Rice yield and relationships to soil properties for production using overhead sprinkler irrigation without soil submergence

Production of irrigated rice (Oryza sativa L.) without conventional soil submergence may increase water use efficiency but risk a decline in rice yield. Spatial variability in rice yield and relationships among yield and soil properties were examined across a 3.3-ha experimental site after uniform crop management. The site was initially consolidated from small, previously puddled parcels of land into eight laser-leveled plots across a 2.1-m elevation gradient in the Philippines. Six crops of rice in rotation with three crops of mungbean (Vigna radiata (L.) R. Wilczek) were then grown from January 2012 to March 2015. The four rice crops in the dry season (crops 1, 2, 4, and 6) were irrigated using an overhead sprinkler system to maintain soil water potential greater than −10 kPa without soil submergence. Soil was puddled and flooded with irrigation for only rice crop 5 grown in the wet season. Mean rice yield with full fertilization was 6.1 Mg ha−1 for crop 2 but only 3.6 Mg ha−1 for crop 6. Yield for rice crop 6 ranged from 1.6 to 5.4 Mg ha−1 across 36 locations in the 3.3-ha site. Rice yield for crop 6 was inversely related to soil bulk density (r = −0.72 and −0.80, P < 0.001), log10 saturated hydraulic conductivity (Ksat) (r = −0.71, P < 0.001), and sand content (r = −0.63, P < 0.001). Yield was directly related to water-filled pore space of soil (r = 0.61, P < 0.001). Field locations with large downward water movement and the correspondingly most aerobic soil conditions were most prone to low rice yields and large yield decline. Rice yield was directly related to anaerobic N mineralization (r = 0.72, P < 0.001) and soil ammonium-N (r = 0.54, P < 0.001) at 16 days after transplanting and to indigenous N supply (r = 0.83, P < 0.001), as determined from rice yield without application of fertilizer N. These relationships suggested lower net N mineralization at locations with the most aerobic soil conditions, as determined by water-filled pore space. Rice yield was, however, not limited by insufficient fertilizer N. Anaerobic N mineralization and indigenous N supply were unrelated to optimum fertilizer N rate calculated from yield gain with fertilizer N, which was determined at 36 locations using paired plots with and without added N. The long rainy season and clayey soil at the site limited triple-cropping systems to two rice crops and only one non-rice crop per year. In such environments where crop rotation is not an option to prevent yield decline for rice grown without soil submergence, periodic soil submergence to increase anaerobic soil conditions would be required to sustain rice yield. Our study suggested that soil bulk density and Ksat, which can increase for successive rice crops grown without puddling and soil submergence, could serve as indicators of field locations prone to yield decline from factors other than water-deficit stress and requiring a period of soil submergence to sustain large rice yields.