HE Shu-qin ,MA Rui ,WANG Na-na ,WANG Shuang ,Ll Ting-xuan ,ZHENG Zi-cheng

1 College of Forestry,Sichuan Agricultural University,Chengdu 611130,P.R.China

2 College of Resources Science,Sichuan Agricultural University,Chengdu 611130,P.R.China

Abstract The loss of N in farmland is an important cause of agricultural non-point source pollution,which seriously impacts the aquatic environment. A two-year (2017-2018) experiment was conducted to investigate the characteristics of runoff and N losses under different tillage practices. Taking downslope ridge planting and cross ridge planting as the experimental treatments,the characteristics of surface runoff,interflow,and N losses in sloping farmlands with yellow soil were studied throughout the maize growth period. As the rainfall increased,the surface runoff and interflow also increased. The surface runoff and N losses in the surface runoff of downslope ridge planting were significantly higher than those of cross ridge planting. The interflow volumes and N losses in the 0-20 and 20-40 cm soil layers of the cross ridge planting were significantly higher than those of the downslope ridge planting. The total N (TN) losses from surface runoff accounted for 54.95-81.25% of the N losses from all pathways. Therefore,we inferred that surface runoff is the main pathway of N losses. Dissolved total N (DTN) was the main form of N loss under different tillage measures,as it accounted for 55.82-94.41% of the TN losses,and dissolved organic N accounted for 52.81-87.06% of the DTN losses. Thus,we inferred that dissolved N is the main form of N loss. Future research must focus on the prevention and control of the N losses during the maize seedling stage to reduce the environmental pollution caused by ammonium N through runoff.

Keywords:nitrogen losses,surface runoff,interflow,maize growth stages,tillage measures

1.lntroduction

With point source pollution under control,non-point source pollution has become the dominant cause of water eutrophication in recent years. Soil N losses are an important cause of non-point source pollution globally(Randall et al.2001;Shortle et al.2012). With increased fertilizer application,pollution of water from agricultural sources is also increasing (Qiao et al.2012;Povilaitis et al.2014). While the proper application of N fertilizer can increase crop yields (Wang G L et al.2014),its excessive application results in fertilizer wastage (Davis et al.2016)and causes nutrient losses and water eutrophication (Cai et al.2014). Reducing the N losses in agricultural soil is necessary to control water eutrophication and mitigate the hazards posed to human health (Ruiz Diaz et al.2010).

The severity of erosion in sloped farmlands appears to be affected by improper tillage practices,and results in the decline of soil fertility and deterioration of agroecological wellbeing,which hinder of agricultural economic development (Lal 1993;Pimentel et al.1995;Swanepoel et al.2015). When the soil erosion rate is up to 50-60%,the agricultural productivity of sloping farmland is reduced significantly (Li et al.2016),and factors such as rainfall and tillage have nonnegligible effects on soil erosion and N losses (Shipitalo et al.2013;Liu et al.2014,2015;Wang J et al.2014). Surface runoff and soil erosion were significantly decreased under conservation agriculture by reduced tillage and crop residue cover (Huang et al.2001;Araya et al.2011). For the use of sloping land,conservation tillage measures,such as monopoly measures and mulching measures,can effectively prevent soil erosion,and the reasonable use of different tillage measures can improve soil structure,improve soil fertility,etc.,and prevent soil erosion(Arnhold et al.2014;Wang et al.2015);meanwhile,irrational tillage measures can accelerate erosion,leading to a decline in soil quality (Mandal and Sharda 2013;Zhu et al.2015;Yang et al.2018). Compared with traditional downslope ridge planting,cross ridge planting reduced sediment volume by 46.9% and runoff volume by 52.9%,although the concentration of nitrogen losses increased (Yang et al.2019). In general,cross ridge planting can effectively control surface runoff and significantly control soil erosion,while soil nitrogen losses are higher under downslope ridge planting conditions,and the effects of different tillage measures on nitrogen losses control are different. Nitrate-nitrogen(NO3--N) and ammonium-nitrogen (NH4+-N) were reported to be essential nutrients for crop growth and were also the main forms of N losses (Wang et al.2010;Shan et al.2015;Hanly et al.2017). In the past,it was believed that the losses of soil N occurred mainly through surface runoff and nutrient enrichment in sediments,while interflow was neglected(Panuska et al.2008). Wang et al.(2012) demonstrated that the interflow may be the main mode of N losses,while some researchers concluded that the local environmental pollution caused by N losses from interflow was consistent with that from surface runoff and sediment production (Chen et al.2012;Gilley et al.2012;Xia et al.2014). Therefore,it is important to study the different pathways of N losses under different tillage measures.

