Nitrogen use efficiency, yield and yield traits of wheat response to slow-releasing N fertilizer under balanced fertilization in Vertisols and Cambisols of Tigray, Ethiopia

Abstract The study was initiated to investigate the effects of slow nitrogen-releasing fertilizer (UREAStabil) on yield, yield components and nitrogen use efficiency indices of bread wheat at Emba Alaje and Hawzien Districts in Tigray, Ethiopia in 2015 at six farmers’ field and arranged in a randomized complete block design with three replications. Treatments were four levels of nitrogen (0, 32, 64 and 96) kg ha−1. The nitrogen source was UREAStabil, which is slow N-releasing fertilizer. Conventional urea at the recommended rate (64 kg N ha−1) was included as a positive control at both sites. A full dose of UREAStabil were applied at planting while prilled urea (conventional urea) was applied in two splits 1/3 at planting and 2/3 at tillering. Application of UREAStabil significantly influenced yield and yield components at both soil types. The highest grain yield was obtained on plots treated with 64 kg N ha−1 in the form of UREAStabil and prilled urea (conventional urea) in Hawzien and Emba Alaje districts, respectively. The highest nitrogen uptake was recorded on plots treated with 64 kg N ha−1in the form of UREAStabil and prilled urea (conventional urea) in Hawzien and Emba Alaje, respectively. Nitrogen uptake, agronomic, physiological and apparent recovery efficiency were significantly influenced by the application of slow-releasing and conventional nitrogen fertilizers at both sites. In Hawzien, the application of slow-releasing N fertilizer in the form of UREAStabil reduces the amount of N used and application time. Hence, it could be concluded application of slow-releasing fertilizer could be used as an alternative source of nitrogen for wheat production in the study site.


PUBLIC INTEREST STATEMENT
Nitrogen fertilizer is one of the main inputs for cereal production systems. Highly soluble N fertilizers like urea may be lost from the soilplant system through leaching, volatilization and denitrification thereby reduce yields and NUE of arable crops. There are different mechanisms to improve the nitrogen fertilizer use efficiency. The use of appropriate N fertilizer, application rate and time are among the main management options to increase N fertilizer use efficiency. In addition, the use of slow N-releasing fertilizers like UREA satbil is an option to alleviate the problem.

Introduction
Nitrogen (N) transformation in soils, climate and sub-optimal fertilizer management practices have led to reduce N use efficiency (NUE) in many instances (i.e. up to 30-50%) (Sutton et al., 2013). These factors have contributed to an increase in N losses such as ammonia (NH 3 ) volatilization, nitrate (NO 3 − ) leaching and nitrous oxide (N 2 O) emissions (IPCC, 2007), which are of economic and environmental concern. There are different mechanisms to improve the nitrogen fertilizer use efficiency. Cropping system, soil and water management, use of appropriate N fertilizer and application rate are the main management options to increase N fertilizer use efficiency (Fageria, 2009). Enhanced-efficiency fertilizers such as those containing nitrification inhibitors (NIs) and urease inhibitors (UIs) have been developed to increase NUE and reduce N losses by increasing the congruence between N supply and crop N demand. This effect is achieved by delaying the bacterial oxidation of ammonium (NIs) or the hydrolysis of urea (UIs). However, the use of these technologies is under debate because there are studies in which yield increases are not observed despite the additional costs (Akiyama et al., 2010).
UREA Stabil is one form of slow nitrogen-releasing urea. The nitrogenous fertilizer UREA Stabil is urea enriched with the inhibitor of urease enzyme NBPT (N-(n-butyl)-thiophosphoric triamide). It is a new fertilizer that was registered in the Czech Republic in 2006. UREA Stabil reduces losses due to volatilization, leaching and denitrification (Mraz, 2007). In Hawzien and Emba Alaje districts, application of N as Urea fertilizer and P as Diammonium Phosphate (DAP) at the rate of 100 kg urea and 100 kg DAP ha −1 is a recommended rate to boost up wheat production and productivity. Nevertheless, wheat production and productivity are not as such satisfactory from year to year. At Hawzien and Emba Alaje districts, N losses from applied urea fertilizer might be through volatilization, denitrification and leaching. Therefore, an appropriate source of N fertilizer, rate and time of application may improve N fertilizer use efficiency of the wheat. However, the effects of slow nitrogen-releasing fertilizer on yield and yield components of wheat at Hawzien and Emba Alaje districts were not investigated. Therefore, this study was initiated to investigate the effects of slow nitrogen-releasing fertilizer (UREA Stabil ) on yield and yield components of bread wheat and to quantify nitrogen use efficiency of wheat in Hawzien and Emba Alaje.

