Modelling Soil Organic Matter Dynamics – Global Challenges - Sustainable Management of Soil Organic Matter - page 80

underlie most SOM transformation models should be tested and shored up
wherever possible. How is this to be done?
Prognosis
Isotope tracing offers the potential for tracking C and N fluxes between
measurable (i.e. real) SOM fractions. Where 15 N and 13 C are employed in
tandem, it should be possible in theory to measure the fluxes between up to
five fractions including the gaseous and solution phases (Arah, 2000a).
This in turn should allow the assumptions (approximations) of effective
spatial homogeneity and first-order reaction kinetics to be examined in
detail. Such work seems to us to be essential.
It calls for the following: (i) development of a fractionation procedure
which partitions soil organic matter into five routinely measurable
fractions, the first gaseous, the second soluble and the other three defined
solely by their method of extraction, such that fractions 1-5 add up to
the total organic content of the sample, elemental and isotopic contents
of each fraction can be determined, and, ideally, fractions 1-4 are
reasonably dynamic over a typical cropping cycle; (ii) incubation of 15 N-
and 13 C-labelled soil samples with SOM fractionation and fraction analysis
at intervals; and (iii) interpretation of the results via some kind of a model
in which (13) C and (15) N fluxes are linked, to derive the actual C and N
fluxes between the measurable (indeed measured) SOM fractions of the
incubation. Sohi et al . (2000) describe such a fractionation procedure; Arah
(2000b) describes such an interpretative model.
Comparison of the results of such an exercise against the constant
first-order-reactivity and effective spatial homogeneity assumptions implicit
in all current SOM transformation models might do one of three things.
It might reveal these assumptions to be unacceptable, necessitating the
development of a new generation of models in which more attention is
paid to soil structural and microbial kinetic parameters. It might throw
up no major contradictions, allowing a relatively seamless shift from the
dependence of existing models on functionally defined, unmeasurable
SOM pools to a dependence on similarly uniform but procedurally defined,
measurable SOM fractions and thereby permitting the development of
predictive soil tests for SOM function. Or, most likely, it might fall
somewhere between these extremes, suggesting circumstances in which
either response is appropriate.