| Agricultural ecosystem is an important emitter of N2 O, CH4 and CO2, accounting for 60%, 50% and 10% of the global N2 O, CH4 and CO2 emissions, respectively. The rainfed farmland area accounts for 80% of the global farmland area. In China, rainfed farmland area account for more than 70% of total land area, mainly in the northern and northwestern regions. The Loess Plateau of northwest China covers more than 600,000 km2, consists of typical semiarid and arid areas with rainfed farming, and provides about 40% of the local food needs. Inherent soil fertility is low in the loess region and nitrogen(N) levels in all soils are particularly low. Nitrogen fertilization and cropping rotation are two important management measures to sustain soil productivity in the rainfed farmland on the Loess Plateau. Since 1980 s, chemical fertilizer addition has been an important measure to improve soil fertility and crop yields in the loess region. However, greenhouse gases(GHGs) emission under nitrogen fertilization and cropping rotation is not yet clear in the rainfed area.Two cropping systems of continuous spring maize and three-year rotation were designed to investigate GHGs emission at the State Key Agro-Ecological Experimental Station in the Loess Plateau(35°12′N, 107°40′E; 1220 m.a.s.l.) in Changwu County, Shaanxi Province, China. For continuous spring maize cropping system, five treatments consist of control(CK), conventional N fertilization rate(Con), optimal N fertilization rate(Opt), optimal N fertilization rate plus nitrification inhibitor(Opt+DCD), and optimal N fertilization rate with slow release urea(Opt+SR). For three-year rotation system, i.e., winter wheat, millet, pea, to explore the changes in soil respiration and Q10 values under each cropping phase. The results were summarized as follows:(1) The cumulative soil CO2 emissions were 35% for 2013, 54% for 2014 greater in N treatment than in CK treatment. Though nitrogen fertilization significantly increased the cumulative soil CO2 emissions(P<0.05), it did decrease evidently the temperature sensitivity of soil respiration(P<0.05). The Q10 values in N treatment were decreased by 27% and 17% compared with CK treatment in 2013 and 2014, respectively. Nitrogen fertilization increased significantly aboveground biomass and root biomass(P<0.05). Root biomass in N treatment was 32% and 123% greater than that in CK treatment of 2013 and 2014, respectively. Nitrogen fertilization had no marked influence on soil temperature or moisture. Root biomass was critical biotical factor for variation of soil respiration under nitrogen fertilization.(2) Compared to Con, the Opt, Opt+DCD, and Opt+SR treatments resulted in a significant decrease in annual cumulative N2 O emission, net greenhouse gas(GWP) emission, and net greenhouse gas intensity(GWPI). The greatest decrease of annual N2 O emissions(48%) occurred in Opt+DCD treatment, followed by Opt+SR(38%) and Opt(28%). N fertilization and heavy rainfall event(>40 mm) were the main factors controlling N2 O emissions. The cumulative N2 O emissions within 10 days after N fertilization accounted for 26% of annual N2 O emissions, and were positively associated with mean soil NO3-N content(P<0.05). The cumulative N2 O emissions induced by heavy rainfall accounted for 6.4% of total annual N2 O emissions in 2013 and 12.5% in 2014, respectively. The urea-derived annual mean N2 O emission factor ranged from 0.12%-0.55%. The soil acted as a small sink for atmospheric CH4. There was no significant difference in CH4 uptake among the N fertilization practices. Compared with Con treatment, GWP was decreased by 31.2%, 52.5% and 45.0% in Opt, Opt+DCD, Opt+SR treatments in 2013, and by 32.8%, 60.5% and 43.0 in 2014(P<0.05); and GWPI was decreased by 32.1%, 48.7% and 43.6% in 2013, and 25.4%, 58.7% and 39.7% in 2014, respectively. In conclusion, nitrification inhibitor was the most effective fertilization practice in the rainfed regions of Loess Plateau.(3) The three optimized N treatments, which saved 20% of N fertilization against the current conventional agricultural N fertilization rate, did not significantly decrease grain yields. The grain ranged from 9.61 to 10.46 Mg ha-1 in 2013 and from 11.41 to 12.23 Mg ha-1 in 2014 in the four nitrogen treatments. Residuals of nitrate nitrogen at the depth of 0-100 cm and 100-200 cm of five treatments ranged from 33.5 to 148.9 mg kg-1 and 24.8 to 92.8 mg kg-1, respectively. The highest accumulation of nitrate nitrogen in profile(0-200 cm) was in the Con treatment(225.9 mg kg-1), followed by 47.2%、48.5% and 45.5% of decrease in the Opt, Opt+DCD and Opt+SR treatments compared to that in the Con treatment, respectively. The residuals of nitrate nitrogen among Opt, Opt+DCD and Opt+SR treatments had no significant difference. The three optimized N managements significantly increased the agronomic efficiency of applied N and partial factor productivity from applied N compared to Con.(4) The soil respiration rate was significantly lower in the winter wheat phase(1.63 μmol m-2s-1) than the millet phase(2.40 μmol m-2s-1) and pea phase(2.21 μmol m-2s-1) from July 2010 to June 2013. However, the Q10 value was significantly higher in the wheat phase(2.76) than in the millet phase(1.85) and pea phase(1.47). The relationship between the Q10 values and soil temperature followed an exponential decay function in the rotation system, and the Q10 value remained stable(1.8) with no obvious variation when the temperature exceeded 15 °C. The Q10 value tended to increase with the increasing soil moisture and declined until the soil moisture reached a threshold of 14.7%. Our results indicated that under the condition of global warming, temperature-respiration empirical models should be parameterized according to crop types in the rotation phases, especially when estimating soil respiration of coldresistant crops. |
PDF Full Text Request