Article | Open

Improving rice production sustainability by reducing water demand and greenhouse gas emissions with biodegradable films

Published online:


In China, rice production is facing unprecedented challenges, including the increasing demand, looming water crisis and on-going climate change. Thus, producing more rice at lower environmental cost is required for future development, i.e., the use of less water and the production of fewer greenhouse gas (GHG) per unit of rice. Ground cover rice production systems (GCRPSs) could potentially address these concerns, although no studies have systematically and simultaneously evaluated the benefits of GCRPS regarding yields and considering water use and GHG emissions. This study reports the results of a 2-year study comparing conventional paddy and various GCRPS practices. Relative to conventional paddy, GCRPSs had greater rice yields and nitrogen use efficiencies (8.5% and 70%, respectively), required less irrigation (−64%) and resulted in less total CH4 and N2O emissions (−54%). On average, annual emission factors of N2O were 1.67% and 2.00% for conventional paddy and GCRPS, respectively. A cost-benefit analysis considering yields, GHG emissions, water demand and labor and mulching costs indicated GCRPSs are an environmentally and economically profitable technology. Furthermore, substituting the polyethylene film with a biodegradable film resulted in comparable benefits of yield and climate. Overall, GCRPSs, particularly with biodegradable films, provide a promising solution for farmers to secure or even increase yields while reducing the environmental footprint.


Agriculture is a major driver of global climate change1. The estimated source strength of agriculture is 5.1–6.1 Pg CO2-equivalents yr−1, contributing 10–12% to total global anthropogenic emissions2. Thus, producing more food to nourish a growing population while minimizing environmental costs, particularly by mitigating greenhouse gas (GHG) emissions, remains one of mankind’s greatest challenges3,4.

One of major agricultural commodities is rice, which is a major staple food for most people on earth and provides more calories for human consumption than any other cereal crop5. It is estimated that global rice production will need to increase by approximately 8–10 million Mg per year or by an annual yield of 1.2–1.5% in the coming decades to meet forecasted food needs6. Of particular concern is that increasing rice production corresponds with an increasing demand for water. Current estimates show that an average of 3000–5000 liters of water is needed for the production of one kilogram of rice, which is approximately 2–3 times greater than the water footprints of other cereal crops, such as wheat or maize7. Increasing urbanization and industrialization, however, are depleting water reserves and limiting the availability of irrigation water in many parts of the world, which particularly threatens the sustainability of irrigated rice systems8. Consequently, various studies are being conducted to explore water-saving technologies for rice production9,10,11,12. On the other hand, irrigated rice cultivation is a significant source of global CH4 emissions13. Also, high N2O emissions have been reported in fields with intermittent irrigation and midseason drainage or the excessive use or overuse of nitrogen fertilizer14,15. Recently, the IPCC16 estimated that rice production accounts for approximately 55% of the worldwide budget of GHG emissions from agricultural soils. Moreover, environmental problems associated with rice production will likely increase in the future because GHG emissions from rice fields and water scarcity (due to increasing frequency and severity of droughts) may increase as a result of climate change17,18. Consequently, rice systems are interrelated with food security, water scarcity and global climate change issues. However, currently available field scale and modeling studies have investigated these aspects of rice systems separately19,20. Thus, an integrated assessment considering water use, GHG emissions, rice productivity and the economic costs of different rice production systems as a basis for mitigation and adaptation strategies is missing.

Globally, China is one of the largest rice producing countries and is the second largest consumer of irrigation water21. Because China’s socioeconomic growth is expected to continue into the next decades, the associated increase in demand for water resources and rich foods can be reasonably projected22. Ground cover rice production systems (GCRPSs) have been developed by considering the diminishing availability of water for agriculture and the ever increasing demand for water in rice cultivation. For GCRPSs, the soil surface is covered with a thin plastic film and the soil moisture content is maintained near saturation with no standing water layer. This novel water-saving management practice has already been adopted in many provinces (more than 4 million hectares) of China23,24,25. Also, plastic mulching is commonly and increasingly used in East Asian countries other than China, such as Korea and Japan, and in Africa and the Middle East. The total area of arable land with plastic mulching is increasing annually by 15–20%26. Due to the increasing importance of GCRPSs in China, economic and environmental assessments are urgently needed. For example, GCRPSs with direct seeding have been reported to reduce water use, decrease rice yield, and increase total GHG (CH4 and N2O) emissions27,28. In contrast, several studies that have investigated GCRPSs with transplanted rice seedlings have indicated that this technology has resulted not only in irrigation water savings but has also ensured equal or greater rice yields24,29. However, the effects of these systems on N2O and CH4 emissions and economic costs have not been studied. Moreover, an obvious shortcoming of current GCRPSs is the use of a common polyethylene mulch film, which degrades extremely slowly and negatively affects the soil health and pollutes the environment. However, synthetic biodegradable polymers have become increasingly available and could provide a solution for overcoming this obstacle30,31.

In this study, we present the results from 2 years of continuous field measurements to assess water use, rice productivity and CH4 and N2O emissions in a Chinese subtropical rice-based cropping system under contrasting technology and management practices. The specific objectives of this study were to (i) determine the annual CH4 and N2O fluxes and their emission factors from different rice cultivation practices; (ii) assess the effects of GCRPSs with different mulching materials (polyethylene or biodegradable film) or different soil water statuses (near saturated or increased water stress) on the GHG emissions expressed on area- and yield-based scales; and (iii) identify a promising management option for maximizing irrigation water savings and yields while minimizing environmental impact.


