Herbicide resistance in problem weeds is now a major threat to global food production, being particularly widespread in wild grasses affecting cereal crops. In the UK, black-grass (Alopecurus myosuroides) holds the title of number one agronomic problem in winter wheat, with the
loss of production associated with herbicide resistance now estimated to cost the farming sector at least £0.5 billion p.a. Black-grass presents us with many of the characteristic traits of a problem weed; being highly competitive, genetically diverse and obligately out-crossing, with
a growth habit that matches winter wheat. With the UK’s limited arable crop rotations and the reliance on the repeated use of a very limited range of selective herbicides we have been continuously performing a classic Darwinian selection for resistance traits in weeds that possess great
genetic diversity and plasticity in their growth habits. The result has been inevitable; the steady rise of herbicide resistance across the UK, which now affects over 2.1 million hectares of some of our best arable land. Once the resistance genie is out of the bottle, it has proven difficult
to prevent its establishment and spread. With the selective herbicide option being no longer effective, the options are to revert to cultural control; changing rotations and cover crops, manual rogueing of weeds, deep ploughing and chemical mulching with total herbicides such as glyphosate.
While new precision weeding technologies are being developed, their cost and scalability in arable farming remains unproven. As an agricultural scientist who has spent a working lifetime researching selective weed control, we seem to be giving up on a technology that has been a foundation
stone of the green revolution. For me it begs the question, are we really unable to use modern chemical and biological technology to counter resistance? I would argue the answer to that question is most patently no; solutions are around the corner if we choose to develop them.
Synthetic evolution of herbicide resistance using a T7 RNAP-based random DNA base editor