Yellow soil is the second largest soil type of dry slope farmland in Sichuan,of which about 103.26 million hectares are arable land,mainly distributed in the eastern part of the Sichuan Basin and the surrounding low and middle mountainous areas. According to the Chinese Bulletin of Soil and Water Conservation in 2018 (MWR 2019),widespread soil erosion exists in this region,and the total area is about 38 606 km2,which represents 12.88% of the upper reaches of the Minjiang River. It has unique rainfall and abundant rainfall in summer. Severe soil erosion in the slope farmland of yellow soil is caused by rainfall.One of the main crops in the slope farmland of yellow soil is maize.The maize growth period coincides with the rainfall-driven erosion period in the study area. One study found that the surface runoff and corresponding N losses increased as the vegetation cover decreased (Wu L et al.2012). Similarly,there may be significant differences in the N losses during different growth stages of maize. Accordingly,we studied the N losses through different pathways with different tillage measures during the growth stages of maize in sloping farmland of yellow soil. The objectives of this study were:(1) to explore the characteristics of surface runoff,interflow and N losses,(2) to determine the main pathway(s) of N losses throughout the maize growth period under different tillage measures,(3) to quantify the pathways and loss ratios of total N (TN),dissolved total N (DTN),NO3--N,and NH4+-N,and (4) to investigate the impact of N losses on the environment in the study area.

2.Materials and methods

2.1.Study area

The study area is situated in Guankou Town,Dujiangyan City (103°36′52′′E,31°00′34′′N),Sichuan Province,which is located in the rainy area of West China. The annual average rainfall is 1 218.4 mm and the rainfall in summer is 670.1 mm,which accounts for 55% of the average annual rainfall. The rainfall in autumn is 304.6 mm,which accounts for 25% of the average annual rainfall. Rainfall levels in 2017 and 2018 were 966.81 and 1 245.93 mm,respectively.The perennial average temperature is 15.1°C,the average spring temperature is 13-21°C and the average summer temperature is 21-29°C. The average yearly sunshine hours are 1 016.9 h. The study area has mountainous topography,the soil type is Argosols according to the Chinese Soil Taxonomy (CRGCST 2001),and its texture is sandy loam. Prior to the experiment,the topsoil (0-20 cm)contained 36.27% sand,44.26% silt,and 19.47% clay. A total of 1 kg of topsoil contained 17.18 g of organic C,0.62 g of TN,52.14 mg of available nitrogen,5.09 mg of available phosphorus,and 50.64 mg of available potassium,and the soil pH was 5.95.

2.2.Experimental design

The experimental runoff plots were 3.864 m×2 m with a slope of 15°,and the actual slope length of the sloping farmland was 4 m. The tillage measures for the runoff plots were downslope ridge planting and cross ridge planting,and each tillage measure was repeated three times for a total of six experimental runoff plots. The ridges were 20 cm high,50 cm wide,and placed 90 cm apart (Fig.1). The experimental runoff plots were separated by concrete ridges to prevent runoff from flowing outward or affecting the adjacent experimental plots. A U-shaped collecting tank with a height of 10 cm was placed at the bottom of the plot slope,and it was connected to the runoff collection barrel using a polyvinyl chloride (PVC) pipe with a diameter of 3.2 cm. This collection barrel collected surface runoff and sediments. Impermeable layers were set at depths of 20 and 40 cm in the soil to collect the interflow,and the interflow of the 0-20 and 20-40 cm layers were collected using the PVC pipe. Four rain gauges were placed around the experimental plots,and the rainfall duration was measured by manual observation.

Fig.1 Design of the experimental plot. Downslope ridge planting (A),cross ridge planting (B) and plot section (C).

The maize used was Ziyu 22,and it was sown on April 19,2017,and April 18,2018. The summer maize was cultivated according to the local conditions using the method of burrowing and planting two plants in a row with a plant spacing of 45 cm and a row spacing of 90 cm. Before the maize was sown,270,90,and 75 kg ha-1of N,P,and K were applied to the soil,respectively,and the soil was top-dressed with 140 kg ha-1of N in the elongation stage of the maize.Each treatment was repeated three times.

