Effects of tillage and cropping sequences on crop production and environmental benefits in the North China Plain

The ever-increasing trend of greenhouse gas (GHG) emissions is accelerating global warming and threatening food security. Environmental benefits and sustainable food production must be pursued locally and globally. Thus, a field experiment was conducted in 2015 to understand how to balance the trade-offs between agronomic productivity and environment quality in the North China Plain (NCP). Eight treatments consisted of two factors, i.e., (1) tillage practices: rotary tillage (RT) and no-till (NT), and (2) cropping sequences (CS): maize–wheat–soybean–wheat (MWSW), soybean–wheat–maize–wheat (SWMW), soybean–wheat (SW), and maize–wheat (MW). The economic and environmental benefits were evaluated by multiple indicators including the carbon footprint (CF), maize equivalent economic yield (MEEY), energy yield (EY), and carbon sustainability index (CSI). Compared with NT, RT increased the EY and MEEY, but emitted 9.4% higher GHGs. Among different CSs, no significant reduction was observed in CF. The lowest (2.0 Mg CO2-eq ha−1 year−1) and the highest (5.6 Mg CO2-eq ha−1 year−1) CF values were observed under MW and SWMW, respectively. However, CSs with soybean enhanced MEEY and the net revenue due to their higher price compared to that of MW. Although the highest CSI was observed under RT-MW, soybean-based crop rotation could offset the decline in CSI under NT when compared to that for RT. These findings suggest that conservation agriculture (CA) could enhance the balance in trade-offs between economic and environmental benefits. Additional research is needed on how to achieve high crop production by establishing a highly efficient CA system in the NCP.


Introduction
The anthropogenic climate change is challenging global crop production and human survival (Guo 2015;IPCC 2018). Agronomic solutions are needed toward the site-specific variations among different regions, and the discrepancy in food supply-demand makes the identification of these solutions an urgent and complex challenge. According to the Intergovernmental Panel on Climate Change (IPCC 2013), agricultural activities emitted almost 20% of the global greenhouse gases (GHGs). The improper use of farming practices (e.g., overuse of N fertilizer, high-intensity tillage) are the major direct source of GHG emissions. In addition, the use of diesel fuel in the process of production, storage, and delivery of agricultural raw materials are indirect sources of GHG emissions (Lal 2004). Therefore, it is necessary to adopt a set of farming practices with high resource use efficiency, and achieve a win-win strategy to ensure food security and offset GHG (direct and indirect) emissions.
Best management practices (BMPs, e.g., conservation agriculture) have been identified as effective options to reduce GHG emissions and maximize economic profits (Janzen et al. 2006;Halvorson et al. 2008). A system-based CA, including minimal tillage, permanent soil cover, and complex crop rotations, has been regarded as one of the systematic solutions for sustainable crop production (Lal 2015). CA reduces indirect emission of GHGs by decreasing the number of field operations and associated cost inputs (West and Marland 2002;Lal 2004). No-till (NT) and crop residue retention (RR), two important principles of CA, can increase soil organic carbon (SOC) sequestration by promoting the formation of soil macro-aggregates and reducing mineralization (Kan et al. 2020a;Kan et al. 2020b;Meng 2020). However, some long-term studies have indicated that conversion to NT has risks of crop yield reduction during the initial stages of implementation (Huynh et al. 2019). Legumebased crop rotation is recognized as a feasible option for nutrient management in CA due to its advantage of biological N fixation (Zeng 2018). However, the impacts of legumebased crop rotation on the trade-offs between economic profits and environmental benefits of a CA system have not been well documented. Therefore, a comprehensive assessment of productivity and carbon (C) emission is imperatively helpful to establish an efficient CA system.
Carbon footprint (CF) is a useful tool to assess the environmental impacts of agricultural activities (van der Werf et al. 2014;Schenck and Huizenga 2014). In addition, CF based on the scales of different energy yield (EY), maize equivalent economic yield (MEEY), carbon sustainability index (CSI), and net benefits were also functional to determine and compare the C efficiency from multiple aspects. CF was widely used to evaluate NT, RR, and strategic tillage practices in some previous studies Zhang et al. 2016;Liu et al. 2021) and to compare different cropping systems (Yang et al. 2014;Sun et al. 2021). Under the CA system, assessing the comprehensive changes in CF, especially from multiple aspects, for the combination of different principles is valuable to optimize the CA system through balancing the trade-offs among cereal production and economic and environmental profits.
The North China Plain (NCP) is one of the major producers of wheat and maize cereals in China. However, the overuse of agricultural inputs such as fertilizers and the ground water in this area has led to severe environmental problems. There are still opportunities to alleviate environmental pressures from agricultural production through the development of cleaner production technologies, such as improved farming practices and crop rotation. CA has been adopted in this area since the 1970s . Similarly, the effects of farming practices on the SOC, yield, and economic and environmental benefits of tillage or crop rotation have been documented for the NCP (Li et al. 2010;Liang et al. 2011;Kan et al. 2020b). Although risks of crop failure under NT have been reported, its environmental benefits and profitability have also been widely reported (Canalli et al. 2020). Nonetheless, the research is lacking for the NCP region with regard to the performance of NT, and how crop rotations interact with it and affect the trade-offs among crop production and economic and environmental benefits. Thus, it is important to identify the interactions between NT and legume-based crop rotation for establishing an optimized CA system in the NCP. The present study was based on the hypothesis that, in the CA system, complex crop rotation, NT, and residue retention promote the balance of economic and environmental tradeoffs, and legume-based crop rotation enhances the net economic benefits. Thus, the objectives of this study were to (1) evaluate the crop yield, SOC, and CSI under tillage systems combined with different cropping sequences (CSs) and (2) determine a comprehensive assessment of crop yields, economic profits, and environmental costs under CA in this region.

