Sunflower response to potassium fertilization and nutrient requirement estimation

Ll Shu-tian , DUAN Yu, GUO Tian-wen, ZHANG Ping-liang, HE Ping , Kaushik Majumdar

1 Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing 100081, P.R.China

2 International Plant Nutrition Institute Beijing Office, Beijing 100081, P.R.China

3 Institute of Resources Environment and Detection Technology, Inner Mongolia Academy of Agricultural and Animal Husbandry Sciences, Hohhot 010031, P.R.China

4 Dryland Agriculture Institute, Gansu Academy of Agricultural Sciences, Lanzhou 730070, P.R.China

5 International Plant Nutrition Institute (IPNI), Gurgaon 122001, India

Abstract Field experiments were conducted in oil and edible sunflower to study the effects of potassium (K) fertilization on achene yield and quality, and to estimate the nutrient internal efflciency (IE) and nutrient requirement in sunflower production. All trials in edible sunflower and 75% trials in oil sunflower showed positive yield responses to K fertilization. Compared with control without K fertilization, the application of K increased achene yield by an average of 406 kg ha–1 for oil sunflower and 294 kg ha–1 for edible sunflower. K application also increased 1 000-achene weight and kernel rate of both oil and edible sunflower. K fertilization improved the contents of oil, oleic acid, linoleic acid and linolenic acid in achenes of oil sunflower,and increased contents of oil, total unsaturated fatty acid and protein in achenes of edible sunflower. The average agronomic efflciency of K fertilizer was 4.0 for oil sunflower and 3.0 kg achene kg–1 K2O for edible sunflower. The average IE of N, P and K under balanced NPK fertilization was 22.9, 82.8, and 9.9 kg kg–1 for oil sunflower, and 27.3, 138.9, and 14.3 kg kg–1 for edible sunflower. These values were equivalent to 45.5, 14.1, and 108.1 kg, and 39.0, 8.0, and 71.7 kg of N, P and K,respectively, in above-ground dry matter required for production per ton of achenes. The average harvest index of N, P and K was 0.47, 0.56 and 0.05 kg kg–1 in oil sunflower, and 0.58, 0.58 and 0.14 kg kg–1 in edible sunflower.

Keywords: sunflower, potassium, yield response, K use efflciency

1. lntroduction

China is the 5th highest sunflower production country in the world following Russia, Argentina, India and America.China now has around 1 million hectares of sunflower with annual achene production of 2.7 million tones. The areas of oil sunflower and edible sunflower are 40 and 60% of the total sunflower area in China, respectively, and about 73%of sunflower area and 79% of achene production are from the northwest region, mainly in Inner Mongolia Autonomous Region (IMAR) and Gansu Province (MOA 2015).

Nutrient management is one of the main factors that influences sunflower achene yield, kernel oil and fatty acid contents. Farmers have traditionally applied organic manure plus nitrogen (N) and phosphate (P) fertilizers, resulting in overuse of N and P but insufflcient potassium (K) input in sunflower production (Tuoet al.2010). This imbalanced fertilization practice not only influences sunflower achene yield but also decreases other nutrient efflciencies(Amanullah and Khan 2010). Insufflcient K application requires plants to acquire K mainly from native soil, leading to mining of native K reserves which may affect sunflower yield and quality.

Sufflcient K supply increases crop growth and productivity as well as promotes the tolerance of crops to environmental stress situations, particularly increasing drought tolerance (Jákliet al.2016). Sunflower plants with high K levels showed greater resistance to drought stress conditions and higher yield and dry matter allocation to the grain fllling process than plants with low K supply(Soleimanzadehet al.2010a). K promotes enzyme activity, which contributes to carbohydrate, protein and fat metabolism and affects sunflower achene quality (Tisdaleet al.1985). Although the amount of K uptake by sunflower plants varied greatly due to different species and growing conditions, it tends to be greater than N or P uptake (Liet al.2009; Jiang 2011).

