Effects of tillage, mulching and N management on yield, water productivity, N uptake and residual soil nitrate in a long-term wheat-summer maize cropping system

Field Crops Research 213 (2017) 154–164

Contents lists available at ScienceDirect

Field Crops Research journal homepage: www.elsevier.com/locate/fcr

Effects of tillage, mulching and N management on yield, water productivity, N uptake and residual soil nitrate in a long-term wheat-summer maize cropping system ⁎

Zhanjun Liua,b, , Zhujun Chena,b, Pengyi Maa, Yan Menga, Jianbin Zhoua,b, a b

MARK

⁎

College of Natural Resources and Environment, Northwest A & F University, Yangling, 712100, China Key Laboratory of Plant Nutrition and the Agri-Environment in Northwest China, Ministry of Agriculture, Northwest A & F University, Yangling, 712100, China

A R T I C L E I N F O

A B S T R A C T

Keywords: Water-saving Nitrogen fertilization Grain yield Water productivity Soil fertility

A thorough understanding of coupled effects of soil management (tillage), mulch and N rate on the wheat-maize system is crucial for achieving sustainable agriculture in the southern Loess Plateau of China. This study was based on a 12-year (2003–2015) field experiment and aimed to evaluate the impact of three wheat-maize systems (S) which varied in terms of tillage, mulch, wheat row spacing and irrigation management (CT, conventional tillage with no mulch; RFM, ridge-furrow with plastic film-mulched ridges and straw-mulched furrows; CTM, conventional tillage with straw mulch) combined with N fertilizer rates (0, 120 and 240 kg N ha−1) on crop yield, water productivity (WP, kg grain per kg of water input), N uptake, residual soil nitrate (RSN) and soil physicochemical properties. Results demonstrated that RFM significantly increased maize yield in comparison with CT in all 12 years, while CTM increased yield in comparison with CT from year 3 onwards. By contrast, wheat yield was not strongly influenced by RFM and CTM from 2004 to 2012 (except for 2008). Maize yields of RFM were significantly higher than those of CTM from the third year onwards. Compared with CT, the other two practices, and more so RFM, showed beneficial effects on crop yield, the amount of stored water, WP, N uptake and RSN. N fertilization significantly increased crop yield, WP and N uptake, while no significant difference was observed between the N120 and N240 treatments. Notably, considerable buildup of RSN to ∼ 490 kg N ha−1 at maize harvest and ∼340 kg N ha−1 at wheat harvest were observed in 0–200 cm soil depth when 240 kg N ha−1 was applied. These results suggest that the conventional N rate of 240 kg N ha−1 is excessive, and risks serious contamination of the groundwater as a result of NO3−-N leaching. The N120 treatment was characterized with considerably lower RSN accumulation after harvest, while maintaining crop yield. Thus, we concluded that the RFM practice with 120 kg N ha−1 application could reduce irrigation water and fertilizer inputs and increase crop land and water productivity, and is a promising strategy for developing sustainable agriculture in the southern Loess Plateau and other areas with similar climate and cropping systems.

1. Introduction

predictions suggest that the Loess Plateau will become drier and warmer in the future (Sun and Ma, 2015). Agricultural irrigation water resource depletion is becoming increasingly severe with the growing population and global climate change (Elliott et al., 2014). Thus, conserving and efficiently utilizing the limited precipitation available is crucial for cropping in the Loess Plateau and other water-limited areas in the world. Numerous studies have demonstrated that soil surface mulch with crop straw or plastic sheets is an effective means of increasing maize or/ and wheat yield and water productivity (WP, kg grain per kg of water input) in the Loess Plateau (Liu et al., 2014; Li et al., 2015; Zhang et al., 2015). Similar beneficial results were also observed in Mexico

The winter wheat-summer maize rotation system has been widely adopted in the southern Loess Plateau of northwest China (Deng et al., 2006). Crop production in this region is primarily constrained by water shortage and nutrient deficiency (Li et al., 2015). Uneven rainfall and high evaporation always result in an imbalance between crop water requirements and water supply thus leading to low grain yields (Gan et al., 2013). The limited annual precipitation is insufficient to support two crops and therefore supplemental irrigation is usually applied to wheat during spring and occasionally to maize if rainfall is scarce during early growth stage (Deng et al., 2006). Climate change

⁎

Corresponding authors at: College of Natural Resources and Environment, Northwest A & F University, Yangling, 712100, China E-mail addresses: [email protected] (Z. Liu), [email protected] (J. Zhou).

http://dx.doi.org/10.1016/j.fcr.2017.08.006 Received 3 January 2017; Received in revised form 8 August 2017; Accepted 8 August 2017 0378-4290/ © 2017 Elsevier B.V. All rights reserved.

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

experiment initiated in 2003 and aimed to (i) examine the combined effects of tillage-mulch-irrigation management and N rate on crop yield, WP, N uptake, and residual soil nitrate; and (ii) investigate the changes in soil physicochemical properties under the systems involving different tillage-mulch-irrigation management and fertilizer N application. The information obtained can be used to optimize N management and enhance crop production for the winter wheat-summer maize rotation system in dry sub-humid areas of northwestern China and other areas with similar climates and cropping systems.

(Govaerts et al., 2006; Verhulst et al., 2011), India (Sharma et al., 2011; Lenka et al., 2013; Balwinder-Singh et al., 2016), and other regions of China such as North China Plain (Huang et al., 2015). Moreover, the findings of studies on crop yields in response to straw mulch are still contradictory, including positive (Sharma et al., 2011; Liu et al., 2014; Zhang et al., 2015), or no obvious (Govaerts et al., 2006; Huang et al., 2015) or negative effects (Fabrizzi et al., 2005; Gao et al., 2009; Naveen-Gupta et al., 2016), even though these experiments were treated with a similar amount of straw (∼ 4500 kg ha−1). Most of those studies were short-term (1- or 2-year) studies, whereas long-term studies are needed to understand the effects of climate variability and changes in soil properties over time as a result of the different treatments. Therefore, long-term field experiments are needed to clarify the response of grain yield to straw mulch and soil management under a wheat-maize cropping system. Ridge-furrow with plastic film or straw mulching system is widely applied in northwest China (Li et al., 2015), Mexico (Verhulst et al., 2011) and India (Sharma et al., 2011). However, the long-term effects of this management practice are still poorly understood (Li et al., 2017; Zhang et al., 2017), and little information is available regarding a comparison of straw mulch and ridge-furrow with plastic film or straw mulching system influencing the wheat-maize rotation system in water-limited areas. Nitrogen (N) deficiency is widespread in soils of the Loess Plateau (Guo et al., 2012), and the crop yield gap in this region is greater than 50% due to the limited supply of water and nitrogen (Mueller et al., 2012). During the past two decades, N fertilizer application has increased drastically in pursuit of higher grain yields, and the excessive N fertilization problem has become serious in the wheat-maize rotation system in northwestern China (Zhang et al., 2015). Previous studies documented that the high N application beyond the needs of maizewheat cropping system would reduce N use efficiency and enhance soil nitrate buildup, and agricultural N management remains a key environmental challenge (Lenka et al., 2013; Yang et al., 2017). Nitrate-N leakage generally occurred when substantial rain or irrigation occur shortly after N application (Jia et al., 2014). In view of the scarcity of irrigation and precipitation, nitrate leaching in cultivated lands of northwestern China has been underestimated or even ignored for many years (Dai et al., 2016). A survey conducted in the Shaanxi province showed that up to 25% of 167 groundwater samples exceeded the WHO drinking water standard (NO3−-N < 10 mg kg−1) in 1998, and up to 54% of 225 groundwater samples exceeded the critical value of 10 mg kg−1 in 2008 (Deng et al., 2008), indicating an increasing risk on the environment. Undoubtedly, these enhanced concentrations of nitrate in the groundwater were associated with the high application of N fertilizer in cultivated lands (Yang et al., 2017). Huang et al. (2015) found that straw mulch could decrease NO3−-N leaching losses while sustaining maize-wheat grain yields in North China Plain. Murphy et al. (2016) also reported positive effect in southern Mexico. Similarly, several studies have been conducted in the Loess Plateau for optimizing N management, however they were either treated without mulching practice (Zhang et al., 2015; Dai et al., 2016) or performed on monoculture cropping systems (Liu et al., 2014; Li et al., 2015). Up to now, an appropriate N management under the systems involving different combinations of tillage, mulch and irrigation management in order to increase crop yield and decrease the yield gaps remains poorly investigated in wheat-maize cropping systems in the Loess Plateau and other water-limited areas. In addition, previous related studies mainly focused on crop yield and WP in response to either mulching or N rate, while little attention has been paid on evaluating soil fertility changes such as soil physicochemical properties (Gao et al., 2009; Verhulst et al., 2011; Sharma et al., 2011; Lenka et al., 2013; Bu et al., 2013; Li et al., 2015; Dai et al., 2016; Yang et al., 2017). A comprehensive evaluation of long-term coupled effects of tillage-mulch-irrigation management and N fertilization on wheat-maize cropping systems in the Loess Plateau remains scarce. Therefore, the present study utilized a long-term field

