The Potential Risk Of Gene Flow From Transgenic Rice To Different Weedy Rice Biotypes And Fitness Of Their Hybrids And Backcross

The development and commercialization of transgenic rice with novel traits may promote rice productivity. As rice is a major food crop in China, the enhancement of rice production is important for national food security. If left unaddressed, the potential biosafety issue over the release and commercial cultivation of transgenic rice may hamper the development and application of this technology. One of the ecological concerns over genetically modified herbicide-resistant crops is the gene flow to their wild/weedy relatives, which might result in a menace to local biodiversity. For this reason, it has to be assessed before their release in field. Crop-to-wild/weedy gene flow mediated by pollen involves the efficiency of hybridization, backcross and fitness of their hybrids. These are the focuses of our research.From 2006 to 2007, the F1 and F2 generation hybrids between transgenic herbicide-resistant rice (Y0003 and 99-1) and Malaysia weedy rice(WR1) as well as Anhui weedy rice (WR2) were backcrossed with corresponding weedy rice to obtain backcross generations, and these experiments were performed in greenhouse. We counted the seed setting rates of the backcross generations, tested the proportion of plants carrying the resistance gene in the backcross generations, and observed the fitness of hybrid and backcross generations in order to comprehensively assess the risk of gene flow from herbicide-resistant rice to weedy rice. The results showed that the hybrid generations were able to backcross with corresponding weedy rice and set seed with the range of 15-60% the seed setting rate. The results also indicated that the seed from reciprocal hybrid F1 generation showed favorable herbicide-resistance, but F2 and backcross generations showed 3:1 and 1:1 segregation proportion of resistance respectively, which followed Mendelism genetic law. The study on fitness showed that the fitness of hybrid and backcross generations was not obviously different from that of their corresponding weedy rice. Most hybrid generations were not largely different from backcross generations in fitness, whereas some hybrid generations were lower than backcross generations in fitness. All above results indicate that the herbicide-resistant gene flow from transgenic rice possibly occurs via hybridization and backcrossing.The manual and natural crossing experiments in the network room and field between 2007 and 2009 were conducted to compare the compatibility and outcrossing rates of transgenic rice with those of selected different weedy rice biotypes and to elucidate the key innate factors causing the different outcrossing rates. Hybrid seed sets from manual crossing procedures between transgenic rice and weedy rice varied from 31.82% to 82.68%. The significant differences in the quantity of germinated donor pollens and pollen tubes entering the weedy rice ovule directly contributed to the different seed sets. The number of transgenic pollen grains adhering to the stigma of weedy rice biotypes was related to the morphological characteristics of their stigma. Results showed that seed sets of manual crosses were negatively correlated to the genetic distances between Y0003 and the different weedy rice biotypes.The natural outcrossing rates varied from 0 to 6.66%o. Plant and panicle height affected outcrossing result. Results showed that the height of transgenic rice Y0003 (144.8cm) is just larger than all the weedy rice biotypes. The weedy rice biotypes much higher than rice Y0003 (Zhanjiang and Kunming) and the ones much lower than rice Y0003 (Inner Mongolia and Heilongjiang) did not outcross at all with transgenic rice Y0003 unlike the Lianyungang biotype whose plant height was similar to rice Y0003. The Guangde, Jinhua and Wuhu biotypes, which are of a similar height to Y0003, showed the highest outcrossing rates. The panicle height exhibited similar effects on outcrossing rate to plant height.The duration of flowering overlap was the key factor influencing natural outcrossing which was determined by the time of first and final flowering, peak flowering, flowering duration and plant batch of flowering overlap with Y0003. The study showed that peak flowering, time of final flowering, flowering duration, and plant batch of flowering overlap between weedy rice biotypes and Y0003 directly influenced the outcrossing rates. Panicle length and number of effective tillers did not influence the outcrossing rates. The peak flowering of most of the weedy rice biotypes occurred between 08:20am and 10:20am except for Wuhu weedy rice (10:20-11:30). Y0003 rice flowered between 10:00 and 13:35, and its peak flowering time (11:10-12:20) overlapped with that of Wuhu weedy rice, which exhibited the second largest outcrossing rate for about 20 minutes. The weedy rice flowering overlaps with the 4 batches of Y0003 showed a positive relationship with outcrossing rate. Florescence batch of 4 transgenic rice plants around weedy rice was significantly related to the outcrossing rates. Florescence batch of rice and weedy rice depended on their flowering duration. The flowering duration of Guangde weedy rice which exhibited the largest outcrossing rate was the longest, and overlapped with all 4 batches of Y0003.The genetic distance between the various weedy rice biotypes and the transgenic rice affected the seed sets of hybrids. The genetic distance represented the compatibility. And the amount of germinated pollen and the numbers of pollen tubes entering the ovule as determined by morphological characteristics of their stigma essentially determined the different outcrossing rates between transgenic rice and the weedy rice biotypes. All these factors in combination affected the outcrossing rate. While the likelihood of gene flow between transgenic rice and weedy rice biotypes was predominantly determined by floral synchronization and subordinately influenced by genetic compatibility and morphological characteristics such as plant height difference and so on.From 2009, fitness components of parents and F1 hybrids were compared among four types of weedy rice Jinhua (JH), Guangde (GD), Yangzhou (YZ), Wuhu (WH) and herbicide-resistant transgenic rice Y0003 under different cultivation patterns, planting proportions and herbicide treatment. We measured plant height, tiller number/plant, stem diameter, flag leaf length, flag leaf width, flag leaf area in vegetative growth, and pollen viability, panicle number/plant, spikelet number/panicle, filled grain number per panicle, filled grain number per plant, seed set of self-pollination, panicle length, seed shattering, 100-grain weight, seed dormancy in reproduction. The results showed that cultivation pattern, weedy rice populations and herbicide treatment significantly affected the trait differences between F1 hybrids and parents.Cultivation pattern significantly affected tiller number, flag leaf length and stem diameter of F1 hybrids in vegetative growth and panicle number/plant, filled grain number per plant and seed shattering of F1 hybrids in reproductive growth. Spraying herbicide reduced the production of hybrid F1. Mixed planting mainly affected the reproductive growth of F1 hybrid. Biotypes of four weedy rice played significant roles in the relations of F1 and parents in all the variables except plant height, flag leaf area, pollen viability, filled grain number per panicle and seed set of self-pollination. Total fitness of weedy rice and F1 were not significantly differet from transgenic rice Y0003 without select pressure (not spraying herbicide and planting individually). While setting select pressure (herbicide spraying, mixed species and cultivation pattern), hybrid F1 's total fitness was significantly lower than rice parents, but had no significant differences from weedy rice parents. Planting proportion and cultivation pattern had no effect on the dormancy of rice, weedy rice and F1 hybrids.From 2009, the fitness of herbicide-resistant transgenic rice Y0003, Guangde weedy rice and their hybrid F1 was compred to assess their survival compatibility under different cultivation and herbicide treatments in the field. We measured plant height, tiller number/plant, stem diameter, and flag leaf area in vegetative growth, and panicle number/plant, spikelet number/panicle, panicle length, filled grain number per panicle, filled grain number per plant, seed set of self-pollination, seed shattering,100-grain weight, pollen viability and seed dormancy in reproduction.The result showed that in Jiangpu field and Pailou plot experiments, planting proportions and cultivation pattern significantly affected the traits of both GD Fi and rice Y0003. In Jiangpu experiment, planting proportion affected tiller number/plant, panicle length, stem diameter, panicle number/plant, filled grain number per plant, and 100-grain weight of GD F1 and rice parent, while cultivation pattern affected flag leaf area, spikelet number/panicle, seed shattering and 100-grain weight of GD F1 significantly. In Pailou plot experiment, planting proportion of GD F1 to rice significantly affected panicle length, and cultivation pattern affected plant height, tiller number/plant, panicle number/plant, and filled grain number per plant significantly. In the case of herbicide spraying, no matter what cultivation pattern or planting proportion were, the values of plant height, stem diameter, spikelet number/panicle and 100-grain weight of GD ?were significantly higher than those of the rice parents, and the values of pollen viability, filled grain number per panicle, filled grain number per plant and seed set of self-pollination were significantly lower than those of rice parent. With the increase of GD F1 number, its leaf area changed from not differently to significantly larger than the rice parent. Seed dormancy of GD F1 from the two sites was much higher than that of rice Y0003.In summary, gene flow from transgenic rice to weedy rice can happen. However, frequency of gene flow to different weedy rice biotypes were significantly different, which was mainly determined by the degree of floral synchronization. Both compatibility determined by genetic distance and morphological characteristics such as plant height difference affected the gene flow frequency too. Although the fitness of hybrids and backcross was not higher than those of the parents, after setting select pressure (cultivation pattern, mixed species, and spraying herbicide), hybrid F1's total fitness was significantly lower than those of rice parents, but had no significant difference from those of weedy rice parents. However, further experiments need to be done to address to the change of ecological fitness of hybrids and backcross in order to assess the environmental safety of transgenic rice accurately.