|Firbank, Les - INST ECOL & HYDROL, UK|
Submitted to: Science
Publication Type: Other
Publication Acceptance Date: September 13, 2000
Publication Date: N/A
Technical Abstract: An article appearing in the same issue of Science as this perspective implies that genetically modified herbicide tolerant (GMHT) crops may be detrimental to seed-eating birds that are dependent upon weed seeds. This effect supposedly occurs because herbicides used in GMHT crops may provide better control of weeds than those used in conventional crops. This conclusion is derived from a modeling exercise involving a weed and a GMHT crop. A large number of questionable assumptions are used in this sugarbeet/lambsquarters model, and errors of logic result from these assumptions. For instance, maintenance of weed seeds as an over-winter food resource can not logically occur in sugarbeet because weed seeds produced during summer are buried beyond reach of birds during harvest of sugarbeet roots in autumn. The model's assumption that weed control in conventional crops can never be greater than that in GMHT crops contradicts many recently published articles in the weed science literature. Furthermore, reasons for adoption of GMHT technology by farmers seems to have more to do with simplifying weed management than obtaining better weed control, although certainly good control of weeds is a characteristic of GMHT systems. Although the model may not simulate food resources for farmland birds realistically, it does provide food for thought regarding trophic- level consequences of new technologies. In summary, too many problems exist with the model for its results to be taken seriously, and national policy decisions on acceptance of GMHT crops based upon results from such a model may be flawed.