Most previous studies focused on sunflower yield responses to K application (Anet al.2007; Cheng and He 2011). Although a few studies focused on salt tolerance influence on sunflower quality measures such as protein,oil content and compositions (Ebrahimianet al.2011), they rarely considered changes in achene quality in response to K fertilization and nutrient requirement estimation.Studies have indicated that N content had the greatest direct effect on achene oil concentration, while Mg, P and N had the primary influences on oleic acid concentration,and K content had the highest indirect effect on fatty acid concentration (Bozkurt and Karacal 2001). While other studies showed that N increased oil content and oil production, P only affected the 1 000-achene weight and N and P did not change fatty acid content (Skarpa and Losak 2008). Therefore, it is necessary to clarify the effect of K fertilization on oil and fatty acid contents of sunflower achenes as well as nutrient requirements for producing a target yield.

The objectives of this research were to study: 1) the effect of K fertilization on sunflower achene yield and quality; and 2) nutrient internal efflciency and nutrient requirements of sunflower.

2. Materials and methods

2.1. Field experiments

A total of 60 fleld experiments on oil sunflower in Jingyuan,Jingtai and Dingxi counties of Gansu Province and 60 fleld experiments on edible sunflower in Wuchuan County, Linhe and Wuyuan cities of IMAR were conducted during 2014 to 2016. There were two treatments in each experiment: 1) the balanced NPK recommendation of Agro Services International (ASI) systematic approach (Hunter 1980; Portch and Hunter 2002), and 2) control without K fertilization, each with no replicates and plot areas of 60 to 120 m2. Oil sunflower varieties used in the experiments were Longkui 2 and Longkui 3. Edible sunflower varieties used in the experiments were Hybrid 318 and Hybrid 3639.All fertilizers were applied basally before planting. Planting dates were April 20 to April 25 in Gansu and May 25 to May 30 in IMAR. The plant density was 50 000 to 60 000 plants ha–1for oil sunflower and 35 000 to 40 000 for edible sunflower. The average rainfall in the study area was 145 and 146 mm during the growing periods in Gansu and IMAR,respectively. Oil sunflower was irrigated with 300 mm water at bud and flowering stage, while edible sunflower was irrigated 240 mm before planting and 60 mm at flowering.The mean temperatures in study areas were 9.4°C in Gansu and 7.7°C in IMAR.

Soil samples were collected before planting, air-dried,ground and passed through a 2-mm sieve. Properties of tested soils are listed in Table 1. At harvest, the achene yield was tested by harvesting all the heads of each plot after removing two edge rows in the plot. The heads were air-dried, threshed and then the achenes were weighted.Five sunflower plants were randomly selected from each plot and separated into stems, leaves, achenes and heads,which were each cut into 3–5 cm pieces, oven-dried at 70°C and then ground to pass through a 0.2-mm sieve for analysis.

2.2. Analytical procedures

Soil organic C and mineral Nwere determined according to the corresponding methods in Sparkset al.(1996). Soil available P was analyzed by Olsen extraction (Olsenet al.1954) and exchangeable K by ammonium acetate (NH4OAc) extraction (Knudsen 1982).Soil pH was determined in a 1:2 soil:water (w/v) mixture(McKeague 1978). Total N, P and K in sunflower achenes and residues were extracted by H2SO4-H2O2digestion, with N in the digested solution determined using the Kjeldahl method, P in the solution determined by the molybdenum antimony–ascorbic acid colorimetric method, and K in the solution determined by flame emission spectrophotometer(Lu 1999). The oil content of sunflower achenes was extracted by a continuous flow extractor using Soxhlet’s method (AOAC 1990), and the fatty acids were transformed to their methyl esters (Metcalfet al.1966) and determined using a gas chromatograph.

Table 1 Properties of tested soils and nutrients applied in the experiments

2.3. Calculation

Yield response to K fertilization (YRK), agronomic efflciency of K fertilizer (AEK), nutrient internal efflciency (IE) and the nutrient requirement (RIE) are calculated below:

Where, Y is achene yield (kg ha–1) in the balanced NPK treatment; Y0is achene yield (kg ha–1) in the K control treatment; Kfis K fertilizer rate (kg K2O ha–1); and Nu is total N, P or K in above-ground dry matter of the NPK treatment(kg ha–1).

2.4. Statistical analysis

The signiflcant differences of achene yields and quality indices between K treatments and K control plots were analyzed by pairedt-test using SAS software. Statistical analyses of yield response, agronomic efflciency, internal efflciency and nutrient uptake were performed by the Microsoft Excel package 2010.

