• Abstract
• Introduction
• Methods
• Results
• Discussion
• Conclusions
• Acknowledgments
• Citations

Introduction
Ohio is ranked 7th nationally in field corn and soybean production (Agricultural Statistics, 1999).  Approximately 3.5 million acres of corn and 4.5 million acres of soybeans are planted in Ohio each year. In the last few years, transgenic field corn acreage (primarily Bt corn) has reached nearly 350,000 acres (pers. comm. H. Willson) and transgenic RR soybeans has exploded to nearly 2 million acres (pers. comm. M. Loux).  While the move to agricultural plant biotechnology has undergone a cooling trend in 2000, regulatory agencies, university researchers, industry, farmers and consumers have been engaged in an information tug of war about the potential benefits, harmful effects, and proper use of transgenic plants. 

One area of concern is the potential impact transgenic field crops have on pest natural enemies (Figure1).  The primary function of Bt corn is to deliver a lethal dose of toxin via plant tissue to immature Lepidopteran or Coleopteran pests as they begin feeding on the plant.  It is not clear if certain non-target organisms, especially natural enemies, are being negatively impacted by either loss of potential prey or feeding on prey that have consumed transgenic tissue.  Schuler et. al. (1999), put forth many arguments and data sets supporting either side of this issue.  The primary function of RR soybeans is to allow glyphosate herbicide to be applied over the crop for weed control.  Therefore, the type of interaction or exposure beneficial insects have to transgenic soybeans is fundamentally different than those in Bt corn fields.  Recently, Buckelew et. al. (2000) concluded that the difference in insect populations in transgenic soybean fields was due more to the effect of weed control and not the herbicide used in those fields. 

The purpose of this study is to provide empirical results on the impact of transgenic and non-transgenic field crops on natural enemy populations in Ohio.

Methods
Thirteen transgenic (6 Bt corn, 1 RR corn, and 6 RR soybean) and eleven non-transgenic (5 hybrid corn and 6 conventional soybean) fields were selected for this study in the northwestern, southwestern, and central parts of the state (Figure2).  Each region had two survey sites; each site contained a pair of transgenic/non-transgenic cornfields and a pair of transgenic/non-transgenic soybean fields.  One pair of cornfields in Clark County actually consisted of a Bt hybrid and a non-Bt RR hybrid.

Sampling in the soybean fields was accomplished using a sweep net and Pherocon AM yellow sticky traps. Weekly sweep net sampling began at the end of June and was discontinued mid August after pod set.  For the first two weeks, six locations in each field were swept; two at 100, 200, and 300 feet from the field edge. Each pair of sweep net locations was separated by at least 100 feet.  From the third week through the end of the study, only four locations in each field were swept; 100 and 300 feet from the field edge. Ten pendulum sweeps were performed at each location.  Any insects caught in the sweep net samples were placed in Ziploc bags for later classification. 

Two Pherocon AM yellow sticky traps were deployed mid July in the same soybean fields monitored with sweep nets.  The traps were attached to posts above canopy level at approximately 100 and 300 feet from the field edge.  The old traps were collected and replaced with fresh ones every week.  The collected traps were refrigerated or frozen until a later date when they would be fully inspected for beneficial insects and arthropods.  Sticky trap sampling was discontinued mid August.

Sampling in field corn also relied upon the use of Pherocon AM yellow sticky traps, which were attached to the stalk of the corn plant near the ear zone.  Placement of the first trap within the field was at least 24 rows into the field, with the second trap placed an additional 100 feet toward the interior of the field. Both sticky traps were changed weekly in each field starting at the beginning of July and ending around mid August. The collected traps were refrigerated or frozen until a later date when they would be fully inspected for beneficial insects and arthropods.  Sticky trap sampling was discontinued the same week sticky trapping was halted in nearby soybean fields, mid August.

Beneficial insects and arthropods captured in sweep net and sticky trap samples were identified to family or species level, with the notable exceptions of the parasitic Hymenoptera, spiders, and mites, which were placed in those general categories.

Differences in the transgenic/non-transgenic field crop natural enemy data were statistically analyzed by technique (sweep net and sticky trap) and by crop (soybean and corn) at each site.  Site data were then combined into regional data, and the regional data then combined to look for study wide effects. Statistical analysis of data includes parametric two-sample t-test and non-parametric Mann-Whitney median tests (Minitab v. 13.2, 2000, State College, PA).

Results
Sweep net samples between transgenic and non-transgenic soybean fields revealed no statistical differences for any of the 14 natural enemy categories compared (2 sample t-test, P>0.05).  Pooling site data into regions revealed no statistical difference for any of the 14 natural enemy categories (2 sample t-test, P>0.05).  Pooling all site data together revealed no study wide statistical difference for any of the 14 natural enemy categories (2 sample t-test, P>0.05, Figure 3). Very few of the analyzed populations were normal in distribution, therefore a Mann-Whitney non-parametric test was run on the sweep net data set by site, region, and over the complete study.  No statistical differences between the 14 natural enemy categories were found (P>0.05).

Pherocon AM yellow sticky trap data between transgenic and non-transgenic soybean fields revealed an increase in the number of spiders in non-transgenic fields at Champaign county (2 sample t-test, P=0.041).  Comparisons of the remaining beneficial insect categories and other sites revealed no other statistical difference (P>0.05).  Pooling site data into regions revealed no statistical difference for any of the 15 natural enemy categories (2 sample t-test, P>0.05).  Pooling all site data together revealed an increase of green lacewing adults in non-transgenic fields (P=0.015), but none of the other14 natural enemy categories (2 sample t-test, P>0.05, Figure 4). Very few of the analyzed populations were normal in distribution, therefore a Mann-Whitney non-parametric test was run on the sticky trap data set by site, region, and over the complete study.  Combining two sites in the central region showed an increase of spiders in non-transgenic fields (P=0.045).  No other statistical differences between the 14 natural enemy categories were found (P>0.05).

Pherocon AM yellow sticky trap data between transgenic and non-transgenic cornfields revealed an increase of Orius sp. in transgenic fields at Hancock county (2 sample t-test, P=0.008).  Comparisons of the remaining beneficial insect categories and other sites revealed no other statistical differences (P>0.05).  Pooling site data into regions revealed no statistical difference for any of the 15 natural enemy categories (2 sample t-test, P>0.05).  Pooling all site data together revealed no study wide statistical differences (2 sample t-test, P>0.05, Figure 5). Very few of the analyzed populations were normal in distribution, therefore a Mann-Whitney non-parametric test was run on the sticky trap data set by site, region, and over the complete study.  No statistical differences between the 15 natural enemy categories were found (P>0.05).

Discussion
Pesticide use among paired transgenic / non-transgenic sites was looked at to see if any trends appeared (Table 1).  Three of the six sites where both RR and conventional soybeans were planted used glyphosate herbicide (burn down or post application).  Two other sites were comparisons of RR soybeans and STS soybeans, which are technically not classified as a transgenic plants even though they were developed using non-conventional breeding techniques.  One pair of Bt/non-Bt cornfields was actually a Bt and a non-Bt RR cornfield where the Bt field was sprayed with Pounce® in the spring.  Another pair of Bt/non-Bt corn fields was treated with Lorsban 15G® insecticide at planting.  All other soybean and cornfields were insecticide free.  Most soybean fields were planted No-till and most corn fields utilized some type of minimum tillage, such as a field cultivator.  Field size ranged from 10 to 82 acres.

Statistically none of the 14 natural enemy categories collected using a sweep net were different between transgenic and non-transgenic fields.  There were 9 categories where transgenic soybean fields actually yielded more beneficial insects, both generalists and specialists.  Weed populations in these fields are a major consideration affecting beneficial insect populations in soybean fields. Informal notes on weed species and abundance were taken on fields surveyed, but translating those notes into how they specifically affect insect abundance is not clear.  According to the sweep net data collected, it would appear that transgenic soybean plants have an overall neutral effect on the beneficial insects identified by this study. 

There were some statistical differences in the data collected from sticky traps in both transgenic and non-transgenic soybean fields.  Populations of spiders at the Champaign county site and green lacewing adults study wide were significantly higher in non-transgenic soybean fields.  Eight of the fifteen beneficial insect categories which include both generalists and specialists, were more numerous in transgenic soybean fields.  Comparing beneficial insect categories between sweep net and sticky trap samples showed Coleogilla maculata (Cmac lady bird beetle), Harmonia axiridis (Multi-colored Asian ladybird beetle), parasitic Hymenoptera, Hoover flies, and Soldier beetles more common in transgenic soybean fields.  Similar to concerns of sweep net sampling in these fields was the presence and distribution of weeds, which may have impacted the species and abundance of arthropods captured in the traps.

Sticky trap data from transgenic and non-transgenic cornfields revealed statistically higher Orius sp. at the Bt corn Hancock county site.  Additionally, no-spot ladybird beetles, green lacewing adults, and mites were more numerous in transgenic cornfields.  The remaining eleven categories of beneficial insects were higher in non-transgenic fields.  Over 2,000 parasitic wasps were collected in non-transgenic cornfields, about 100 more than transgenic cornfields.  Given the direct impact Bt corn has on European corn borer populations, various life stages of which are parasitized by several families of Hymenoptera, the effect of Bt corn on these organisms is minimal compared to conventional hybrid corn.  This suggests the possibility that alternative hosts (prey) in these fields may be able to support them.

Conclusions
The purpose of this study was to address concerns that transgenic field crops, corn and soybeans, may negatively impact beneficial insect populations.  The results of this survey do not appear to support that assertion.  In fact, out of the fifteen different categories of beneficial insects and arthropods, there are only a few instances where any statistical difference between transgenic and non-transgenic field crops could be detected.  There are instances where specific beneficial insects, both generalists and specialists, were found in greater abundance in transgenic or non-transgenic fields. Based on the data collected, no negative effects on beneficial arthropods can be directly associated with transgenic soybean and corn crops in Ohio.

Acknowledgments 
We would like to thank the Champaign, Clark, Darke, Hancock, Miami, Van Wert, and Wood county extension agents who helped identify the fields necessary for this study.

Citations
Buckelew, L., L. Pedigo, H. Mero, M. Owen, & G. Tylka. 2000. Effects of weed management systems on canopy insects in herbicide resistant soybeans. Journal of Economic Entomology. 93(5), 1437-1443. 

Minitab v. 13.2, 2000, State College, PA.

Schuler, T., G. Poppy, B. Kerry, & I. Denholm. 1999. Potential side effects of insect-resistant transgenic plants on arthropod natural enemies. Trends In Biotechnology. V. 17, 210-216.