Agriculture is the source of a number of gases of nitrogen (N) that are of considerable interest due to their importance in atmospheric pollution and global warming (Harrison and Webb, 2001) Ammonia (NH3) is the only significant gas-phase alkaline species in the atmosphere and it is critical in the formation of a variety of aerosols, which represent a significant contribution to regional haze (Wu et al., 2008). NH3 deposition also contributes to large scale N eutrophication and acidification of ecosystems (van Breemen et al., 1982; Fangmeier et al., 1994; van der Eerden et al., 1998). While the main source of NH3 input to the atmosphere is the volatilization from animal wastes (Kirchmann et al, 1998), croplands contribute about 23% of the global annual emission (Bouwman et al., 1997) and are among the most uncertain components of the NH3 budget (Battye et al., 2003). On the other hand, nitrous oxide (N2O) is the third most significant greenhouse gas and nearly 310 times more potent than carbon dioxide in global warming potential (Forster et al., 2007). Natural sources of N2O are soils and oceans, and the anthropogenic increase is mainly caused by accelerated emissions from agricultural soils, which account for near 25% of the global annual emission (Ehhalt et al., 2001).
NH3 emissions from croplands are related mainly to ammoniacal fertilizer and manure applications (Freney et al., 1983). While soil-derived NH3 has been long recognized, exchange of NH3 in plant canopies has also been reported (Sutton et al., 1993b, 1995, 1998). N2O emissions from arable lands are mainly produced by microbial-mediated nitrification and denitrification processes (Bremner, 1997). Consequently, soil N2O production rates depend on the interaction of soil physical properties, meteorological conditions, carbon (C) and N availability, and agricultural management (Davidson, 1991).
At present, several studies have been conducted to assess NH3 and N2O emissions from different combinations of soil, crop cover, and agricultural management (Lemon and van Houtte, 1980; Sutton et al., 1993a, 2000; Dobbie et al., 1999; Goossens et al., 2001; Rochette et al., 2004; Dusenbury et al., 2008). However, much of the emissions represent specific conditions that can hardly be extrapolated or, simply, measurements contain a high degree of variation in both space and time that cannot be relied on. In fact, it is the high temporal and spatial variability of the soil water, N and C, which pose a great challenge for measuring and modeling N trace gases. This variability stems from changing environmental and nutrient availability conditions, soil anisotropy, and the scale at which complex biological, chemical and physical processes are taking place in the soil (Grant et al., 2006; Chen et al., 2008). For example, spatial variability in N2O typically causes coefficient of variation from 70 to 380% to be measured at a spatial scale of several meters within field plots (Folorunso and Rolston, 1984; Mathieu et al., 2006), while temporal variability in N2O is also large with coefficients of variation greater than 150% (Grant and Pattey, 2003).
Gaseous N fluxes are usually measured by surface chambers or micrometeorological techniques. Usually, measurements using surface chambers are well adapted to resolving topographic and treatments effects. However, their low temporal resolution and small spatial scales are of limited value for long-term estimates (Grant and Pattey, 2003). On the contrary, measurements using micrometeorological techniques have a high temporal resolution but a low spatial resolution in the field. Therefore, micrometeorological techniques are well adapted to provide long-term estimates of gaseous N fluxes at the landscape scale, but have limited ability to resolve topographic and treatment effects on N trace gas emissions (Grant and Pattey, 1999, 2003). Whichever the technique chosen to monitor gases of N from cropping systems, practical and financial reasons reduce the measurements to few locations and for limited periods of time, which precludes field-based observations from being scaled-up to reliable seasonal and regional estimates.
