it covers 5 million hectares which represent 80% of arable land in the country. Unfortunately, yields are low and variable due to rainfall scarcity and variability [1]. Climatic change pattern in Mediterranean region would increase drought frequency with high fluctuation in precipitation [2]. In fact, at the Saïs region of Morocco, the average rainfall is passed from 600 mm per year during the period 1960-1980 to 440 mm for the period 1981-2000 [3]. By 2020, the climate change scenarios in Morocco in relation with agricultural production include a reduction in grain yield of cereals estimated at 50% in dry years and 10% in normal years [4]. In Morocco, agriculture is facing challenges to produce for a population increasingly growing and to be more efficient in the use of water given the competition from other non-agricultural sectors [4].
Under these conditions, to increase and stabilize production there has been a tendency to adopt application of high rates of nitrogen under irrigated conditions [5]. Indeed, compared to non-irrigated treatment, an increase in grain yield for bread wheat of 419, 389 and 281% respectively for the full supplemental irrigation (FSI), 2/3 rd of the FSI and 1/3 rd of the FSI has obtained in northern Syria [6]. The effect of supplemental irrigation is more remarkable as it is made in the early stages of crop development (early tillering) for varieties with low tillering and heading stage for varieties with high coefficient of tillering. In both cases, the yield increases up to 100% [7].
Also, nitrogen use of cereals is the second factor after water, which determines the level of intensification of the production. It is one of the most difficult practices in the dry areas. Indeed, the addition of high amounts of nitrogen in dry years and this element deficiency in wet years often cause considerable falls in cereal yields [8]. Water and nitrogen use should be carefully managed to avoid losses [9]. In the North of France, on the new variety of wheat Recital, with 170 kg N ha -1 gave an increase of 87% of yield comparatively to control (0 kg N ha -1 ). While the old variety Cappelle is indifferent to the dose of nitrogen applied, but its efficiency is low [10].
A field trial was set up in experimental stations with the objectives to stabilize and increase the production of bread wheat in Moroccan rainfed area of the Saïs through better management of supplemental irrigation (application stage), the nitrogen dose and the genotypes with the aim of increasing the efficiency of their use.

Experimental Site
A field trail was conducted in Saïs region at the Experimental Station Douyet (National Agricultural Research Institute, Morocco) in the province of Fez (34°2'N, 4°50'W, altitude: 416 m), for two years 2007-08 and 2008-09.

Treatments
Sowing was done by a seeder on 26 December for the two years of the trial at a dose equal to 160 kg ha -1 . The three factors tested were: Water regime (I 0 ; rainfed, I 1 ; irrigated (60 mm) The third genotype is an Advanced Lineage II (LA II) from the Moroccan Observation Nursery (INRA), its origin is a cross between of Moroccan variety Nesma and Australian Lineage (Nesma x 2/14-2/7/ Ias 20 x 5/H567.71/6/Bow"S"), they are resistant to the Cecidomyie, Rusts and Septoria.
The experiment design was a split-plot with 3 replications, the water regime in main plots while genotype and nitrogen were allotted to sub-plots. The experimental area was divided into 135 plots of 5 m x 3.5 m each, with an alley of 2 m between adjacent plots and 8 m between main plots. Spray irrigation was used and was controlled by the location of several graduated cylinders at the irrigated plots. The final results showed that irrigation water doses were made between 55 and 65 mm.
Analysis of nitrogen soil before installation of the test showed that the soil was rich in this element (40 ppm NO 3 -).

Sampling and Analysis
Measurements at physiological maturity included grain and straw yield, biomass, 1000 kernel weight, kernel number per unit area, harvest index and soil moisture at sowing and physiological maturity. The gravimetric method [12] was used to determine volumetric water content for soil sample taken by hand auger from 0-20, 20-40 and 40-60 cm profile depths (the virtisols and hand auger used does not allow penetration into the deeper horizons more than 60 cm). Water use efficiency (WUE) for grain or dry matter is calculated: WUE = Grain or dry matter / ET, (kg ha -1 mm -1 ).
To quantify water consumption throughout the crop cycle, we used the water balance equation [13] that actual evapotranspiration (ET) is calculated as: ET = P + I -R -D ± ∆S P: Rainfall, I: Irrigation, ∆S: Change in stocks of water in the soil. R: Runoff, D: Drainage, were neglected. SPSS 17.0 and SNK (Student-Newman-Keuls) was used for analysis of variance and mean comparison.

