???????????????????????????????????

Abstract

Cultivars ‘Ishikari’ and ‘Hoshinoyume’ showed a stable high split-hull paddy rate regardless of location or year, while the rate was relatively low in ‘Kirara 397’ and ‘Yukihikari’. In both ‘Hoshinoyume’ and ‘Kirara 397’, split-hull paddies began to appear 3–4 weeks after emergence of the panicle neck (i.e. after the dough ripe stage) and reached a peak 5–7 weeks after that stage. In normal paddies (=without split-hull paddies), Trigonotylus caelestialium produced more pecky rice grains at the flowering to milk-ripe stage or at the milk ripe stage, then decreased as the growing stage proceeded. Most of these pecky rice grains were damaged at the top part of the grain. However, the occurrence of pecky rice grains did not decrease remarkably with growth of the ear in split-hull paddies. Most pecky rice grains were injured on the side of the grain, and it was also verified that each injury site coincided with the split part of the paddy. These results indicate that T. caelestialium fed from the split part or joint (particularly the top of the joint) of the hull. Furthermore, this bug had difficulty in attacking grains as the growth of the ear progressed, although it was able to attack hardened grains (i.e. after the yellow ripe stage) in the case of split-hull paddies.