| Summary: | In many regions of the world, the crop, oilseed rape (Brassica napus), is giving rise to populations of volunteer weeds in fields and feral plants outside fields, both of which can retain crop genes and hybridize with compatible wild relatives. Feral oilseed rape has received global attention as a means by which genetically modified (GM) traits may persist in the environment. There are still major uncertainties, however, over the long term environmental and economic consequences of its persistence and invasiveness, particularly in relation to GM coexistence and environmental risk assessment. This thesis presents a demographic study of feral oilseed rape over an 11 year period from 1993 to 2004 within a 500 km2 area of Tayside (Scotland). The number of feral oilseed rape populations increased almost five-fold during a period when the number of fields and total area cropped with oilseed rape decreased. Ferals did not usually remain at the same location for more than one or two years, and did not spread by gradual movement out from the sites of initial colonization. They persisted and spread in the region by occurring at different places each year, most likely through long range dispersal. Transport corridors hosted higher densities than farmland, in which ferals were more prevalent in areas having a high density of oilseed rape crops. The insect communities associated with feral oilseed rape and a related ruderal plant charlock were compared to gauge the potential ecological impact of ferality. Ferals did not appear to compete with charlock but provided an additional host for those invertebrate species already living on charlock. They also had the potential to function as a bridge for crop pests between growing seasons. At current levels of feral oilseed rape there are unlikely to be any issues related to coexistence (i.e. ferals will bring a negligible contaminant to crops), but feral oilseed rape can persist and flower outside the range of cropped oilseed rape plants. It has become part of the native weed and wildflower community, but to date has had no major ecological impact. The long term demographic changes in feral oilseed rape that were found in the 11 year study could not have been predicted from the initial early years when there were few populations or from prior estimates of risk carried out at small spatial scales. A long term approach is therefore needed at realistic scales for successful ecological risk assessment. The Tayside study could provide a baseline and model for assessing the ecological impact of new GM traits such as cold tolerance or insect resistance.
|
|---|
