Agriculture Reference
In-Depth Information
2 days prior to the measurement period, which considerably reduced the proportion
of N
2
O emitted, vs. immediately prior to measurement. In contrast, prior wetting
had no effect on the mole fraction in the Early Successional system—further illus-
trating presumed differences in denitrifier communities.
Measuring denitrification rates in the field is difficult (Groffman et al. 2006).
Fluxes of N
2
cannot be assessed directly because the comparatively small amount
of N
2
produced by denitrification cannot be readily differentiated from the very
large atmospheric background. Consequently, we must rely on inference from labo-
ratory incubations (e.g., Robertson and Tiedje 1985, Weier et al. 1993) or mass
balance techniques where denitrification rates are assumed to be the difference
between total N inputs (e.g., N fixation, fertilization, and deposition) and measur-
able outputs (e.g., leaching, harvest, and erosion).
In a novel approach to directly measure denitrification
in situ
, Bergsma et al.
(2001) used an
15
N-gas nonequilibrium technique to measure simultaneous fluxes
of N
2
O and N
2
from an MCSE soil planted to winter wheat. The N
2
O mole fraction
ranged from <0.004 to 0.14, with an average of 0.008 ± 0.004 when both gases were
above the detection limit, showing that N
2
O production is only a small fraction of
total N gas production (N
2
O + N
2
). If generalizable, this suggests that N
2
emissions
from U.S. Midwest row crops may, on average, be substantially greater than N
2
O
emissions. If so, then extrapolation suggests that total losses of N from denitrifica-
tion in these systems are of similar importance to the hydrologic losses discussed
earlier (Fig. 9.7). Although similar findings have been estimated from mass balance
approaches elsewhere (e.g., Gentry et al. 2009), N losses from denitrification in
KBS LTER soils remain highly uncertain.
Nitrous Oxide (N
2
O) Emissions
Nitrous oxide is produced primarily by denitrifying and nitrifying bacteria; other
sources appear unimportant in agricultural soils (e.g., Robertson and Tiedje 1987,
Crenshaw et al. 2008). The extent to which N
2
O is produced by denitrifiers vs.
nitrifiers in agronomic systems is a source of controversy; knowing this could
be valuable for designing N
2
O mitigation strategies. Nitrous oxide isotopomer
analysis (the intramolecular distribution of the
14
N and
15
N isotopes; Ostrom and
Ostrom 2011) is the only current field-based technique that can unambiguously
differentiate between these two processes without significantly altering micro-
bial activity. Ostrom et al. (2010a) measured isotopomer site preference to show
that denitrification is the dominant source of N
2
O in the Mown Grassland (never
tilled) community following tillage of a subplot; in that experiment, denitrifica-
tion accounted for 53-100% of the N
2
O produced over a diurnal cycle. Using
the same approach in no-till wheat of the Resource Gradient Experiment, Ostrom
et al. (2010b) showed that denitrification dominated regardless of N fertilizer rate
(0, 134, and 246 kg N ha
−1
).
Nitrogen availability, on the other hand, is the single best predictor of over-
all N
2
O emissions across both unmanaged and cropped ecosystems at KBS LTER
(Gelfand and Robertson 2015, Chapter 12 in this volume) as elsewhere (Matson and
Vitousek 1987; Bouwman et al. 2002a, b). In cropped systems, paired comparisons
