World Food Policy Volume 2, Number 1, Spring 2015 | Page 41

World Food Policy

was 1.5 ton/ha (see Figure 1).4 Also, note
that half of the rice area in SSA is upland,
where the yield is substantially lower than
in lowland paddy fields (Balasubramanian
et al. 2007). Thus, it seems reasonable
to assume that if new technology is not
introduced and production is carried out
under rain-fed conditions, the lowland
paddy yield will range from 1.0 to 1.5 ton/
ha. We also hypothesize that the average
rice yield has increased in SSA primarily
due to the introduction of Asian-type
improved rice technology.5

We focus on lowland rice, not
upland rice, primarily because the prospect
for a large improvement of the yield is
much greater for lowland rice than upland
rice. We also did not encounter upland rice,
such as NERICA (new rice for Africa), in
our study sites except in Uganda. Kijima,
Sserunkuuma, and Otsuka. (2006; 2008;
2011) found that NERICA is potentially a
high-yielding one, but sensitive to rainfall
and that the rate of discontinuation of
NERICA adoption is also high, indicating
that NERICA was grown in unsuitable
areas or that the sustainable management
was not well understood by farmers. Also,
the NERICA yield is exceptionally high in
Uganda compared with other countries
in SSA (Otsuka and Larson 2013). The
tentative conclusion of this study is that
upland rice is not particularly promising,
even though there were great expectations
for the impact of NERICA on the upland

rice yield.

The Green Revolution in Asia
is alternatively called the seed-fertilizer
revolution because the engine of growth
was the development and diffusion
of fertilizer-responsive, high-yielding
modern varieties (MVs) of lowland
rice (David and Otsuka 1994). It is also
important to realize that the paddy fields
were bunded and leveled almost without
an exception in Asia when the rice Green
Revolution began.6 Bunding is needed to
store water in the paddy fields to reduce
the weed growth, whereas leveling is
necessary for the even growth of rice plants
and germination of directly broadcasted
seeds. In other words, these production
practices are essential for water and weed
control a nd healthy growth of lowland
rice plants. Draft animals or tractors are
usually used for bunding and leveling,
but they are often not used in SSA, as will
be shown shortly. No less important than
these production practices is the straightrow transplanting, which provides space
for weeding. Instead of transplanting,
direct seeding can be adopted without
sacrificing yield if paddy fields are bunded
and leveled well and if herbicide is used.
Herbicide, however, may not be available
or may be too expensive, even if available
in SSA. In the African setting, direct
seeding is not generally recommended
and transplanting is the generally
preferred option. A major contribution

4

Nearly half of the paddy fields were irrigated in Asia but the difference in yield between rain-fed and
irrigated areas was not large before the advent of MVs.
5
This is consistent with the results of a review of rice farming in SSA by Balasubramanian et al. (2007).
6
There is no clear evidence on the prevalence of bunding and leveling in paddy fields in Asia in the
1960s and 1970s. My argument is based on interviews with rice scientists who worked in Asia in the
1970s.

40