Site description
The field experiment, which was initiated in 2015 at Wuqiao Experimental Station of China Agricultural University (37°36′ N, 116°21′ E) in Hebei Province, was cultivated with the winter wheat-summer maize cropping system for a long history until 2017. The region is a temperate continental monsoon climate, with an average annual temperature of 13.1 °C and average annual rainfall of 531 mm. The rainfall is unevenly distributed during the year, and 60-70% of the annual rainfall is concentrated between June and August. The soil type is Aquic Cambosol of a silt-loam texture (Zhao et al. 2020), containing 11.2% clay, 18.6% sand, and 70.2% silt. Baseline soil properties in the 0-20 cm layer were 9.52 g kg −1 SOC, 1.01 g kg −1 TN, 44.6 mg kg −1 available P, 92.20 mg kg −1 available K, and 7.51 pH.

Experimental design
The field experiment was laid out according to a split-plot design with three replications. The main plots consisted of two tillage practices (started from October 2015), with rotary tillage (RT) and no-till (NT). The subplots consisted of four different CSs (started from October 2017), including maize-wheat-soybean-wheat (MWSW), soybean-wheat-maize-wheat (SWMW), soybean-wheat-soybean-wheat (SW), and maize-wheat-maize-wheat (MW) (Fig. 1). Winter wheat (cv. Jimai 22) was seeded at a rate of 300 kg ha −1 . The planting densities of summer maize (cv. Longpin 208) and soybean (cv. Zhonghuang 13) were 60,000 and 250,000 plants ha −1 , respectively. In treatments with RT, soil tillage was conducted with a depth of 10 cm by rotary machine only before seeding of winter wheat, and no tillage operation was conducted before seeding of summer maize and soybean. Residue management differed between winter wheat and summer maize seasons. Wheat residues were mulched on the soil surface, while maize and soybean residues were chopped into small pieces (5-10 cm long) using a straw shredder. All plots received the same type of chemical fertilizers. For wheat, the application ratio of N/P/K as base fertilizers is 243:104:107 kg ha −1 . In addition, 75 kg N ha −1 was applied at the jointing stage. For maize and soybean, the application of N-P 2 O 5 -K 2 O were 180-60-60 and 108-60-60 kg ha −1 as base fertilizers, respectively. All crops are irrigated 75-80 mm before sowing. In addition, maize and soybean are irrigated for 70-80 mm in the growing season according to drought conditions. Detailed information on management practices were shown in previous studies ).

