Using Gaussian plume– exponential approach to simulate the pollen-mediated gene flow (PMGF) of maize in the landscape scale: a case study for Puzih City, Chiayi County

• Main Authors: Sheng-Hua Yu, 余昇驊
• Other Authors: Bo-Jein Kuo
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/beqd2z

碩士 === 國立中興大學 === 農藝學系所 === 102 === The genetically modified maize (GM maize) is the wind-pollinated crop. It is reported sure that the pollen-mediated gene flow (PMGF) is the major way in maize gene flow. Until now, there is no evidence to prove that the GM food is harmless in human health. Because no GM crops can legally planted in the open field in Taiwan, it is important to model the PMGF of maize and simulate the coexistence measure before the GM crops are allowed to be planted in Taiwan. To investigate the influence of gene flow of maize in Taiwan, our study assessed the maize pollen dispersal experiments in the worst-case for modeling the PMGF between pollen donor and recipient field at Puzih in 2009 and 2010. The donor field used the commercial variety named Black pearl and the recipient field used Tainan No. 23. Both varieties were the popular commercial varieties in Taiwan. At harvest, we sampled the tassels at the recipient field and quantified the cross-pollination (CP) rate in terms of the proportion of contaminated kernels by the Xenia effect. To predict the CP % in the real agricultural landscape, our study used the Gaussian plume model (GPM) to calculate the relative pollen concentration in the gridded by means of inputting the hourly mean wind speed (m/s) and the hourly wind direction of instantaneous wind speed. Finally, we used the exponential model to transform the relative pollen concentration to CP %. Our study called this estimated method as Gaussian－exponential approach. According to the 10-fold cross-validation, the correlation coefficient was greater than 0.85 and the RMSE was lower than 0.03, confirming the ability of the model fitting. Our study used the Gaussian－exponential approach to predict maize PMGF in the landscape scale, and simulate three coexistence measures, e.g. implemented isolation perimeters, separated border rows and clustering planted GM maize. However, the average field area of the agriculture landscape in Taiwan was too small to implement the isolation perimeter as the coexistence measure. Our study suggested that implemented a complex coexistence measure by clustering GM field at the downwind area and harvested GM-facing rows in 20m separately. If the remained area of the non-GM field was larger than 1 ha, theoretically, the average CP % of the non-GM field could be lower than the EU threshold 0.9%.