Landscape effects of transgenic cotton on non-target ants and beetles

Transgenic crops producing toxins from Bacillus thuringiensis (Bt) can be planted in the same field for many years, and many insects exploiting such crops must disperse to other habitats to persist. Accordingly, effects of transgenic crop farming could accumulate through time and affect insect populations across agricultural landscapes. We monitored the population density of seven ant genera and beetle families and of rare ants and beetles in 84 non-cultivated sites abutting agricultural fields in Central Arizona. We assessed the short-term (during planting year) and long-term (over 5–6 years) landscape effect of farming Cry1Ac cotton on ant and beetle density in non-cultivated sites, in addition to several local and regional variables. Landscape variables (e.g., sequence of crops planted in neighbouring fields, crop diversity, and abundance) were more frequently associated with insect density than local variables (e.g., plant productivity and diversity in non-cultivated sites). In the short-term, use of Bt relative to non-Bt cotton in neighbouring fields was positively associated with density of one ant and two beetle groups in non-cultivated sites. However, acreage of Bt cotton located within 1 km from non-cultivated sites had more negative effects than acreage of non-Bt cotton on density of one ant and one beetle group. In the long-term, the proportion of years that Bt cotton was planted in neighbouring fields was positively associated with ant density but not beetle density. Results suggest that the farming of Bt cotton in neighbouring fields frequently resulted in positive short- and long-term landscape effects on ants and beetles in non-cultivated sites, while Bt cotton planted farther away had less frequent negative short-term impacts.