3. Results

3.1. Yield response to K fertilization

Achene and straw yield, head diameter, 1 000-achene weight and kernel rate data are presented in Table 2. For oil sunflower, the average achene yield obtained with K fertilization was 3 951 kg ha–1and the maximum attainable yield was 5 520 kg ha–1. More than 2/3 of the trials had yields between 3 000 and 5 000 kg ha–1. Statistical analysis by pairedt-test showed signiflcant (P＜0.001) differences in achene yield, straw yield, head diameter, 1 000-achene weight, and kernel rate between plots with and without K fertilization. On average, potassium application increased achene yield by 406 kg ha–1, straw yield by 721 kg ha–1,head diameter by 0.9 cm, 1 000-achene weight by 4.2 g,and kernel rate by 2.7%, compared with the control without K fertilization. Among the 60 trials, 46 showed increases of achene yield, 45 showed increases of head diameter, 47 showed increases of 1 000-achene weight and 44 showed kernel rate increases with K fertilization.

For edible sunflower, the average achene yield with K fertilization was 3 703 kg ha–1, the maximum attainable yield was 6 169 kg ha–1, and more than half of the trials had achene yields between 3 000 and 4 000 kg ha–1. On average,potassium fertilization signiflcantly increased achene yield(P＜0.001) by 294 kg ha–1, straw yield (P＜0.001) by 622 kg ha–1, 1 000-achene weight (P＜0.05) by 4.9 g, and kernel rate(P＜0.05) by 0.6%. All 60 trials showed increases of achene yield, 39 of the 60 trials showed increases of 1 000-achene weight and 43 of the 60 trials showed increases of kernel rate by K fertilization.

Agronomic efflciency of K (AEK) is the achene yield response per unit of K2O fertilizer applied, and reflects the efflciency of applied potassium fertilizer. The AEKfor oil sunflower was in the range of 0.0 to 12.1 kg achene kg–1K2O, with an average of 4.0 kg achene kg–1K2O, while 60%of the observations were between 2 and 10 kg achene kg–1K2O, and 75% of the observations were less than 6.8 kg achene kg–1K2O. The AEKfor edible sunflower was in the range of 0.2 to 8.2 kg achene kg–1K2O, with an average of 3.0 kg achene kg–1K2O, while 63% of the observations were between 2 and 6 kg achene kg–1K2O, and 75% of the observations were less than 3.9 kg achene kg–1K2O (Fig. 1).

3.2. Achene quality response to K fertilization

Compared with the K omission plots, K fertilization signiflcantly increased the contents of oil, total saturated fatty acids and several unsaturated fatty acid in achenes of both oil and edible sunflower (Table 3). For oil sunflower,the contents of oleic acid, linoleic acid and linolenic acid in achenes were signiflcantly increased by K application,while crude protein content was not signiflcantly affected.For edible sunflower, K fertilization did not signiflcantly influence the contents of oleic acid or linoleic acid, but it increased the crude protein content in achenes. These results indicated that K fertilization improved the quality of sunflower achenes.

3.3. Nutrient uptake and efficiency

Nutrient uptakeNutrients in above-ground dry matter (DM)under balanced NPK treatments are presented in Tables 4 and 5. N, P and K in above-ground DM varied greatly,ranging from 69.7 to 324.8 kg N ha–1, 21.5 to 153.9 kg P ha–1, and 134.8 to 793.5 kg K ha–1in oil sunflower; and from 52.8 to 281.6 kg N ha–1, 11.5 to 68.1 kg P ha–1, and 130.2 to 493.2 kg K ha–1in edible sunflower. N and P in achenes were much higher than in other parts for both oil and edible sunflower. N in leaves was higher than in stems and heads,especially in oil sunflower, while P was evenly distributed in different parts of straw, and K in stems was much more than in leaves and heads.

Table 2 Comparison of achene yield, straw yield and some agronomic traits of sunflowers between K fertilization treatment (+K)and control treatment without K fertilization (–K)

Fig. 1 Frequency distribution of agronomic efflciency of K (AEK) for oil sunflower (left) and edible sunflower (right).

