Responses of soil faunal food webs to pesticide seed treatments

Responses of soil faunal food webs to pesticide seed treatments

Summary

The use of crop seeds pretreated with pesticides is a common agricultural practice; however, little is known about how pesticide seed treatments affect soil food webs and their associated agroecosystem services. We conducted a two-year field experiment to examine the effects of insecticidal-fungicidal seed treatments on soil food web diversity, structure, and function. Corn (2013) and soybean (2014) with and without pesticide seed treatments were planted as a two-year rotation in a completely randomized design with five replications. We measured soil mesofaunal communities, litter decomposition and crop response over the course of the growing season each year of the study.

Data from the first year of the study were processed and analyzed in 2014. Our preliminary results suggest no community-level shifts in soil mesofaunal composition. At the population-level, however, we observed differential responses to the seed treatment in several taxa, including an unexpected increase in abundance of a targeted pest species, corn rootworm (Coleoptera: Chrysomelidae: Galerucinae: Diabrotica). Rates of aboveground litter decomposition were unaffected by the pesticide seed treatment. Corn biomass early in the summer was significantly greater in the pesticide seed treatment plots compared to the controls; however, this early season advantage did not result in greater grain yields at harvest. In 2015, we will finish processing the 2014 field data, complete the isotopic analysis of the mesofaunal food web, present our results at multiple scientific conferences, and prepare manuscripts for submission to peer-reviewed journals.

Objectives/Performance Targets

We conducted a two-year (corn- soybean rotation) field experiment addressing this objective in 2013 and 2014. All 2013 samples have been processed and analyzed. The samples collected in 2014 will be processed and analyzed in 2015.

This objective was refined from the original proposal to be more succinct with the emerging literature in this field. All proposed data are still being collected.

2. Determine wheather pesticide treated seeds foster soil food webs with less diversified consumer-resource links.

Soil mesofaunal samples collected for Objective 1 have been processed and sorted to family. After we finish processing the 2014 samples, all 2013 and 2014 samples will be dried, weighed and ground for isotopic analysis at the University of New Hampshire Stable Isotope Laboratory (2015). These data will then be used to model and compare changes in the soil food web using the IsoWeb model developed by Kadoya et al. (2012).

3. Determine the effects pesticide seed treatments have on agriculturally important ecosystem serives including decomposition and nitrogen cycling.

To meet this objective, litter bags and cation/anion resin strips were deployed during both the 2013 and 2014 field seasons. Litter bags from 2013 were processed and analyzed. The 2014 litter bags will be processed and analyzed in 2015. All resin strip extracts were frozen immediately after extraction and will be processed this coming year to determine the effects of seed treatment on nitrogen cycling. Crop growth, including plant height and seedling biomass, along with leaf chlorophyll content and yield data were also collected during both field seasons.

Accomplishments/Milestones

All samples collected during the 2013 field season were processed in 2014 (reported in the 2013 Annual Report). This involved extracting, counting, and identifying all soil fauna collected. These data were then analyzed using appropriate univariate and multivariate analyses. The major results from these analyses are presented below.

Data collected in 2013 show little evidence of community-level differences in soil mesofaunal composition. Mesofaunal species richness was not affected by the pesticide seed treatment (ANOVA, p > 0.05). Community composition was further analyzed using per-MANOVA and illustrated using non-metric multidimensional scaling (Fig. 1a,b). The results of this analysis further suggest community-level composition was unaffected by the pesticide seed treatment across all sampling time points (Fig. 1a; perMANOVA, Treatment, p = 0.6646). Interestingly, we detected clear shifts in community composition during the growing season (Fig. 1b; Time, p = 0.0050) with the early summer sampling period (post-planting) being associated with higher abundances of Acari Oribatidae (detritivore), Diplura Japygidae (omnivore), Acari Mesostigmata (omnivore), and Coleoptera Staphylinidae (predator). Our ability to detect seasonal shifts in mesofauna community composition suggests that our sampling strategy is adequate to detect treatment effects if these were to occur.

At the population-level, however, we did observe treatment effects on the abundances of several mesofauna taxa (Fig. 2).  For the sake of brevity, we present only the results from the mid-season sampling period (tasseling). Unexpectedly, a targeted pest species, larva of corn rootworm (Coleoptera: Chrysomelidae: Galerucinae: Diabrotica), increased in abundance in treated plots compared to the control (t-test, Galerucinae, p = 0.0341). These results may be due to pesticide resistance in this pest population or possibly reductions in natural enemies. Further research is needed to understand the mechanisms underlying this result. Interestingly, we also found that the density of a mostly predatory mite population, Acari Mesostigmata, was negatively impacted by the seed treatment (t-test, Mesostigmata, p = 0.0576). It is important to note that these results only include a single field season and the cumulative effects of this management practice over the two-year rotation may provide more insight into the effects pesticide seed treatments have on the soil food web community.