Description of the study areas
The studies were conducted in Eastern Zone of Tigray Region, on Cambisols of Hawzien district and on Vertisols of Emba Alaje in 2015 main cropping season (Dargie et al., 2018). Soils of Hawzien District are one of the most degraded soils in the Tigray Region (Northern Ethiopia), which are very low in soil organic matter content and macro-nutrients such as N, P and K (EthioSIS, 2014). The dominant soil type in the district is Cambisols and the dominant soil type in Emba Alaje is Vertisols (Tigray finance and economic bureau [TFEB], 1995). The main soil properties of the six experimental sites are presented in Table 1.

Soil properties 2.3. Experimental design, treatments and procedures
The experimental design was a randomized complete block design (RCBD) with three replications. In all sites, the preceding crop was wheat or barley. Plot sizes were 4 m X 3 m. The spacing between plots and blocks was 50 and 100 cm, respectively. The spacing between wheat plant rows was 20 cm. A disease-resistant, early maturing and relatively high yielding varieties called Pica flor (kakaba) and kingbird were chosen for Hawzien and Emba Alaje, respectively.
Four levels (0, 32, 64 and 96) kg ha −1 of nitrogen were tested. The nitrogen source was UREA Stabil which is slow N-releasing fertilizer. Treatments were applied at planting (sowing). Positive control of prilled urea (conventional urea) was included with a split application of 1/3 of 64 kg N ha −1 at planting and 2/3 of N was applied at the tillering stage in the form of urea. Phosphorus in the form of triple superphosphate (TSP) was applied at the rate of 46 kg ha −1 as (P 2 O 5 ). Potassium in the form of chlorite potash (KCl) was applied at the rate of 100 kg ha −1 . Sulfur in the form of CaSO 4 was applied at the rate of 30 kg S ha −1 . P, K and S fertilizers were applied as a basal application at planting to all plots.

Data collection and data analysis
Agronomic data were collected following the standard procedures. Plant samples were collected at maturity for tissue analysis. Analysis of variance (ANOVA) and correlation were carried out using the SAS software program using SAS version 9.1.3 (SAS, 2002). A significant difference among treatment means, agronomic, physiological and apparent recovery efficiency were assessed using the least significant difference (LSD) at 0.05 level of probability (Gomez & Gomez, 1984). Normality and homogeneity of variance were checked using Anderson-Darling and Bartlett test, respectively, using SAS software (SAS, 2002).

Plant height
Results revealed that plant height was significantly (P < 0.001) affected by rates of nitrogen at all sites (Table 2). Plant height increased with increasing rates of nitrogen at all sites. At Suluh site, differences in plant height were significant, but at Atsela and Ayba sites, differences in plant height were significant only between the control and the other treatments. This may be due to the sufficient nitrogen in the soil (Table 1) and an abundant amount of rainfall. At Suluh site, the differences in plant height were significant on plots received the same rates of N (64 N kg ha −1 ), but from different sources (conventional Urea and UREA Stabil ) and it was higher for plots received N as UREA Stabil . At all sites, 32 kg N ha −1 had also no statistically significant difference with recommended conventional urea (64 N kg ha −1 ). This indicates UREA Stabil is superior to conventional urea in increasing plant height at Suluh site which might be due to the slow release of N that could accessible to plant throughout the growing period. The shortest plant height was recorded on plots that received 0 kg N ha −1 at all sites.
The results obtained from Suluh site in Hawzien were in line with the finding of Khan et al. (2013) stated that plant height of wheat was high on plots received urea treated with urease and nitrification inhibitors than the conventional urea only. The results obtained from Atsela and Ayba sites in Emba Alaje were in line with the findings of Espindula et al., (2014) who stated that there was no significant difference at (P < 0.05) between wheat plant heights on plots treated with different rates of urea and urea +NBPT. Mengel and Kirkby (2001) reported significant increments in plant height due to the application of high nitrogen rate.

