The aus panel were 150 cultivars selected and sent by the International Rice Research Institute (IRRI) and while a panel of 353 cultivars from Brazil and Japan were sent from Sao Paulo University. Both panels were assessed by buried herbicide method [1] . In both screening four cultivars Azucena, Bala, Black Gora and IR 64 were used as check varieties.

The experiment was set up in greenhouse at the end of spring and the beginning of the summer season. The experiment had a randomized complete block design with four replications. The experimental box was set up linearly along the length of the box set up in the greenhouse. A box of 200 cm length, 81.5 cm width and 35 cm depth was positioned on the bench. The box was prepared with plywood wrapped with clean plastic. Non-woven fabric which was folded up the sides of the 5 cm was used to cover the bottom of the box. The box was filled up with soil (mixture of builder’s sand 50% and top soil 50%, verve). Theta probes were placed at 15 cm and 25 cm (at the herbicide layer) to measure the soil moisture content throughout the experiment. Diuron herbicide at 100 mg per plant was buried at 25 cm depth using the method of herbicide application [1] . Full strength Yoshida’s nutrient solution [20] was applied to saturate the soil of the box. Nutrient solution was applied once throughout the experiment. Seeds were directly sown at the 1.5 cm depth using 5 × 5 cm spacing. Water was applied to germinate the seeds, and after germination watering was stopped. The growing of rice plants in the greenhouse and score of the herbicide injury symptoms scale from 0 to 5 were recorded [1] Where 0 indicates no symptoms, 1 indicates 5% to 15% leaf area affected, 2 indicates substantial leaf 15% - 50% leaf area affected, 3 indicates >50% leaf injury and noticeable leaf death, 4 indicates 15% - 50% leaf area dead, 5 indicates virtual to complete plant death (>50% leaf area). Figure 1 demonstrates the steps of the experimental set up. The herbicide shoot injury symptoms first appeared at day 20 after sowing and were recorded approximately every other day until 47 days after sowing (DAS).

The experiment was conducted in the same greenhouse during the monsoon in the same condition as maintained in the previous experiment. The experiment was designed to use a similar methodology described as above and applying a randomized complete block design with four replications. The box dimension was bigger due to bigger population size, being 460 cm long with the same width and depth. Diuron herbicide was prepared according to each plant having 100 mg of diuron. Approximately 20 pieces of filter paper (90 cm × 23 cm) were saturated with diuron herbicide was placed on same depth mentioned in the previous methodology part. Mixture of top soil and builder’s sand (50%, 50%) were used in this experiment. Full strength Yoshida’s nutrient solution [20] was applied to saturate the soil of the box. After sowing the seeds, water was applied to germinate the seeds, after germination watering was stopped. The other procedure was as the same the previous set up. The score of visible leaf injury was taken six times depending on the symptoms development ( Figure 2 ).

The soil moisture content throughout the experiment is presented in Figure 3 . The soil water content started to reduce in the box at 6 DAS according to theta probe reading. In the herbicide layer, the soil humidity decreased steadily from around 25% but the decrease accelerated from about 17 days to reach 5% by the end of the experiment indicating that plants were taking up water from soil around the herbicide layer.

The progression of the mean herbicide score (HS) throughout the experiment is presented in Figure 4 . The herbicide symptoms were first recorded at day 20 after sowing where 13 plants had developed symptoms. The herbicide injury symptoms were recorded on eighth occasions from day 20 to day 47 after sowing. On each occasions ANOVA shows the proportion of variation explained by genotypes was highly significant (P < 0.01) and were in the range of 31.4% to 35.6% of variation explained with F values in the range of 1.5 to 1.8 and which did not change much with time. The highest discrimination between the genotypes were observed in the day 47 after sowing. The frequency distribution of herbicide score on day 47 is displayed on Figure 5 . The variation of mean herbicide score of the accessions on day 47 is shown in Figure 6 and Figure 7 . Out of four check varieties Black Gora had a mean score 1.75 with Azucena 0.5 on day 47. Bala and IR 64 did not show any symptoms even at the last scoring day after sowing. For 70 cultivars there were no symptoms even after 47 days. For the rest their herbicide score on day 47 are presented in Figure 6 and Figure 7 . The mean herbicide score of Azucena and Black Gora during the screen shown in Figure 8 . Among the accessions Narikel Jhupi (more than 4 HS), Pannitti (More than 3 HS) showed the highest herbicide score compared to the other cultivars. Kele, kalia, Dheki shaita, Dharia boalia, Boteshwar, Noroi, Kalu A 30, Dharia, Koi Murali, Dhali Khama, Baulan showed higher herbicide score than the Black Gora while Tepi Boro, Simul Khuri, Seenetti, Lenza Murali and Gerdeh showed the same score as Black Gora ( Figure 6 ). Bathuri, Garia, ARC 12071, Kangro, Bawoi, Gulmurali, Jambali Bussa, Madhabsail 741, N22 and Mahlar 346, Santhi sufaid 207 showed very low score (less than 0.5) even on the last scoring day ( Figure 7 ). Anglais, ARC 10317, ARC 10955, Aus 196, 257, 41, 439, Bailam, Bans 4, Baranboro, Biranj, Chengri, Chengri Murali, Chiadi Naki, Chikon shoni, DA 16, DA 18, Dangar, Dhal kachi, Dhala bhadoi, Fulkati, they did not show any symptoms even at the end of the experiment.

The result of the theta probe readings throughout the experiment are presented in Figure 9 . The soil water content at the herbicide layer (25 cm) was above 25% at the start of the experiment, after that the moisture level slowly dropped steadily to reach around 10% by the end of the experiment. At 22 days after sowing (DAS) symptoms started on twenty cultivars and data were collected every third day. On the first scoring day cultivars 6-V-7, Aulha Esav, Arabiya Mochi, Aus 61, Black Gora, Col/Miyazaki/1963, gaisen Mochi, IPSL 2070, Ishiya Kushi Mochi, Kang youho, Linha 2459, Nourin Mochi 4, Oosumi, Owri 79, Piemonte IAMG N4, Saitama Senshou, Senshou, Senshou Ibaragi, Sonobe Mochi, Touzam Mochi, Vermelho de Pernambuco, Yonoyuki Mochi had plants which showed symptoms. The overall average symptoms were very low at the beginning and rose to near 2.5 at the end of the scoring at 34 days after sowing ( Figure 10 ). On each day of measurement, cultivar had a significant effect on herbicide score as detected by one way analysis of variance (ANOVA). The F value from ANOVA increased from day 22 to day 34 ( Figure 11 ) and it was highest (1.54) in the 34 DAS. It suggested that this was the day where it showed the most discriminating between the cultivars. At this time the ANOVA indicated that genotype explained 34% (R² value) of the variation within the population. The frequency distribution of herbicide score on day 34 is displayed on Figure 12 . The result revealed that by 34 DAS IAC 202, Santa America (IAC), Vermelho de Pernambuco, Nourin 5, Urasan1, Saitama Senshou, Ootama and Chioda Wase showed the highest herbicide score (more than 4 HS) than the other cultivars ( Figure 13 ). These cultivars showed higher herbicide score than the Black Gora, the deep rooted check cultivars [1] [18] . It suggested that those cultivars are deep rooted accessions among the tested cultivars. A total of 29 accessions showed very low herbicide score (below 1) at the end of the scoring day. Among the all tested cultivars Araure 4, BRS Formoso, Centauro, Cimmaron, CT 9682, Eika Ine, FL00447, Fonaiap, IAC120, IR 8, IRGA 409, Iwate Ryoon 1, Koshihikari, Kurohige, Linea 17 and Nipponbare did not show any symptoms even at the last scoring day. Cultivars which had 3 and more than 3 for herbicide score at day 34 are chosen to show in Figure 13 and Figure 14 . Around 211 cultivars showed less than 3 herbicide score. Out of four check varieties Black Gora has shown a maximum score of 4.5 on day 34 after sowing and Azucena has shown maximum 2 on end of the score. The variation of the symptoms development of these two check cultivars throughout the experiment are shown in Figure 15 . The checks Bala and IR 64 did not show any symptoms throughout the experiment.