Soil Fertility as a Contingent Rather than Inherent Characteristic - Principles of Sustainable Soil Management in Agroecosystems - page 144

6.3 HIGH CROP PRODUCTIVITY FROM
“POOR SOILS” IN MADAGASCAR
In the mid-1990s, some anomalous results at variance with established principles of soil
science were encountered in Madagascar by the first author, and these eventually led
him into communication with the other authors. While being the director of the Cornell
International Institute for Food, Agriculture and Development (CIIFAD), Uphoff
became involved with the implementation of a conservation-and-development project
funded by the US Agency for International Development. Its objective was to help save
the rain forest ecosystems and their biodiversity within Ranomafana National Park.
Food-insecure Malagasy households living and farming in the peripheral zone
around the park needed alternatives to their use of slash-and-burn methods to grow rice
( Oryza sativa ) within the nearby rain forest. Their conventional practices used on irri-
gated patches of land outside the Park gave them low paddy rice yields of only 2 Mg ha -1 .
This made shifting cultivation within the rain forest necessary for households to feed
themselves—unless their lowland rice yields could be substantially increased.
An evaluation of the soils in the peripheral zone around the park, done by the
Agronomy Department of North Carolina State University, had concluded that these
were some of the least fertile soils that had been analyzed at NC State (Johnson
1994). Systematic sampling of soils to a depth of 1.5 m had shown them to be very
acidic (pH 3.8-5.0), with low or very low cation exchange capacity (CEC) in all hori-
zons. Iron (Fe) and aluminum (Al) toxicity were serious and common in acid soils
throughout the peripheral zone. Most constraining, available P was very low, usually
in the range of 3-4 mg kg -1 (ppm), less than half the 10 mg kg -1 usually considered
as a minimum for productive cropping.
It was thus unexpected that, on the same land and with the same rice varieties
that had been giving them very low yields, farmers using different agronomic prac-
tices taught to them by a Malagasy nongovernmental organization were able to raise
their average yields fourfold, to 8 Mg ha -1 . The alternative methods, known as the
System of Rice Intensification (SRI), had been synthesized by a French priest who
had lived and worked for 34 years in Madagascar (Laulanié 1993; see also Stoop et
al. 2002; Uphoff 2007). After such results were achieved for three consecutive years,
it was unlikely that they occurred by chance. Also, the same methods were used for
measuring both SRI and conventional yields, so the relative increase was hard to
dismiss, even if the absolute yield levels might be disputed.
This quadrupling of yields did not require the use of new, high-yielding varieties,
or the purchase and application of chemical fertilizers, or the use of agrochemical
crop protection. Moreover, farmers were reducing their applications of irrigation
water, keeping their fields just moist or alternatively wetted and dried, rather than
continuously flooded. Even labor requirements, greater at first while new techniques
were being learned, could be reduced once farmers had mastered them (Barrett et
al. 2004).
Increased production was achieved, counterintuitively, with many fewer plants.
SRI methods reduced plant population densities per square meter by 80% to 90%:
young seedlings (<15 days) were carefully and quickly transplanted just one per hill
in a widely spaced square pattern, 25 × 25 cm apart (16 plants m -2 ). Weeds were