Agriculture Reference
In-Depth Information
proportion is clay-associated (Calderón et al. 2011) and some may be charcoal
(Janik et al. 2007).
The intraaggregate SOC fraction obtained by breaking apart aggregates to
release the SOC protected by aggregation showed that only 38% was corn-derived
C (Table 5.6). Thus, 62% of the intraaggregate fraction was more than 10 years old.
A C:N ratio of 15.2 indicates that accumulation of microbial biomass lowered the
C:N ratio from that of the interaggregate fraction. The clay fraction was enriched
in SOC relative to the silt fraction and 21% of its C was corn-derived in contrast to
18% in the silt fraction. This indicates that the SOM components of these fractions
are primarily composed of materials greater than 10 years old, with a mixture of
both young and old SOM components.
Data from this 800-day incubation confirm δ
13
C field data in that the interaggre-
gate fraction lost nearly half its C during the 800 days (Table 5.6). But there was
an increase in the proportion of intraaggregate C, indicating the formation of new
aggregates during the incubation. The drop in corn-derived C in the silt fraction dur-
ing incubation showed turnover in this fraction even though most of it was old. Harris
et al. (1997) used δ
13
C to calculate the MRT of corn-derived vs. noncorn, older C.
The MRT of the corn-derived interaggregate C was 3.9 years vs. 19.7 years for the
noncorn C (Table 5.6), showing that although older, the noncorn residues were still
decomposing. The corn-derived C in the intraaggregate fraction, with an MRT of
11.4 years, is part of the slow SOM pool as defined by incubation and curve fitting
for kinetic analysis (Paul et al. 2001b). The 34.4-year MRT of the noncorn C in the
interaggregate fraction was older than that of the slow pool determined by incubation
(Table 5.6). The MRTs of the corn and noncorn C were determined using δ
13
C, as
described in Harris et al. (1997). The MRTs of the corn-C in the silt and clay fractions
(10.9 and 16.5, respectively) reflected the length of time that corn was grown (10
years), while the noncorn C MRT for both fractions was greater than 40 years.
A comparison of Tables 5.2 and 5.6 shows that much longer MRTs are obtained
for the soil C from
14
C dating than from incubation and δ
13
C analysis following a
C
3
-C
4
plant switch. The radiocarbon dates represent the accrual of SOM over a
pedogenic (soil formation) time scale. The δ
13
C values are a function of the length
of time since native C
3
plants were replaced by C
4
corn. Both dates are correct,
but the history of the switch and the turnover of the SOC pools must be taken into
account when interpreting and modeling tracer data (Andrén et al. 2008).
The pool size and MRT data in Table 5.5 were used to mathematically model
the emission of CO
2
from the Conventional system in 1994 (Fig. 5.5). The timing
of the CO
2
flux was influenced by rainfall and temperature in the field. Predicted
CO
2
emission based on the data in Table 5.5 and using the CENTURY model cor-
responded well to measured values during the summer (Paul et al. 1999a). This
shows the reliability of the concepts used for modeling and the measurements of
pool sizes and their MRTs (Basso et al. 2011) based on long-term incubation and
tracers. However, the model did not accurately predict field fluxes for the fall sea-
son. This is likely because the model assumes that once harvest is complete, the
residue is available for decomposition, whereas in fact aboveground residues are
not immediately incorporated into soil; there is also a physical conditioning period
before decomposition that is not adequately represented in the models.