Table 3 Comparison of oil, fatty acid and protein contents between K fertilization (+K) and K control plots (–K)

Table 4 Plant N, P and K accumulation in achenes, leaves, stems, heads and above-ground plant dry matter (DM), and nutrient harvest index at maturity stage of oil sunflower in NPK treatments (n=60)

On average, N, P and K in above-ground DM were 179.4 kg N ha–1, 55.2 kg P ha–1and 425.9 kg K ha–1for oil sunflower, 147.7 kg N ha–1, 30.5 kg P ha–1and 268.7 kg K ha–1for edible sunflower, and produced achene yields of 3 951 kg ha–1and 3 703 kg ha–1, respectively. The nutrient harvest index (HI), i.e., nutrient accumulation in achenes as a proportion of nutrient accumulation in above-ground DM, was similar for N and P, but much lower in case of K(Tables 4 and 5). The HI of N, P and K in oil sunflower under balanced NPK fertilization was 0.24–0.63, 0.35–0.88, and 0.02–0.11 kg kg–1, with an average of 0.47, 0.56 and 0.05 kg kg–1, respectively. The HI of N, P and K in edible sunflower under balanced fertilization was 0.39–0.74, 0.41–0.74, and 0.07–0.21 kg kg–1, with an average of 0.58, 0.58 and 0.14 kg kg–1, respectively.

Nutrient internal efficiency and requirementNutrient internal efflciency (IE) represents the ability of nutrient uptake from soil and applied fertilizers to contribute to the economic yield, and is represented by achene yield (kg) per unit of nutrient (kg) in above-ground DM. The reciprocal of IE (RIE) represents the amount of nutrient in above-ground DM required for producing a unit of sunflower achenes, and is usually expressed as kg nutrient per ton of achenes. The average IE of N, P, and K for oil sunflower was 22.9, 82.8,and 9.9 kg kg–1, equivalent to RIE of 45.5, 14.1 and 108.1 kg t–1, respectively; or an N:P:K ratio of 3.23:1:7.67 in plant DM. The mean IE of N, P, and K for edible sunflower was 27.3, 138.9, and 14.3 kg kg–1and the corresponding RIE was 39.0, 8.0 and 71.7 kg t–1, respectively; with an N:P:K ratio of 4.88:1:8.96 in plant DM (Table 6).

Table 5 Plant N, P and K accumulation in achenes, leaves, stems, heads and above-ground plant dry matter (DM), and nutrient harvest index at maturity stage of edible sunflower in NPK treatments (n=60)

4. Discussion

4.1. Effect of K fertilization on sunflower achene yield and quality

K fertilization signiflcantly increased sunflower achene and straw yield as well as individual yield components (Table 2),which is consistent with other studies (Lewiset al.1991;Ayubet al.1999; Ahmadet al.2001; Amanullah and Khan 2010; Wanget al.2012; Baiet al.2016, 2017). Some studies also demonstrated that K application had positive effects on 1 000-grain weight and kernel rate, but not on head diameter (Amanullah and Khan 2010; Ertiftik and Zengin 2016). There were great variations in achene yield, straw dry matter and agronomic traits such as head diameter,1 000-achene weight and kernel rate. These variations might be largely caused by differences in sunflower varieties, soil conditions and management practices, because most of the experiments were conducted in farmer’s flelds and each farmer had their own management practices which reflected a broad range of soil, climate and management conditions.These experiments were conducted under balanced nutrient supply and a uniform sampling methodology, so the results can provide valuable information for sunflower nutrient management in the main sunflower areas of northwest China.

K fertilization also improved sunflower product qualities that were reflected in the increase of some quality indices listed in Table 3. The K fertilization effects on contents of oil and fatty acid in achenes of sunflowers were through indirect influences, different from the direct effect that N had on achene oil concentration and oleic acid content(Bozkurt and Karacal 2001). Sunflower is very sensitive to K, so K deflciency results in low oil yield because of both low grain yield and low oil concentration in grains (Abbadiet al.2008). Since oil sunflower in Gansu and edible sunflower in IMAR were both planted in drought conditions with limited precipitation and water supply, K application could alleviate the drought stress and improve oil content and oil yield (Soleimanzadehet al.2010b). Sunflower plants with high levels of K showed higher resistance to drought stress conditions and higher yield and dry matter allocation to the grain fllling process (Soleimanzadehet al.2010a).

Achene quality responses to K fertilization in oil sunflower in this study show some differences from research in Pakistan. In those studies, the application of K had signiflcant effects on achene oil concentrations in both spring and autumn seasons for oil sunflowers, but had no signiflcant influence on the composition of unsaturated fatty acids in the spring season, while it substantially increased achene linoleic acid concentration in the autumn season(Ahmadet al.2001). These different effects of K fertilization on sunflower product quality may be due to differences in the varieties, soil conditions and climate conditions in the two studies (Wittet al.1999).