Spike length
Spike length is one of the yield attributes of wheat that contributes to grain yield. Crops with higher spike length could have higher grain yield. Spike length was significantly (P < 0.01) influenced by the rate and source of nitrogen fertilizer at all sites. At Suluh site in Hawzien district, there was an increasing trend in spike length with N rates. The highest and lowest spike length were recorded on plots treated with 96 kg N ha −1 as UREA Stabil (7.38 cm) and on control plots, 0 kg N ha −1 (6.24 cm), respectively (Table 2). At Atsela and Ayba sites in Emba Alaje district, the highest and lowest spike length were recorded on plots treated with 64 kg N ha −1 as UREA Stabil (8.04 cm) and 0 kg N ha −1 (7.08 cm), respectively. The differences were higher for plots treated with N as UREA Stabil than plots treated with conventional urea at the same rates (64 kg N ha −1 ). Plots received 32 kg N ha −1 had higher spike length than 64 kg N ha −1 in the form of conventional urea.

Number of effective tillers
Number of effective tillers contributes a lot to the total biomass gain and other yield attributes. The analysis of variance showed there was a highly significant difference (P < 0.01) due to the difference in nitrogen rates. The highest numbers of effective tillers were recorded for plots treated at 32 kg N ha −1 and at 64 kg N ha −1 of UREA Stabil at Suluh in Hawzien and at Atsela and Ayba site in Emba Alaje, respectively (Table 3). The lowest numbers of effective tillers were recorded for plots received zero nitrogen at all sites and this might be due to the role of N in accelerating vegetative growth of plants. The results were in agreement with Abdullatif et al. (2010) reported shows increasing in the number of effective tillers with nitrogen fertilization. Haileselassie et al. (2014) and Abdollahi Gharekand et al. (2012) also reported that nitrogen fertilization has a significant effect on the effective number of tillers of wheat.

Number of seeds per spike
The analysis of variance showed a significant difference (P < 0.01) on the number of seeds per spike due to the effects of sources and rates of nitrogen. There was an increasing trend in the number of seeds per spike across UREA Stabil treatments at all sites. The lowest and highest value were recorded at zero nitrogen treatment and 96 kg N ha −1 as UREA Stabil respectively at both sites (Table 3). At Suluh site differences in the number of seeds per spike was significant for the same rates of nitrogen (64 kg N ha −1 ) but from different sources (UREA Stabil and conventional urea) ( Table 3). The values were higher for plots treated with UREA Stabil than plots treated with conventional urea at equal rates. Number of seeds per spike for plots treated with 32 kg N ha −1 UREA Stabil was statistically significant on plots received 64 kg N ha −1 as conventional urea at Hawzien. This might be due to the existence of NBPT in UREA Stabil which has the capacity to inhibit or slow down the nitrification process and made N to be accessible for the rest of the growing period. Similar results have been reported for the influence of urease and nitrification inhibitors on yield and yield components including the number of seed per spike, plant height, spike length, number of tillers, TSW of wheat by Zaman et al. (2009).

Thousand seed weight
Analysis of variance showed that nitrogen rates and sources significantly influence the thousand seed weight. The highest thousand seed weight was recorded for plots treated at 64 kg N ha −1 UREA Stabil and the lowest was at zero nitrogen at Suluh in Hawzien. The thousand seed weight obtained at plots treated with 32 kg N ha −1 UREA Stabil was statistically the same with plots received 64 kg N ha −1 as conventional urea at Hawzien. On the contrary, N rates and sources had no significance effect on the thousand seed weight at Atsela and Ayba in Emba Alaje. Numerically, thousand seed weight was higher in Emba Alaje than in Hawzien. This may be due to the genetic potential of the wheat variety used, the environmental stress especially rainfall and differences in grain filling period. The results recorded from Suluh site in Hawzien district were in line with Abdollahi Gharekand et al. (2012), Khan et al. (2013) and Xu et al. (2002) reported that urease and nitrification inhibitors influenced the thousand seed weight of wheat. Espindula et al. (2013) also reported that plants received urea had lower thousand grain mass than those which received urea with inhibitors.

