No abstract

No abstract

Rice-based cropping systems are the foundation of food security in countries of Southeast Asia, but productivity of such systems has declined with deterioration in soil quality. These systems are different from other arable systems because rice is grown under submergence, and this may require a different set of key soil attributes for maintenances of quality and productivity. A minimum dataset was screened for assessing quality of soils belonging to three Soil Orders (Inceptisols, Entisols and Alfisols) by using statistical and mathematical models and 27 physical, chemical and biological attributes. Surface soils were collected from farmers’ fields under long-term cultivation of rice–potato–sesame cropping systems. Most of the attributes varied significantly among the Soil Orders used. Four or five key attributes were screened for each Soil Order through principal component and discriminate analysis, and these explained nearly 80% and 90% of the total variation in each Soil Order dataset. The attributes were dehydrogenase activity (DHA), available K, cation exchange capacity (CEC) and pHCa for Inceptisols; organic C, pHCa, bulk density, nitrogen mineralisation (Nmin) and β-glucosidase for Entisols; and DHA, very labile C, Nmin and microbial biomass C for Alfisols. Representation of the screened attributes was validated against the equivalent rice yield of the studied system. Among the selected key soil attributes, DHA and CEC for Inceptisols, organic C for Entisols, and Nmin and very labile C for Alfisols were most strongly correlated with system yield (R2 = 0.45, 0.77 and 0.78). Results also showed that biological and chemical attributes were most sensitive for indicating the differences in soil quality and have a strong influence on system yield, whereas soil physical attributes largely varied but did not predict system yield.

No abstract