From the literature review and the data in Table 15.1, we selected three major factors to distinguish areas prone to soil N2O consumption at the global scale. These factors are nitrogen availability (or nitrogen limitation), soil wetness and temperature. These appear to be the major controls that are consistently observed in nearly all studies and are therefore most suitable to delineate global areas with possible N2O sink activity.
We used 0.5° x 0.5° resolution maps for both inputs and outputs. Surface nitrogen inputs from atmospheric nitrogen deposition, biological N2 fixation, nitrogen fertilizer and animal manure were taken from Bouwman et al. (2005) (Fig. 15.2). Surface nitrogen inputs were grouped into five classes: nitrogen loading of <5 kg/ha/year is assigned a value of 5, 5-10 kg/ha/year a value of 4, 10-15 kg/ha/year a value of 3, etc. (Fig. 15.2). This classification is largely based on the threshold of 10 kg N/ha/year, above which changes to the nitrogen cycle may occur in sensitive ecosystems (Bobbink et al., 1998), and the assumption that the sink strength increases with increasing nitrogen availability.
For soil wetness, we used an index obtained from the annual net precipitation (annual precipitation minus evapotranspiration) from Van Drecht et al. (2003) divided by the total soil water-holding capacity for the top meter of soil (from Batjes, 1997; Batjes, 2002). The maps of classified surface nitrogen inputs and wetness index were added and areas with no month with mean temperatures >5°C were excluded. Hence, we assume N2O uptake is negligible below 5°C, which is in accordance with the work of Ryden (1983).
By adding the numbers of both classified maps, we obtain a maximum value of 10 and a minimum of 2. These were classified into 'high' (values >8), 'moderate' (values = 7-8) and 'low' (values <7) (Fig. 15.3). As expected on the basis of the literature review, nitrogen limitation occurs in areas remote from agricultural and industrial activity, mainly in northern latitudes, and semiarid and arid regions. Ecosystems in northern latitudes are known to be nitrogen-limited due to low atmospheric nitrogen deposition and the absence or low activity of biological nitrogen fixation (Crews, 1999; Sprent, 1999). When the soil wetness index is combined with this map, the arid and semiarid areas are excluded, and the major regions prone to high N2O sink activity are located mainly in northern regions (Fig. 15.3). According to this quick
ho OJ VJ
inventory, ~8% of the global land area has a 'high', and another 8% has a 'moderate', likeliness of N2O uptake (Table 15.2).
Our results should be interpreted as a first assessment of the likeliness of sinks for N2O at the Earth's surface. By adding the numbers of the two classified maps, we implicitly assigned equal weights to them. In reality, however, one factor may be more important than another and the relative importance of different factors may vary from soil to soil. Moreover, we ignored other important direct or indirect factors influencing soil N2O consumption, such as soil drainage conditions, soil fertility indicators and soil pH. These were not included in this inventory because of the difficulty in comparing the importance of these factors to surface nitrogen inputs, temperature and soil wetness. Soil wetness is also not always directly related to precipitation. This holds, for example, for histosols, which are generally poorly drained organic soils with a high water table. Also, mineral soils may have hydromorphic properties and impeded drainage because of the occurrence of impermeable layers or layers impeding air exchange.
Maps of areas with occurrence of soils with hydromorphic properties and classified soil pH (Batjes, 1997; Batjes, 2002) are presented in Fig. 15.2, but were not used in the compilation of Fig. 15.3. However, these factors could be used to modify the general patterns of Fig. 15.3.
Our inventory is a quick method to identify areas with conditions where N2O uptake may occur during prolonged periods, during a part of the year or where there may be large interannual variability with some years showing net N2O uptake and others indicating zero or small emissions. We recognize that such inventories are fraught with potential scaling errors. For example, for soil
Table 15.2. Percentage of global land area in classes of probability of N2O sink activity calculated from data presented in Fig. 15.3.
properties, we use those of the dominant soil within each 0.5° x 0.5° grid cell. In reality there is large heterogeneity at smaller scales that is not represented in the data used. A more sophisticated way to investigate the probability of N2O sinks on the smaller scale is by using mechanistic models that account for all the climate, soil and hydrological factors influencing the ecosystem's carbon and nitrogen cycling (and nitrogen limitation). Our results indicate that such studies should focus on the northern regions of North America, Europe and Asia.
Was this article helpful?