CONCERNS about genetically engineered crop plants centre on three conj
ectural risks: that transgenic crop plants will become weeds of agricu
lture or invasive of natural habitats; that their engineered genes wil
l be transferred by pollen to wild relatives whose hybrid offspring wi
ll then become more weedy or more invasive; or that the engineered pla
nts will be a direct hazard to humans, domestic animals or beneficial
wild organisms (toxic or allergenic, for example). Here we describe an
experimental protocol for assessing the invasiveness of plants. The o
bject is to determine whether genetic engineering for herbicide tolera
nce affects the likelihood of oilseed rape becoming invasive of natura
l habitats. By estimating the demographic parameters of transgenic and
conventional oilseed rape growing in a variety of habitats and under
a range of climatic conditions, we obtain a direct comparison of the e
cological performance of three different genetic lines (control, kanam
ycin-tolerant transgenics and herbicide-tolerant transgenic lines). De
spite substantial variation in seed survival, lines were less invasive
and less persistent than their conventional counter arts.