Weather Conditions
In 2007-2008 crop year total rainfall throughout the season was 334 mm; 115 mm before sowing and 219 mm during the test period ( Figure 1). While the 2008-2009 crop year recorded 750 mm of rainfall; 388 mm before sowing and 362 mm between sowing and physiological maturity. Rainfall during the last year was well distributed throughout the crop cycle ( Figure 1).

Grain Yield
The 2007-2008 crop year was a relatively dry year, the water supply had a highly significant effect on grain yield. Indeed, supplemental irrigation at 21 according to the Zadoks scale and 59 according to the Zadoks scale have resulted in improvements in grain yield respectively from 59.43 and 90.96% compared to no-irrigated (Table 2). Increased yield noticeably support the findings in Asia [14,15] and in the semi-arid region of Settat (Morocco) [7,16] in favor of the potential for conjunctive use of supplemental irrigation. Nitrogen dose did not affect the grain yield. The low yields in 2007-2008 compared to the potential of the region and the effect of two successive drought years (2006)(2007)(2008) have limited samples the nitrogen supply and we would suggest that the balance of nitrogen from previous years was sufficient to produce yields recorded. Genotype and the interactions did not show significant effects.
Since the genotype had no significant effect on most of the parameters measured only the results of the water regime and nitrogen dose will be presented in the Tables (2, 3, 4, 5, and 6).  (Table 2). Similar results were found in France [10] and in Morocco [17]. The addition of 168 kg N ha -1 had no significant effect on grain yield of durum wheat (variety Karim) compared to an intake of 126 kg N ha -1 [17]. While, the dose 170 kg N ha -1 gave a yield higher than 87% compared to the doses 0 N ha -1 for wheat Recital [10]. Statistically supplemental irrigation and genotype had no significant effect on grain yield. However, the two water regimes (irrigated at 59 according to the Zadoks scale and irrigated at 21 according to the Zadoks scale) gave an average of 49.38 qx ha -1 , higher than 3 qx ha -1 compared to no-irrigated. The interactions did not show significant effects.

Water Use Efficiency
Supplemental irrigation had a significant effect on the water use efficiency (WUE) for grain in 2007-08. Irrigated at 59 according to the Zadoks scale showed an efficiency of 11.89 kg ha -1 mm -1 , higher respectively from 40.3 and 49.8% compared with irrigated at 21 according to the Zadoks scale and no-irrigated regime (Table 3). Similar results were observed for WUE for total biomass. In dry year, a relatively small amount of irrigation water applied at strategic growth stages of cereals could realize an important increase in WUE of rainfed wheat [18,19].
WUE for grain and total dry matter in the rainy season 2008-2009 was influenced by the nitrogen dose. Statistical analyses showed that N 4 and N 3 allow the highest WUE for grain (6.43 kg ha -1 mm -1 ), higher respectively from 41.9% for N 0 and 19.5% for N 2 and N 1 ( Table  3). These results corroborate those found in northern Syria in the case of supplemental irrigation [20]. Indeed, the dose 100 kg N ha -1 allowed the highest WUE for the grain (2.2 kg ha -1 mm -1 ), followed by the doses 150 and 50 kg N ha -1 (1.4 kg ha -1 mm -1 ) and finally the dose 0 kg N ha -1 (0.7 kg ha -1 mm -1 ). The difference in water use efficiency between these 2 years was due mostly to the difference in grain yield. In each growing season, Means followed by different letters are significantly different at 5% level of probability

Harvest Index
Statistical analyses revealed in 2007-2008 season a significant effect of water regime on harvest index (HI). The results are similar to those obtained for grain yield. In fact, the rainfall regime gave the lowest HI with an average of 30.71%, followed respectively by the regime irrigated at 21 according to the Zadoks scale with an increase average of 9.34% and the regime irrigated at 59 according to the Zadoks scale with an increase average of 26.12% (Table 4). Climatic conditions during the development of kernel have a direct effect on HI. Irrigation especially at 59 according to Zadoks scale, realized upon the cessation of rains has allowed remobilization of the photoassimilates from source to sink.
In 2008-2009 crop year, the water regime has no influence on this component unlike to nitrogen dose. Statistical analyses revealed that the higher HI was obtained by N 4 with an average of 38.98. While, N 0 have registered the lower HI with an average of 30.25 (Table 4). The difference in performance for different doses of nitrogen was more influenced by the amount of dry matter produced rather than the translocation of photoassimilates to the grain as the case for the dry year. In each growing season, Means followed by different letters are significantly different at 5% level of probability

Kernel Weight
The 1000 kernels weight (KW) in 2007-2008 season have been influenced by the water regime. The highest KW (36.15) was obtained with regime irrigation at 59 according to the Zadoks scale, while the other two water regimes showed values statistically equal (26.26 g) ( Table 5). Even under irrigation in the 59 according to the Zadoks scale did not express the highest potential weight of 1000 kernels. However, irrigation at this stage helped cushion the negative effect of hot winds (Chergui) that prevalent in the region at the end of the crop cycle. For genotype and nitrogen dose, the effect was not significant on this component.  (Table 5). In contrast, in France for fifteen wheat genotypes tested the average for KW was equal to 43.1 g for nitrogen dose 0 kg ha -1 and an average of 40.9 g for dose 170 kg N ha -1 . Unlike the KW of five genotypes VM014, Audace, Etoile de Choisy, Sidéral and Camp Rémy are indifferent to the nitrogen dose applied [10]. Rainfed regime gave a KW equal to 35.34 g followed by irrigated regime at 59 according to the Zadoks scale with kW 36.39 g. The regime irrigated at 21 according to the Zadoks scale is statistically similar to the two regimes. This difference was not very important to influence the final grain yield. Genotype and the different interactions had no effect on this component. ) ( Table 6). In general, the number of grains per unit area was determined by the population spike only, rather than the other components that have not recorded any significant differences. This result confirms those found in the semi-arid region of Settat (Morocco) [16].
Number of grains per unit area has been influenced in 2008-09 by the nitrogen dose. N 1 gave the lowest number (12408 kernels m -2 ), while N 4 gave the highest number (14303 kernels m -2 ). Although N 1 gave a number of grains per unit area small than N 0 , it was offset by the kernel weight for a high yield of grain.

CONCLUSION
With climatic change and rainfall scarcity, supplemental irrigation could be an opportunity to enhance wheat grain yields and respond to population food demand. In fact, during the dry year 2007-2008, water regime greatly influences the production of bread wheat in the Saïs region of Morocco. Minimal Irrigation (60 mm) at 59 according to the Zadoks scale increase grain yield by 91% comparatively to rainfed regime. While, during years of heavy rainfall as 2008-2009, supplemental irrigation at 21 or 59 according to the Zadoks scale gives an increase of 3 qx ha -1 even if this increase was not statistically significant. This gain compensates amply the irrigation's cost. For nitrogen, during 2007-2008, no significant effect was recorded, while during 2008-2009, 120 kg N ha -1 gave the highest grain yield because of more rainfall availability. Genotypes and also the different interactions have no significant effects on grain yield during both years. This experiment should be reconducted in different soils and climatic conditions for extension of findings. The introduction of efficient genotypes for water and nitrogen uses should be of great interest. Responses to supplemental irrigation, nitrogen and genotypes can be modeled in order to use this model in other areas of Morocco and in the Mediterranean region.