Grain yield
At Suluh site in Hawzien, there was an inconsistent increment in the amount of grain yield with rates of N. The highest grain yield was recorded from plots treated with 64 kg N ha −1 as UREA Stabil (1708.33 kg ha −1 ) and the lowest was recorded from plots that received 0 kg N ha −1 (1102.73 kg ha −1 ) ( Table 5). Grain yield obtained from plots treated with 64 kg N ha −1 in the form of conventional urea was significantly different (P < 0.05) from grain yield obtained from plots treated with the same rates of N as UREA Stabil . Higher grain yield was obtained from plots treated with N as UREA Stabil . In addition to this, grain yield obtained from plots treated with 32 kg N ha −1 as UREA Stabil was not significantly different from 64 kg N ha −1 in the form of conventional urea. This might be due to the slow-releasing effect and low volatilization of N when applied as UREA Stabil which could contribute to the higher grain yield. Khan et al. (2013) reported a higher grain yield of wheat when N was applied as super urea (urease plus nitrification inhibitor-treated urea).
At Atsela and Ayba sites in Emba Alaje, there was an increasing trend in grain yield with N application rates. Haileselassie et al. (2014) also reported that an increasing rate of nitrogen fertilization increased grain yield of wheat. The highest grain yield was recorded for plots that received 64 kg N ha −1 (5467.9 kg ha −1 ) as conventional urea. This might be due to the application of UREA Stabil only at planting while N as convectional urea was applied through the split application. Therefore, split application of conventional urea (1/3 at planting and 2/3 at tillering stage) has an advantage to obtain higher grain yield at Atsela and Ayba sites in Emba Alaje district. Nelson et al. (2014) reported that split-N applications as urea, Urea+NBPT generally resulted in greater wheat yields than the application of N at planting. Hirzel et al. (2010) also reported that the highest grain yield of durum wheat was recorded for plots treated with a split application of urea. From these and results in Table 4, it is possible to predict that split application of N as UREA Stabil could result in higher grain yield than conventional urea at Atsela and Ayba sites in Emba Alaje. Application of full dose of UREA Stabil at planting may increase the loss due to denitrification and leaching at Emba Alaje district. This was in line with Picone et al. (2014) reported that the denitrification rate in wheat crop was observed when N fertilizer was applied and the rain was more frequent and intensive.
The mean grain yield at Suluh site in Hawzien was lower compared with that of Atsela and Ayba sites in Emba Alaje. This could be due to differences in wheat variety used, soil fertility status and amount of rainfall received. The fertility status of soils of Atsela and Ayba sites in Emba Alaje is better than that of Suluh site in Hawzien (Table 1). The amount of rainfall was also higher at Emba Alaje (960.4 mm) than that received at Hawzien (371.9 mm) during the cropping season (Dargie et al., 2018). As a result of these, grain yield was not as such satisfactory at the Suluh site in Hawzien.

Straw yield
At Suluh site in Hawzien, straw yield showed an increasing trend even though there are some inconsistencies. The lowest and highest straw yield were obtained from control plots with no nitrogen treatment (2259.3 kg ha −1 ) and from plots received 64 kg N ha −1 (3400.0 kg N ha −1 ) in the form of UREA Stabil , respectively. There were no statistically significant differences in straw yield between plots treated with the same rates of N as conventional urea and UREA Stabil (Table 4). Espindula et al. (2013) reported that Urea+ NBPT had significant effect on straw yield of wheat. Khan et al. (2013) also reported that maximum straw yield was obtained from plots treated with super urea (urease plus nitrification inhibitor-treated urea).
At Atsela and Ayba sites in Emba Alaje, straw yield showed consistently increasing trend with N rates that applied as UREA Stabil (Table 4). The highest and lowest straw yields were recorded from plots treated with 64 kg N ha −1 as conventional urea (7810.1 kg ha −1 ) and from zero nitrogen, respectively. Abebe (2012) and Haileselassie et al. (2014) reported that wheat straw yield increased with N rates. For the same rates of N (64 kg N ha −1 ) applied as UREA Stabil and conventional urea, higher straw yield was obtained from plots to which conventional urea was applied with spilt application. This clearly shows the importance of split application of N at Atsela and Ayba sites in Emba Alaje district.