2.3.Surface runoff and interflow collection and measurements

For each treatment,surface runoff,interflow,and sediment yield samples were collected in buckets during natural rainfall events. The surface runoff and interflow were measured using the volumetric method. After approximately 3 h of settling of the surface runoff and interflow samples,250 mL of the supernatant was placed into clean polyethylene bottles,to which 98% sulfuric acid (H2SO4) was added to reduce the pH to ≤2 in order to suppress microbial activity.The treated samples were stored at 4°C for subsequent chemical analysis.

All the samples were filtered with 0.45-μm filter paper before analysis. TN and DTN in the runoff were determined by potassium persulfate oxidation and a discontinuous chemical analyzer (CleverChem 380). Dissolved organic nitrogen (DON) and granular N values were obtained by a difference subtraction method. NO3--N and ammoniumnitrogen (NH4+-N) were also measured using a discontinuous chemical analyzer (Lu 2000).

2.4.Statistical analysis

The experimental data were statistically analyzed using DPS 11.0. Analysis of variance was used to determine the relationships between the different treatments. All the results were deemed statistically significant at P＜0.05.

3.Results

3.1.Surface runoff and interflow

This study found that the volumes of surface runoff and interflow increased as the rainfall increased,and the surface runoff varied under the different tillage measures (Fig.2).As the maize growth period advanced,the rainfall amount also increased. The maximum rainfall amount occurred in the maize maturity period,which was related to the climate of the study area. In general,the surface runoff of the downslope ridge planting was higher than that of the cross ridge planting during the maize season. The total surface runoff volume was the highest at the stage of maize maturity and the lowest at the stage of maize elongation.However,individual surface runoff at the stage of maize elongation was much higher than that of the maize seeding stage. The total surface runoff amounts of the maize growth stages were in the following order:maturity stage＞tasseling stage＞seedling stage＞elongation stage. There were also differences between the surface runoff and the interflow for the different tillage measures,however the interflow volumes showed the similar characteristics in the soil layers of 0-20 and 20-40 cm (Figs.3 and 4). The interflow volume of cross ridge planting was much higher than that of downslope ridge planting during the maize growth period. The single highest interflow volume was observed on July 31,2018,during the maize maturity stage;whereas,the highest surface runoff volume was observed on August 17,2017,during the maize maturity stage. The total interflow volumes were also in the following order:maturity stage＞tasseling stage＞seedling stage＞elongation stage.

Fig.2 The characteristics of surface runoff under different tillage measures and rainfall amounts during maize growth stages. The error bars refer to the standard error of the mean (n=3). Different lowercase letters indicate significant differences at the 0.05 level during a given date. ＊ indicates significant differences at the 0.05 level under different tillage measures.

3.2.N losses in surface runoff and interflow

The N losses in surface runoff varied significantly with different tillage measures (Fig.5). The N losses in surface runoff of the downslope ridge planting were higher than those of the cross ridge planting during the maize season.The total N losses in surface runoff were the highest at the stage of maize maturity and the lowest at the stage of maize elongation. However,individual N losses in surface runoff at the stage of maize elongation were much higher than those of the maize seeding stage. The maximum N losses in surface runoff occurred at tasseling maturity,which were influenced by the nitrogen concentration. The total N losses in surface runoff of the maize growth stages were in the following order:maturity stage＞tasseling stage＞seedling stage＞elongation stage.

Fig.5 The characteristics of total nitrogen losses in surface runoff under different tillage measures during maize growth stages.The error bars refer to the standard error of the mean (n=3). Different lowercase letters indicate significant differences at the 0.05 level during a given date. ＊ indicates significant differences at the 0.05 level under different tillage measures.

Fig.3 The characteristics of interflow (0-20 cm) under different tillage measures and rainfall amounts during maize growth stages.The error bars refer to the standard error of the mean (n=3). Different lowercase letters indicate significant differences at the 0.05 level during a given date. ＊ indicates significant differences at the 0.05 level under different tillage measures.