Seasonal and annual CH4 fluxes

During the rice-growing seasons, the CH4 emissions were highly dependent on the water regime and soil moisture conditions (Fig. 1). In the CP system, the CH4 emissions continued to increase with rice growth, except for the midseason drainage in July, and peaked in mid-August (up to 9.0–18.5 mg C m−2 h−1) before decreasing thereafter. For the raised beds under GCRPSs, substantial CH4 emissions occurred but never exceeded 4.0 mg C m−2 h−1 and decreased to negligible when the soil was drained for harvesting. Compared with the raised beds, CH4 emissions from plant-free furrows were substantially lower (see Supplementary Table S2). Seasonal CH4 emissions varied across the rice-growing seasons and cultivation practices but were not significantly different between the N fertilized (+N) and unfertilized (−N) plots in each cultivation practice (Table 1). For the CP system, the seasonal CH4 emissions ranged from 71.3 to 85.5 kg C ha−1, averaging 80.2 kg C ha−1. Compared with the CP, the seasonal CH4 emissions were significantly reduced by 64% (P < 0.05), 73% (P < 0.05) and 81% (P < 0.05), on average, in the GCRPSsat, GCRPSbio and GCRPSlow treatments, respectively. For all GCRPS treatments, the seasonal average CH4 emissions were lower in the GCRPSlow (15.5 kg C ha−1) than in the GCRPSsat (29.1 kg C ha−1) and GCRPSbio (21.8 kg C ha−1) treatments (see Supplementary Table S2). During the fallow periods, the soil CH4 uptake prevailed in all treatments. However, the soils occasionally served as a weak source of CH4. The cumulative CH4 fluxes across the fallow periods ranged from −0.25 to −1.02 kg C ha−1, without any significant treatment effects.

Figure 1: Seasonal variations of methane (CH4) fluxes for different rice cultivation practices fertilized using two nitrogen application rates during the period of 2012–2014.
Figure 1

Vertical bars indicate the standard errors of three replicates. The legends in panels (a), (c), (e) and (g) also apply for the panels of the same row, respectively. Definitions of the treatment codes are referred to the footnotes of Table 1 and the text.

Table 1: Seasonal and annual cumulative fluxes of methane (CH4, in kg C ha−1) and nitrous oxide (N2O, in kg N ha−1) and the direct emission factor of applied nitrogen (EFd, in %) for the different rice cultivation practices fertilized using two nitrogen application rates during the rice-fallow systems of 2012–2014.

The average annual CH4 emissions from the different treatments ranged from 10.9 kg C ha−1 yr−1 (GCRPSlow+N) to 83.5 kg C ha−1 yr−1 (CP+N) (Table 1). For all GCRPS treatments, the substitution of biodegradable film for polyethylene film (GCRPSbio) or increasing water stress (GCRPSlow) did not significantly influenced the annual CH4 emissions compared to the GCRPSsat. However, plastic film mulching (GCRPSsat, GCRPSbio and GCRPSlow) reduced the average annual CH4 emissions by 73% compared to CP (P < 0.05).

Seasonal and annual N2O emissions and their direct emission factors

The N2O emissions during the rice-growing seasons varied depending on the soil water status and N application rate, with peak emissions occurring within 1-2 months following fertilizer applications (Fig. 2). Although all urea-fertilized treatments showed comparable seasonal patterns, the magnitude and duration of peak emissions were greatly affected by the different cultivation practices. During the 2012 rice-growing season, the magnitude of N2O peak emissions in the CP+N plots (473 μg N m−2 h−1) was significantly lower than the magnitude in the GCRPS+N plots (727–951 μg N m−2 h−1) (P < 0.05). For the 2013 rice-growing season, the high emission (800–1200 μg N m−2 h−1) period following fertilization lasted approximately one month for the CP+N plots and approximately two months for the GCRPS+N plots. Relative to the CP+N treatment, therefore, the GCRPS+N treatments generated significantly greater N2O emissions during both growing seasons. Similar to the CH4 emissions, the N2O emissions from plant-free furrows were lower than those from the raised beds (see Supplementary Table S2). Across the rice-growing seasons, the lowest seasonal N2O emissions, with an average of 1.47 kg N ha−1, were observed for the CP+N plots, followed by the GCRPSsat+N (2.26 kg N ha−1), GCRPSbio+N (2.64 kg N ha−1) and GCRPSlow + N (2.98 kg N ha−1) plots, respectively.

Figure 2: Seasonal variations of nitrous oxide (N2O) fluxes for different rice cultivation practices fertilized using two nitrogen application rates during the period of 2012–2014.
Figure 2

Vertical bars indicate the standard errors of three replicates. The legends in panels (a), (c), (e) and (g) also apply for the panels of the same row, respectively. The arrows indicate the fertilization dates for each annual rice-fallow system. Definitions of the treatment codes are referred to the footnotes of Table 1 and the text.

Substantial N2O emissions also occurred during the fallow periods, particularly in April 2013 and April 2014 following heavy rainfall events. Thus, the application of N fertilizer not only results in higher N2O emissions during the growing season but also shows a significant legacy effect during the fallow period. Across the fallow periods, seasonal N2O emissions ranged from 0.37 kg N ha−1 for GCRPSsat-N to 2.29 kg N ha−1 for CP+N. Combining the total N2O emissions from the rice-growing and fallow periods, the CP, GCRPSsat, GCRPSbio and GCRPSlow resulted in annual average N2O emissions of 0.96, 0.80, 0.78 and 1.27 kg N ha−1 yr−1 from the -N plots, and 3.47, 3.78, 3.40 and 4.71 kg N ha−1 yr−1 from the urea-fertilized (+N) plots. The differences between –N and +N plots were statistically significant (P < 0.05) (Table 1).