2. Methods and materials 2.1. Site description The field experiment was conducted at the Agricultural Research Station of Northwest Agriculture & Forestry University (34°18′ N, 108°04′ E, and 520 m above the sea level) in Yangling, Shaanxi, China. The site is located in the southern Loess Plateau. A temperate and dry sub-humid climate prevails in this region and the mean annual temperature and precipitation are 12.9 °C and 583, respectively (1957–2015). About 65% of the precipitation occurs between June and September, and the potential evaporation is 1400 mm. The precipitation during the experimental years (2003–2015) is shown in Fig. 1. The winter wheat and summer maize rotation system is the major local cropping sequence, with rotary tillage for wheat and no-till for maize. Summer maize is usually sown in early June and harvested in early October, and winter wheat is sown in early or mid October and harvested in early June of the following year. The soil at the experimental site is classified as a calcareous Eum-Orthic Anthrosol (Udic Haplustalf in the USDA system), which is the typical soil in the area. Selected chemical properties of the 0–20 cm soil layer at the experimental site in 2003 were as follows: pH 8.25, 15.2 g kg−1 organic matter, 0.91 g kg−1 total N, and available N, P and K concentrations were 32.3 mg kg−1, 17.2 mg kg−1 and 169 mg kg−1, respectively. 2.2. Experimental design The long-term field experiment was established in 2003 and is still on-going with a maize-wheat rotation. Three tillage-mulch systems have been widely adopted in the study area, and these three systems were compared in main plots in this experiment: CT (conventional tillage with no mulch), RFM (ridge-furrow with plastic film-mulched ridges and straw-mulched furrows), and CTM (conventional tillage with straw mulch). Three fertilizer N rates were compared in sub-plots: 0 (control), 120 (moderate), and 240 kg N ha−1 (high; conventional N

Fig. 1. Precipitation during the 12 consecutive maize-wheat rotation cycles (2003–2015). “2003–2004” represents summer maize in 2003 (from June to September) and winter wheat in 2004 (from October in 2003 to May in 2004), and so on.

155

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

2.3. Sampling and measurements

rate used by farmers). The experimental design was a split plot with four replicates. Each sub-plot was 4 m × 4.5 m, with 1.5 m buffers between the treatments to prevent nutrient movement from one plot to another. Details of the 3 system treatments are as follows:

To prevent the edge effect, sampling was performed in the central area (3.5 m × 4.0 m) of each subplot. Soil samples were collected from each subplot in 20 cm increments to a depth of 200 cm using a soil auger, and plant samples were also collected from each subplot at maturity. Considering the disturbance of the plots and analytical costs, we did not collect soil and plant samples every year. In addition, in each plot of CT and CTM, soil scores were sampled randomly in the middle of two planting rows, while in RFM plots, soil scores were taken randomly from both ridges and alternating furrows. The present study mainly summarized the data obtained previously, although soil and plants were not sampled simultaneously. Soil sampling was conducted in 5 wheat-maize rotation cycles (2003–2006, 2008–2009 and 2012–2013) and plant sampling was performed in 3 consecutive wheat-maize rotation cycles (2008–2011). Five maize plants were randomly selected at maize maturity and cut at the stem base, separated into fractions (including leaves, stems, tassels, husks, cobs and kernels), chopped, and dried to a constant weight at 70 °C. At wheat maturity, 1 m2 (1 m × 1 m) area of wheat plants were collected and separated into fractions (leaves, stems, sheath, glume and spike-stalk), chopped, and dried at 70 °C until constant weight. Each plant fraction was ground separately for N analysis, which was determined by the Kjeldahl digestion method as described in Lu (2000). The N accumulation (kg ha−1) in plant fractions was calculated by multiplying fraction N concentration (%) and fraction dry matter (t ha−1). To investigate soil fertility changes after long-term implementation of the cropping system treatments and N rates, 5 cores (5.0 cm diameter) were taken randomly from each subplot and well-mixed into a composite sample following the maize harvest in 2015. The samples were immediately transported to the laboratory, where one sub-sample was used for measuring soil water content (SWC), and another subsample was air-dried at room temperature for chemical analysis. The selected properties including SWC, pH (soil:water, 1:2.5), soil organic matter (SOM; K2Cr2O7-external heating method), total N (microKjeldahl method), available N (2 M KCl extraction), available P (0.5 M NaHCO3 extraction) and available K (1.0 M CH3COONH4 extraction) were determined following the methods described by Lu (2000). The middle 4 rows (4.0 m in length) of summer maize and 8 rows of winter wheat per plot were harvested to determine maize and wheat grain yields. Wheat yield (adjusted to a moisture content of 12.5%) and maize yield (adjusted to 14% moisture content), which were recorded on plot basis, were converted to kg ha−1 for statistical analysis.

(1) CT: Conventional tillage: prior to planting each wheat crop, the soil was tilled to 20 cm using a disk harrow. All crop residues were removed at the time of harvest. (2) RFM: Conventional tillage combined with ridge-furrows system (tilled for planting wheat after maize harvest with a disk harrow to 20 cm after which new ridges were formed; the plastic film mulched on the ridges applied previously was removed totally, and the remaining non-degraded wheat residue on the furrows applied during maize growing season was incorporated by the tillage operation); the ridges (30 cm wide × 15 cm high) rebuilt before planting wheat were mulched with plastic film and the furrows (30 cm wide) were mulched with maize residue. The wheat was sown in the furrows. After wheat harvest, the plastic mulch on the ridges was replaced and maize was sown in the furrows and mulched with wheat residues. 2250 kg ha−1 of sun-dried crop residue was applied to each crop, and the maize (all except the cobs) or wheat residues were chopped into ∼5 cm lengths and applied 30 days after seedling emergence. (3) CTM: Conventional tillage as for CT; maize/wheat residues were applied across the soil surface 30 days after seedling emergence as for RFM, but the amount applied was higher (4500 kg ha−1). Maize production is always constrained by water limitation during the early growing stage, although most precipitation (∼ 65%) occurred during maize season (Bu et al., 2013). Therefore, in the CT treatment, flood irrigation of 40 mm was employed during the winter wheat (Triticum aestivum L.) season, and the irrigation applied during the maize (Zea mays L.) season varied from 0 to 60 mm yr−1, depending on the amount of rainfall occurring during the early growing season. Compared with no mulch, reduced irrigation was applied under RFM and CTM treatments which were considered as water-saving management practices, and thus wheat was not irrigated, while maize received one-half the rate of the CT treatment. Wheat (Xiaoyan-22) was sown in early October and harvested in early June of the following year. The seeding rate in all treatments was 225 kg ha−1. The row spacing in the CT and CTM treatments was 20 cm and each plot contained 21 rows. In the RFM treatment, 2 rows of wheat were sown in each furrow and no wheat was planted on the ridges, and resulting in an average row spacing of 30 cm. Fertilizer types were urea and superphosphate for N and phosphorus (P), respectively, and the rate of P fertilizer was 80 kg P2O5 ha−1. No K fertilizer was applied due to the high intrinsic K supply (> 150 mg kg−1) in the soil of the study site. Fertilization is performed in each subplot individually. Basal fertilizers including all N and P were evenly broadcast onto the soil surface and immediately incorporated into the plowed soil (0–20 cm depth) manually using a rake. Summer maize (Shaandan-902 until 2007; then Zhengdan-958) was used as a test crop and was planted by hand immediately after the wheat harvest without tilling the soil. Summer maize is usually sown in early June and harvested in early October. The row spacing was 60 cm in all treatments and the plant density was 57,780 plants ha−1, corresponding to local common plant density in the region. One maize row was planted in each furrow of the RFM treatment. No N fertilizer was applied at planting, and N fertilizer was split-applied, using 1/3 at the seedling (three leaves) stage and 2/3 at the booting stage. Topdressing of N was performed with a hoe to incorporate N fertilizer into the soil near the seedlings. In addition, in each plot of RFM and CTM, straw residue on the top of fertilization point was removed and re-mulched using a hoe during fertilizer N application.

