An image analysis technique for assessing resistance in rice cultivars to root-feeding chironomid midge larvae (Diptera : Chironomidae)

Computer-based image analysis was used to assess 44 rice cultivars for root vigour and resistance to larvae of the rice bloodworm, Chironomus tepperi Skuse, and to determine the relationship between bloodworm density and plan t damage in two cultivars, Amaroo (highly susceptible) and Koshihikari (sli ghtly resistant). Cultivars assessed for resistance included 28 with report ed resistance to other rice pests, 14 past or current Australian commercial cultivars, and 2 experimental cultivars. Screening was conducted under con trolled conditions using plants sown in replicated pairs of plastic trays. C. tepperi larvae were introduced into one tray of each pair, while the oth er tray was maintained as a control. After 8 days exposure plants were remo ved from the trays and their profile root areas measured using image analys is. Residual maximum likelihood analysis was used to remove the error compo nents associated with black effects and low repeatability, and to separate the influence of seedling root vigour from other aspects of resistance. Cul tivars with the highest root vigour included Mas, Pokkali, Taichung Native 1 and TKM6, whilst those with the most residual root tissue after bloodworm exposure included Pokkali, HR19, Mas and TKM6. The cultivars that showed t he highest resistance independent of vigour were (from highest to lowest) Y RK3, Calrose, HR19 and WC1403. YRK3, an experimental cultivar, is estimated to lose approximately 63% of potential root growth when exposed to C. tepp eri larvae under our experimental conditions, and this level of resistance is insufficient to allow this cultivar to be grown in the field without che mical protection. Further screening work is required to locate stronger sou rces of resistance. Maximum plant damage to Amaroo was achieved at 80 final instar C. tepperi l arvae per container (153 cm(2) area), however maximum damage to Koshihikari plants of the same age was not achieved at 160 larvae per container, the h ighest pest density evaluated. This result suggests that differences in the relationship between pest density and plant damage may assist in confirmin g bloodworm resistance identified in screening trials at fixed pest densiti es. (C) 2000 Elsevier Science B.V. All rights reserved.