Averaged across two rice-growing seasons, the mean N2O EFd values were 1.38%, 1.63% and 1.78% for GCRPSsat, GCRPSbio and GCRPSlow, respectively, and 0.88% for CP. The seasonal N2O EFd for GCRPSs were significantly higher than those for CP (P < 0.05), while no significant differences were found among the GCRPS practices. When the emission factors were estimated based on the annual N2O emissions, the EFd varied from 1.14% to 2.83% for all rice cultivation practices, with a mean of 2.00% for GCRPS and 1.67% for CP, respectively, with no significant difference between them (Table 1).

Rice yields, nitrogen and irrigation water use efficiencies

Averaged across the 2 years and cultivation practices, the rice grain yields were significantly higher in the +N plots (6.88–7.57 Mg ha−1) than in the −N plots (5.91–6.37 Mg ha−1) (P < 0.05) (Table 2). Compared with the CP+N, the grain yields increased by 9.9% (P < 0.05), 5.4% and 10.1% (P < 0.05) on average in the GCRPSsat+N, GCRPSbio+N and GCRPSlow+N plots, respectively, with a mean value of 8.5%. Among the GCRPS+N plots, the substitution of biodegradable film for polyethylene film (GCRPSbio) or further increasing the water stress (GCRPSlow) did not significantly affected grain yield relative to the GCRPSsat.

Table 2: The characteristics of the grain (in Mg ha−1) and straw (in Mg ha−1) yields and N uptake (in kg N ha−1) of aboveground biomass (i.e., grain+straw) as well as the estimated nitrogen use efficiency (NUE, in %) for the different rice cultivation practices fertilized using two nitrogen application rates during the rice-growing seasons of 2012 and 2013.

Although the N contents in the rice plants at maturity were not significantly different among the treatments, the mean NUE across the two growing seasons was clearly higher for GCRPSs (25.5–36.8%) than CP (18.1%) (P < 0.05) (Table 2). Across the rice-growing seasons, the average irrigation water demand decreased as follows: CP (753 ± 47 mm) > GCRPSbio (360 ± 17 mm) = GCRPSsat (344 ± 28 mm) > GCRPSlow (119 ± 5 mm). The IWUE, which is an important indicator for the water use efficiency of a crop production system, was calculated by dividing the grain yield with the amount of irrigation water supplied. The values of IWUE were 0.91, 2.20, 2.01 and 6.36 kg grain m−3 water for CP+N, GCRPSsat+N, GCRPSbio+N and GCRPSlow+N, respectively. Thus, compared with the CP treatment, the GCRPS treatments used 64% less irrigation water and the IWUE improved by 286% (P  <  0.05). Among the different GCRPS treatments, the GCRPSlow treatment was ideal because it required the lowest amount of irrigation water and had the highest IWUE.

Total CH4 and N2O emissions

Similar to the area-scaled CH4 emissions, the yield-scaled growing season CH4 emissions were consistently lower for the GCRPSs than for the CP system (P < 0.05). Also, the trends and magnitudes of the effects of GCRPSs on the yield-scaled N2O emissions during the growing season relative to the CP were comparable to the effects of the area-scaled N2O emissions. Integrating CH4 and N2O emissions across growing seasons resulted in 115 to 692 kg CO2-eq Mg grain−1 season−1 (or 739 to 4672 kg CO2-eq ha−1 season−1) for all rice cultivation practices (see Supplementary Figs S6 and S7).

Averaged over the 2-year study, the annual CH4 and N2O emissions across all cultivation practices ranged from 229 to785 kg CO2-eq Mg grain−1 yr−1 (or from 1449 to 5410 kg CO2-eq ha−1 yr−1) (Fig. 3). Most emissions occurred during the rice-growing seasons, during which 65–93% of the total annual emissions occurred. Averaged across years and N application rates, the total CH4 and N2O emissions from GCRPSs compared to the CP system were reduced by 54% (P < 0.05), 60% (P < 0.05) and 59% (P < 0.05), when expressed on a grain yield basis (or by 49%, 58% and 55% when expressed on an area basis) in the GCRPSsat, GCRPSbio and GCRPSlow treatments, respectively. Among the GCRPS practices, no significant differences were observed in the annual CH4 and N2O emissions.

Figure 3: Annual area- and yield-scaled carbon dioxide (CO2) equivalents of methane (CH4) plus nitrous oxide (N2O) emissions for different rice cultivation practices fertilized using two nitrogen application rates during the period of 2012–2014.
Figure 3

Mean represents the mean values of the two years. Vertical bars indicate the standard errors of three replicates in each rice cultivation practice. The area- and yield-scaled CO2 equivalents of CH4+N2O emissions for each N application rate followed by same letter are not significant at P < 0.05. Definitions of the treatment codes are referred to the footnotes of Table 1 and the text.


Flooded rice systems are a significant source for anthropogenic GHG emissions because they emit substantial amounts of CH48,13. Consequently, rice systems have higher total GHG emissions than any other major crop system, such as wheat or maize32, allowing for substantial mitigation. Our study shows that the introduction of GCRPSs significantly reduced the annual CH4 emissions, which were dominated by emissions during the rice-growing season (Table 1). The large CH4 mitigation potential of GCRPSs is mainly driven by the improved aeration of the topsoil because the soil water statuses (70–85% WFPS) under GCRPS (see Supplementary Figs S3, S4 and S5) prevent the development of strictly reducing conditions. Consequently this reduced topsoil methanogenesis and increased the CH4 consumption by methane-oxidizing bacteria8,33. Measurements of the redox potential (Eh), which are rarely performed in conjunction with WFPS measurements at the seasonal scale4, support this interpretation because the average value was significantly higher for GCRPSs (88–210 mV) than for the CP (27 mV) (see Supplementary Figs S2, S3, S4 and S5).