Treatments from 2013 were maintained when soybean was planted in 2014. We used no-tillage management practices to minimize soil disturbance and the same isoline of soybean with and without the pesticide seed treatment was assigned to the same treatment plots as were used in 2013. All litter bags were deployed immediately after soybean planting and were collected at similar crop developmental stages and dates as in 2013:  post-planting, flowering, and harvest. Soil faunal communities were sampled, extracted and preserved for later identification. Crop growth metrics, including growth stage, aboveground biomass, leaf chlorophyll content, and yields were also measured in a manner similar to the previous year.

Beginning this winter, all 2014 mesofaunal, litter bag, and resin strip samples will be processed and analyzed. We will present these data at farmer and scientific meetings this coming summer and fall.

2. Determine whether pesticide treated seeds foster soil food webs with less diversified consumer-resource links.

To meet this objective, we collected soil fauna during both our 2013 and 2014 field seasons. These samples are currently being processed, which involves separating specimens by family, desiccating the specimens, pulverizing the biomass, and encapsulating the sample in tin. To date, all 2013 samples have been identified to family and the 2014 samples are almost completed. We anticipate sending the prepared samples to the UNH Stable Isotope Laboratory this spring 2015 for isotopic analysis. The data we receive from the laboratory will then be used to determine if the consumer-resource links in these food webs changes with the use of pesticide seed treatments.

3. Determine the effects pesticide seed treatments have on agriculturally important ecosystem services including decomposition and nitrogen cycling.

To quantify community-level functions, including rates of above ground cover crop decomposition and nitrogen cycling in the rhizosphere, we deployed both a litter bag study and cation/anion resin strips each field season. Cash crops were also used as phytometers, which is a measure of the crop’s physiological responses to the environment. We quantified crop seedling biomass, growth stages, crop dimensions, chlorophyll leaf content, and grain yields.

Litter bags were deployed in both field seasons directly after planting. Recovery of the bags corresponded with the three soil sampling periods: post-planting, flowering, and harvest. After all soil fauna were removed from bags using Berlese-Tullgren funnels, all remaining cover crop biomass was carefully removed from the litter bags, oven dried, weighed, and incinerated to determine ash free dry mass remaining. To date, the 2013 litter bag study analysis is complete, and the 2014 samples have been processed. Our 2013 data suggest that rates of aboveground litter decomposition were unaffected by the pesticide seed treatment (ANOVA, Treatment, p = 0.9219).

Concentrations of ammonium and nitrate in rhizosphere soils were monitored with cation and anion resin strips. Resin strips were buried two weeks prior to each soil sampling period. Upon removal, potassium chloride was used to extract all ammonium and nitrate present on each strip. This solute was then promptly frozen. Colorimetric assays of all samples, including the 2013 and 2014 field seasons, are slated for 2015. 

In 2013, we found an early season growth benefit of the pesticide seed treatment, but this advantage did not result in a grain yield benefit at harvest (Fig. 3a,b ). Specifically, corn seedlings in treated plots had significantly greater biomass during their second vegetative stage (Fig 3a; ANOVA, Treatment, p = 0.0100). At harvest, however, there was no difference in grain yields between the treatments (Fig 3b; ANOVA, Treatment, p = 0.2963). We expect that the nitrogen data from the resin strips will provide further insight into these results.

• Figure 2
• Figure 3
• Figure 1

Impacts and Contributions/Outcomes

Our first year of data suggests that pesticide seed treatments can differentially affect key soil mesofaunal populations but may have little effect on overall community composition or the soil food web’s capacity to decompose aboveground plant litter. Interestingly, we observed an increase in corn rootworm abundance in the presence of pesticide seed treatments. Also, we observed no grain yield benefit from the pesticide seed treatment. Taken together, these results raise questions regarding the efficacy and utility of this common agricultural practice. Data from our second field season will provide additional data regarding the longer-term effects of pesticide seed treatments and greater insight into the consequences for the soil food web and associated decomposition and nutrient cycling processes. Together these data will provide growers with new knowledge that can potentially influence the agricultural management practices they employ.

In August 2014, we were invited to share the results of this research in a symposium on rhizosphere interactions at the Annual Meeting of the Ecological Society of America in Sacramento, CA. Our talk entitled “Effects of pesticide seed treatments on rhizosphere food web composition and function in agroecosystems” was attended by approximately 60 scientists, students, and agricultural professionals. This project has also provided research opportunities for two undergraduate students at the University of New Hampshire. These students were mentored by Atwood and Smith and conducted independent research projects examining specific effects of the treatments on soil food web composition and weed community response. These students then presented their results at the 2014 UNH Undergraduate Research Conference in Durham, NH. Upon completion of the remaining data analyses in 2015, we plan to present the final results of this project at a variety of farmer meetings and scientific conferences, as well as in peer-reviewed journals.

Collaborators: