Four field experiments were carried out in Tamil Nadu, India, to deter
mine effects of internal drainage (percolation rate) on growth and yie
ld formation in rice, in interaction with nitrogen (N) management. Gra
in yield response to drainage was positive at all N levels in all expe
riments. Yields (averaged over N treatments) increased by 14% (Expt. 1
), 10% (Expt. 2), 25% (Expt. 3) and 22% (Expt. 4) in response to drain
age. Higher yield responses to N application were found under well dra
ined than poorly drained soil conditions. In late-season crops under p
oor drainage, grain yields decreased in response to high doses of N. T
he positive yield response to drainage was always associated with incr
eased number of filled grains per panicle (+13 to +17%), increased tra
nslocation of stored reserves, increased root biomass at harvest (+20
to +36%), increased N concentration in roots and increased root activi
ty (as measured by alpha-naphthylamine oxidation). Of the yield increm
ent resulting from drainage, 25 to 35% could be accounted for by incre
ased translocation of stem reserves in Expt. 1, 5 to 25% in Expt. 2 an
d 30 to 40% in Expt. 3. Yield increase in response to drainage was not
always associated with more green leaf area, leaf longevity, leaf N c
ontent, crop N uptake, nor with a larger dry matter production at give
n levels of the leaf N pool and global radiation, as quantified with t
he help of a calibration factor f(sv). A literature review is presente
d leading to two hypotheses to explain the effect of drainage on yield
: (1) a possible potassium deficiency is expressed more in reduced (lo
w-redox) soil due to a lowering of K availability per se; or to increa
sed K demand associated with maintaining the root's 'oxidizing power';
(2) improved root condition resulting from an increase in soil redox
potential induces a prolonged synthesis and transport of cytokinins in
roots, resulting in an extended photosynthetic activity, enhanced sin
k strength of grains and more translocation and deposition of carbohyd
rates in the grains. These hypotheses are not mutually exclusive. They
were not explicitly tested in this study. (C) Elsevier Science B.V.