2.4. Calculation methods 2.4.1. Residual soil nitrate According to Dai et al. (2016), the amount of residual soil nitrate (RSN, kg N ha−1) in each soil layer can be calculated as: RSN = Ti × Di × Ci/10

(1)

Where Ti is the soil layer thickness (20 cm), Di is the soil bulk density (g cm−3), Ci is the soil nitrate-N concentration (mg N kg−1) of the corresponding layer, and 10 is the conversion coefficient. 2.4.2. Water productivity Water productivity (WP) calculated as the ratio of grain yield (kg ha−1) to water input (rain plus irrigation, mm) over the study period. In this study, WP was estimated by the following equation: WP = Grain yield/(irrigation + rain) (kg mm−1 ha−1)

(2)

Grain yield is the sum of maize yield and wheat yield over the studied 10 consecutive maize-wheat cropping system (2003–2013). The amount of irrigation plus rain over the study period (from maize season in 2003 to wheat season in 2013) is calculated as water input. 156

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

2.4.3. Fertilizer N use efficiency Fertilizer N use efficiency included the N uptake efficiency (NupE), N fertilizer productivity (NfP) and N use efficiency (NUE), which were calculated as described by Guo et al. (2014). The formulas were as follows: NupE = Total N uptake/N application rate (kg kg

−1

)

Table 2 Wheat grain yield (kg ha−1) as influenced by system and N rate over twelve years. Year

CT

(3)

NfP = Grain yield/N application rate (kg kg−1)

(4)

NUE = Grain yield/Total N uptake (kg kg−1)

(5)

2004 4843a 2005 4952a 2006 4278a 2007 4437a 2008 3263c 2009 3130a 2010 3017a 2011 4034a 2012 2906a 2013 1120b 2014 1984b 2015 3745b Mean 3463b CV(%) 36.9a Multi-way ANOVA analysis System (S) F = 8.13 N rate (N) F = 870 Year (Y) F = 146 S×N F = 1.82 S×Y F = 3.27 N×Y F = 7.64 S×N×Y F = 1.43

2.5. Statistical analysis SAS for Windows 8.1 was used to perform analysis of variance (ANOVA). Comparisons of treatment means were carried out using a Duncan’s multiple range test at P < 0.05. Sigma Plot 10.0 (Aspire Software International, Ashburn, VA) was used for the illustrations. 3. Results 3.1. Crop production 3.1.1. Maize yield Maize yield showed a high interannual variability, and both system (S) and N rate (N) significantly affected maize yield across the 12-year study (Table 1). Significant year × S and year × N interactions were found while no effect of S × N and year × S × N interactions was observed (Table 1). Compared with CT, the RFM treatment significantly increased maize yield over the entire period. By contrast, the CTM treatment did not substantially increase maize yield until the third experimental year. Maize yields under RFM were considerably higher than that of CTM in most cases. Averaged across years, maize yield was highest for RFM followed by CTM and CT. RFM and CTM showed less variation in maize yield across time when compared with CT, whilst maize yield under RFM varied less than CTM (Table 1).

Systems CT

2003 5733b 2004 8161b 2005 4572c 2006 2918c 2007 3960c 2008 5145c 2009 4099b 2010 5720c 2011 5104c 2012 6403c 2013 4690c 2014 4090c Mean 5188c CV(%) 31.8a Multi-way ANOVA analysis System (S) F = 165 N rate (N) F = 300 Year (Y) F = 165 S×N F = 1.73 S×Y F = 5.06 N×Y F = 9.22 S×N×Y F = 1.29

CTM

N0

N120

N1240

6548a 8995a 5507a 4514a 5389a 6515a 5572a 7742a 6698a 7629a 7840a 5827a 6545a 22.4c

6105ab 8648ab 5216b 3922b 4170b 5830b 5048a 7304b 6063b 7015b 6817b 5294b 6030b 26.5b

5470b 7770b 4823b 2642b 2833b 4160b 4270b 5177b 4335b 4905b 4812b 4620b 4800b 32.1a

6317a 9062a 5248a 4255a 5443a 6890a 5186a 7640a 6871a 8088a 7109a 5425a 6468a 23.8b

6599a 9045a 5224a 4457a 5242a 6730a 5264a 7949a 6659a 8054a 7174a 5166a 6494a 24.8b

RFM

CTM

N0

N120

N1240

4780a 4726a 4032a 3967a 4508a 3315a 3248a 4317a 2670a 1713a 2318a 4292a 3657a 32.2b

5005a 4926a 3860a 3917a 3719b 2985a 2979a 3764a 2835a 1358b 1978b 3955ab 3425b 37.2a

3361b 3041b 2373b 2415b 2113b 1276c 1848b 1807b 1216b 723b 1535b 2382b 2027b 41.6a

5580a 5964a 4928a 5156a 4764a 3834b 3752a 5054a 3635a 1761a 2447a 4802a 4307a 32.2b

5586a 5600a 4867a 4786a 4612a 4321a 3644a 4906a 3560a 1703a 2328a 4565a 4212a 32.5b

P < 0.001 P < 0.001 P < 0.001 P = 0.125 P < 0.001 P < 0.001 P = 0.054

Applying N fertilizer significantly increased maize yield compared to the unfertilized treatment, while no significant difference was found for maize yield between N120 and N240 from 2003 to 2014 (Table 1). A similar trend was recorded for the average maize yields across the 12year period. Additionally, the N120 and N240 treatments showed equivalent variation in maize yield, which varied less across time when compared with N0. 3.1.2. Wheat yield As shown in Table 2, wheat grain yield also showed a high interannual variability and was significantly influenced by tillage-mulch system and N rate over the 12-year period. N fertilization showed a significant positive effect on wheat yield from 2004 to 2015, but the yield differences in winter wheat between tillage-mulch systems were only statistically significant in four out of twelve years. Significant year × S and year × N interactions were observed, while no significant S × N and year × S × N interactions was observed. With the exception of 2008, the RFM treatment significantly increased wheat yield from 2013 to 2015, while the CTM treatment showed a non-significant influence on wheat yield across the study period when compared with CT. Averaged across years, wheat yields under different tillage-mulch systems were ranked as RFM > CTM ≈ CT, and their variation followed the order CT ≈ CTM > RFM. N fertilization strongly increased wheat yield compared to the unfertilized treatment, while no significant difference was observed for wheat yield between N120 and N240 except for one year out of twelve (2009). Similar trends were recorded for their averaged wheat yields across the study period. In addition, wheat yields under N120 and N240 varied less across time when compared with N0, although the variation in wheat yield under N120 was equal to that of N240.

N rates RFM

N rates

Means within a row for mulch and N fertilization factors separately that are followed by the same letter are not significantly different at P ≤ 0.05. N0, N120, N240 = 0, 120, 240 kg N ha−1; CT, conventional tillage with no mulch; RFM, ridge-furrow with plastic film-mulched ridges and straw-mulched furrows; CTM, conventional tillage with straw mulch.

Table 1 Maize grain yield (kg ha−1) as influenced by system and N rate over twelve years. Year

Systems

P < 0.001 P < 0.001 P < 0.001 P = 0.143 P < 0.001 P < 0.001 P = 0.126

Means within a row for mulch and N fertilization factors separately that are followed by the same letter are not significantly different at P ≤ 0.05. N0, N120, N240 = 0, 120, 240 kg N ha−1; CT, conventional tillage with no mulch; RFM, ridge-furrow with plastic film-mulched ridges and straw-mulched furrows; CTM, conventional tillage with straw mulch.

3.1.3. Relative yield There were cumulative benefits of maize relative yields from the straw retention at 2.25 t ha−1 (RFM) and 4.5 t ha−1 (CTM) over a 12year period, regardless of N rate. Notably, maize relative yields of RFM/ 157

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

Fig. 2. Relative yield of the treatments (RFM/CT and CTM/CT) for the 0, 120 and 240 kg N ha−1 during 12 consecutive maize-wheat growing cycles. *significant at P ≤ 0.05; **significant at P ≤ 0.01.

CT and CTM/CT in the absence of N fertilizer application increased significantly over time (Fig. 2). Similar trends were observed but not statistically significant for wheat relative yields, whilst the effects were negative during half of the 12 years (Fig. 2). When 120 and 240 kg N ha−1 were applied, the relative yields of RFM/CT and CTM/ CT were remarkably higher in 2006 and 2013 during maize season, but in 2008 and 2013 during wheat season.

Table 3 The amount of stored water (0–200 cm) at maize harvest as affected by system and N rate over five years. Systems CT

N rates

0 120 240 Mean RFM 0 120 240 Mean CTM 0 120 240 Mean Multi-way ANOVA analysis System (S) F = 1.33 N rate (N) F = 4.34 Year (Y) F = 131.1 S×N F = 1.77 S×Y F = 0.54 N×Y F = 1.85 S×N×Y F = 1.14

3.2. The amount of stored water in the 0–200 cm soil profile The amount of stored water (ASW) in the 0–200 cm soil depth determined after maize (Table 3) and wheat (Table 4) harvest showed a high interannual variability over the study period. No significant year × S and S × N and year × S × N interactions were observed during both maize and wheat seasons. Tillage-mulch systems did not affect ASW at maize harvest, while the RFM treatment significantly increased ASW in 0–200 cm soil depth at wheat harvest in 2006, 2009 and 2013. By contrast, N fertilization exhibited a remarkable and negative impact on ASW within 0–200 cm soil depth in most cases (Tables 3 and 4). Interestingly, the observations of ASW in the 0–200 cm soil profile at wheat harvest were appreciably lower than that observed at maize harvest.

2003

2004

2005

2008

2012

530a 525a 523a 526A 536a 534a 517a 529A 509a 530a 516a 518A

502a 481b 462b 482A 502a 467b 480b 483A 515a 485b 489b 496A

535a 544a 537a 538A 539a 559a 546a 548A 553a 550a 543a 549A

470a 431b 438b 446A 509a 431b 415b 451A 496a 446b 450b 464A

460a 442b 441b 448A 469a 459a 459a 463A 462a 452a 462a 459A

P = 0.274 P < 0.05 P < 0.01 P = 0.149 P = 0.707 P = 0.134 P = 0.353

Upper-case and lower-case letters indicate comparisons between three mulching methods and between three N rates, respectively. Treatment means followed by the same letter are not significantly different at P ≤ 0.05.

3.3. Water productivity WP of the wheat-maize rotation system calculated from June 2003 158

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

3.5. Residual soil nitrate

Table 4 The amount of stored water (0–200 cm) at wheat harvest as affected by system and N rate over five years. Systems

N rates

CT

0 120 240 Mean RFM 0 120 240 Mean CTM 0 120 240 Mean Multi-way ANOVA analysis System (S) F = 4.38 N rate (N) F = 89.1 Year (Y) F = 67.3 S×N F = 2.43 S×Y F = 2.33 N×Y F = 6.46 S×N×Y F = 0.55

2004

2005

2006

2009

2013

348a 322a 327a 332A 345a 297b 309b 317A 359a 305b 305b 323A

387a 306b 320b 338A 380a 321b 311b 337A 369a 292b 310b 324A

409a 344b 365b 373B 479a 383b 362b 408A 427a 386b 368b 394AB

335a 299b 285b 306B 377a 299b 303b 326A 369a 302b 310b 327A

379a 356b 354b 363B 413a 378b 368b 386A 394a 369b 368b 376AB

As shown in Fig. 4, the accumulation of RSN in 0–200 cm soil showed a high interannual variability over the studied 3 wheat-maize rotation cycles. Significant year × S, year × N, S × N and year × S × N interactions were found at maize harvest, while no significant year × S and S × N interactions was observed at wheat harvest. Tillage-mulch systems showed significant influence on RSN accumulated in the 0–200 cm soil at maize harvest (Fig. 4). Moreover, the RFM treatment exhibited the highest RSN in 0–200 cm soil under N120 and N240, while considerably higher RSN values were observed in soils treated with CTM under N0 (Fig. 4A). However, RFM and CTM usually received appreciably lower RSN contents in the 0–100 cm soil layer at wheat harvest (Table S2), although there was no significant difference in RSN accumulation in 0–200 cm soil when compared with CT (Fig. 4B). N fertilization significantly increased RSN accumulation in the 0–200 cm soil layer after crop harvest, which declined in the order N240 > N120 > N0 (Fig. 4). Averaged across seasons, the RSN following maize harvest was 97.3 and 208 kg N ha−1 in the 0–100 cm soil layer, and 66.0 and 285 kg N ha−1 in the 100–200 cm soil layer at 120 and 240 kg N ha−1 rates, respectively (Table S2). At wheat harvest, 68.5 and 175 kg ha−1 RSN accumulation in the 0–100 cm layer and 28.3 and 170 kg N ha−1 in the 100–200 cm soil layer were recorded corresponding to N applications of 120 and 240 kg N ha−1 (Table S2).

P < 0.05 P < 0.01 P < 0.01 P = 0.059 P = 0.068 P < 0.01 P = 0.813

Upper-case and lower-case letters indicate comparisons between three mulching methods and between three N rates, respectively. Treatment means followed by the same letter are not significantly different at P ≤ 0.05. Table 5 Effects of system and N rate on water productivity (WP) of wheat-maize over the ten-year period (June 2003-June 2013). Treatment

Systems CT RFM CTM N rates N0 N120 N1240

Total rainfall (mm)

Irrigation (mm)

Water consumption (mm)

WP (kg mm−1 ha−1)

6457 6457 6457

805 202 202

7235 6632 6640

12.2c 15.3a 14.2b

6457 6457 6457

403 403 403

6816 6844 6847

9.56b 15.7a 15.6a

3.6. N fertilizer use efficiency NfP and NupE of maize under RFM practice were significantly increased by 22% and 26%, respectively, in comparison with CT, averaged across three years (Table 6). Similar trends were observed in CTM treatment. No regular changes were found for NUE, although NUE of maize was significantly different among the CT, RFM and CTM treatments. On the contrary, RFM and CTM showed no significant effect on NfP, NupE and NUE of wheat in 2009, 2010 and 2011 (Table 6). In addition, NfP and NupE under a high N rate of 240 kg N ha−1 were always significantly lower than that of the low N rate of 120 kg N ha−1, while the N application rate had no consistent effect on NUE, regardless of maize and wheat in three years (Table 6).