Soil organic carbon storage
Soil samples were obtained at 0-10, 10-20, 20-30, and 30-50 cm depths from each plot near harvesting of each crop (wheat, maize, and soybean) after a complete cycle of MWSW and SWMW in June 2020. After air-drying, soil samples were gently grinded and passed through a 2-mm sieve. The concentration of SOC was determined using a K 2 Cr 2 O 7 -H 2 SO 4 oxidation procedure (Bao 2008). To account for differences in soil bulk density, the SOC storage (Mg ha −1 ) was calculated on equivalent mass basis by using Eqs. (1) and (2) (Ellert and Bettany 1995). The SOC sequestration rate (kg CO 2 -eq ha −1 year −1 ) was determined by using Eq. (3).
where M soil, i is the soil mass per unit area in the ith layer (Mg ha −1 ). BD i is the bulk density of the ith layer (g cm −3 ). T i is the thickness of the ith layer (m). conc i is the concentration of SOC in ith layer. conc i + 1 is the concentration of SOC in i + 1th layer. M o, i is the designated equivalent mass of each layer (i.e., the maximum soil mass). i is the soil depth interval, and i = 1, 2, 3, and 4 for 0-10, 10-20, 20-30, and 30-50 cm layer, respectively. The numbers 1000 and 0.001 are unit conversion coefficients. SS2015 and SS2020 are the SOC storage at the beginning of the experiment and the SOC storage at the harvest of winter wheat in 2020, respectively, and 4 represents the duration of the experiment. The 44/12 is the coefficient that converts C to CO 2 .

Crop residues and carbon inputs from crop residues
Crops grain yield were measured in a quadrant of 2 m 2 for winter wheat and summer soybean in each plot, and double rows of 10 m for summer maize in each plot. The amount of crop residue was calculated as the ratios of grain/straw and root/straw (Zhao et al. 2018). The C input included aboveground residues and belowground roots. The C input of the aboveground residues was calculated by using Eq. (4) and that of the belowground C input by Eq. (5) (Zhao et al. 2018).
where GY represents the grain yield of each season. WC represents the water content of the grain, which is 0.125, 0.135, and 0.103 for wheat, maize, and soybean, respectively. GS ratio represents grain/straw, and wheat, maize, and soybean are 0.846, 0.955, and 0.524, respectively. RS ratio represents root/straw, and wheat, maize, and soybean are 0.2, 0.28, and 0.196, respectively. The conversion coefficient for crop biomass to C content is 0.45 (Fang et al. 2007).

Net revenue
Net revenue was calculated based on the economic value from grain minus total inputs during crop production. The prices of all grain and inputs were calculated according to the actual local conditions ). The total inputs involved farming, harvesting, fertilizer, herbicide, and irrigation. The economic profit was calculated by Eq. (6) (Ray et al. 2016).

Carbon footprints
International Organization for Standardization (ISO) has defined the CF as the sum of GHGs directly and indirectly produced by crops during a single life cycle and can be expressed in carbon dioxide equivalents (CO 2 -eq) (ISO 2013). CF consisted mainly of off-farm input production and on-farm application. The system boundary was from sowing (including seedbed preparation) to the harvesting of crops. The involved inputs, practices, or inventories are shown in Fig. 2. The N content in aboveground crop residues was calculated with Eq. (7). N 2 O emission in this study was calculated by using inputs of N fertilizer, crop residues, and emission factors. Therefore, the direct N 2 O emission was calculated by using Eq. (8) and indirect N 2 O emission by Eq. (9). The total emission of N 2 O was calculated by Eq. (10).
where F S is the N content of aboveground crop residues and roots. F N is the N fertilizer application. C si is the C inputs from the aboveground residue. C ri is the C inputs from the belowground root. i is wheat, maize, or soybean. N C(i) is the N content of different crops. σ 1 , σ 2 , and σ 3 are the emission coefficients of N inputs, volatilization of N fertilizer, and N leaching, respectively. FRAC GASF represents the fraction of N fertilizer volatilized (NH 3 and NO X -N). FRAC LEACH represents the fraction of N leaching. The value of 44/28 and 298 are the coefficients of N 2 to N 2 O and N 2 O to CO 2 -eq, respectively (Table 1).
where AI represents the amount of each item of input (Table 2). EF represents its emission coefficient (Table 1). i represents different agricultural inputs.
where ∆SOC is the mean annual C change in the 0-30 cm soil layer.
To unify the average annual crop yield for crops, the actual yield was converted into EY and MEEY. The CF per kilogram of EY (CF EY ) (kg CO 2 -eq GJ −1 year −1 ) was calculated by Eqs. (13) and (14).
where E Y (GJ ha −1 ) represents energy yield. Y g and Y s are the yield of grain and straw. E g and E s are the calorific value of yield and straw. The grain calorific values of wheat, maize, and soybean were 16.3, 16.3, and 20.9 MJ kg −1 , respectively. The straw calorific values of wheat, maize, and soybean were 14.6, 14.6, and 15.1 MJ kg −1 , respectively (Chen 2002).
where Y (kg ha −1 ) is the yield, P (US $ kg −1 ) is the price, and i is wheat and soybean. The local prices of wheat, maize, and soybean are 0.38, 0.3, and 0.88 US $ kg −1 , respectively.
where C output is the average C produced by biomass, which includes grain, straw, and root. The average C produced by straw and root was calculated in Eqs. (4) and (5), respectively. C emission was calculated by Eq. (12).