Table 6 Internal efflciency (IE) and the nutrient requirement (RIE) for N, P and K at maturity stage of sunflower with NPK treatments(n=60)

Generally, an increase of N supply signiflcantly increases protein content and decreases oil content for edible oil crops(Huanget al.1985; Hu 2005). However, some studies have found K fertilization increased both oil and protein contents in achenes of edible sunflower (Maet al.2015) and oil sunflower (Qianet al.2015). These flndings are similar to the results of our study, where K application increased oil contents in achenes of both oil and edible sunflower,and increased achene protein content of edible sunflower but did not affect achene protein content of oil sunflower compared with K omission plots (Table 3). This information suggests that K promotes metabolism and transformation of carbohydrates and then affects the contents of oil and protein in achenes simultaneously. Therefore, there is no signiflcant relationship between the protein and oil contents of sunflower achenes in the current study.

4.2. Nutrient efficiency and requirement

The recommended NPK treatments, where the three macronutrients were balanced fertilization in the farmer’s fleld, were used to calculate IE of N, P and K and the corresponding RIE, reflecting the balanced nutrient uptake and avoiding the influence of unbalanced nutrient supply on nutrient uptake and achene yield. There were variations of IEs that may be derived from the sunflower varieties, crop management, soil and climatic conditions (Wittet al.1999;Andrianasoloet al.2014; Kianiet al.2016). However, that the IE of P was high while the IE of K was low reflected the characteristics of sunflower nutrition, and was similar to the results of other studies of 51.3 to 140.8 (avg. 84.7) kg kg–1P, and 7.3 to 28.0 (avg. 11.5) kg kg–1K for oil sunflower (Liet al.2009, 2010; Cheng and He, 2012; Wanget al.2014;Ren 2015), and 75.8 to 169.5 (avg. 111.1) kg kg–1P, and 6.9 to 16.3 (avg. 10.8) kg kg–1K for edible sunflower (Liet al.2009; Baiet al.2016, 2017; Duanet al.2018).

The average RIE of oil sunflower was 45.5 kg N t–1, 14.1 kg P t–1and 108.1 kg K t–1with N:P:K ratio of 3.23:1.0:7.67,which was higher than RIE of edible sunflower of 39.0 kg N t–1, 8.0 kg P t–1and 71.7 kg K t–1with N:P:K ratio of 4.87:1.0:8.96 (Table 6), suggesting that oil sunflower required more nutrients (especially P) than edible sunflower.These results are similar to other studies (Liet al.2009;Jiang 2011) that suggested oil sunflower requires more N, P and K than edible sunflower for producing a unit of achene yield. This higher requirement is most probably due to the higher achene oil content of oil sunflower that needs more nutrients than edible sunflower, because of the direct effects of N and P and indirect influences of K on achene oil concentration and fatty acid contents (Bozkurt and Karacal 2001). Studies have also indicated that as the amount of K in sunflower increases, the oil content also increases (Ertiftik and Zengin 2016).

The relationships between achene yield and N, P or K in above-ground DM under NPK treatments represent the yield response to balanced nutrient uptake (Fig. 2).These relationships indicate that achene yield is not always linearly related to nutrient uptake in plants, since the nutrient requirement estimation based on these relations may overestimate the actual nutrient requirement under conditions of luxurious uptake of nutrients which would result in low calculated IEs. So the estimation of nutrient uptake requirements should be based on models like QUEFTS(Quantitative Evaluation of the Fertility of Tropical Soils)(Janssenet al.1990), there are not sufflcient nutrient uptake data in the low level of achene yield to establish such a model. Generally, nutrient uptake in above-ground DM is linearly related to crop economic yield up to 60 to 70% of maximum yield (Wittet al.1999; Byjuet al.2012; Xuet al.2015). Therefore, the average RIE estimated based on balanced nutrient uptake under NPK treatments in this study can be used as a reference for fertilizer recommendation in sunflower production at yields between 3 000 and 4 000 kg ha–1, which is 60 to 70% of maximum yield and generally representative of local farmer’s yields in sunflower production areas in Gansu and IMAR.