Values within the same column followed by different letters are significantly different at P ≤ 0.05. N0, N120, N240 = 0, 120, 240 kg N ha−1; CT, conventional tillage with no mulch; RFM, ridge-furrow with plastic film-mulched ridges and straw-mulched furrows; CTM, conventional tillage with straw mulch.

3.7. Soil physicochemical properties Four soil variables including SWC, SOM, total N and available N differed significantly among the different tillage-mulch systems (Table 7). Compared to CT, RFM and CTM strongly enhanced soil moisture; the RFM treatment substantially increased available N; and the CTM treatment markedly increased SOM and total N. N fertilization only significantly affected total N and available N out of the 7 determined properties (Table 7). Applying N fertilizer significantly increased total N, while no significant difference was observed between the N120 and N240 treatments. As expected, available N declined in the order N240 > N120 > N0, and the interaction between tillage-mulch system and level of N was only significant (P < 0.05) on available N.

to June 2013 was significantly influenced by tillage-mulch system and N fertilization (Table 5). WP was highest for RFM followed by CTM and CT. Applying N fertilizer significantly increased WP when compared with the unfertilized treatment, while there was no significant difference in WP between the N120 and N240 treatments (Table 5).

3.4. N uptake Tillage-mulch systems significantly affected seasonal N uptake by maize, but not by wheat, over the 3 consecutive wheat-maize rotation cycles studied (Fig. 3). RFM showed the highest N uptake during maize seasons (Fig. 3), and the total N uptake over a 6-season period was ranked as RFM > CTM ≈ CT (Table S1). Applying N fertilizer substantially increased seasonal N uptake by maize and wheat plants, while the differences in seasonal N uptake between the N120 and N240 treatments were not statistically significant over the studied seasons (Fig. 3). The S × N, S × Y and S × N × Y interactions were significant on seasonal N uptake by maize, but not wheat.

4. Discussion 4.1. Grain yield Mulching always significantly increases maize and wheat yields from the first experimental year when ∼300 mm water were irrigated (Huang et al., 2015; Li et al., 2017), but other researchers have also reported yield reduction in maize or wheat under straw mulch conditions without irrigation (Fabrizzi et al., 2005; Gao et al., 2009). In our study, RFM and CTM showed no effect on wheat yield during the first 159

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

Fig. 3. Seasonal N uptake by maize (A) and wheat (B) influenced by system (S) and N rate (N) over 3 consecutive rotation cycles (2008–2011) (kg N ha−1). Values followed by the same small letters within each N fertilization rate are not significant different at P ≤ 0.05, and the different upper-case letters indicate the significant difference between three systems. ns, no significant difference; * significant at P ≤ 0.05; **significant at P ≤ 0.01.

negative quadratic model (Liu et al., 2014; Dai et al., 2016). In this way, the non-significant differences in maize and wheat yields between N120 and N240 over the 12-year period indicated that 240 kg N ha−1 crop−1 application may be overused in wheat-maize rotation systems in the Loess Plateau. The less variation of maize and wheat yields in the fertilized-N treatments suggested that applying N fertilizer is important for stabilizing crop productivity in agriculture. In addition, the few remarkable increase of relative yield (Fig. 2) may be probably due to the dramatic scarcity (maize season in 2006 and 2013, and wheat season in 2008) or uneven distribution (wheat season in 2013) of precipitation occurring in the corresponding seasons (Fig. 1 and Table S3). The cumulative trends were occasionally interrupted by differences in weather across years, which confirmed that crop productivity in the Loess Plateau is primarily driven by annual rainfall and its distribution (Guo et al., 2012). Notably, the clear cumulative benefits from the straw retention at 2.25 (RFM) and 4.5 t ha−1 (CTM) observed in maize relative yield implied that mulching is a more practical option in maize rather than wheat growing in the Loess Plateau.

four trial years, and CTM did not significantly increase maize yield until the third trial year, while RFM practice showed significant increase of maize yield across the entire study period. The inconsistent yield response to mulch may be attributed to the large differences in the amount of irrigation among those experiments (Ram et al., 2013). Abebe et al. (2016) found that temperatures during the maturity phase negatively affected maize grain yield. Thus, the more beneficial soil condition created by RFM that cooler in the hot summer months during maize growing season observed by Chen et al. (2015), could explain the fact that maize yield benefited from the first trial year under RFM. By contrast, the non-significant increase of maize yield under CTM in the first two years is probably associated with the low N fertility that occurred when the soil was covered with mulch residue (Huang et al., 2015). With an exception of 2008, wheat yields under RFM and CTM were equivalent to that of CT from 2004 to 2012. This is discrepant with the finding of Gao et al. (2009), who found yield reduction in a winter wheat and summer fallow system, although similar straw amounts of 4500 kg ha−1 was applied. The conflicting findings might be ascribed to the different cropping systems. The positive impact of CTM on soil moisture (Table 7) may be offset by the decrease of soil N availability and temperature, thus resulting in no effect on wheat yield over a 12-year period compared with CT (Table 2). By contrast, the relatively more favorable soil condition created by RFM that warmer in the cool winter months during wheat growing season observed by Chen et al. (2015) may explain the significant increase of wheat yield during the last three years. Compared to CTM and CT, the significant increase and less variation of maize and wheat grain yields observed in RFM suggested that RFM practice is more beneficial for enhancing and stabilizing crop production in wheat-maize rotation systems. N application always increases crop yield, whereas excessive N fertilization results in yield reduction (Agegnehu et al., 2016). The relationship between maize/wheat yield and N rate is well-fitted by a

4.2. The amount of stored water and water use efficiency Owing to climate change, increasing water productivity has never been urgent in the Loess Plateau (Sun and Ma, 2015). The amounts of stored water (0–200 cm) in soils treated with RFM and CTM were equivalent at maize harvest (Table 3), but were appreciably higher at wheat harvest when compared with CT (Table 4), which was ascribed to the rainfall distribution across the year (Table S2). Contrary to the finding that RFM practice might not be sustainable in the long term as its increased grain yield is associated with soil water depletion in deeper (> 100 cm) soil layers (Wang et al., 2015), our results showed equal or even slightly higher amount of stored water in 100–200 cm soil 160

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

Fig. 4. Residual soil nitrate accumulation (0–200 cm) affected by system (S) and N rate (N) over 3 maize-wheat rotation cycles. Values followed by the same small letters within each N fertilization rate are not significant different at P ≤ 0.05, and the different upper-case letters indicate the significant difference between three systems. ns, no significant difference; *significant at P ≤ 0.05; **significant at P ≤ 0.01.

Table 6 N uptake efficiency (NupE, kg−1), N fertilizer productivity (NfP, kg−1) and N use efficiency (NUE, kg−1) of maize and wheat at maturity as affected by system (S) and N rate (N) over a 3 consecutive maize-wheat growing cycles. Systems

Summer maize CT

RFM

CTM

Winter wheat CT

RFM

CTM

N rates

2008

2009

2010

NupE

NfP

NUE

NupE

NfP

NUE

NupE

NfP

NUE

120 240 Mean 120 240 Mean 120 240 Mean F test (S) F test (N) F test (S×N)