The N losses in the 0-20 cm interflow under different tillage measures were significantly different (Fig.6). The N losses in the 0-20 cm interflow under cross ridge planting were much higher those of downslope ridge planting during the maize season. The single highest N losses in the 0-20 cm interflow was observed on August 17,2017,during the maturity stage. The N losses in the 20-40 cm interflow exhibited some similar characteristics to the 0-20 cm interflow. The N losses in the 20-40 cm interflow under cross ridge planting were also much higher than those of downslope ridge planting during the maize season(Fig.7),and the total N losses in the 20-40 cm interflow of the maize growth stages were in the following order:maturity stage＞tasseling stage＞elongation stage＞seedling stage. The highest N losses in the 20-40 cm interflow was observed on July 31,2018,during the maturity stage.

Fig.6 The characteristics of total nitrogen losses in interflow (0-20 cm) under different tillage measures during maize growth stages. The error bars refer to the standard error of mean (n=3). Different lowercase letters indicates significant differences at 0.05 level during date. ＊ indicates significant differences at 0.05 level during different tillage measures.

Fig.7 The characteristics of total nitrogen losses in interflow (20-40 cm) under different tillage measures during maize growth stages. The error bars refer to the standard error of the mean (n=3). Different lowercase letters indicate significant differences at the 0.05 level during a given date. ＊ indicates significant differences at the 0.05 level under different tillage measures.

3.3.Forms of N losses in surface runoff and interflow,and the loss ratio

By comparing the N losses from surface runoff and interflow throughout the maize growth period,the main forms of N losses were determined (Table 1). The main forms of N losses in the surface runoff were DTN (55.82-92.86%) and same phenomena were observed in the 0-20 cm interflow(83.02-92.22%) and 20-40 cm interflow (65.96-94.41%).Among the different forms of DTN,the main forms were DON (56.93-87.06%),0-20 cm interflow (53.01-91.23%),and 20-40 cm interflow (52.81-86.80%). Under different tillage measures,the losses of NO3--N and NH4+-N were small. The losses of DTN,NO3--N,and NH4+-N of the surface runoff and interflow during different stages were in the following order:maturity stage＞tasseling stage＞seedling stage＞elongation stage. Simultaneously,we found that the NO3--N concentration in the surface runoff during the maize seedling stage was much higher than the prescribed NO3-concentration (10 mg L-1),and the NH4+-N concentration was higher than that prescribed by standard III (1 mg L-1) of the environmental quality standard for surface water (GB3833-2002). The concentrations of NO3--N in the 0-20 cm interflow and the 20-40 cm interflow conformed to standard III (20.0 mg L-1),while the concentration of NH4+-N exceeded standard III (0.5 mg L-1) of the groundwater standard (GB/T 14848-2017). This high NH4+-N concentration posed a risk to human health. The losses of DTN,NO3--N,and NH4+-N in surface runoff throughout the maize growth period were greater under downslope ridge planting than under cross ridge planting,while the interflow exhibited the opposite trend. DTN,NO3--N,and NH4+-N suffered the greatest losses in both the surface runoff and interflow during the maize maturity stage.

Fig.4 The characteristics of interflow (20-40 cm) under different tillage measures and rainfall amounts during maize growth stages.The error bars refer to the standard error of the mean (n=3). Different lowercase letters indicate significant differences at the 0.05 level during a given date. ＊ indicates significant differences at the 0.05 level under different tillage measures.

Surface runoff was the main pathway of the N losses,accounting for 54.95-81.25% of TN losses under different tillage measures (Table 2). The surface runoff N loss ratio reached the maximum during the maize seedling stage.The N loss ratio in the 0-20 cm interflow was higher than that in the 20-40 cm interflow under the downslope ridge planting,except during the maize elongation stage. Under cross ridge planting,the N loss ratio in the 0-20 cm interflow was higher than that in the 20-40 cm interflow,except during the maize tasseling stage.

Table 2 Proportions of nitrogen in different loss pathways within total nitrogen (%)

