Modelling Soil Organic Matter Dynamics – Global Challenges - Sustainable Management of Soil Organic Matter - page 39

the model is sufficiently constrained to solve for parameter values (Arah,
2000).
Collins et al . (2000) used C mineralization data, total C and acid
hydrolysis to estimate amounts and turnover rates for a three-pool model
(representing active, slow and passive fractions) for five long-term experi-
ments in central USA. They demonstrated that the amount of crop-derived
carbon, determined from field sampling and 13 C natural abundance
methods, was well correlated with the mean residence time of the slow pool
determined from laboratory fractionations. The relationship was consistent
across soil depth increments (up to 1 m) within sites, but differed distinctly
between forest- and prairie-derived soils.
For both approaches, measuring the modellable or modelling the
measurable, the derivation and testing of SOM models has benefited sub-
stantially from the increased application of isotopic methods, particularly
the use of the stable isotopes 13 C and 15 N and radioactive 14 C. Isotope
dynamics can be coded explicitly into models, including appropriate
isotope discrimination coefficients. In particular, the use of natural
abundance 13 C has been valuable in quantifying SOM dynamics following
land use change or changes in agricultural management (e.g. Balesdent
et al . 1987; Six et al ., 1998; Collins et al ., 2000). The method relies on the
difference in 13 C natural abundance between plants with different photo-
synthetic pathways (usually contrasting C 3 versus C 4 plants) so that the
relative mix of SOM derived from a particular vegetation can be quantified.
Thus, where land use changes have occurred at a known point in time, for
example C 3 forest conversion to C 4 cropland, the rate of loss of the original
forest-derived SOM and the contribution of new crop-derived SOM can be
inferred from changes in the 13 C signature of the soil. The approach can be
extended to make inferences about the dynamics of SOM isolates as well as
whole soil. Since land use or management changes involving shifts in major
vegetation types are relatively widespread and occur in nearly all types of
climatic and soil conditions, the approach is well suited for testing models
for general use and regional applications. However, the inferences based on
stable isotope data introduce additional sources of variability and error that
need to be accounted for (Veldkamp and Weitz, 1994), and accurate mea-
sures of pre-disturbance soils and the amount and type of post-disturbance
plant residue inputs are required.
Radiocarbon dating of SOM and tracer methods utilizing the 14 C
enrichment from above-ground bomb testing have been utilized in conjunc-
tion with simulation modelling of SOM dynamics (e.g. Jenkinson et al .,
1987; Trumbore et al ., 1995), although to a lesser extent than 13 C, due in
part to the greater expense and more limited availability of 14 C analysis
facilities. Radiocarbon dating of SOM and SOM fractions has had a major
role in the definition of recalcitrant pools in SOM models (Falloon and
Smith, 2000). Initialization of the inert pool fraction in the Roth-C model is