The Temporal And Spatial Variability And Emission Flux Of Soluble Nitrous Oxide And Methane In Shanghai River Network

Nitrous oxide (N2O) and methane (CH4) are important greenhouse gases, which directly or indirectly influence future climate change. The inventories of global N2O and CH4, however, still left significant uncertainties. With the process of urbanization in the whole global dimension, human being's living and development extremely affect global carbon and nitrogen cyclings, causing the exponential increases of carbon and nitrogen loads in the river ecosystem and eutrophication and black-odor of river water. Consequently, the potential to produce and emit N2O and CH4 of river ecosystem has been stimulated.This paper was supported by the project of Nitrous oxide production mechanism and emission flux at Yangtze delta river net, funded by State Natural Science Fund and National Water Project:The research and demonstration of the application of extraneous sources interdiction, engineering restoration, and in-situ multilevel ecological purification technology in urban black-odorous river. The whole river network of Shanghai was selected as the research object and multi-disciplinary synthetic study methods were used in this study, including environmental geochemistry, biogeochemistry, physical geography and so on. Based on field samplings, laboratory analysis, in situ measurements and incubation experiments, this paper was aimed to systematically study the spatial and temporal variabilities of solube N2O and CH4 concentrations and saturations, gas production and exchange across sediment-water inferface, the effects of primary environmental factors on gas concentrations and the preliminary calculation of N2O and CH4 total emission across air-water interface of Shanghai river network. Main conclusions were as followings.(1) The total variation ranges of N2O and CH4 concentrations in river water were 0.26±0.01～89.91±45.50μg N·L-1 and 0.74±0.64～1367.68±186.69μg C·L respectively. The total variation ranges of N2O and CH4 saturations in river water were 108.74±5.75～21355.11±10808.41% and 95.95±49.70～64704.78±13886.65%, and the surface water had a overall high saturation; In February and April, the N2O, CH4 concentrations and saturations were significantly higher than those in other months; the highest concentrations and saturations of N2O and CH4 in river network appeared at downtown area and industrial area, like SQ, BS and JD.(2) N2O and CH4 concentrations in river water were significantly correlated with each other; Both N2O and CH4 concentrations had significant negative correlations with water temperature, air temperature, pH, ORP and DO respectively, and had significant positive correlations with NH4+-N, TP, DOC and DIC respectively; N2O concentration had a uniform positive correlation under different DO concentration, and this positive correlation could be described by regression formula:y 0.713015*e0298754*x (R2=0.430, P<0.001). However, the rapid increase of N2O concentration with N03--N could only be observed under low DO concentration ranging from 0～1.2 mg·L(3) The results of CART model indicated that NH4+-N, water temperature, DO, N03--N, TP and pH were primary environmental that could be used to classify N2O concentration in river water, and NH4+-N, N03--N, DO, ORP and SO42-correspondently were the classification factors of CH4 concentrations. CART model functioned ideally under the poor-quality water condition and high N2O, CH4 concentrations to classify N2O and CH4. GAM model could explain 72.31% of total variation of N2O concentration in which NH4+-N was attribute to 42.99% so that primary environmental factors could be used by GAM model to well fit N2O concentration in river water; However, GAM model could only explain 51.20% of total variation of CH4 concentration which unable GAM model to exactly fit or predict CH4 in river water. Generally, GAM model was effective under the good-quality water condition.