Data analysis
Normal distribution and homogeneity of variance were initially tested on the data set. Statistical analysis was conducted using SPSS software 23.0 (SPSS Inc., Chicago, IL, USA). Analysis of variance (ANOVA) was conducted to test the effects of treatments by assuming tillage and cropping sequences as fixed factors with the least significant difference (LSD) at the P < 0.05 level for means separation. Repeated ANOVA was used to test the effects of tillage practices and cropping sequences on the grain yield of wheat, maize, and soybean. The average values of the selected CFrelated parameters were standardized to show their performance through radar diagrams (Ladha et al. 2016). Origin Pro 2017 was used to design figures.

Grain yield
EY was significantly affected by tillage practices and CSs (P < 0.05, Table 3), while a non-significant interaction was observed between tillage practices and CSs. RT-MW and NT-SW were observed with the highest and lowest EY, respectively. Among different CSs under RT, the EY of MWSW, SWMW, and SW were 15.2%, 17.2%, and 19.5% lower than that of MW, respectively. Similarly, the EY of MWSW, SWMW, and MW were 25.7%, 12.8%, and 28.7% higher than that of SW under NT, respectively. Similarly, tillage and CSs were observed to affect MEEY individually (P < 0.05), but with no significant interaction (Table 3). Under RT, SW and SWMW were observed to have the highest and lowest MEEY. Under NT, the highest MEEY was observed under MWSW, which was higher than those under MWSW, SW, and MW by 9.9%, 2.6%, and 18.1%, respectively.

Net revenue
The net revenue of each CS was higher under RT than that under NT (Table 4), which is 11.6%, 20.2%, 35.4%, and 34.1% higher under RT than that under NT for MWSW, SWMW, SW, and MW, respectively (Table 4). RT-SW and NT-MW were observed with the highest and lowest net revenue, respectively. RT-SW was 35.4% higher than that under NT-SW, and NT-MW was 25.4% lower than that under RT-MW. Among RT, SW (the highest) was 17.8% higher than that under SWMW (the lowest). Among NT, MWSW (the highest) was 22.2% higher than that under MW (the lowest). Similarly, input/output economic ratios were higher under RT than that under NT for the four CSs (Table 4). However, the input/output economic ratio of MWSW was almost the same under RT and NT. RT-MW and NT-SW were observed with the highest and lowest input/output economic ratio,  respectively. The input/output economic ratio under RT-MW was 12.1%, 8.8%, and 2.8% higher than that under MWSW, SWMW, and SW, respectively. The highest input/output economic ratio under NT was observed for MWSW, which was 10.0%, 13.8%, and 6.5% higher than that under SWMW, SW, and MW, respectively.

SOC storage
Tillage and CSs were observed to affect SOC storage and SOC sequestration rate at 0-30 cm depth individually (P < 0.05, Fig. 3), but without significant interaction between tillage and CSs. The highest SOC storage and SOC sequestration rate were both observed under

Comprehensive assessment of economic and environmental benefits
Among the assessed indicators for different CSs under RT, MW has the highest GY and EY due to increased grain yield, as well as the highest CSI and ∆SOC. SW had the highest MEEY and net revenue. For NT, the highest and lowest GHG emissions were observed under SWMW and SW, respectively. Furthermore, NT-MWSW has a higher MEEY and net revenue than others. NT-MW had the highest GY, EY, ∆SOC, and CSI (Fig. 6).