4.Discussion

4.1.Surface runoff and interflow

The tillage measure is an important factor affecting surface runoff and interflow (Wang J et al.2014;García-Díaz et al.2017). Sloping farmland is the main cause of soil and water losses. Soil and water losses often lead to significant soil nutrient losses,which further reduce soil fertility (Yakutina et al.2015;Zhang et al.2015;Luo et al.2017). Our study area was located in the rainy area of West China,which is a unique climatic and geographical unit in the western margin of Sichuan Basin that has higher rainfall than the surrounding areas (Liu et al.2018). This rainfall causes significant soil and water losses in the study area. Both surface runoff and interflow volume reached their maximum values at maize maturity.The runoff volume was mainly affected by the rainfall,and the runoff was at its peak when the maximum rainfall occurred. The surface runoff volume under the downslope ridge planting was higher than that under the cross ridge planting throughout the maize growth period. The cross ridge planting could have retained the surface runoff and reduced both the erosion of soil on the slope and the collection of runoff,thereby minimizing the yield and discharge of cross ridge planting. However,the furrow direction was the same as the flow direction of the surface runoff,thereby accelerating the flow speed,increasing the surface runoff,and causing serious soil and water losses (Yan et al.2014;Ford et al.2015). Cross ridge planting can significantly increase the 0-20 and 20-40 cm interflow volumes. Thus,cross ridge planting can decrease the erosion rate of runoff and increase the surface infiltration to increase the flow rate in soil (Morillas and Gallardo 2015;Morbidelli et al.2018). We found that the 0-20 cm interflow volume was larger than the 20-40 cm interflow volume,and this may be attributed to the decreased interflow production in each layer of soil. During the interflow of red soil in sloping farmland,the interflow volume at 30 cm was greater than that at 60 cm,and the deeper the soil layer,the longer the lag and tailing times of the interflow (Xie et al.2015;Mo et al.2016). Our previous findings showed that the surface runoff and the interflow (at 0-20 cm and 20-40 cm) were affected not only by rainfall intensity,but also by the plant height of maize (Yang et al.2016),and similar results were reported by Liu Z D et al.(2012) and Yan et al.(2015). One reason for this could be that the surface runoff and interflow were affected by the randomness,complexity,and diversity of rainfall patterns and the redistribution effect of the maize canopy on rainfall (Martello et al.2015).

4.2.N losses in surface runoff and interflow

The N dissolved in surface runoff or interflow migrates from the soil to water,where it is adsorbed onto the surface of soil particles (Panuska et al.2008;Wu X Y et al.2012;Yang et al.2014;Cao et al.2015). Many previous studies have shown that both surface runoff and interflow are vital pathways for runoff N transport (Silva et al.2005;Zheng et al.2017).Our results showed that the N losses in surface runoff of the downslope ridge planting were higher than those of the cross ridge planting during the maize growth period,which can be the result of surface runoff and rainfall. The single highest N loss value in surface runoff was observed on July 16,2017,and was mainly affected by the rainfall intensity,which was 36.24 mm h-1. The single highest N losses in the 0-20 cm interflow and 20-40 cm interflow were observed on August 17,2017,and July 31,2018,respectively,and were mainly affected by the rainfall amounts,which were 94.93 and 105.10 mm,respectively. Previous studies have found that there is a linear correlation between the runoff and N losses (Kothyari et al.2004;Jia et al.2007),and Liu Y et al.(2012) and Wang G Q et al.(2014) showed that there is a significant positive correlation between the runoff and N loss. Our results showed that the total N losses in surface runoff and 0-20 cm interflow were first decreasing and then increasing during the maize season,and the total N losses were the minimum at elongation stage of maize.However,the total N losses in 20-40 cm interflow showed a trend of gradual increase with the advancement of the maize growth period. Both surface runoff and interflow showed that the N losses at the maize maturity stage were significantly higher than those at other maize growth stages. This is primarily due to the differences between rainfall and maize growth period. With the development of the maize growth period,the rainfall intensity,rainfall time,raindrop kinetic energy,and surface runoff gradually increased in our study areas. From our measurements,the leaf area index and plant height of maize displayed rising-and-falling trends,and higher values were found at the tasseling stage of maize for the cross ridge planting and downslope ridge planting.The effects of the maize canopy on rainfall interception and stem confluence gradually increased. In addition,the enhancement of the soil consolidation ability of the maize plant roots leads to increased soil stability and erosion resistance,reduced infiltration and increased surface runoff(Garcia-Estringana et al.2010;Wu et al.2018). We found that the N losses in the 0-20 cm interflow were higher than in the 20-40 cm interflow. This could be attributed to the influence of the interflow.

