The effect of toposequence position on soil properties, hydrology, and yield of rainfed lowland rice in Southeast Asia

Abstract

A large proportion of rainfed lowland rice in Southeast Asia is grown in gently sloping areas along toposequences with differences in elevation of a few meters. These small differences in elevation can lead to differentiation in soil properties and hydrological conditions, which in turn may affect crop performance and yield. It may be appropriate to replace blanket crop management recommendations in rainfed areas with toposequence-specific management recommendations. However, thorough statistical analyses of the relationships between toposequence position and field and crop conditions are lacking. In this paper, we statistically analysed the effect of toposequence position on soil properties, hydrological conditions, yield, and yield increase due to weed control and/or fertilizer management in rainfed areas in four villages in Indonesia and Thailand each in 2000–2002. Differences were substantial in field hydrology (average depth of ponded surface water and of groundwater, number of days without ponded surface water), exchangeable K, organic C, and clay content depending on toposequence position. There were also differences in other soil properties, including N, P, CEC, pH, sand, silt, bulk density, yield, and the magnitude of yield increase due to intensive weed control and/or recommended fertilizer application, but these effects were not consistent across countries, seasons, and years. The hypothesis that toposequence position would be a useful recommendation domain for weed control and fertilizer recommendations was not supported by our statistical results. The reasons why toposequence position has an inconsistent statistical effect could be (1) that the variability of the field conditions is larger among villages than among toposequence positions, and/or (2) that farmers already respond to differences in field conditions in their prevalent management practices, thus masking the effects of toposequence-specific variation on yield. Our findings suggest that despite the large toposequence effects on soil nutrient and water availability, weed and fertilizer management recommendations should be field-specific and time-specific rather than toposequence-specific.