|Hellmich Ii, Richard|
Submitted to: Environmental Entomology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/17/2007
Publication Date: 2/1/2008
Citation: Prasifka, J.R., Hellmich II, R.L., Dively, G.P., Higgins, L.S., Dixon, P.M., Duan, J.J. 2008. Selection of Nontarget Arthropod Taxa for Field Research on Transgenic Insecticidal Crops: Using Empirical Data and Statistical Power. Environmental Entomology. 37(1):1-10.
Interpretive Summary: Transgenic crops have been genetically modified to have several beneficial traits for agriculture. Some crops produce substances that kill certain destructive crop-feeding insects, but do not harm animals other than insects. Researchers test for any possible negative effects on other insects (called ‘nontargets’), including beneficial species that pollinate plants or feed on pests. If there is little evidence to suggest which beneficial or non-pest insects might be affected, researchers may attempt to examine many different groups. Looking at many groups or species makes it more likely that not enough good information will be collected to detect problems with any of the nontargets. One way to improve this type of research is to focus on fewer groups of insects and study them more closely. Groups of nontarget insects can be chosen based on how common and reliably collected they are in a crop, which can be determined from previous research. Analysis of several previous nontarget studies in corn suggests that there are several good candidate nontargets from groups of insects that are beneficial in different ways. This process for selecting nontarget insects to study remains flexible, allowing some insects that are uncommon or difficult to collect to be studied if they are judged important to humans. This information is useful for industry, government, and academic stakeholders interested in testing for potential nontarget effects of transgenic crops.
Technical Abstract: One of the possible adverse effects of transgenic insecticidal crops is the unintended decline in the abundance of nontarget arthropods. Field trials designed to evaluate potential nontarget effects can be more complex than expected because decisions to conduct field trials and the selection of taxa to include are not always guided by the results of laboratory tests. Also, recent studies emphasize the potential for indirect effects (adverse impacts to nontarget arthropods without feeding directly on plant tissues), which are difficult to predict because of interactions among nontarget arthropods, target pests, and transgenic crops. As a consequence, field studies may attempt to monitor expansive lists of arthropod taxa, making the design of such nebulous studies more difficult and reducing the likelihood of detecting any positive or negative effects that might be present. To improve the taxonomic focus and statistical rigor of future studies, existing field data and corresponding power analyses may provide useful guidance. Analysis of control data from several nontarget field trials employing repeated measures designs suggests that while detection of small effects may require considerable increases in replication, there are taxa from different functional groups that are sampled effectively using standard methods. The use of empirical data and statistical power to guide selection of taxa included in nontarget trials reflects scientists’ inability to predict complex interactions, but does not preclude employing scientific judgment. Rather, it seeks to more objectively define the focus of future field studies, while allowing a few taxa of special interest to be included because of their perceived value to humans.