Broad spectrum pesticide application alters natural enemy communities and may facilitate secondary pest outbreaks

Abstract

Background: Pesticide application is the dominant control method for arthropod pests in broad-acre arable systems. In Australia, organophosphate pesticides are often applied either prophylactically, or reactively, including at higher concentrations, to control crop establishment pests such as false wireworms and earth mite species. Organophosphates are reported to be disruptive to beneficial species, such as natural enemies, but this has not been widely assessed in Australian systems. Neither has the risk that secondary outbreaks may occur if the natural enemy community composition or function is altered.

Methods: We examine the abundance of ground-dwelling invertebrate communities in an arable field over successive seasons under rotation; barley, two years of wheat, then canola. Two organophosphates (chlorpyrifos and methidathion) were initially applied at recommended rates. After no discernible impact on target pest species, the rate for chlorpyrifos was doubled to elicit a definitive response to a level used at establishment when seedling damage is observed. Invertebrates were sampled using pitfalls and refuge traps throughout the experiments. We applied measures of community diversity, principal response curves and multiple generalised linear modelling techniques to understand the changes in pest and natural enemy communities.

Results: There was large variability due to seasonality and crop type. Nevertheless, both pest (e.g., mites and aphids) and natural enemy (e.g., predatory beetles) invertebrate communities were significantly affected by application of organophosphates. When the rate of chlorpyrifos was increased there was a reduction in the number of beetles that predate on slug populations. Slugs displayed opposite trends to many of the other target pests, and actually increased in numbers under the higher rates of chlorpyrifos in comparison to the other treatments. Slug numbers in the final rotation of canola resulted in significant yield loss regardless of pesticide application.

Discussion: Organophosphates are a cost-effective tool to control emergent pests in broad-acre arable systems in Australia. We found risks associated with prophylactic application in fields under rotation between different crop types and significant changes to the community of pests and natural enemy. Disrupting key predators reduced effective suppression of other pests, such as slugs, and may lead to secondary outbreaks when rotating with susceptible crops such as canola. Such non-target impacts are rarely documented when studies focus on single-species, rather than community assessments. This study represents a single demonstration of how pesticide application can lead to secondary outbreaks and reinforces the need for studies that include a longer temporal component to understand this process further.

Keywords: Community ecology; Organophosphate; Pesticide; Pestsuppression; Secondary outbreak.

Figure 1. Site layout.

Site layout indicating insecticide treatments and invertebrate sampling transects (blocks) in relation to overall yield from the 36 ha field (pooled data from 2003–2007).

Figure 2. Species richness.

(A) Change in total species richness over time, total of 115 taxa. The grey bars represent the sampling times and the red dashed lines represent the application of pesticides associated with each treatment. (B) Total proportional species turnover for each time point through the study period. (C) Mean rank shifts. Note for (B) and (C) the initial sampling and spray event is not present, as each point represents the change from the previous sampling event. The first spray event is immediately before the beginning of these panels, however.

Figure 3. Selected species abundances.

Abundances (untransformed count data) through time for selected pests and natural enemies displaying important responses in the analyses (see Figs. 4–6 and Table 3). Pests (A) slugs, (B) earth mites and (C) earwigs, and the predatory beetles: (D) carabids and (E) staphylinids.

Figure 4. Principal response curves.

(A) Principal response curve for the natural enemy community. The left y-axis shows the Effect size. The position on the right y-axis reflects the weighting of the species to the overall response. The 0 line reflects the untreated control. (B) Principal response curve for the pest community. The left y-axis shows the Effect size. The position on the right y-axis reflects the weighting of the species to the overall response. The 0 line reflects the untreated control.

Figure 5. Beneficial community.

Effect size through time on the natural enemy community identified in this study. Each species grouping has had individual generalised linear models performed on abundance. The effect size is relative to the control, and the different colours represent the species contribution to that effect size, at that point in time. The dashed vertical lines represent the application of the pesticide treatments. (B) The overall community response to the application of the treatments through time. For both (A) and (B), the control is represented by the 0 line, and the treatments correspond to the deviance from the control. The three dashed vertical lines represent the application of the pesticide treatments. (C) The proportional contribution of each species to the overall deviance.

Figure 6. Pest community.

Effect size through time on the pest community identified in this study. (A) Each species grouping has had individual generalised linear models performed on abundance. The effect size is relative to the control, and the different colours represent the species contribution to that effect size, at that point in time. The three dashed vertical lines represent the application of the pesticide treatments. (B) The overall community response to the application of the treatments through time. The control is represented by the 0 line, and the treatments correspond to the deviance from the control. The four dashed vertical lines represent the application of the pesticide treatments. (C) The proportional contribution of each species to the overall deviance.

Figure 7. Yield.

Yield (tonnes per hectare) per crop type and season, treatment and experimental transects (blocks 1–3).