0.97a 0.49b 0.73B 0.95a 0.65b 0.80A 0.99a 0.58b 0.79A * ** **

48.5a 25.5b 37.0C 61.7a 30.4b 46.1A 57.7a 27.5b 42.6B ** ** *

47.2a 48.6a 47.9B 60.1a 48.7b 58.4A 55.3a 47.5b 51.4B ** ** **

0.71a 0.37b 0.54C 0.91a 0.44b 0.67B 0.94a 0.49b 0.72A ** ** **

36.7a 19.1b 27.9C 50.0a 24.6b 37.3A 42.8a 21.9b 32.4B ** ** *

51.7a 51.7a 51.7A 55.0a 55.8a 55.4A 45.4a 44.8a 45.1B * ns ns

0.99a 0.48b 0.73C 1.22a 0.68b 0.95A 1.09a 0.52b 0.81B ** ** ns

55.9a 28.4b 42.2B 66.6a 36.7b 51.6A 68.3a 34.2b 51.2A ** ** ns

56.4a 58.8a 57.6B 54.7a 53.6a 54.2B 62.7a 65.1a 63.9A ** ns ns

120 240 Mean 120 240 Mean 120 240 Mean F test (S) F test (N) F test (S×N)

2009 1.09a 0.71b 0.91A 1.19a 0.64b 0.92A 1.16a 0.48b 0.82A ns ** ns

29.3a 18.6b 24.0A 34.9a 18.6b 26.7A 31.5a 16.7b 24.1A ns ** ns

26.8a 26.1a 26.4A 30.2a 29.2a 29.7A 26.9a 34.9a 30.9A ns ns ns

2010 1.57a 0.83b 1.20A 1.74a 0.69b 1.22A 1.63a 0.82b 1.23A ns ** ns

30.1a 14.1b 22.1A 31.8a 16.6b 24.2A 31.7a 14.7b 23.2A ns ** ns

19.1a 17.4a 18.2A 18.2a 24.0a 21.1A 19.4a 17.9a 18.6A ns ns ns

2011 1.01a 0.47b 0.74A 1.05a 0.56b 0.81A 0.91a 0.47b 0.70A ns ** ns

40.2a 20.2b 30.2A 48.5a 22.1b 35.3A 37.4a 18.9b 28.2A ns ** ns

39.5a 43.3a 41.4A 45.7a 39.8a 42.7A 40.9a 39.5a 40.2A ns ns ns

Upper-case and lower-case letters indicate comparisons between three mulching methods and between three N rates, respectively. Treatment means followed by the same letter are not significantly different at P ≤ 0.05. ns, no significant difference; * Significant at P ≤ 0.05; ** Significant at P ≤ 0.01.

161

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

Table 7 Soil physicochemical properties (0–20 cm) as influenced by system (S) and N rate (N) after maize harvest in 2015. Systems

N rates

SWC (%)

pH

SOM (g kg-1)

Total N (g kg-1)

Available N (mg kg-1)

Available P (mg kg-1)

Available K (mg kg-1)

CT

0 120 240 Mean 0 120 240 Mean 0 120 240 Mean F test (S) F test (N) F test (S×N)

16.6a 17.3a 17.3a 17.1B 18.3a 18.3a 18.7a 18.4A 17.9a 18.1a 19.0a 18.3A ** ns ns

8.00a 8.08a 8.17a 8.08A 7.92a 7.96a 8.10a 8.00A 8.05a 8.10a 8.15a 8.10A ns ns ns

14.8a 15.1a 15.5a 15.1B 15.0a 15.3a 15.3a 15.2B 15.9a 17.1a 17.8a 16.9A ** ns ns

0.90a 0.95a 0.97a 0.94B 0.92a 0.97a 1.00a 0.96B 0.98a 1.09ab 1.16a 1.08A ** ** ns

6.74b 8.44b 24.6a 13.3B 3.97c 21.8b 85.1a 33.6A 7.08b 13.8ab 24.6a 15.2B ** ** *

38.8a 39.1a 43.4a 40.4A 33.9a 40.9a 44.1a 41.7A 39.1a 43.0a 43.4a 41.9A ns ns ns

90.8a 96.3a 96.4a 94.5A 98.5a 99.5a 102a 100A 96.6a 104a 106a 102A ns ns ns

RFM

CTM

Upper-case and lower-case letters indicate comparisons between three mulching methods and between three N rates, respectively. Treatment means followed by the same letter are not significantly different at P ≤ 0.05. ns, no significant difference; * Significant at P ≤ 0.05; ** Significant at P ≤ 0.01.

for the prevention of N leaching in wheat-maize cropping system. Furthermore, the appreciably higher RSN accumulation at 0–200 cm soil depth under RFM than CTM confirmed that the plastic coverage could protect the fertilizer from leaching (Ruidisch et al., 2013). Rainfall is an important factor controlling NO3−-N leaching, and soil wetting front could reach a depth beyond 2 m in extremely wet year in the southern Loess Plateau (Liu et al., 2010; Jia et al., 2014). Therefore, the sharp decrease in RSN within the 0–200 cm soil profile at maize harvest in 2012 may be attributed to the exceptionally high rainfall occurring in 2011 (Fig. 1). Interestingly, the amounts of RSN accumulated in 1 m soil under RFM and CTM were slightly lower than CT at wheat harvest (Table S2), which is probably due to the enhanced soil moisture conditions created by mulch (Table 4), resulting in nitrate movement downwards. The relatively low N uptake and the increased soil N mineralization caused by the higher soil temperature during maize season may explain the regular decrease in RSN in the 0–200 cm soil profile from the maize to wheat season. As expected, RSN increases rapidly with increases in the quantity of N applied (Fig. 4). Averaged across seasons, the considerable buildup of RSN to the extent of > 190 kg N ha−1 in 100 cm soil depth were observed at harvest when 240 kg N ha−1 was applied (Table S2), which was appreciably higher than the value of 90–100 kg N ha−1 accepted in Europe as environmentally safe following crop harvest (Hofman, 1999). Liu et al. (2003) noted that the accumulation of NO3−-N below 100 cm depth (beyond the reach of most roots) was the main pathway for N losses in the wheat-maize cropping system. RSN accumulated in the 100–200 cm soil profile after crop harvest was > 170 kg N ha−1 at the conventional N rate of 240 kg N ha−1 (Table S2), suggesting a serious N leaching and environmental risk. On the contrary, the N120 treatment was characterized with considerably lower RSN accumulation in both 0–100 cm and 100–200 cm soil layers at harvest, while maintaining an acceptably high crop yield. Furthermore, Lenka et al. (2013) noted that RSN should not exceed 150 kg N ha−1 in the 0–100 cm soil layer at harvest, or decrease crop response to added N fertilizer. Therefore, farmers’ N rate of 240 kg N ha−1 must be properly reduced and the large indigenous N supply should be taken into account due to its high N-fertilizer substitution. The significant increase of N uptake by maize but not by wheat (Fig. 3) under RFM and CTM practices was mainly associated with the corresponding grain yield (Table 1 and 2), as crop production is wellcorrelated with N uptake (Agegnehu et al., 2016). As expected, N fertilization significantly increased N uptake compared to the unfertilized treatment. This confirmed the previous findings of Scholberg et al. (2000), who discovered that lower N supply appeared to reduce crop N uptake, in turn hampering full canopy development and crop yield.

depth under RFM compared with CT (data not shown). This is probably due to the greater reduction in water loss via evaporation (Pabin et al., 2003). N fertilization always substantially decreased the amount of stored water (0–200 cm) at harvest, which could be attributed to the dramatic transpiration and water depletion caused by increased maize and wheat plant development (Tables 1 and 2). Similar finding was reported by Li et al. (2015). The equivalent amount stored soil water (0–200 cm) between N120 and N240 observed at maize harvest in 2003 and 2005 may be related to the exceptionally high rainfall during the late stage of the corresponding seasons (Table S3). The water consumption of wheat was much higher than the precipitation received during the growing season in the Loess Plateau (Huang et al., 2003), which may explain the sharp decrease in the amount of stored water in 2 m soil depth at wheat harvest. Many researchers focused on improving WP of grain in water-limited areas (Gan et al., 2013). Similar to previous reports (Liu et al., 2014; Li et al., 2015), this study also illustrated that RFM and CTM substantially increased both total yield and WP over the ten years (June 2003–June 2013), which was mainly attributed to the retardation of soil evaporation and enhancement of rainwater infiltration. The more favorable soil conditions created by RFM being cooler in hot summer months and warmer in cool winter months (Chen et al., 2015) could enhance crop production (Tables 1 and 2) and soil moisture (Table 4), resulting in more beneficial effects on WP compared to CTM. Notably, the RFM and CTM treatments saved ∼ 600 mm in irrigation from 2003 to 2013 (Table 5), which was partially compensated by a more efficient utilization of the available soil moisture in our study area. This is of great practical significance because the Loess Plateau is predicted to become drier and warmer in the future. N fertilization markedly increased WP, which was probably due to its capacity for increasing crop production and relieving water limitation under dry and poor water supply (Gonzalez-Dugo et al., 2010). Similar result was observed by Liu et al. (2014), who also found a significant negative quadratic correlation between WP and N rate. Thus, the non-significant differences in WP between N120 and N240 implied that N fertilizer application of 240 kg N ha−1 crop−1 probably exceeds the N requirement of the wheat-maize rotation system in our study area.

