Studies On Soil Fertility And Carbon Sequestration And Mitigation Under Long-Term Fertilization In Chinese Double Rice-Cropping Systems

Food security is one of the major challenges in the modern world. However, the possible negative impacts on soil quality and ecological environment are often ignored in modern agricultural production. Therefore, research about the long-term effects of agricultural measures concerns the sustainable development of agriculture in the future. A20-year field experiment began in1990was carried out in a long-term fertilizer experimental grid at Taoyuan Agro-ecological Experimental Station. It is located at the bottom of slope in a typical hilly agricultural area in Hunan Province, China (28°55’N,111°30’E; altitude:92.2-125.3m), where cropping regime is dominated by the double rice-cropping systems. The primary objective was to examine the effect of long-term fertilization on the trends of crop yields and soil fertility, carbon sequestration in topsoil, and characteristics of methane (CH4) and nitrous oxide (N2O) emissions and ammonia (NH3) volatilization in the typical double rice-cropping systems. In this long-term fertilizer experiment, a randomized block experiment with three replicates was established, with six different fertilizer treatments. The treatments included inorganic fertilizer [nitrogen and phosphorus fertilizer (NP), nitrogen and potassium fertilizer (NK), and balanced mineral fertilizer (NPK)], combined inorganic/organic fertilizers at full and reduced rate (FOM and ROM), and no fertilizer application as a control (CK). The detailed results were summarized as follows:1. Over20years, compared with control, the grain yields with imbalanced fertilizer application (NP and NK) were increased by9.2-64.2%and13.6-30.6%in early-and late-rice, respectively, while balanced fertilizer application (NPK) increased the grain yields by90.2%and46.4%. Long-term combined inorganic/organic fertilizer (FOM and ROM) application significantly increased grain yields, which enhanced by89.6-117.4%in early-rice and47.6-53.5%in late-rice as compared to CK. The mean yields in both of early-and late-rice linearly increased with increasing of integrated fertility index. Compared with the initial soil in1990, soil fertility showed a significant downward trend without any fertilizer application (CK) in double rice-cropping systems (P<0.05). The IFI averaged0.644-0.679in the plots with long-term imbalanced fertilizer application, but there was no significant difference in comparison with the initial soil. Compared with the control as well as initial soil, in contrast, long-term balanced mineral fertilizer application and combined inorganic/organic fertilizer application significantly increased IFI,0.723-0.787, which had reached up to the high grade of soil fertility. The rice yields in NP, NPK, FOM, and ROM plots showed significant and positive time trends for early rice-cropping seasons, while there were no any trends in P-omitted plots (N and NK) both in the double rice-cropping seasons. The much lower levels of available soil P in CK and NK plots suggested that soil P deficiency is an important factor limiting increased yields in double rice-cropping systems.2. In1990-1996,1997-2005and2006-2009, the contents of soil nitrogen in the fertilization treatments were in surplus conditions (except for CK). However, the amounts of nitrogen surplus varied from year to year, mainly associated with the input of nitrogen element through inorganic/organic fertilizer and the output through uptake by rice plants in the double rice-cropping systems. The amounts of nitrogen surplus in FOM and ROM treatments were higher than NP and NPK treatments in all stages. Compared with CK, phosphorous deficit got worse towards the end of the long-term experiment without phosphorous application in NK treatment. Instead, the apparent balances of soil phosphorous in the treatments with phosphorous application (NP, NPK, FOM and ROM) were in surplus conditions. The highest amount of phosphorous surplus was in FOM treatment from1990to2005, while it was in NP2006-2009. The amount of phosphorous surplus was higher in NP than NPK in all stages, because phosphorous uptake by rice plants declined with imbalanced fertilizer application in NP. The potassium deficits in NP and CK plots showed negative time trends in these three stages, mainly because the potassium uptake was limited in these plots. The apparent balances of soil potassium varied from year to year. The amount of potassium surplus was increased with increasing potassium input in FOM treatment in all stages. The balances of soil potassium showed deficit in ROM treatment, but the amount of deficit was decreased in the three stages. The apparent balances of soil potassium showed deficit in NK and NPK treatments in1990-1996, while they were surplus and the amount gradually increased from1997to2005. In addition, the amount of potassium surplus was much higher in NK than NPK, because potassium uptake by rice plants declined with imbalanced fertilizer application in NK.3. The net ecosystem carbon balance was estimated by the changes in topsoil (0-20cm) organic carbon (SOC) density over the10-yr period1999-2009. Annual increase rate of SOC averaged0.49-0.80g C kg-1yr-1. Annual topsoil SOC sequestration rate was estimated to be0.96t C ha-1yr-1for the control and1.01-1.43t C ha-1yr-1for the fertilizer plots. The topsoil SOC density averaged36.4-48.2t C ha-1in the double rice-cropping systems in2009.4. The cumulative NH3volatilizations in the fertilizer plots ranged between12.8and27.3kg N ha-1for early-rice season and between17.3and32.7kg N ha-1for late-rice season, or accounted for9.2-33.6%and17.8-32.2%of the applied N, respectively. The NH4+concentration in floodwater, originated mainly from urea hydrolysis, is a predominant factor to NH3losses in the double rice-cropping systems. NH3volatilization rates increased nonlinearly with an increase of NH4+contents in surface water (P<0.01). Long-term fertilization significantly increased NH3volatilization, except for no difference between the control and ROM plots in early-rice season. Compared with the NPK, the cumulative NH3volatilizations tended to be increased with long-term imbalanced mineral fertilizer application in NP and NK treatments, and also can be enhanced by additional organic fertilizer amendment in the FOM plots. Nonetheless, NH3loss was declined when part of the mineral fertilizers in NPK were substituted by organic fertilizer (ROM). The increased NH3volatilization can be ascribed to the significantly decreased N uptake by rice for the NP and NK plots.5. Rice paddy field is the important source of CH4. During the early-rice growing season under continuous flooding, CH4fluxes were gradually increased in the early stage and stepped down in the late stage. The CH4fluxes dramatically ascended after late-rice transplanting in July. The highest CH4fluxes were observed on days to1-week after rice transplanting then gradually decreased to background levels in the late-rice season. Particularly, a remarkable peak of CH4flux was observed approximately1-2weeks after late-rice transplanting for the plots with combined inorganic/organic fertilizer applications. Substantial CH4emission was observed in late-rice growing season,113-157%greater than those in early-rice season (paired t-test, P<0.001). In the non-rice winter seasons, all field treatment soils acted as small net sink or source of CH4to the atmosphere, which is largely due to drainage. Annual mean CH4emissions ranged from621kg CH4ha-1for the control to1175kg CH4ha-1 for the FOM plots. Long-term inorganic fertilizer application increased CH4emissions by4.4-27.8%as compared to CK. Long-term combined inorganic/organic fertilizer application remarkably increased the CH4emissions by89.8%(FOM) and73.9%(ROM) under continuous waterlogging.6. During the double-rice growing seasons, N2O emissions from paddy fields were negligible. Substantial N2O emission was observed in the non-rice growing period, although no fertilizer was applied in the winter season. Annual N2O emission averaged1.15-4.11kg N2O-N ha-1in the double rice-cropping systems. Seasonal dynamics of N2O fluxes were insignificantly influenced by long-term fertilizer application, while annual N2O emissions were greatly affected by fertilization. Compared to CK, annual N2O emissions were increased by257%in FOM and193%in ROM under continuous waterlogging. Long-term inorganic fertilizer application increased N2O emission by68-158%in comparison with CK.7. On average, the net annual global warming potential (GWP) and greenhouse gas intensity (GHGI) were estimated12587-26066kg CO2-equivalent ha-1yr-1and1.35-2.06kg CO2-equivalents kg-1grain yield yr-1in the double rice-cropping systems. Compared with the control, inorganic fertilizer application slightly increased the net annual GWP, while they were remarkably increased by combined inorganic/organic fertilizer application due to the greatly increased annual CH4emissions during the flooded rice seasons. The GHGI was lower for the NP and NPK plots and higher for the FOM and ROM plots.First, in conclusion, the trends of crop yield were significantly affected by long-term fertilization and soil fertility. Second, the apparent balances of soil nutrients are not only decided by the practice of fertilizer management, but also influenced by the level of nutrient uptake by plants in the long-term experiment. Third, topsoil C sequestration generally increased in the double rice-cropping systems, and it can be further enhanced by additional organic fertilizers application in comparison with inorganic fertilization. Fourth, NH3volatilizations were not only affected by the rate of fertilizer application, but also influenced by fertilization method. Nitrogen uptake by plants can be enhanced by balanced fertilization and thereby NH3losses were decreased in the double rice-cropping systems. Finally, the seasonal dynamics characteristics of CH4and N2O were identical in the double rice-cropping systems. Compared to CK, long-term inorganic fertilizer application tended to increase CH4emissions during the flooded rice season and significantly increased N2O emissions from drained soils during the non-rice season. However, GHGI can be decreased by balanced inorganic fertilizer application, particularly with P fertilizer supplement. Nonetheless, net GWP and GHGI can be remarkably increased by organic fertilizer application mainly due to the greatly increased CH4emissions during the flooded rice seasons.In order to simultaneously achieve high crop productivity and low greenhouse gas intensity, we proposed some agricultural management strategies in the double rice-cropping systems, including balanced fertilizer management (particularly P fertilizer supplement), midseason drainage instead of continuous waterlogging, and crop residue incorporated into soil in the non-rice drained season rather than in the flooded rice season. However, the practical results by these strategies need yet to be proved by further experiments in Chinese double rice-cropping systems.