Kinetic Genetic Analysis And Prediction Using Molecular Markers Of Plant Height In Hybrid Rice

Rice is one of the most important foodstuff crops. Its height is considered as an important agronomic traits, affecting both the potential yield and the lodging resistance. It is important for rice breeding to research the heritance of rice height and to predict hybrid rice height effectively. Only rice height in maturity was conducted on heritance analysis in most previous researches, and the results were much discrepancy for different materials or environments.It is one of the most important measures to increase rice yield by utilization of heterosis. However, selection of parents in traditional breeding was much manpower-comsuming, time-comsuming, having a low efficiency and being affected by environment. This was the challenge for effectively using hererosis. In earlier stage, many methods such as morphological index, biochemical index and so on, were put forward and applied in heterosis predicting. Although these measures had some insightful suggestion, they could not be applied to breeding practise for low accuracy. Molecular marker technology was invented in 1980s, it was a great milestone for heterosis prediction. Unfortunately, there are also many discrepancies in conclusions for different materials, marker types and environments.1.Heritance analysis on rice height at different stagesIn the present research, two half-diallel sets cultivated in Beibei of Chonging and Luzhou of Sichuan respectively, had been conducted heritance analyze of rice height at different stages. The results were as follows.1) Both mean of F₁ height and accumulation height for 10 days are higher than those of their parents for two diellel sets in two location. The results were changable for different parent materials and environments.2) Rice height was attributed to genes which have addictive effect (V_A) and dominant effect (V_D), but the influential degree of them varied among materials, and among environment. The interaction of environment have a certain influence on plant height, and gradually decreased with the height development. the gene in addictive effect and the gene in dominant effect both have some new performances at different development, especially mainly in expression of gene in addictive effect..3) Significant positive HMP and HBP of rice height in two set were showed at different periods, and the value increased gradually with the development progress. In different development stages, both the HMP and the HBP had some new performances, and their values and directions varied with cultivate loci and materials.2. Prediction of F₁ height of hybrid rice by GD based on molecular markerThe relationship between parental genetic distance based on molecular marker and F₁ plant height was analysised for providing a novelty method to prediction heterosis. The results were as follows.1) Positive loci (PL) accounted for 1/5 in total loci, and effect-increasing loci (IL) accounted for 1/2 in PL. The rate of PL and IL were different for varing materials , but environment showed no significant effect.2) The rate (%) of non-environment loci, non-materials loci, effect-opposite loci (in two sets) and non-time type were 40.72%-79.57%, 1.41%-11.71%, 45.45%-93.93%, 67.58%-89.13%.3) Correlationships between general GD and F₁ height were low, and had no significance in most cases. Those for positive GD were higher, being significantly related in most cases. But their absolute value was still low and attributed to materials, environment and marker types. For GD based on effect-increasing loci, Correlationship with rice height were higher than those for PL, and all reached extreme significant level, from 0.6501 to 0.8687, and their value have been affected by materials, and showed no effect on environment.4) Correlation with F₁ height, GD based on effect-increasing loci of non-environment, non-materials and non-time were respectinely 0.5472-0.8401,0.2972-0.7867, 0.3775-0.8368, and all was lower than correlation between F₁ height and GD based on all IL.5) Prediction of F₁ height in different environment by IL was high, with a mean of 0.5651, being lower than correlation coefficient between F₁ height and GD based on non-environment IL, which also reached positive extreme significant level. The prediction effects were affected by different materials and geographical environments.6) Predictions between two sets had great variations in the value and direction of correlationship among two sets in two environments, and absolute value of them was lower than correlation coefficient between F₁ height and GD based on non-materials IL.