An increase in the N2O emissions was observed in the GCRPSs compared to the CP across the rice-growing seasons (Table 1), which confirmed the results of previous studies27,34. As also observed in other studies35,36,37, the N2O emissions were largely influenced by the soil temperature, WFPS, and mineral N concentrations. Thus, lower N2O emissions for the CP can be explained by the strongly reductive soil conditions, which hamper the microbial oxidation of NH4+ to NO3 by nitrification. Such prevailing strong anaerobic conditions not only allow denitrification becoming substrate limited, but also support the complete reduction of oxidized mineral N compounds (NO3, NO2, NO and N2O) to the final denitrification end product N236. Potter et al.38, Dobbie et al.39, and Weller et al.8 performed model simulations or field measurements and showed that the higher N2O fluxes generally occurred when the soil water contents were ≥60% WFPS, with peak emissions often occurring in a range of 70–90% WFPS. The WFPS in the GCRPSs generally ranged from 70–85%, which provided optimal environmental soil conditions for N2O emissions. Apart from the soil water status, the higher N2O emissions in the GCRPSs also resulted from the integrative effects of soil temperature and N fertilization. For example, the higher magnitude or longer duration N2O emission peaks in the GCRPSs relative to the CP were mainly observed during the first 1–2 months following fertilizer application (i.e., the period during which the plastic film used for GCRPSs resulted in the highest increases in topsoil temperatures (+2.2–3.7 °C) compared to the CP (see Supplementary Figs S2, S3, S4 and S5). Thus, the N2O emissions in the GCRPSs were stimulated by the high mineral N availability following urea application (see Supplementary Figs S2, S3, S4 and S5) and the optimal soil water content as well as increased soil temperature. All of these factors strongly stimulated soil microbial processes such as mineralization and coupled nitrification-denitrification, fueling N2O emissions35,36. On the other land, Liu et al.25 provided evidence from soil 15N isotope profiles that GCRPSs have a high potential to reduce NH3 volatilization and nitrate leaching. These N losses are both sources of indirect N2O emissions. Based on several studies investigating effects of plastic film mulching on NH3 volatilization and nitrate leaching from paddy fields, we estimated that due to the use of plastic film NH3 emissions are reduced by approx. 38% while nitrate leaching is reduced by approx. 21%40,41. Fertilizer N losses due to NH3 volatilization for the conventional paddy systems have been reported to be in the range of 9–42%, with a mean value of 25%42. As our present fields were fertilized with 150 kg N ha−1, in GCRPSs NH3 volatilization was thus likely reduced from 37.5 to 23.3 kg NH3-N ha−1. Likewise, fertilizer N losses due to nitrate leaching for conventional rice paddies have been estimated to be in the range of 2–10%, which depended on soil properties and water percolation rates41,43. Accordingly, for our field study nitrate leaching losses ranging from 3–15 kg NO3-N ha−1 (mean: 9 kg NO3-N ha−1) can be expected for the conventional paddy, while this value would be reduced to 7.1 kg NO3-N ha−1 for the GCRPSs. Using the IPCC default value for indirect N2O emissions due to NH3 volatilization (0.010) and nitrate leaching (0.0075)44, this translates to a reduction of indirect N2O emissions for GCRPS system by (37.5*0.38)*0.010 + (9*0.21)*0.0075 = 0.16 kg N2O-N ha−1 or 35% as compared to conventional paddy system. That is, indirect N2O emissions from GCRPSs are lower, while direct N2O emissions are higher as compared to the CP. So far, however, there are only a few studies reporting on indirect N2O emissions or considering both, i.e., direct and indirect N2O emissions from crop cultivation. It is evident that approaches to quantify indirect N2O emissions should be revisited and that indirect emissions should be included in estimates of N2O losses from different agroecosystems.

It should be noted that although our seasonal N2O measurements supported previous findings (e.g., GCRPS generally increased N2O emissions during the rice-growing season), we observed no significant differences in annual N2O emissions between the GCRPS and CP treatments (Table 1). This is because the N2O emissions from CP during the fallow period off-set the increased N2O emissions from the GCRPS at the beginning of the rice-growing season. Liu et al.45 has reported that water regime during the rice-growing season plays an important role on N2O emissions during the following upland period. They also observed that while N2O emissions from flooded fields during the rice-growing season were minor, substantially higher N2O emissions were occurring during the following fallow period, which were even higher than those from fields managed with water-saving technologies. These authors explained the higher N2O emissions during the fallow period in CP with priming effects on soil organic carbon mineralization following the switch from anaerobic to aerobic conditions45. This explanation was supported by the present finding that soil respiration (CO2) in CP was generally increased in the fallow season as compared to GCRPSs (see Supplementary Fig. S8). Therefore, one can only reinforce the general necessity for measurements spanning years that include fallow periods for obtaining the representative GHG emissions.

