Abstract. Arthropod predators and parasitoids provide valuable ecosystem services in agricultural crops by suppressing populations of insect herbivores. Many natural enemies are influenced by non-crop habitat surrounding agricultural fields, and understanding if, and at what scales, land use patterns influence natural enemies is essential to predicting how landscape alters biological control services. Here we focus on biological control of soybean aphid, Aphis glycines Matumura, a specialist crop pest recently introduced to the north-central United States. We measured the amount of biological control service supplied to soybean in 26 replicate fields across Michigan, Wisconsin, Iowa, and Minnesota across two years (2005)(2006). We measured the impact of natural enemies by experimentally excluding or allowing access to soybean aphid infested plants and comparing aphid population growth over 14 days. We also monitored aphid and natural enemy populations at large in each field. Predators, principally coccinellid beetles, dominated the natural enemy community of soybean in both years. In the absence of aphid predators, A. glycines increased significantly, with 5.3-fold higher aphid populations on plants in exclusion cages vs. the open field after 14 days. We calculated a biological control services index (BSI) based on relative suppression of aphid populations and related it to landscape diversity and composition at multiple spatial scales surrounding each site. We found that BSI values increased with landscape diversity, measured as Simpson's D. Landscapes dominated by corn and soybean fields provided less biocontrol service to soybean compared with landscapes with an abundance of crop and non-crop habitats. The abundance of Coccinellidae was related to landscape composition, with beetles being more abundant in landscapes with an abundance of forest and grassland compared with landscapes dominated by agricultural crops. Landscape diversity and composition at a scale of 1.5 km surrounding the focal field explained the greatest proportion of the variation in BSI and Coccinellidae abundance. This study indicates that natural enemies provide a regionally important ecosystem service by suppressing a key soybean pest, reducing the need for insecticide applications. Furthermore, it suggests that management to maintain or enhance landscape diversity has the potential to stabilize or increase biocontrol services.

Aim  Coccinellid beetles are important predators that contribute to pest suppression in agricultural landscapes. Since the introduction of the exotic coccinellids Coccinella septempunctata L. and Harmonia axyridis Pallas into the USA, several studies have reported a decline of native Coccinellidae in agroecosystems. We aimed to investigate the influence of landscape composition on native and exotic coccinellid abundance within soybean fields.Location  Iowa, Michigan, Minnesota and Wisconsin.Methods  As part of a 2‐year study (2005–06) on the biological control of the soybean aphid, Aphis glycines Matsumura, we examined coccinellid communities in 33 soybean fields using yellow sticky card traps. Landscape heterogeneity and composition were measured at multiple spatial scales ranging 1–3.5 km from focal soybean fields where coccinellid sampling took place.Results  Exotic species made up 90% of the total coccinellid community in Michigan soybean fields followed by Wisconsin (84%), Minnesota (66%) and Iowa (57%). Harmonia axyridis was the dominant exotic coccinellid in all states comprising 45–62% of the total coccinellid community, followed by C. septempunctata (13–30%). Two additional exotic species, Hippodamia variegata (Goeze) and Propylea quatuordecimpunctata (L.) were also found in the region. Overall, the most abundant native coccinellid was Hippodamia convergens Guerin‐Meneville; however, its abundance varied across the region, comprising 0% (Michigan) to 28% (Iowa) of the total coccinellid community. Landscape structure significantly influenced the composition of coccinellid communities in soybean agroecosystems. We found that native coccinellids were most abundant in low‐diversity landscapes with an abundance of grassland habitat while exotic coccinellids were associated with the abundance of forested habitats.Main conclusion  We propose that grassland dominated landscapes with low structural diversity and low amounts of forested habitat may be resistant to exotic coccinellid build‐up, particularly H. axyridis and therefore represent landscape‐scale refuges for native coccinellid biodiversity.

No abstract

Abstractn the north central United States, populations of the exotic soybean aphid, Aphis glycines Matsumura (Hemiptera: Aphididae), are highly variable across space, complicating effective aphid management. In this study we examined relationships of plant nutrients, landscape structure, and natural enemies with soybean aphid abundance across Iowa, Michigan, Minnesota, and Wisconsin, representing the range of conditions where soybean aphid outbreaks have occurred since its introduction. We sampled soybean aphid and its natural enemies, quantified vegetation land cover and measured soybean nutrients (potassium [K] and nitrogen [N]) in 26 soybean sites in 2005 and 2006. Multiple regression models found that aphid abundance was negatively associated with leaf K content in 2005, whereas it was negatively associated with habitat diversity (Simpson's index) and positively associated with leaf N content in 2006. These variables accounted for 25 and 27% of aphid variability in 2005 and 2006, respectively, suggesting that other sources of variability are also important. In addition, K content of soybean plants decreased with increasing prevalence of cornsoybean cropland in 2005, suggesting that landscapes that have a high intensification of agriculture (as indexed by increasing corn and soybean) are more likely to have higher aphid numbers. Soybean aphid natural enemies, 26 species of predators and parasitoids, was positively related to aphid abundance; however, enemyto-aphid abundance ratios were inversely related to aphid density, suggesting that soybean aphids are able to escape control by resident natural enemies. Overall, soybean aphid abundance was most associated with soybean leaf chemistry and landscape heterogeneity. Agronomic options that can ameliorate K deficiency and maintaining heterogeneity in the landscape may reduce aphid risk. KeywordsAphis glycines, biological control, landscape ecology, plant chemistry, potassium Disciplines Agriculture | Entomology CommentsThis article is from Environmental Entomology 39 (1) Soybean aphid natural enemies, 26 species of predators and parasitoids, was positively related to aphid abundance; however, enemy-to-aphid abundance ratios were inversely related to aphid density, suggesting that soybean aphids are able to escape control by resident natural enemies. Overall, soybean aphid abundance was most associated with soybean leaf chemistry and landscape heterogeneity. Agronomic options that can ameliorate K deÞciency and maintaining heterogeneity in the landscape may reduce aphid risk.

The occurrence of Alfalfa mosaic virus (AMV) has increased in Wisconsin soybean fields in conjunction with the establishment of the soybean aphid (Aphis glycines). Field studies were conducted to determine the seasonal incidence of AMV-infected plants, progression of symptom severity caused by AMV, and the effect of AMV on soybean yield and seed quality. An isolate of AMV, collected from field-grown soybean, was introduced into plots by mechanical inoculation of plants at the V1 growth stage. The goal of the experiment was to achieve an incidence of AMV-infected plants of 0, 50, and 100% in 2002, and 0, 10, 25, 50, 75, and 100% in 2003. Severity of foliar symptoms was rated, and leaf samples were collected for serological assay (enzyme-linked immunosorbent assay [ELISA]) to estimate the incidence of AMV-infected plants from growth stages VC to R5. The maximum incidence of AMV-infected plants was 49% at growth stage R5, yet detection of the virus by ELISA dropped dramatically thereafter in both years. Incidence of AMV-infected plants accounted for 53 to 67% of the variability observed for severity of foliar symptoms in 2002 and 2003, respectively. Maximum yield loss ranged from 32% in 2002 to 48% in 2003 based on the difference in yield between noninoculated plots and plots with the highest incidence of AMV-infected plants. Incidence of AMV-infected plants explained 31% of the variation in yield in 2002 and 26% in 2003. An AMV incidence of 30% or greater was required for yield loss caused by AMV. Results of this study suggest that AMV has the potential to lower soybean yield and warrants further study.