Striving for sustainable pest management in no-till, field-crop systems: Understanding the role of insecticidal seed treatments

Striving for sustainable pest management in no-till, field-crop systems: Understanding the role of insecticidal seed treatments

Summary

The aim of this study is to examine whether a common insecticidal seed treatment (thiamethoxam) poses non-target risks to natural enemies in no-till corn and soybeans. In particular, we are exploring the potential for this seed treatment to travel up food chains from plants to slugs to natural enemies, potentially disrupting biological control. In 2012, we followed-up our previous laboratory studies by conducting two field experiments, one in corn and one in soybeans, comparing plots that were planted with thiamethoxam-treated or untreated seed. While we are still processing the data from these experiments, the preliminary data are informative. In soybeans, thiamethoxam seed treatment reduced early-season pitfall catches of some predatory taxa, increased pitfall catches of slugs, and undermined stand establishment. Early season predation on sentinel caterpillars was also marginally reduced in treated plots. We did not observe similar effects in corn; however, early season flooding in that field site complicated data collection and interpretation. For the outreach component of this project, we shared our laboratory results and preliminary field results with farmers at several field days, including one hosted by the Pennsylvania No-Till Alliance. In addition, we shared our results with researchers at several scientific meetings, participating in symposia related to insecticidal seed treatments and IPM. In 2013, we will finish processing data from the 2012 field experiments, complete a final laboratory experiment, and complete our outreach efforts.

Objectives/Performance Targets

• We completed a soybean laboratory experiment on this question in 2011. A corn laboratory experiment was planned for 2012 but postponed to 2013 (see Accomplishments/Milestones).

• We conducted soybean and corn field experiments addressing this objective in 2012; preliminary results are in and remaining predator samples will be processed in 2013.

• We gave several presentations to farmer and extension groups in 2012. Work on videos and publications is in progress and will be completed in 2013.

Accomplishments/Milestones

In 2011, we conducted laboratory experiments to determine whether neonicotinoid seed treatments can travel up the food chain from soybean seedlings to slugs to their insect natural enemies. The results of this experiment were reported in the 2011 Annual Report. In 2012, we had hoped to conduct similar laboratory experiments using corn as the focal plant. However, weather conditions in 2012 made it difficult to collect an adequate number of beetles and slugs of the appropriate size for these experiments. We now plan to complete the corn experiment in 2013.

This year we conducted two large field experiments, one in soybeans and one in corn, to examine the influence of neonicotinoid seed treatments on natural enemies and the predation services they provide. Each experiment comprised corn or soybean plots planted with untreated or high-rate thiamethoxam-coated seeds (n = 6/treatment, plots = 0.25 acre), arranged in a randomized block design. Within these experiments, we measured agronomic characteristics including crop establishment, plant damage, and yield. In addition, we monitored slugs weekly through the season using refuge traps (6/plot), and also measured slug activity-density at three times during the season using pitfall traps (4/plot). These pitfall samples also gave insight into predator activity-density. Predator abundance was also measured using quadrat samples on three sample dates (4 samples/plot, 0.10m2 quadrats). We measured predation services by deploying sentinel waxworm caterpillars (10/plot) and weed seed cards (50 seeds/card; 3 cards/plot) at three times during the season. Predation on sentinel caterpillars was measured over two, 12-hr periods to capture day versus night predation. Seed cards were left out for 48 rain-free hours. We had originally planned to use sentinel slugs to measure slug predation, but the method of enclosing slugs in a salt moat turned out not to be feasible in a field setting.

Results from the soybean and corn experiments were analyzed with ANOVA in PROC MIXED (SAS), with block as a random factor and seed treatment as a fixed factor. In cases with repeated measures, time was included as a repeated factor. We predicted that seed treatments would adversely affect ground-dwelling predators, with potentially cascading effects on slug populations and crop performance. We also predicted that this effect would be strongest early in the season when the concentration of insecticide in plant tissues would be high. Here we present preliminary results from these experiments. We are still processing pitfall samples and quadrat samples, so our data on predator activity-density is limited to pitfall data from the first sample date. We plan to finish processing these samples in 2013.

Soybean field experiment

In soybeans, thiamethoxam seed treatment reduced early season pitfall catches of ground beetles by 29% (P = 0.045) and rove beetles by 37% (P = 0.03) (Fig. 1). No significant differences were seen for ants, wolf spiders, sheet web spiders, harvestmen, or predaceous beetle larvae. A non-significant trend toward reduced catches in the treated plots was seen for parasitic wasps (P = 0.12) and other spiders (P = 0.10) (Fig. 1). Consistent with the decrease in some predator groups, early season predation on sentinel caterpillars was marginally reduced in treated plots (P = 0.09; Fig. 2). Later in the season, there were no significant main effects of seed treatment on attack of sentinel caterpillars (July: P = 0.89; August: P = 0.81), but in August, there was a significant Treatment*Time effect (P = 0.04), apparently driven by reduced predation in the treated plots at night (Fig. 3). Predation on weed seeds did not differ by treatment on any of the three sample dates (Treatment P = 0.28; Treatment*Time P = 0.19). Activity-density of slugs measured via shelter traps did not differ between treatments (P = 0.25); however, on the first sample date, pitfall trap captures of slugs in treated plots were 69% higher than controls (P < 0.01; Fig. 4). Seed treatment had mixed effects on slug damage to soybean seedlings and appeared to alter the pattern of slug feeding on seedlings. Seed treatment reduced damage to soybean leaves (P = 0.03), but marginally increased the percentage of seedlings killed by slugs (P = 0.09). Perhaps related to this latter point, soybean establishment was reduced by 19% where seed treatment was present (P < 0.001; Fig. 5). In spite of the reduced stands in the treated plots, final yield did not differ significantly between treatments (P = 0.39; Fig. 6). Interestingly, yield did appear to be more variable where seed treatment was present (Fig. 6).

