Fig. 12.1: Mean cumulative N2O-N emission and mean NO3-N content of the top soil (0 - 25 cm) in the first and second experimental year (all treatments included)
The mean values of nitrate-N can be separately correlated with the cumulative N2O emission. Since the nitrate level was higher in the second year than in the first, it is not reasonable to calculate mean values over both years. If the two years are treated separately, relative good positive correlations (r2= 0.44 and 0.68) for the first and second experimental year are found, respectively (Fig. 12.1.1). This observation points again to denitrification as the major source of N2O emissions and has already been found in numerous studies (e.g. Ruser et al., 2001; Sehy et al., 2003).
12.1.2 WFPS
Fig. 12.2: Mean cumulative N2O-N emission and mean water-filled pore space (0 - 25 cm) in the first and second experimental year.
The mean WFPS values for both the single experimental years as well as for the whole study show a good negative relationship with the cumulative N2O emission (r2= 0.64 for the first year; r2= 0.48 for the second year; r2= 0.71 for both years, Fig. 12.1.2). Usually positive correlations are reported for non-irrigated systems, but in irrigated systems like vegetable production the WFPS is usually higher than the threshold value for strongly increased N2O production. Further increases in WFPS are known to decrease the N2O/N2 ratio. The reason is an elevated reduction potential in the soil due to decreased oxygen levels because water serves as a diffusion barrier for oxygen.
12.2 N2O mitigation strategies
12.1.3 Carbon availibility
A temporary C-limitation of N2O production is probably less common than N-limitation, but has been seen in other ecosystems like temperate forest soils (Perez, 2010), humid tropical forest soils (Garcia-Montiel et al., 2003) and compacted grassland (Abbasi and Adams, 1999). There are many observations in our study which point to the major influence of carbon as a parameter controlling N2O emissions. For example the correlation of N2O and CO2, the increased emission after incorporation of carbon rich substrate and the 32fold increase in N2O flux rates after the addition of glucose (Chapter 7). One possibility cited in literature to find out more about carbon availability is the measurement of dissolved organic carbon (DOC). DOC is defined as the organic carbon fraction in solution that passes through a 0.45 µm filter. DOC is the most mobile and active cycling organic carbon fraction as compared to the fraction immobilized in organic compounds (Bolan et al., 2011). In our study, DOC was analyzed for the first experimental year. However, due to practical constrictions, soil was frozen before extraction. That was probably the reason why the data was not very plausible. Furthermore, the total amount of DOC does not necessarily reflect the in situ availability of DOC in the soil (Sehy et al., 2004). For example, DOC in macropores might not be used for denitrification because of the good aeration status (Zsolnay, 2003). But even though DOC was not a good predictor for N2O fluxes, many other indications show that carbon availability is of very high importance in vegetable production (see also Chapter 7).
12.2 N
2O mitigation strategies
In Chapter 6 it has been shown that even though IPCC emission factors for N2O emissions from vegetable fields are within the range proposed by the IPCC (0.3 - 3 %); absolute N2O emission can be >10 kg N2O-N ha−1 yr−1. If the higher emissions from the tractor compacted areas are taken into account, these values would even increase to>12.5 kg N2O-N ha−1yr−1 (see Chapter 5). The main reasons for these high emissions include the
high fertilizer N input as well as the high amounts of organic nitrogen and carbon which are provided by plant residues and usually incorporated into the field after harvest.
Tab. 12.4: Different strategies for the mitigation of N2O emissions from intensive vegetable production: per- centage reduction of annual N2O emissions in % and kg CO2-equ as well as converted to km of a passenger car (102 g CO2-consumption (100 km)−1) for the first and second years of measurement. An asterix (*) indicates a significant reduction in N2O emissions by a certain mitigation strategy in our study.
Several strategies have been tested in regard to their efficiency in reducing the elevated N2O emissions from intensive vegetable production and are summarized in Table 12.4. For the desynchronization of C- and N-input, the cumulation for a whole year is not described in Chapter 7, but was achieved by combining data of the chard period and the preceding measurements on the same site, starting from the planting of cauliflower in 2009. These reduction potentials are shown expressed in percentage and were then converted into CO2 equivalents (equ), using the conversion factor of 296 which is the Global Warming Potential of N2O on a 100 year time frame (IPCC, 2001). To make these numbers more vivid, the kg CO2-equ ha−1were also converted to kilometers which can be driven by car with this amount. Therefore, a relatively economical car was chosen with a consumption of 102 g CO2 (100 km)−1. It must be kept in mind that these numbers refer to only one ha and that about 106 ha of Germany are covered with intensive vegetable production (Statistisches Bundesamt, 2010).
12.2.1 Fertilizer reduction
In Chapter 6 it has been shown that fertilizer reduction is a very effective tool in the reduction of total N2O emission from vegetable cropped sites. This strategy functions, as it has been described in many studies before,
12 General Discussion
via the concentrations of mineral N in the soil. When only comparing the three fertilization levels and the control as described in Chapter 6, a strong correlation was found between the mean nitrate content of the top soil (0 - 25 cm) and the cumulative N2O emission on an annual basis in the first and second experimental years (r2= 0.98 and 0.89 respectively).
However, in vegetable cropping systems a relatively high N-fertilizer input, as well as a rather high soil mineral N level, is necessary for effective production. A fertilizer reduction by 20 % beneath the level prescribed by the German Target Value System led to lower lettuce yields in 2008. N-fertilizer reduction is therefore relatively restricted by plant demand. In the second year, no yield reduction was observed for further fertilizer reduction. For farmers however, the monetary aspect is of most importance. Since yield declines were observed for one out of two years, it should not be recommended to farmers as a mitigation option. For our data it can be concluded that farmers can be recommended to apply N-fertilizer according to the German Target Value System. After all, we measured a reduction potential of up to 17 %, but also no reduction in yields for the second year of measurement.
12.2.2 Addition of DMPP
Addition of DMPP has been discussed as a mitigation option for N2O emissions in detail in Chapter 8. With a reduction potential of 45 and 40 % it is even more effective than fertilizer reduction and no decrease in total yields was observed. However, it is questionable whether a large-scale application of an additional chemical compound is recommendable from an ecological point of view. DMPP has been tested in several model trials (Andreae, 1999; Roll, 1999). Roll (1999) reported that no indication was found for example for acut oral or subcutaneous toxicity in rats and no skin irritations in rabbits. However, chromosome aberrations were found in vitro in mammal cells. No further attention was given to this aspect however, since all in vivo tests were negative. He summarized that no toxic effects could be expected from DMPP. Andreae (1999) summarized that also no ecotoxicology was detected, for example no toxicity was found in fish, daphnia or earthworms. However, they could not prove biological degradation of DMPP in aqueous systems, but argued that DMPP had been seen to be degraded in soil and would not infiltrate into deeper soil layers. All in all, DMPP was judged as "not harmful to the environment" according to the European criteria EEC 93/21 (Andreae, 1999).
However, in both studies no long-term experiments were included and the use of model organisms does not exclude that further adverse effects could arise for other organisms. Up to the present, empiric studies have shown that obviously the ammonium monooxygenase is inversely inhibited by DMPP. Yet the mechanism is not fully understood. At least no inhibition of the structurally similar methanmonooxygenase has been found which might have negated the lower N2O emissions by lower CH4 oxidation (Weiske et al., 2001). Of course, it is very difficult to study the effects of inhibition of nitrification on the complete ecosystem. Many correlations exist between the organisms of an environment and in such complex systems it is almost impossible to evaluate the effects on the complete system.