RESUMEN
A series of rotation experiments at five sites over four years has explored the environmental and agronomic implications of growing herbicide tolerant oilseed rape and sugar beet. This paper reports on the population dynamics of volunteer rape (Brassica napus). The experiments compared four winter oilseed rape (WOSR) cultivars: a conventional cultivar (Apex) and three developmental cultivars either genetically modified (GM) to be tolerant to glyphosate or glufosinate, or conventionally bred to be tolerant to herbicides of the imidazolinone group. Seed losses at harvest averaged 3575 seeds m(-2) but ranged from less than 2000 up to more than 10000 seeds m(-2). There was a rapid decline in seed numbers during the first few months after harvest, resulting in a mean loss of seeds of 60%. In subsequent seasons, the seedbank declined much more slowly at four of the five sites (ca 20% per year) and the models predicted 95% seed loss after approximately 9 years. Seed decline was much faster at the fifth site. There were no clear differences between the four cultivars in either the numbers of seeds shed at harvest or in their subsequent persistence. The importance of the persistence of GM rape seeds, in the context of the coexistence of GM and non-GM crops and the role of good management practices that minimize seed persistence, are discussed.