Selection of ideotype to increase yield potential of GM and non-GM rice cultivars.

Using classical breeding, plant breeders envision a plant type whose yield they aim to enhance by selecting for individual traits to create model/ideal plants or ideotypes. To achieve this, those factors restricting yield need to be identified and controlled through the use of new technologies to achieve the desired ideotype. This study aimed to determine the ideotype of seven genetically modified (GM) and non-GM rice (Oryza sativa L.) cultivars. Field experiments were carried out in three isolated regions in the north of Iran under the Iranian bio-safety standard protocol. Four of the GM cultivars carried the cry1Ab gene in the vegetative stage while three non-GM cultivars served as the control. R2 values showed that five, six and seven variables in Sari, Amol and Rasht regions accounted for 63 %, 52 % and 74 % of paddy yield variation, respectively. In the same three regions, paddy yield variation due to white heads accounted for 28.38 %, 8.45 % and 3.95 % of the total variation in paddy yield, respectively. The total estimated variation in paddy yield in Sari, Amol and Rasht was 1810.50, 2377.6 and 2176.47 kg ha-1, respectively. Average data over the three regions indicated that highest loss in paddy yield was observed in non-GM 'Nemat', 'Khazar' and 'Tarom Hashemi'. GM cultivars derived from 'Khazar' showed significantly lower paddy yield loss than the non-GM parent. Dead heart, a condition that occurs in the vegetative stage in which the stem borer larva enters the stem and feeds on the growing shoot, causing the central shoot to dry, as well as white heads, which is a condition in which whole ear heads of adult plants become dry and yield chaffy grains, in all three regions were important variables contributing to paddy yield loss. In the future, producing GM rice resistant to striped stem borer with an active promoter in the reproductive growth stage might allow farmers to reduce a significant part of paddy yield loss resulting from white heads, which is directly negatively correlated with filled spikelets per panicle (R2 = -0.57**), in order to achieve an ideotype.