(4) N2O production rate in water body of QP and SQ were 0.10±0.67 ng N·h-1 and 0.97±0.65 ng N·h-1, respectively, and were significantly lower than N2O emission rate across sediment-water interface, which were 3.86±1.93 ng N·h-1 and 12.49±6.25 ng N·h-1. CH4 emission rate across sediment-water interface of QP and SQ were 0.079±0.0084μg C·h-1 and 7.04±0.48μg C·h-1, respectively. The oxidation effect of oxygenic overlying water made the CH4 production rate extremely low. Sediments were main sources of N2O and CH4 in overlying water of QP and SQ. The emission flux of N2O and CH4 in QP and SQ were 4.02±2.01μg N·m-2·h-1 and 13.01±6.51μg N·m-2·h-1, and 83.66±11.22μg C·m-2·h-1 and 7335.48±495.79μg C·m-2·h-1, respectively. The significant differences of N2O and CH4 emission flux across sediment-water interface were one of the primary reasons that could explain spatial variations of N2O and CH4 concentrations in river network.(5) Oxygen comsuption of both QP and SQ were higher under oxygen-rich condition than those under oxygen-poor condition; because of the high organic matter concentration, oxygen comsuption of SQ sediment were higher than those of QP sediment under both oxygen-rich and oxygen-poor conditions, which made the overlying water in persistant oxygen-deficit condition; under the condition of NH4+-N or NH4+-N/NO3--N addition, oxygen comsuption of both QP and SQ sediment increased evidently, and also increased with addition gradient; the acceleration of nitrification process in surface sediment layer caused by NH4+-N diffusion from overlying water might be the main reason to explain the increase of oxygen comsuption across the sediment-water interface.(6) Under oxygen-poor condition, both N2O emission fluxes across sediment-water interface of QP and SQ were higher than those under oxygen-rich condition; NH4+-N and NH4+-N/NO3--N addition significantly increased N2O emission fluxes across sediment-water interface of QP and SQ sediments. DO and NH4+-N concentration in overlying water were most important factors that could influence N2O emission across sediment-water interface; DO was the only factor that could affect CH4 emission fluxes across sediment-water interface of QP and SQ sediments.(7) Under natural condition N2O concentrations of sediment core water in top 0-1 cm of QP and SQ sediments ranged from 1.02±0.64～2.04±1.45μg N·L-1 and 2.23±0.85～3.40±1.07μg N·L-1, and both of them were higher than N2O concentrations in QP and SQ surface water; Oxygen-poor condition, NH4+-N addition and NH4+-N/NO3--N addition significantly stimulated N2O production in top sediment; Under high NH4+-N and low DO condition, nitrification processing at the surface layer of sediment was the main mechanism for producing N2O. Under natural condition CH4 concentrations of sediment core water in top 0～1 cm of QP and SQ sediments ranged from 49.48±12.94～54.14±21.48μg N·L-1 and 6941.25±2647.55～7465.17±1234.68μg C·L-1; oxidation effect in top 0～1 cm of sediment significantly decreased CH4 concentration in sediment core water; the extremely high organic matter concentration of QP and SQ sediments disabled the inhibiting effect of sulfate-reduction and nitrate-reduction on CH4 formation, and made sediments the main sources of CH4 in overlying water.(8) N2O and CH4 emission fluxes across water-air interface of river network ranged from 0.05±0.003～48.12±6.42 mg N·h·m-2 and -0.04±0.01～579.25±86.20 mg N·h-1·m-2, respectively. Generally, river network is the source of atmospheric N2O and CH4; High water temperature promoted N2O and CH4 emission; N2O and CH4 emission fluxes in urban and industrial areas like SQ, BS, JD were significantly higher; N2O emission flux of river network was slightly lower than that in sea water of Yangtze estuary, and CH4 emission flux of river network was equivalent to those of presentative wetland ecosystems.(9) The year total N2O and CH4 emissions of Shanghai river network were 7.23 Gg N·yr-1 and 58.22 Gg C·yr-1, respectively; year total N2O emission accounted for 0.66% of the global N2O emission from river ecosystem,indicating the inaccuracy in current estimation on global N2O emission from river ecosystem; year total CH4 emission of river network accounted for 6.62% of CH4 emission from freshwater wetland nationwide,corrobating the importance of river ecosystem as the source of atomsperic CH4; the river network in urbanized area was the potential important source of atomsperic N2O and CH4, and deserved more concerns and attentions.