Crop yields and economic profits
The data presented herein indicate that crop yields of wheat, maize, and soybean were significantly affected by tillage and CS treatments, but non-significant interactions were observed between tillage and CSs. This trend was similar to that reported by Hemmat and Eskandari (2004). However, significant interactions between tillage and crop rotation have been reported in crop yield during dry years (Ray et al. 2012;Huynh et al. 2019). Results of the present study also show that, despite differences among tillage practices, the EY of MW was higher than that of soybean involved in CSs. This trend may be due to the lower photosynthesis and also photorespiration of soybean than those of maize (Zhang and He 2020). The MEEY of CSs including soybean were higher than that of MW. Therefore, establishing a CA system based on soybean rotation could increase the MEEY, mainly because of the compensatory effects from the high price over the low crop yield of soybean. However, different results have been reported from previous studies in different regions (Uzoh et al. 2019;Sun et al. 2021). Decreases in crop yield under NT were observed in the present study. Zhang et al. (2013a) obtained the lowest yield of NT wheat among the four tillage treatments for a 3-year field study. This trend may be attributed to soil compaction caused by continuous NT, resulting in limited root growth (Kan et al. 2020c) and increased incidence of weeds (Arvidsson et al. 2014). Similarly, a previous study also reported that a lower input/output economic ratio observed under NT may be due to somewhat lower crop yield. The decline in crop yield under NT may be caused by the reduced number of mechanical operations and agricultural inputs surpassed the cost-saving compensation of adopting NT (Peng and Zhang 2006). It is generally believed that increased diversification of crop rotation systems could enhance the agronomic and environmental benefits of NT (Yang et al. 2014). However, the present study leads to the conclusion that the benefit of SW and MW was higher than that of MWSW and SWMW because of the relatively lower grain yield of soybean (Canalli et al. NT, no-till; MWSW, maize-wheat-soybean-wheat; SWMW, soybean-wheat-maizewheat; SW, soybean-wheat; MW, maize-wheat. Different lowercase letters represent significant difference at P < 0.05 between treatments 2020). However, soybean-based CA can improve the economics of cropping systems compared to maize because of its higher economic price.

SOC storage
Previous studies (combination of field experiments, laboratory incubation, and meta-analysis) suggested that with the increase in experiment duration, the mineralization of SOC decreased under NT, and the SOC could be sequestered more stably . To clearly assess the cumulative effect of the tillage practices and cropping sequences on SOC, the SOC concentrations in 2015 and 2020 were used to calculate the SOC storage. The results presented in this study showed a higher SOC storage under RT than that under NT, and the highest SOC storage and sequestration rate were observed under RT-MW. Most of the previous studies indicated an enhancement in SOC storage under conservation tillage (Snyder et al. 2009;Kan et al. 2020a), mainly due to reduced surface disturbance, increased soil cover, and enhanced input of substrates (Sindelar et al. 2015). The debate also exists if SOC is enhanced or merely redistributed within the soil profile by conservation tillage (Powlson et al. 2014). Some studies have suggested that NT increased SOC in surface soils, but RT (or plow tillage) increased SOC in deeper soils (Lal 1997;Blanco-Canqui and Lal 2008). The data from the present study verifies the results of previous studies that an increase in SOC in the subsoil is also essential to an objective assessment on SOC storage (Huggins et al. 2007). The results presented herein indicate that SOC sequestration under RT was higher than that under NT when considering storage in the subsoil. The reason for the lower SOC storage and sequestration under NT may be the reduced input of total substrates (relatively lower crop yield) from aboveground biomass input, roots, root exudates, etc.
The SOC content also differed significantly among CS treatments. Soybean-based crop rotation strongly impacted the accumulation of SOC. The highest SOC sequestration in the presented study was observed in WM. Similar results were reported in a previous study in the NCP (Shen et al. 2018). The SOC sequestration of MW was slightly higher than that of SW. Results from the present study also show that soybean-based crop rotation may reduce the SOC sequestration capacity of CA when compared to maize in NCP. However, most of the previous studies reported that the decomposition of crop residues could be accelerated by soybean compared to maize, and thus increase SOC content (Huggins et al. 2007;Schmer et al. 2020). The differences may be induced by the reduction of C input (King and Blesh 2018) and the short experimental duration (Shrestha et al. 2013). Thus, optimizing the field operations and continuous use of NT for a longer duration may offset the negative impacts of soybean-based crop rotation on SOC in the CA system.

