Modelling Gene Flow between Fields of White Clover with Honeybees as Pollen Vectors

The portion-dilution model is a parametric restatement of the conventional view of animal pollination; it predicts the level of pollinator-mediated gene dispersal. In this study, the model was applied to white clover (Trifolium repens) and its most frequent pollinator, the honeybee (Apis mellifera). One of the three parameters in the portion-dilution model is the mean number of flowers a pollinator visits in one foraging bout. An alternative method to estimate this parameter was developed that was not depending on pollinator hive-seeking behaviour. The new estimation method, based on nectar collection, seems to be a good alternative, when reliable observation on visiting behaviour of pollinators is not possible. The gene flow in white clover was modelled. Where fields were assumed to be well separated, and only a low fraction of bees travelled between fields, the gene flow was estimated to be 0.7%, but subjected to large uncertainty. In a worst case scenario with adjacent fields—one with a genetically modified (GM) T. repens cultivar and the other with a conventional T. repens cultivar—and where all arriving bees were expected to transfer GM pollen, the median gene flow was modelled to be 7% with an estimated 95% percentile of 70%. The results show that the European Union threshold limit of 0.9% GM admixture for food and feed will likely be exceeded at times and especially organic farmers that do not accept GM admixture and often have clover and clover–grass fields might face challenges with admixture of GM.