4.3. Residual soil nitrate, n uptake and N fertilizer use efficiency Nitrate leaching mainly occurred during the maize season due to its coincidence with the most (60–70%) rainfall (Jia et al., 2014). RFM and CTM exhibited considerably higher amounts of RSN accumulated in 2 m soil than that of CT at maize harvest, indicating a promising option 162

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

References

RFM and CTM treatments are an effective means of increasing N fertilizer use efficiency due to their significantly higher NupE and NfP for maize. Similar to the finding of Li et al. (2017), the high N rate of 240 kg N ha−1 substantially decreased NupE and NfP compared to the lower N rate of 120 kg N ha−1 regardless of tillage-mulch systems over 3 consecutive maize-wheat growing cycles. Therefore, farmers’ N rate of 240 kg N ha−1 should be decreased to improve N fertilizer use efficiency.

Abebe, A., Pathak, H., Singh, S.D., Bhatia, A., Harit, R.C., Kumar, V., 2016. Growth, yield and quality of maize with elevated atmosphere carbon dioxide and temperature in north-west India. Agric. Ecosyst. Environ. 218, 66–72. Agegnehu, G., Nelson, P.N., Bird, M.I., 2016. Crop yield, plant nutrient uptake and soil physicochemical properties under organic soil amendments and nitrogen fertilization on Nitisols. Soil Tillage Res. 160, 1–13. Singh, Balwinder, Humphreys, E., Gaydon, D.S., Eberbach, P.L., 2016. Evaluation of the effects of mulch on optimum sowing date and irrigation management of zero till wheat in central Punjab, India using APSIM. Field Crops Res. 197, 83–96. Bu, L.D., Liu, J.L., Zhu, L., Luo, S.S., Chen, X.P., Li, S.Q., Hill, R.L., Zhao, Y., 2013. The effects of mulching on maize growth, yield and water use in a semi-arid region. Agric. Water Manage. 123, 71–78. Chen, Y.L., Liu, T., Tian, X.H., Wang, X.F., Li, M., Wang, S.X., Wang, Z.H., 2015. Effects of plastic film combined with straw mulch on grain yield and water use efficiency of winter wheat in Loess Plateau. Field Crops Res. 172, 53–58. Dai, J., Wang, Z.H., Li, M.H., He, G., Li, Q., Cao, H.B., Wang, S., Gao, Y.J., Hui, X.L., 2016. Winter wheat grain yield and summer nitrate leaching: long-term effects of nitrogen and phosphorus rates on the Loess Plateau of China. Field Crops Res. 196, 180–190. Deng, X.P., Shan, L., Zhang, H.P., Turner, N.C., 2006. Improving agricultural water use efficiency in arid and semiarid areas of China. Agric. Water Manage. 80, 23–40. Deng, L., Wang, W.K., Yang, X.T., Li, Y., 2008. Spatial variability of nitrate content of groundwater in Guanzhong Basin. J. Arid. Land Resour. Environ. 22, 152–155 (Chinese with English abstract). Elliott, J., Deryng, D., Müller, C., Frieler, K., Konzmann, M., Gerten, D., Glotter, M., Flörke, M., Wada, Y., Best, N., Eisner, S., Fekete, B.M., Folberth, C., Foster, L., Gosling, S.N., Haddeland, I., Khabarov, N., Ludwig, F., Masaki, Y., Olin, S., Rosenzweig, C., Ruane, A.C., Satoh, Y., Schmid, E., Stacke, T., Tang, Q.H., Wisser, D., 2014. Constraints and potentials of future irrigation water availability on agricultural production under climate change. Proc. Nat. Acad. Sci. U. S. A. 11, 3239–3244. Fabrizzi, K.P., García, F.O., Costa, J.L., Picone, L.I., 2005. Soil water dynamics, physical properties and corn and wheat responses to minimum and no-tillage systems in the southern Pampas of Argentina. Soil Tillage Res. 81, 57–69. Gan, Y.T., Siddique, K.H.M., Turner, N.C., Li, X.G., Niu, J.Y., Yang, C., Liu, L.P., Chai, Q., 2013. Ridge-furrow mulching systems–an innovative technique for boosting crop productivity in semiarid rain-fed environments. Adv. Agron. 118, 429–476. Gao, Y.J., Li, Y., Zhang, J.C., Liu, W.G., Dang, Z.P., Cao, W.X., Qiang, Q., 2009. Effects of mulch, N fertilizer, and plant density on wheat yield wheat nitrogen uptake, and residual soil nitrate in a dryland area of China. Nutr. Cycl. Agroecosyst. 85, 109–121. Gonzalez-Dugo, V., Durand, J.L., Gastal, F., 2010. Water deficit and nitrogen nutrition of crops a review. Agron. Sustain. Dev. 30, 529–544. Govaerts, B., Sayre, K.D., Deckers, J., 2006. A minimum data set for soil quality assessment of wheat and maize cropping in the highlands of Mexico. Soil Tillage Res. 87, 163–174. Guo, S.L., Zhu, H.H., Dang, T.H., Wu, J.S., Liu, W.Z., Hao, M.D., Li, Y., Syers, J.K., 2012. Winter wheat grain yield associated with precipitation distribution under long-term nitrogen fertilization in the semiarid Loess Plateau in China. Geoderma 189–190, 442–450. Guo, Z.J., Zhang, Y.L., Zhao, J.Y., Shi, Y., Yu, Z.W., 2014. Nitrogen use by winter wheat and changes in soil nitrate nitrogen levels with supplemental irrigation based on measurement of moisture content in various soil layers. Field Crops Res. 164, 117–125. Hofman, G., 1999. Nutrient Management Legislation in European Countries NUMALEC Report. Concerted Action. (Fair6-CT98-4215). Huang, M.B., Dang, T.H., Gallichand, J., Goulet, M., 2003. Effect of increased fertilizer applications to wheat crop on soil-water depletion in the Loess Plateau China. Agric. Water Manage. 58, 267–278. Huang, M.X., Liang, T., Wang, L.Q., Zhou, C.H., 2015. No-tillage and fertilization management on crop yields. and nitrate leaching in North China Plain. Ecol. Evol. 5, 1143–1155. Jia, X.C., Shao, L.J., Li, P., Zhao, B.Q., Gu, L.M., Dong, S.T., Bing, S.H., Zhang, J.W., Zhao, B., 2014. Effect of different nitrogen and irrigation treatments on yield and nitrate leaching of summer maize (Zea mays L.) under lysimeter conditions. Agric. Water Manage. 137, 92–103. Kahlon, M.S., Lal, R., Ann-Varughese, M., 2013. Twenty two years of tillage and mulching impacts on soil physical characteristics and carbon sequestration in Central Ohio. Soil Tillage Res. 126, 151–158. Lenka, S., Singh, A.K., Lenka, N.K., 2013. Soil water and nitrogen interaction effect on residual soil nitrate and crop nitrogen recovery under maize-wheat cropping system in the semi-arid region of northern India. Agric. Ecosyst. Environ. 179, 108–115. Li, S.X., Wang, Z.H., Li, S.Q., Gao, Y.J., 2015. Effect of nitrogen fertilization under plastic mulched and non-plastic mulched conditions on water use by maize plants in dryland areas of China. Agric. Water Manage. 162, 15–32. Li, C.J., Wang, C.J., Wen, X.X., Qin, X.L., Liu, Y., Han, J., Li, Y.J., Liao, Y.C., Wu, W., 2017. Ridge-furrow with plastic film mulching practice improves maize productivity and resource use efficiency under the wheat-maize double-cropping system in dry semi-humid areas. Field Crops Res. 203, 201–211. Liu, X.J., Ju, X.T., Zhang, F.S., Pan, J.R., Christie, P., 2003. Nitrogen dynamics and budgets in a winter wheat-maize cropping system in the North China Plain. Field Crops Res. 83, 111–124. Liu, W.Z., Zhang, X.C., Dang, T.H., Ouyang, Z., Li, Z., Wang, J., Wang, R., Gao, C.Q., 2010. Soil water dynamics and deep soil recharge in a record wet year in the southern Loess Plateau of China. Agric. Water Manage. 97, 1133–1138. Liu, C.A., Zhou, L.M., Jia, J.J., Wang, L.J., Si, J.T., Li, X., Pan, C.C., Siddique, K.H.M., Li,

