New publication: Factors explaining variability in rice yields in a rain-fed lowland rice ecosystem in Southern Cambodia – University of Copenhagen

16 June 2016

New publication: Factors explaining variability in rice yields in a rain-fed lowland rice ecosystem in Southern Cambodia

This news is redirected by The Secretariat for Development Cooperation at SCIENCE from Science Direct | Elsevier.

Authors: Proyuth Ly, Lars Stoumann Jensen, Thilde Bech Bruun, Andreas de Neergaard from Department of Plant & Environmental Sciences, University of Copenhagen.


There is a growing body of literature documenting higher productivity of the System of Rice Intensification (SRI) than conventional practices; however, few studies have been conducted to explore the factors explaining this higher productivity. This paper investigated key factors influencing yields in a rain-fed lowland rice ecosystem in Cambodia under farmer’s SRI and conventional management practices (CMP). Rice yields from 70 plots with recorded management practices (36 plots under farmer’s SRI and 34 plots CMP) were measured. Composite soil samples (210 in total) were collected from the harvested plots and analyzed for physical and chemical properties. Stepwise multiple regression analysis was performed to identify important predictors explaining rice yield variability. Variables contributing significantly to yield variation included: number of cattle (access to farmyard manure [FYM]), farmer’s SRI practices, planting density and soil organic carbon (SOC). These four variables explained 39% of variance in rice yield. Keeping other variables constant, rice yields were increased by 14% (458 kg ha−1) by shifting from CMP to farmer’s SRI practices. The change of one unit in cattle (head), hill density (hill m−2) and SOC (g kg−1) results in an increase of rice yield by 5% (153 kg ha‐1), 2% (66 kg ha‐1) and 9% (289 kg ha‐1), respectively. The higher rice yield obtained by transplanting with narrower spacing suggests that the wider spacing recommended by SRI is not likely advisable, especially in nutrient-poor soil. Although the application of manure did not emerge as a key variable, three out of the four key variables (farmer’s SRI, access to FYM, and SOC) are directly linked to its use. Farmer’s SRI was strongly associated with the use of FYM, and the plots that received more FYM and belonged to farmers with more cattle had higher total soil N. Total soil N was positively correlated with SOC. The results highlighted the important role of livestock in crop-animal integration and the contribution of animal manure to increase agricultural productivity within smallholder farmers on infertile sandy Cambodian soils.