The harvest indexes of N and P were substantially more than K for both oil and edible sunflower. This was indicated by a great portion of K in sunflower straw, especially in stem, and only a small portion in achenes. This result is consistent with results of other studies that reported 50 to 70% of plant N, 60 to 80% of plant P, and 5 to 10% of plant K in sunflower achenes (Liet al.2009; Baiet al.2017; Duanet al.2018). Studies of other crops that harvest seeds/grains, such as rice (Wittet al.1999; Xuet al.2015) and wheat (Chuanet al.2013), also showed that 60 to 80% of plant N and P were in grains, while less than 20% of plant K was in grains, and more than 80% of plant K was in residues.Therefore, recycling of these crop residues would greatly influence soil K balance and K recommendation levels for the subsequent crops.

Fig. 2 Relationship between total N, P and K in above-ground dry matter (DM) and achene yield at harvest under balanced use of N, P, K in oil sunflower (A–C) and edible sunflower (D–F).

4.3. Differences between oil sunflower and edible sunflower

K application signiflcantly improved achene yield, straw yield, head diameter, 1 000-achene weight and kernel rate for both oil and edible sunflower (Table 2), but the average yield response and agronomic efflciency of K fertilizer was higher for oil sunflower than for edible sunflower (Fig. 1) at the similar average soil exchange K and K fertilizer rates in both experimental sites. This may be related to the differences of nutrient efflciency and nutrient requirement between oil sunflower and edible sunflower (Table 6). Oil sunflower had lower nutrient IE and higher nutrient requirement than edible sunflower, indicating that oil sunflower needed more nutrient supply than edible sunflower. Therefore, oil sunflower had a higher yield response to K fertilization than edible sunflower.

There were also some differences in achene quality response to K application. For fatty acid contents, K application increased the contents of oleic acid and linoleic acid in achenes of oil sunflower but did not signiflcantly affect these two fatty acid contents in achenes of edible sunflower compared to controls without K fertilization (Table 3). This inconsistent effect of K fertilization is probably due to the different sunflower varieties used or different soil and climatic conditions. Oil sunflower in Gansu has a longer growing period than edible sunflower in IMAR and the average temperature in the experimental sites is typically higher in Gansu than in IMAR. In addition, the tested soils in IMAR had higher saline and chloride contents than the tested soils in Gansu (Table 1) which may also affect contents of oil and fatty acids in achenes of sunflower (Chenet al.2007). A study in a high-salt soil showed that K fertilization increased oleic acid content and decreased linoleic acid content (Qianet al.2015), which was different from the results of our study.

K fertilization increased the crude protein contents of edible sunflower achenes but did not change the protein content of oil sunflower achenes (Table 3). Another study in edible sunflower indicated that K fertilization at 50 to 450 kg K2O ha–1increased crude protein content and had a signiflcant relationship with K rates (Maet al.2015), similar to the results of current study. However, some studies in oil sunflower showed that the crude protein content of sunflower achenes was also increased by K application (Huanget al.1985; Qianet al.2015) which is different from the results of our work.

5. Conclusion

Potassium application increased achene yield by an average of 406 kg ha–1for oil sunflower, and 294 kg ha–1for edible sunflower. K fertilization improved the quality of achenes by increasing contents of oleic acid and linoleic acid contents in oil sunflower, while increasing protein contents in achenes of edible sunflower. The agronomic efflciency of K fertilizer was 4.0 kg kg–1K2O for oil sunflower and 3.0 kg kg–1K2O for edible sunflower. The N, P and K in above-ground dry matter required for producing a ton of achenes was 45.5, 14.1 and 108.1 kg for oil sunflower, and 39.0, 8.0 and 71.7 kg for edible sunflower, respectively. The K harvest index was 0.05 kg kg–1in oil sunflower and 0.14 kg kg–1in edible sunflower.Therefore, oil sunflower had a higher yield response to K fertilization than edible sunflower because oil sunflower needs a higher nutrient supply than edible sunflower. The achene quality response to K fertilization was different between oil and edible sunflower. The lower K harvest index underscores the importance of residue recycling in soil K maintenance and fertilizer K recommendations.

Acknowledgements

The authors thank the flnancial support from the International Plant Nutrition Institute of USA (IPNI-2015-CHN-C14) and the earmarked fund for China Agriculture Research System(CARS-14).