Harvest index
Harvest index was computed as the ratio of grain yield to the total above-ground dry biomass yield. The analysis of variance revealed that rates and sources of N did not significantly affect (P > 0.05) harvest index of wheat at all sites. A similar result was reported by Espindula et al. (2013). The results found in contrary to Espindula et al., (2014) reported that Urea+NBPT had a significant effect on the harvest index of wheat than urea.

Nitrogen uptake and nitrogen use efficiency indices
Application of N significantly influenced uptake, agronomic, physiological and apparent recovery efficiency of wheat at both sites. Nitrogen uptake was increased with a rate of up to 64 kg N ha −1 then starts to decline in Hawzien (Figure 1). On contrary in Emba Alaje grain, straw and total uptake of wheat increased with rate. The highest uptake of wheat was retrieved on plots that received UREAStabil than conventional urea at the same rate (64 kg N ha −1 ). The highest agronomic efficiency of 9.08 kg −1 (32 kg N ha −1 ) and apparent recovery of 55% was obtained at 64 kg N ha −1 as UREA Stabil and physiological efficiency of 60.28 kg −1 was obtained at 64 kg N ha −1 as conventional urea at Hawzien district. The highest agronomic efficiency of 22.2 kg −1 , physiological efficiency of 93.7 kg −1 , apparent recovery of nitrogen 59.7% was obtained at 64 kg N ha −1 as conventional urea at Atsela and Ayba sites in Emba Alaje district.

Correlation among agronomic parameters of wheat
Correlation analysis among growth parameters, yield-related traits and grain yield from all sites is presented in Tables 6 and 7. The correlation analysis revealed that there was a significant and positive correlation between grain yield and yield-related agronomic parameters of wheat at all sites. At Suluh site in Hawzien district, grain yield was significantly correlated with plant height (r = 0.625), number of effective tillers (r = 0.456), thousand seed weight (r = 0.365), straw yield (r = 0.513), harvest index (r = 0.688) ( Table 6). Correlation between grain yield and spike length (r = 0.289), and number of seed per spike (r = 0.203) were not significant. Straw yield was positively correlated with almost all agronomic parameters except thousand seed weight (r = 0.152), grain filling period (r = 0.129) and harvest index (r = −0.244) at Suluh site in Hawzien. At Ayba and Atsela sites in Emba Alaje district grain yield was significantly correlated with plant height (0.771), spike length (r = 0.732), number of effective tillers (r = 0.751), number of seed per spike (r = 0.661), straw yield (r = 0.891) and thousand seed weight (r = 0.636) ( Table 7). Applying higher rates of nitrogen delays heading and maturity which ultimately increase grain yield. Under normal condition plants need more time to efficiently utilize the   resource which is available in the soil and this may ultimately affect grain and straw yield. Thousand seed weight, effective tillers, number of seed per spike also affected significantly and positively grain yield. Grain yield was negatively correlated with harvest index (r = −0.140). Straw yield was positively correlated with all phenological, growth and yield traits of wheat at Ayba and Atsela sites in Emba Alaje district (Table7). Jaglan et al. (1997) and Abebe (2012) reported that grain yield was significantly and positively correlated with plant height, straw yield, thousand seed weight and biomass of wheat.

Conclusion
Application of different rates of UREA Stabil and conventional urea significantly affected most of the crop parameters tested, such as growth parameters, yield and yield components at Hawzien and Emba Alaje. Straw yield showed an increasing trend even though there were some inconsistencies. The lowest and highest straw yield were obtained from control plots with no nitrogen treatment (2259.3 kg ha −1 ) and from plots received 64 kg N ha −1 (3400.0 kg N ha −1 ) in the form of UREA Stabil , respectively, at Suluh site in Hawzien. At Suluh site in Hawzien, the highest grain yield was obtained from plots treated with 64 kg N ha −1 as UREA Stabil and the lowest was recorded from plots received 0 kg N ha −1 . Grain and straw nitrogen uptake of wheat were increased with N rate in Emba Alaje, but not in Hawzien. Nitrogen use efficiency indices were influenced by nitrogen rates and sources at both sites. Grain and straw yields had a positive correlation with most of the parameters tested. Slow-releasing nitrogen fertilizer UREA Stabil could be used as an alternative source of nitrogen fertilizer for wheat production at Hawzien. Further study should be done on solubility, time of application and placement of slow-releasing nitrogen fertilizer which is UREA Stabil .

Funding
The authors received no direct funding for this research.