4.4. Changes in soil fertility Soil physicochemical properties are important indicators of soil fertility (Murphy et al., 2016). The non-significant changes in SOM and total N concentrations in soils treated without N fertilization suggested that the soil sustained a low level equilibrium under the low-input farming system of the Loess Plateau. The enhanced soil moisture observed in soils treated with RFM and CTM practices (Table 7) is beneficial for promoting earlier germination and plant establishment (Bu et al., 2013). Similar finding was reported by Kahlon et al. (2013), who ascribed the substantially increased SOM and total N concentrations under CTM to the extra nutrient input and the enhanced more favorable soil conditions created by straw incorporation. As expected, long-term N fertilization significantly increased soil total N and available N, and the equivalent values of total N observed in N120 and N240 treatments indicated no significant effect of the extra 120 kg N ha−1 on soil total N after 12-year repeated application of 240 kg N ha−1. This is probably due to the low levels of soil organic carbon. Guo et al. (2012) found significant increase in SOM under N fertilized treatments after 25 years, while our study revealed no significant difference in SOM between all treatments, which was mainly associated with the differences in experimental period and annual precipitation (Gan et al., 2013). 5. Conclusion RFM and CTM significantly increased total grain yield and saved ∼ 600 mm of irrigation water over the 12-year study period. Compared to CTM, the RFM treatment showed more favorable impacts on crop grain yield, WP, N uptake and RSN. N fertilization substantially increased wheat-maize production across the entire period. The N120 treatment with low N input showed equivalent crop grain yields, WP and N uptake when compared with the N application rate of 240 kg N ha−1 by farmers. A considerable amount of RSN accumulated in the 0–200 cm soil profile was observed after crop harvest under N240 treatment, indicating a large environmental risk of NO3−-N leaching loss, while the opposite was true of N120. In addition, the current fertilizer management only NP fertilizers applied could lead to an imbalance in soil nutrients, and managers in this region should pay more attention to the balanced fertilization. Overall, the data show RFM practice with the N rate of 120 kg ha−1 as a promising strategy for developing sustainable maize-wheat cropping systems in the southern Loess Plateau and other areas with similar agro-meteorology in the world. Acknowledgements This work was funded by the National Key Research and Development Program of China (2016YFD0200104), the Shaanxi Province Science Foundation for Youths (2016JQ4012), and the Fundamental Research Funds for the Central Universities (2452016065). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.fcr.2017.08.006. 163

Field Crops Research 213 (2017) 154–164

Z. Liu et al.

Nitrogen stress effects on growth and nitrogen accumulation by field-grown tomato. Agron. J. 92, 159–167. Sharma, P., Abrol, V., Sharma, R.K., 2011. Impact of tillage and mulch management on economics, energy requirement and crop performance in maize-wheat rotation in rainfed subhumid inceptisols, India. Eur. J. Agron. 34, 46–51. Sun, C.F., Ma, Y.Y., 2015. Effects of non-linear temperature and precipitation trends on Loess Plateau droughts. Quat. Int. 372, 175–179. Verhulst, N., Sayre, K.D., Vargas, M., Crossa, J., Deckers, J., Raes, D., Govaerts, B., 2011. Wheat yield and tillage-straw management system × year interaction explained by climatic co-variables for an irrigated bed planting system in northwestern Mexico. Field Crops Res. 124, 347–356. Wang, X.K., Li, Z.B., Xing, Y.Y., 2015. Effects of mulching and nitrogen on soil temperature water content, nitrate-N content and maize yield in the Loess Plateau of China. Agric. Water Manage. 161, 53–64. Yang, X.L., Lu, Y.L., Ding, Y., Yin, X.F., Raza, S., Tong, Y.A., 2017. 2017. Optimising nitrogen fertilisation: a key to improving nitrogen-use efficiency and minimising nitrate leaching losses in an intensive wheat/maize rotation (2008–2014). Field Crops Res. 206, 1–10. Zhang, S.L., Gao, P.C., Tong, Y.A., Norse, D., Lu, Y.L., Powlson, D., 2015. Overcoming nitrogen fertilizer over-use through technical and advisory approaches A case study from Shaanxi Province, northwest China. Agric. Ecosyst. Environ. 209, 89–99. Zhang, F., Zhang, W.J., Li, M., Zhang, Y., Li, F.M., Li, C.B., 2017. Is crop biomass and soil carbon storage sustainable with long-term application of full plastic film mulching under future climate change? Agric. Syst. 150, 67–77.

F.M., 2014. Maize yield and water balance is affected by nitrogen application in a film-mulching ridge-furrow system in a semiarid region of China. Eur. J. Agron. 52, 103–111. Lu, R.K., 2000. Analytical Methods of Soil and Agricultural Chemistry China. Agricultural Science and Technology Press, Beijing, China. Mueller, N.D., Gerber, J.S., Johnston, M., Ray, D.K., Ramankutty, N., Foley, J.A., 2012. Closing yield gaps through nutrient and water management. Nature 490, 254–257. Murphy, R.P., Montes-Molina, J.A.M., Govaerts, B., Six, J., van Kessel, C., Fonte, S.J., 2016. Crop residue retention enhances soil properties and nitrogen cycling in smallholder maize systems of Chiapas, Mexico. Appl. Soil Ecol. 103, 110–116. Gupta, Naveen, Yadav, Sudhir, Humphreys, E., Kukal, S.S., Singh, Balwinder, Eberbach, P.L., 2016. Effects of tillage and mulch on the growth, yield and irrigation water productivity of a dry seeded rice-wheat cropping system in north-west India. Field Crops Res. 196, 219–236. Pabin, J., Lipiec, J., Włodek, S., Biskupski, A., 2003. Effect of different tillage systems and straw management on some physical properties of soil and on the yield of winter rye in monoculture. Int. Agrophysics. 17, 175–181. Ram, H., Dadhwal, V., Vashist, K.K., Kaur, H., 2013. Grain yield and water use efficiency of wheat (Triticum aestivum L.) in relation to irrigation levels and rice straw mulching in North West India. Agric. Water Manage. 128, 92–101. Ruidisch, M., Bartsch, S., Kettering, J., Huwe, B., Frei, S., 2013. The effect of fertilizer best management practices on nitrate leaching in a plastic mulched ridge cultivation system. Agric. Ecosyst. Environ. 169, 21–32. Scholberg, J., Mcneal, B.L., Boote, K.J., Jones, J.W., Locascio, S.J., Olson, S.M., 2000.

164