Carbon footprints
CF is defined as the sum of GHG emissions and elimination over the whole production process expressed in CO 2 -eq based on life cycle assessment using a single impact category (ISO 2013). Different functional units (e.g., CF EY , CF MEEY , and CF E ) were used to show the trade-offs between GHG emissions and other ecosystem services (Sun et al. 2021). Liu et al. (2016) reported that reducing tillage intensity and increasing crop diversity can effectively decline  Fig. 6 Performance of selected parameters for different tillage practices and cropping sequences using radar chart. GY, grain yield; EY, energy yield; MEEY, maize equivalent economic yield; ∆SOC, the mean annual SOC change in 0-30 cm soil depth; CSI, carbon sustainability index; CF A , the CF per unit of area; CF EY , the CF per kg of energy yield; CF MEEY , the CF per kg of maize equivalent economic yield; CF E , the CF per unit of total revenue; RT, rotary tillage; NT, no-till; MWSW, maize-wheat-soybean-wheat; SWMW, soybeanwheat-maize-wheat; SW, soybean-wheat; MW, maize-wheat the CF of crop production and increase SOC sequestration. However, several studies have reported the opposite results (Lal 2003;Gregorich et al. 2005;Blanco-Canqui and Lal 2008). Among the eight treatments in the present study, the minimum and maximum values of CF were observed under RT-MW and NT-SWMW, respectively. The consistent results for the three CF indicators (CF EY , CF MEEY , and CF E ) lead to the conclusion that the CF of RT was lower than that under NT, and soybean-based crop rotation could effectively reduce the CF E in the CA system even with a relatively low yield of soybean.
Among CF inventories and components, SOC sequestration made the largest contribution to offset GHG emissions, which did not increase with the increase in crop diversity under NT. This trend was mainly because of the limitation of the short experiment duration for crop rotation. A previous study concluded that SOC dynamics along with the duration of the experiment and the robust results need a longterm assessment (Wang et al. 2021). Combining the tillage and CSs, the lowest CF was observed under RT-MW even with the highest consumption of fertilizer and fuel (i.e., the indirect GHG emission). Such a trend may be attributed to higher yield and SOC sequestration under RT-MW (Yang et al. 2017). CA with soybean-based crop rotation, as shown in the present study, may reduce farming inputs (the indirect GHG emission) and increase economic profit, but enhance the risks of decreasing in crop yield.

Limitations, implications, and perspectives
The present study assessed the trade-offs between crop productivity and environmental impacts of NT with soybeanbased crop rotation in the CA system using multiple CF indicators. Among several limitations in data collection and analysis is the fact that N 2 O emissions and the emission factor were directly and indirectly affected by N application rates and environmental factors (Castanheira and Freire 2013). N 2 O emission of different planting systems was compared in this study. However, this study did not consider the changes in N 2 O emission factors under different precipitation conditions, which may affect the accuracy of CF calculation. In addition, some other gas emissions (i.e., nitric oxide, C monoxide, and ammonia) also affected some complex atmospheric reactions (Monson and Holland 2001), and these were not calculated. This trend was mainly because the effect on the C budget was generally too small to have a significant impact on the results. Furthermore, due to the limitation of the small area study, the calculation of the manufacture, storage, and transportation of agricultural inputs was not accurate, and the price of farm products also fluctuated between years. This limitation affected the absolute values of CF under different treatments. The comparisons and relative changes among treatments were creditable. Therefore, this study provides a useful reference for the comprehensive economic and environmental assessment of CA in the NCP.
Previous studies have predicted that global demand for food will double over the next 50 years, posing significant challenges to environmental sustainability and increasing food supply without compromising environmental hazards (Tilman et al. 2002). There have been many studies on how CA improves crop yield and ecosystem services, but few studies on how legume crops will interact with other CA principles (e.g., NT or RR) on the environmental footprint. Conventional agriculture requires excessive amounts of chemicals to boost crop productivity. However, the combination of legume crops in CA can reduce the application of N fertilizer due to its symbiotic nitrogen fixation. We promoted CA (including NT, soybean-based crop rotation, and straw mulching) with the best agronomic measures to meet the environmental challenges trying to establish a suitable case.
The analysis presented herein shows that the combination of soybean-based crop rotation with NT offers great potential to improve the economic and environmental efficiency in the NCP. It will promote regional sustainable agriculture development. The present study has revealed a future potential to sustain food supply and enhance crop diversity by increasing soybean yield in the CA system with the aim to increase agricultural sustainability. Future research priorities include in-depth studies on how to realize the potential crop yield under the CA system under on farm conditions.

Conclusions
The results presented support the following conclusions: 1. Despite the risks of declining grain yield in NT (i.e., average of 14.4% lower than that under RT), the fewer inputs and the combined benefits from soybean-based crop rotation could balance the trade-offs with environmental and economic profits. 2. N 2 O emission was the main contributor to CF, indicating a high potential to reduce GHGs by optimizing fertilizer application and related field management practices. 3. In combination with soybean-based crop rotation, CA could reduce CF due to less input of N fertilizer and more N fixation. 4. The lower yield under NT partly offsets its environmental benefits. 5. The long-term economic and environmental sustainability of agricultural production in the NCP can be improved by establishing a CA system that combines straw return as mulch with suitable crop diversity (especially leguminous crops).