Corn field experiment

The corn field experiment suffered from several set-backs including severe flooding shortly after crop emergence and very high weed pressure in many of the plots. These factors complicated data collection and interpretation to some degree, but for the most part the experiment was completed as planned.

The first planned pitfall sample in corn could not be completed because many plots were under several inches of water. A pitfall sample was completed as soon as plots had dried out sufficiently, which was in early June, ~5 weeks after planting. Catches of predators in pitfall traps from that sample did not differ significantly, though catches of sheet-web spiders and rove beetles were marginally reduced where seed treatment was present (sheet-web spiders, P = 0.08; rove beetles, P = 0.07). Similarly, predation on sentinel caterpillars did not differ by treatment on any of the three sample dates (Main treatment effect: May, P = 0.18; June, P = 0.26; July, P = 0.53), nor did disappearance of weed seeds differ by treatment (Treatment P = 0.49, Treatment*Time P = 0.61). Slug activity-density as measured under shelter traps and in the first pitfall sample also did not differ by treatment (shelter traps: P = 0.83; pitfall traps: P = 0.13). Corn plants were heavily damaged by pests (slugs and cutworms) early in the season, and plant populations were generally on the low side (~26,000 plants/acre at V3), but seed treatments did not significantly improve or reduce stand establishment (P = 0.41). Consistent with this, the percentage of seedlings severed by cutworms at three early growth stages was similar between treatments (P = 0.74), as was the severity of damage to seedlings by slugs (P = 0.59). Corn yield did not differ by treatment (P = 0.54).

This year we shared our results widely with farmers and extension professionals through presentations in diverse venues. We gave two extension presentations at Penn State’s Russell E. Larson Agricultural Research Farm, one in June as part of a collaborative field day on “Strategies for Soil Health and Nutrient Conservation,” and one in July as part of an annual “Agronomic Weed and Entomology Field Day.” Perhaps most significantly, in October we presented at the request of the Pennsylvania No-Till Alliance at their “Maintaining Soil Health Field Day” in Loganton, PA. Through this outreach event we reached dozens of no-till farmers and also field scouts who often advise farmers on pest management practices. In addition to sharing our preliminary results, at these events we educated farmers generally about the important role played by natural enemies in biological control of crop pests, using live predator specimens whenever possible.

We also continued to gather video footage of ground beetles influenced by neonicotinoid poisoning, and collaborated with another graduate student to take video footage of common natural enemies attacking prey. We plan to edit these videos and make them available via the Penn State Entomology website so that farmers and others can better appreciate the importance of predatory insects. We will also continue working on two farmer-friendly publications on early-season IPM in corn and soybeans, for completion in 2013.

• Figure 3. Effect of thiamethoxam seed treatment on predation on sentinel caterpillars in soybeans in August (T = treated with thiamethoxam, U = untreated). Predation was measured for 12 hours in daytime (AM-12hr) and 12 hours overnight (PM-12hr).
• Figure 1. Effects of thiamethoxam seed treatment on pitfall catches of select predators ~3 wk after planting soybeans (T = treated with thiamethaxam at 0.152 mg/seed, U = untreated). Asterisks indicate significant differences at P = 0.05.
• Figure 4. Effect of thiamethoxam seed treatment on captures of slugs in pitfall traps ~3 wk after planting soybeans (T = treated with thiamethoxam, U = untreated).
• Figure 5. Effect of thiamethoxam seed treatment on soybean establishment at three early growth stages: VC (cotyledon), V1 (one true leaf), and V3 (three true leaves) (T = treated with thiamethoxam, U = untreated).
• Figure 6. Effect of thiamethoxam seed treatment on soybean yield.
• Figure 2. Effect of thiamethoxam seed treatment on predation on sentinel caterpillars ~3 wk after planting soybeans (T = treated with thiamethoxam, U = untreated). Predation was measured for 12 hours in daytime (AM-12hr) and 12 hours overnight (PM-12hr).

Impacts and Contributions/Outcomes

Although the results from this project are not yet complete, the preliminary results shed some light on the place of neonicotinoid seed treatments in IPM for no-till corn and soybeans. In the lab, slugs were able to transfer neonicotinoids from soybean plants to ground beetles, suggesting a possible risk of food chain transfer of these common insecticides. Results from the field were variable. In soybeans, results were generally consistent with the possibility of trophic movement of insecticides. This was reflected in reduced early season pitfall captures of certain predators (particularly ground beetles, known slug predators), a trend toward reduced early season predation on sentinel caterpillars, increased pitfall captures of slugs, and reduced stands. In corn, results from the field were inconclusive with few statistical differences in treated versus untreated plots. Seed treatments did not increase or decrease yield in either experiment. While there is still a fair amount of data from these experiments yet to process, these preliminary results suggest that the perceived benefits of seed treatments are far from certain in slug-prone corn and soybean fields under no-till management. Furthermore, our results suggest that in at least some cases, seed treatments carry a risk of disrupting populations of predatory insects.

In 2012 we shared our research widely with other researchers and extension professionals through presentations at scientific meetings. These included presentations to entomologists at the meeting of the Eastern Branch of the Entomological Society of America in March, and the national meeting of the ESA in November, where we presented talks in symposia on “Ecological considerations of the rising use of systemic insecticides” and “Risks and benefits of global expansion of transgenes and insecticide seed treatments in field crops.” Also, we shared our research findings with slug and snail experts at the Annual Meeting of the American Malacological Society in June. Through these presentations we have helped to move forward the scientific conversation about the costs and benefits of insecticidal seed treatments in field crops.

Collaborators: