The insignificant differences in most of the crop production
indices (Table 3) may be seen as a result of the similar nitrogen
availability under the two water management strategies as indicated
by the almost identical N mass balance for the CF and AWD
soil treatments (Table 2). Only fertilizer N denitrification loss
(Table 2) and below-ground plant biomass (Table 3) were significantly
affected by water management (Table 2), but even under
AWD conditions the fertilizer loss by denitrification was still
negligible. In our experiment, however, the lack of significant yield
increase under AWD may have several explanations. First, the soils
were equally fertile at the time of sowing in both treatments as
a result of the flooding season (SeptembereOctober) prior the start
of this experiment (NovembereFebruary). The fact that fertilizerderived
N only accounted for 21% of the N uptake by the plants in
both water regimes (Table 1) support the notion that the soil was
rich in N, probably deposited during the preceding flooding period.
Second, the imposed AWD treatment only led to a moderate degree
of soil desiccation which may have prevented nitrificationdenitrification
to become as active as seen in other studies
(Khind and Ponnamperuma, 1981; Cabangon et al., 2004). Finally,
assuming that differences between the two water regimes are
induced through nutrient availability, these differences must be
most pronounced during the crops where nutrient limitation is
more likely to occur. Unless compensated for by site-specific N
fertilization, we thus predict that water management impacts on
crop N limitation may be more pronounced during the second or
third crop after the preceding annual flooding season due to
depletion of the flooding-derived plant available N.
The significant higher below-ground biomass found in the AWD
treatment (Table 3) has been reported before (Zhang et al., 2009)
and could indicate that AWD irrigation may lead to increases in the
organic matter content of the rice paddy soil. However, this may be
counteracted by a stimulation of the degradation of organic matter
in soils undergoing wetting and drying cycles (Inubushi andWada,
1987). Clearly, more long-term studies are needed to resolve how
adoption of AWD irrigation will impact soil organic matter content
and composition.
The present study is the first 15N-based study to address the
effects of AWD in intensive rice production systems of the Mekong
delta and indicates that the AWD practice only marginally affects
fertilizer N losses and the associated mass balance of fertilizer N.
The AWD irrigation practice can thus be considered safe with
regard to soil N retention as compared to the conventional CF
practice in alluvial Mekong delta soils. However, we did not address
any long-term effects, positive or negative, of changing from CF to
AWDirrigation, but this should clearly be investigated in the future.
Specifically, more research should be directed to the long-term
agronomic fate of the fertilizer N that accumulated below the