While several studies have evaluated the effects of GCRPS on yield and environmental parameters based on short-term measurements individually, to our knowledge, this is the first study in which multiple goals (water use, GHG emissions and rice yields) have been assessed in a single study. In this study, up to 753 m3 ha−1 of irrigation water was used for the CP treatment, which is typical for irrigated rice systems in Asia5. However, our study shows that implementing GCRPSs not only reduces irrigation water demand by 52–84%, but also increases the average rice yield by 8.5% compared with the CP system. Consequently, the IWUE nearly doubled (from 0.91 kg grain m−3 to 2.01–6.36 kg grain m−3) from the CP to the GCRPSs. Also, the mean NUE across the rice-growing seasons increased from 18.1% under CP to 25.5–36.8% under the GCRPSs (Table 2). The increased yields that occurred in the GCRPSs can be explained by the following factors. a) Long-term flooding of paddy soils can result in high concentrations of toxic reduction products, such as Fe2+, H2S and organic compounds, which seriously affect root growth46,47. Irrespective of N fertilization, improved aeration and increased soil Eh result in greater root biomass at soil depths of 0–10 cm and 10–20 cm and considerable root biomass at soil depths of 20–40 cm in GCRPSs relative to CP25. b) In accordance with the conclusions from previous studies47,48,49, greater root biomass and deeper rooting depths improve crop nutrient acquisition. c) Higher soil temperatures at the beginning of the growing season support crop development. d) Environmental nutrient losses due to leaching or NH3 volatilization are reduced in GCRPSs compared to CP25, increasing the availability of N for the crops and resulting in higher NUEs. This observation is an important finding and could be more attractive in adapting to water and food shortages due to climate change and population growth18,50.

In this study, the area and yield-scaled annual GHG (CH4 and N2O) emissions ranged from 1449 to 5410 kg CO2-eq ha−1 yr−1 and from 229 to 785 kg CO2-eq Mg grain−1 yr−1, respectively (Fig. 3), which fell within the range reported by Adviento-Borbe et al.51 for rice-fallow systems in the USA (658–7126 kg CO2-eq ha−1 yr−1 and 91–874 kg CO2-eq Mg grain−1 yr−1). As relatively low CH4 and N2O emissions can be obtained from rice systems through careful water management, the area and yield-scaled annual GHG emissions expressed in CO2 equivalents could be reduced by 54% and 58% on average, respectively, by converting rice paddies from CP into GCRPSs. However, it is noteworthy that in view of the longer atmospheric lifetime of N2O (121 years) as compared to CH4 (12.4 years) and the detrimental effect of N2O on the stratospheric O3 layer16, the current trend of increased N2O emissions from GCRPSs, specifically for GCRPSlow should be considered while developing mitigation strategies, although GCRPSs do reduce CH4 emissions substantially. On the other hand, a full evaluation of the mitigation potential of GCRPS must consider its influence on soil carbon stocks and effluxes. Although, in general, soil conditions with higher aerobic status and increased soil temperature would have stimulated organic matter mineralization and consequently decreasing soil organic C stocks, our recent and thorough regional scale evaluations have shown that the conversion from CP to GCRPS results in greater soil organic carbon (SOC) concentrations and storage25. This is mainly because GCRPS practices increase the above- and below-ground carbon inputs and improve the physical protection of soil organic matter against microbial degradation25. Overall, these results represent a win-win situation for agronomic and environmental goals because lower water use and GHG emissions can be obtained without compromising rice yields.

Further cost-benefit analysis suggests that GCRPS is a viable option for rice production from both environmental and economic points of view because of its monetary benefits associated with the GHG mitigation and because the water saving and yield increase outweigh the additional costs associated with labor, manufacture and purchasing the film (see Supplementary Table S3). Based on our economic assessment, the net benefits of implementing GCRPSsat, GCRPSbio or GCRPSlow instead of using CP is in the range of $39.8–251.5 ha−1 yr−1, which is in order of GCRPSlow > GCRPSsat > GCRPSbio (see Supplementary Table S3). However, one major drawback of GCRPSsat or GCRPSlow is their disposal, which could result in soil contamination and environmental pollution26. Although farmers manually collect the polyethylene plastic films in the GCRPSsat after harvesting the rice plants, significant amounts of the plastic film remain in the field and accumulate in the soil profile, which creates a serious environmental problem30. Inspiringly, this study shows that non-decomposable polyethylene film can be successfully replaced with the biodegradable film Ecoflex® (GCRPSbio). When using this type of film material, the positive effects of GCRPSbio include the following: a) the mitigation of soil GHG emissions, b) a strong reduction in the demand for irrigation water and c) the stimulating effect on rice yields remain comparable to that observed for GCRPSsat (Tables 1 and 2). Although revenues were higher for GCRPSsat and GCRPSlow than for GCRPSbio, we recommend using GCRPSbio because it avoids soil and landscape pollution resulting from the use of polyethylene plastic films30. This result is a novel finding that has not been reported in previous studies of GCRPS and will hopefully encourage the wider use of biodegradable films for rice production and other crop production systems. Due to the relatively high price for biodegradable films, however, governments should conceive effective policies and incentives such as providing subsidies for farmers to introduce the GCRPSbio technology and management practice.


Study site and field treatments

From 2012–2014, field experiments were conducted on a farm (32°07′13″ N, 110°43′04″ E, 440 m above sea level) located at the Agricultural Bureau in Fangxian County, northwest of Hubei Province, China. The climate at the site is defined as northern subtropical monsoon52. The study region is a typical mountainous agricultural area, with one crop harvest per year. Paddy rice is the dominant crop, and fields remain fallow during the winter period. The topsoil has a silt loam texture and a pH (in water) of 6.0, and more soil properties were shown in the Supplementary Table S1. The daily precipitation and average air temperature during the experimental period are shown in Supplementary Fig. S1.

To assess management opportunities for reducing GHG emissions and water use while optimizing rice grain yield, four common rice cultivation practices and two nitrogen fertilizer application rates were tested. The resulting eight treatments were replicated three times using a randomized complete block design, i.e., the number of experimental plots was 24. Each plot was 9.0 m wide × 10.0 m long, and adjacent plots were separated by concrete ridges (40 cm width) and an impermeable film that was inserted into the soil to a depth of 90 cm.

The following four rice cultivation practices were used:

(1) Conventional paddy (CP), which is the local traditional paddy rice production system. In this system, plots were flooded between seedling transplantation until midseason, when the fields were drained for approximately 7 days. Following this period, fields were flooded again until they were drained approximately three weeks before harvesting the rice.

(2) Ground cover rice production system under nearly saturated soil water content conditions (GCRPSsat). These plots were separated into five raised beds (1.56 m width × 9.4 m length) surrounded by furrows (0.20 m width × 0.15 m depth) that were filled with water. However, no standing water was allowed in the raised beds. The raised beds were covered with regular polyethylene plastic film (1.70 m width and 0.005 mm thickness) with holes to allow for transplanting the rice seedlings.

(3) Ground cover rice production system with biodegradable films (GCRPSbio). In this treatment, water management was comparable to that described for (2) GCRPSsat, but Ecoflex® (BASF, Germany) biodegradable film, which can be metabolized by soil microorganisms and is almost completely decomposed within a growing season, was used31.

(4) Ground cover rice production system with regular polyethylene film under lower soil water content conditions (GCRPSlow) compared to (2) GCRPSsat. In this treatment, water management was identical to that described for (2) GCRPSsat until the rice-regreening stage, which occurs approximately two weeks following transplanting. Following this initial period of near saturation, the soil water content was reduced to approximately 80% of that of the GCRPSsat treatment by monitoring the soil moisture content.

For each rice cultivation practice, two nitrogen application rates were examined: (a) urea applied once before rice transplanting at a common rate of 150 kg N ha−1 (+N), and (b) no synthetic nitrogen fertilizer application (−N). To ensure that neither phosphate (P) nor potassium (K) limited crop growth, all plots received basal fertilization at application rates of 45 kg P2O5 ha−1 and 45 kg K2O ha−1.

The hybrid rice variety Yixiang 3728, which is a cultivar typically grown in the study region, was used for all of the tested rice cultivation practices. In 2012–2014, rice seedlings were transplanted on May 8, 2012, and April 28, 2013, and harvested on September 16, 2012, and September 10, 2013, respectively. After harvest, rice straw was completely removed and all fields were kept fallow in the winter season, which was in agreement with local practice.

Measurements of CH4 and N2O fluxes

The fluxes of CH4 and N2O from the paddy rice-fallow systems were measured using the static vented chamber-based technique21. To account for the effects of micro-topography for plots managed as GCRPSs, two sizes of stainless steel frames, 65 cm × 90 cm × 15 cm and 20 cm × 30 cm × 20 cm (width × length × height), were inserted into the raised bed (accounting for 87% of the total area) and furrow (accounting for 13% of the total area) soils of each plot, respectively. For CP treatment, only one type of frame with dimensions of 65 cm × 90 cm × 15 cm was used in each replicated plot. The frames, that were positioned at least 1.5 m from the edges of the plots, were inserted into the soil to a depth of 15 cm, i.e., nearly reaching the compact plough pan layer. Some small holes were drilled in the frame below the soil line to allow for lateral water movement and root growth. Board walks were used to access the chambers and prevent soil disturbance. The planting density of rice crops between the inside and outside of the frame was similar. During the winter, all plots were drained and remained fallow, and the large chamber frames (i.e., 65 cm × 90 cm × 15 cm) remained in place to obtain flux measurements. The insulated chambers based on the type of frames (i.e., an area of 65 cm × 90 cm and a height of 100 cm and a 20 cm wide × 30 cm long × 30 cm tall) were used for gas sampling. For these chambers, two circulating fans were installed inside of the chamber headspace to facilitate mixing of chamber air and thus inhibiting the formation of gas concentration gradients. Also, a hole of 2 cm diameter was fitted in the top panel of sampling chambers, which could be left open when placing the chamber on the frame to prevent the build-up of over pressure within the chamber. Once the chamber was in place, this hole was connected to a pressure balance tube21.

Gas flux measurements were conducted three times per week during the experimental periods between 09:00 am and 11:00 am. Gas samples (40 ml) were taken from the chamber headspace at equal time intervals of 0, 10, 20, 30 and 40 min after covering by using polypropylene syringes fitted with three-way stopcocks. These samples were all analyzed within 6 hr of sampling by using a gas chromatograph (GC, Agilent 7890 A, Agilent Technologies, Santa Clara, CA, USA) equipped with a flame ionization detector for detecting CH4 at 200 °C and an electron capture detector for detecting N2O at 330 °C21. The CH4 and N2O fluxes were determined using linear or nonlinear regressions of gas concentrations versus the chamber closure time, as described in detail by Wang et al.53.

Auxiliary measurements

During the experimental periods, we also measured the amounts of irrigation water, soil redox potentials (Eh), floodwater depths, soil volumetric water content, soil temperature, soil ammonium (NH4+) and nitrate (NO3) concentrations, and aboveground biomasses (see Supplementary Information).

Data processing and statistical analyses

The data were further processed for calculating total CH4 and N2O emissions, direct emission factor (EFd) of applied-N, irrigation water use efficiency (IWUE) and the fertilizer N-use efficiency (NUE) (see Supplementary Information).

To determine differences in GHG (CH4 and N2O) emissions among treatments during different observation periods (e.g., growing season) in the randomized complete block design, the SPSS 19.0 software (SPSS China, Beijing, China) was used with least significant difference tests with a P-value < 0.05. The cumulative CH4 and N2O emissions, their CO2 equivalents and grain yields due to main effects, such as rice cultivation practices, N fertilizer rates, year, blocking, rice cultivation practice × N fertilizer rate, year × rice cultivation practice × N fertilizer rate and block × rice cultivation practice as a random effect, were analyzed using Linear Mixed Models. The repeated measures ANOVA was used to test the effects of treatment on GHG emissions and environmental variables at a given period (e.g., peak emission period, growth stage).

Additional Information

How to cite this article: Yao, Z. et al. Improving rice production sustainability by reducing water demand and greenhouse gas emissions with biodegradable films. Sci. Rep. 7, 39855; doi: 10.1038/srep39855 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    , & Non-CO2 greenhouse gases and climate change. Nature. 476, 43–50 (2011).

  2. 2.

    et al. Agriculture. In:, Climate Change 2007: Mitigation. Contribution of work group III to the fourth assessment report of the Intergovernmental Panel on Climate Change (eds , , , , ), Cambridge, UK and New York, NY, USA, pp. 497–540 (Cambridge University Press 2007).

  3. 3.

    , , & More food, low pollution (Mo Fo Lo Po): a grand challenge for the 21st century. J. Environ. Qual. 44, 305–311 (2015).

  4. 4.

    et al. Reducing greenhouse gas emissions, water use, and grain arsenic levels in rice systems. Global Change Biol. 21, 407–417 (2015).

  5. 5.

    , , & Rice and water. Adv. Agron. 92, 187–237 (2007).

  6. 6.

    , , & Crops that feed the world 7: rice. Food Secur. 4, 7–24 (2012).

  7. 7.

    et al. eds Water-wise rice production. Proceedings of the International Workshop on Water-wise Rice Production 8-11 April 2002. Los Baños, Philippines. Los Baños (Philippines), International Rice Research Institute, pp. 356 (2002).

  8. 8.

    et al. Greenhouse gas emissions and global warming potential of traditional and diversified tropical rice rotation systems. Global Change Biol. 22, 432–448 (2016).

  9. 9.

    & Field water management to save water and increase its productivity in irrigated lowland rice. Agr. Water Manage. 49, 11–30 (2001).

  10. 10.

    , , , & Water-saving and high-yielding irrigation for lowland rice by controlling limiting values of soil water potential. J. Integr. Plant Biol. 49, 1445–1454 (2007).

  11. 11.

    , , , & Seasonal variations of CH4 and N2O emissions in response to water management of paddy fields located in Southeast China. Chemosphere. 89, 884–892 (2012).

  12. 12.

    et al. More rice with less water-evaluation of yield and resource use efficiency in ground cover rice production system with transplanting. Eur. J. Agron. 68, 13–21 (2015).

  13. 13.

    , , & Global estimations of the inventory and mitigation potential of methane emissions from rice cultivation conducted using the 2006 Intergovernmental Panel on Climate Change Guidelines. Global Biogeochem. Cycles. 23, GB2002 (2009).

  14. 14.

    et al. Methane and nitrous oxide emissions from rice paddy fields as affected by nitrogen fertilisers and water management. Plant Soil. 196, 7–14 (1997).

  15. 15.

    et al. Greenhouse gas fluxes and NO release from a Chinese subtropical rice-winter wheat rotation system under nitrogen fertilizer management. J. Geophys. Res: Biogeosciences. 118, 623–638 (2013).

  16. 16.

    Intergovernmental Panel on Climate Change (IPCC). In , , , , , , , , , (Eds) Climate Change 2013: the Phycial Science Basis, Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA, p. 1535 (Cambridge University Press, 2013).

  17. 17.

    , & Increased greenhouse-gas intensity of rice production under future atmospheric conditions. Nature Clim. Change. 3, 288–291 (2013).

  18. 18.

    et al. Multimodel assessment of water scarcity under climate change. Proc. Natl. Acad. Sci. USA. 111, 3245–3250 (2014).

  19. 19.

    et al. Net annual global warming potential and greenhouse gas intensity in Chinese double rice-cropping systems: a 3-year field measurement in long-term fertilizer experiments. Global Change Biol. 17, 2196–2210 (2011).

  20. 20.

    et al. Simulating the impact of water saving irrigation and conservation agriculture practices for rice-wheat systems in the irrigated semi-arid drylands of Central Asia. Agr. Forest Meteorol. 214-215, 266–280 (2015).

  21. 21.

    et al. Water-saving ground cover rice production system reduces net greenhouse gas fluxes in an annual rice-based cropping system. Biogeosciences. 11, 6221–6236 (2014).

  22. 22.

    , , , & Balancing water resource conservation and food security in China. Proc. Natl. Acad. Sci. U.S.A. 112, 4588–4593 (2015).

  23. 23.

    et al. Saving water with Ground Cover Rice Production Systems (GCRPS) at the price of increased greenhouse gas emissions? In Water-Wise Rice Production (eds , , , , , ), International Rice Research Institute, Los Baños, pp. 197–206 (Philippines, 2002).

  24. 24.

    et al. Ground cover rice production systems are more adaptable in cold regions with high content of soil organic matter. Field Crop. Res. 164, 74–81 (2014).

  25. 25.

    et al. Ground cover rice production systems increase soil carbon and nitrogen stocks at regional scale. Biogeosciences. 12, 4831–4840 (2015).

  26. 26.

    , , & Plastic mulching in agriculture-Friend or foe of N2O emissions ? Agric. Ecosyst. Environ. 167, 43–51 (2013).

  27. 27.

    et al. Sattelmacher, B. Fluxes of methane and nitrous oxide in water-saving rice production in north China. Nutr. Cycling Agroecosyst. 77, 293–304 (2007).

  28. 28.

    et al. Effects of soil water content on growth, tillering, and manganese uptake of lowland rice growth in the water saving ground cover rice production system (GCRPS). J. Plant Nutr. Soil Sci. 170, 7–13 (2007).

  29. 29.

    et al. Ground cover rice production system increases yield and nitrogen recovery efficiency. Agron. J. 104, 1399–1407 (2012).

  30. 30.

    & Biodegradation of agricultural plastic films: a critical review. J. Polym. Environ. 15, 125–150 (2007).

  31. 31.

    , , & Ecoflex® and Ecovio®: Biodegradable, performance-enabling plastics. Adv. Polym. Sci. 245, 91–136 (2012).

  32. 32.

    , , , & An agronomic assessment of greenhouse gas emissions from major cereal crops. Global Change Biol. 18, 194–209 (2012).

  33. 33.

    & The effect of water management based on soil redox potential on methane emission from two kinds of paddy soil in Japan. Agric. Ecosyst. Environ. 107, 397–407 (2005).

  34. 34.

    , , , & Effect of soil water status and mulching on N2O and CH4 emission from lowland rice field in China. Biol. Fertil. Soils. 39, 215–217 (2004).

  35. 35.

    et al. Exchange of greenhouse gases between soil and atmosphere: interactions of soil physical factors and biological processes. Eur. J. Soil Sci. 54, 779–791 (2003).

  36. 36.

    , , , & Nitrous oxide emissions from soils: how well do we understand the processes and their controls? Phil. Trans. R. Soc. B. 368, 20130122 (2013).

  37. 37.

    , , & Characteristics of annual nitrous and nitric oxide emissions from major cereal crops in the North China Plain under alternative fertilizer management. Agric. Ecosyst. Environ. 207, 67–78 (2015).

  38. 38.

    , , & Process modeling of controls on nitrogen trace gas emissions from soils world-wide. J. Geophys. Res. 101, 1361–1377 (1996).

  39. 39.

    , & Nitrous oxide emissions from intensive agricultural systems: variations between crops and seasons, key driving variables, and mean emission factors. J. Geophys. Res. 104, 26891–26899 (1999).

  40. 40.

    & Model estimation of ammonia volatilization from a paddy rice field after application of a surface film-forming material. J. Agr. Sci. 148, 95–100 (2010).

  41. 41.

    , , , & Effect of plastic-film mulching on leaching of nitrate nitrogen in an upland field converted from paddy. Paddy Water Environ. 2, 67–72 (2004).

  42. 42.

    et al. Polyolefin-coated urea decreases ammonia volatilization in a double rice system of Southern China. Agron. J. 105, 277–284 (2013).

  43. 43.

    , , , & Nitrogen and phosphorus leaching losses from paddy fields with different water and nitrogen managements. Paddy Water Environ. 9, 333–342 (2011).

  44. 44.

    Intergovernmental Panel on Climate Change (IPCC). IPCC Guidelines for national greenhouse gas inventories. Hayama, Institute for Global Environmental Strategies (2006).

  45. 45.

    , , & Effects of water regime during rice-growing season on annual direct N2O emission in a paddy rice-winter wheat rotation system in southeast China. Sci. Total Environ. 408, 906–913 (2010).

  46. 46.

    , & Yield formation in rice in respond to drainage and nitrogen application. Field Crop. Res. 51, 65–82 (1997).

  47. 47.

    et al. Alternate wetting and moderate drying increases rice yield and reduces methane emission in paddy field with wheat straw residue incorporation. Food Ener. Secur. 4, 238–254 (2015).

  48. 48.

    , & Root based approaches to improving nitrogen use efficiency in plants. Plant Cell Environ. 32, 1272–1283 (2009).

  49. 49.

    Steep, cheap and deep: an ideotype to optimize water and N acquisition by maize root systems. Ann. Bot. 112, 347–357 (2013).

  50. 50.

    , , & Global water resources: vulnerability from climate change and population growth. Science. 289, 284–288 (2000).

  51. 51.

    et al. Optimal fertilizer nitrogen rates and yield-scaled global warming potential in drill seeded rice. J. Environ. Qual. 42, 1623–1634 (2013).

  52. 52.

    , , , & Organically fertilized tea plantation stimulates N2O emissions and lowers NO fluxes in subtropical China. Biogeosciences. 12, 5915–5928 (2015).

  53. 53.

    et al. Comparison between static chamber and tunable diode laser-based eddy covariance techniques for measuring nitrous oxide fluxes from a cotton field. Agr. Forest Meteorol. 171–172, 9–19 (2013).

Download references


This work was funded by the National Nature Science Foundation of China (51139006, 41305129 and 41321064).

Author information


  1. State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, P.R. China

    • Zhisheng Yao
    • , Xunhua Zheng
    •  & Chunyan Liu
  2. Institute for Meteorology and Climate Research, Atmospheric Environmental Research, Karlsruhe Institute of Technology, D-82467 Garmisch-Partenkirchen, Germany

    • Zhisheng Yao
    •  & Klaus Butterbach-Bahl
  3. College of Earth Science, University of Chinese Academy of Sciences, Beijing 100049, P.R. China

    • Xunhua Zheng
  4. College of Resource and Environmental Science, China Agricultural University, Beijing 100193, P.R. China

    • Shan Lin
    •  & Qiang Zuo


  1. Search for Zhisheng Yao in:

  2. Search for Xunhua Zheng in:

  3. Search for Chunyan Liu in:

  4. Search for Shan Lin in:

  5. Search for Qiang Zuo in:

  6. Search for Klaus Butterbach-Bahl in:


Z.S.Y., X.H.Z., S.L. and Q.Z. designed the experiments. Z.S.Y carried out the experiments and performed the analyses. Z.S.Y., C.Y.L. and K.B substantially contributed to interpreting the results and writing the paper.

Competing interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to Zhisheng Yao.

Supplementary information


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Creative CommonsThis work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit