Using a zero-inflated model to assess gene flow risk and coexistence of Brassica napus L. and Brassica rapa L. on a field scale in Taiwan

Abstract Background The cropping area of genetically modified (GM) crops has constantly increased since 1996. However, currently, cultivating GM crops is associated with many concerns. Transgenes are transferred to non-GM crops through pollen-mediated gene flow, which causes environmental problems such as superweeds and introgressive hybridization. Rapeseed (Brassica napus L.), which has many GM varieties, is one of the most crucial oil crops in the world. Hybridization between Brassica species occurs spontaneously. B. rapa grows in fields as a weed and is cultivated as a crop for various purposes. Both B. rapa weeds and crops participate in gene flow among rapeseed. Therefore, gene flow risk and the coexistence of these two species should be studied. Results In this study, field experiments were conducted at two sites for 4 years to evaluate gene flow risk. In addition, zero-inflated models were used to address the problem of excess zero values and data overdispersion. The difference in the number of cross-pollination (CP) events was nonsignificant between upwind and downwind plots. The CP rate decreased as the distance increased. The average CP rates at distances of 0.35 and 12.95 m were 2.78% and 0.028%, respectively. In our results, zero-inflated negative binomial models were comprehensively superior to zero-inflated Poisson models. The models predicted isolation distances of approximately 1.36 and 0.43 m for the 0.9% and 3% threshold labeling levels, respectively. Conclusions Cultivating GM crops is prohibited in Taiwan; however, the study results can provide a reference for the assessment of gene flow risk and the coexistence of these two species in Asian countries establishing policies for GM crops.