4.3.Forms of N losses in surface runoff and interflow,and the loss ratio

Studies have shown that surface runoff is the main pathway through which N is lost,and it is mainly lost in the form of NO3--N that falls under DTN (Xing et al.2016;Wang et al.2017). Another study also found DTN in the runoff N losses for the losses of the main forms,but the small proportion of DTN in the dissolved N,as NO3--N,was the main component of inorganic N (Jiao et al.2010). The results of our study are consistent with these reports,which may be caused by the application of organic fertilizer before planting the maize.We also found that the concentration of NO3--N in the surface runoff during the maize seedling stage was substantially higher than that prescribed (10 mg L-1) in the environmental quality standard for surface water (GB3833-2002). This high concentration is the result of the application of basal fertilizer before planting. Although the soil contained a high NO3--N content,nitrification by bacteria and oxidation changed the soil characteristics,thereby resulting in the soil NO3--N being easily leached and not easily adsorbed by the soil. The concentration of NO3--N conformed to standard III (20.0 mg L-1) of the groundwater standard (GB/T 14848-2017) in interflows of the 0-20 and 20-40 cm layers,while the NH4+-N concentration exceeded the prescribed limit set by standard III (0.5 mg L-1),thereby posing a risk to human health. This may be the result of the adsorption of NH4+ions by negatively charged soil particles as they moved with the runoff. During rainfall,splashing raindrops and runoff continue to destroy the large grain aggregates in the soil.This increases the infiltration of runoff and disintegrates the large grain aggregates into small grain aggregates with a large specific surface area and strong adsorption capacity,thereby causing the losses of NH4+ions.

N losses via surface runoff were significantly higher than those of the interflows in the 0-20 and 20-40 cm layers. The N loss ratio of surface runoff reached the maximum value at the stage of maize seedling. This could be related to the properties of the yellow soil,which had a sticky texture (44.26% silt,and 19.47% clay) and the leakage rate was slow;thus,the proportion of total soil nutrient losses through the interflow was very small in our study. Another reason was the lesser vegetation cover during the maize seedling stage that reduced the interception effect of maize plants on runoff,causing more surface runoff. The N loss ratio in the 0-20 cm interflow was generally higher than in the 20-40 cm interflow;however,during the maize elongation stage of the downslope ridge planting and the maize tasseling stage of the cross ridge planting,the N loss ratios in the 20-40 cm interflow were greater than in the 0-20 cm interflow.This could be the result of the absorption of a large amount of N by the maize plant roots during the elongation and tasseling stages,which affected the N losses in the interflow.

Our results showed the maize yields were 9 338.21 and 8 316.34 kg ha-1of cross ridge planting and downslope ridge planting,respectively,in 2017,and the maize yields were 9 841.67 and 8 885.42 kg ha-1of cross ridge planting and downslope ridge planting,respectively,in 2018. In general,N losses in 2017 were higher than in 2018. Therefore,N losses have a certain effect on maize yield,but further studies are needed. Our study can provide a theoretical basis for soil erosion control and nutrient management in the study area,in order to maintain,or even increase,crop yields while minimizing environmental costs in Chinese agricultural production systems (Zhu and Chen 2002).Compared with the downslope ridge planting,cross ridge planting enhanced soil stability while reducing soil erosion,but the interflow pollution risk of the cross ridge planting should also be concerned at the stage of maize maturity.

5.Conclusion

We found that both surface runoff and interflow reached their maximum values at maize maturity. Surface runoff of the downslope ridge planting was higher than that of the cross ridge planting during the maize growing season,while the opposite was true for interflow. The main pathway of N losses was surface runoff,which accounted for 54.95-81.25% of the TN losses under different tillage measures. The main form of N losses through surface runoff and interflow was DON. The focus must be placed on the losses of NO3--N in the surface runoff during the maize seedling stage. The losses of ammonia-nitrogen (NH4+-N) in surface runoff and interflow and its environmental impact should not be underestimated. Runoff is an important source for the replenishment of water for rivers and lakes,so it plays an important role in the hydrological cycle of an entire basin. Controlling the losses of N from the source in sloping farmlands would not only improve the nutrient utilization rate but also effectively reduce non-point source pollution. This could provide a new method for the sustainable utilization of sloping farmlands in the mountainous yellow soil areas in southeastern China.

Acknowledgements

The authors sincerely acknowledge the financial support provided by the National Natural Science Foundation of China (41271307) and the Key R&D Project of Sichuan Province,China (2019YFS0463).

Declaration of competing interest

The authors declare that they have no conflict of interest.