Using stink bug migration behavior for physical exclusion (2024)

Search for other works by this author on:

Article history

Abstract

Introduction

Stink bugs belong to the family Pentatomidae (Hemiptera), with over 4,000 species worldwide (Panizzi et al. 2000). While some of these species are predacious and considered beneficial (primarily the subfamily Asopinae), most stink bugs are phytophagous and polyphagous, and include major economic pests of soybean, cotton, wheat, and tree fruits (Panizzi et al. 2000, McPherson 2018). In the United States, stink bugs are increasingly important pests, costing upwards of $9.4 million annually in cotton alone, which has been attributed to a decrease in the use of broad-spectrum insecticides (Greene et al. 2006, Williams 2012). In central Washington tree fruit production, stink bugs have historically been sporadic, secondary pests (Beers et al. 1993). However, stink bugs have become more consistent and challenging tree fruit pests with the adoption of narrow-spectrum and selective pesticides. Interestingly, during a peak year in the late 1990s, stink bug fruit injury increased in both commercial organic and conventionally managed orchards (McGhee et al. 1996, Krupke et al. 2001). This suggests that the lack of broad-spectrum sprays is not the only cause of increased stink bug damage, and further research into the biology and ecology of stink bugs is needed.

Stink bugs are a particularly difficult pest to manage due to their polyphagous nature and ability to disperse, sometimes great distances (Wiman et al. 2014, Lee and Leskey 2015). Most agricultural pest stink bug species are considered a landscape-level concern as they can move from native vegetation to crops or between crop types in response to changing food resource availability (McPherson and McPherson 2000, Tillman et al. 2009). In both cotton and apple, it has been suggested that stink bugs remain on native vegetation for the majority of their life cycle, only moving into crops when the native host plants senesce in late summer (Borden et al. 1952, Toscano and Stern 1976, McGhee 1997). Due to this association with native vegetation, economic injury from native stink bugs is most common in orchards near uncultivated and riparian areas (Mundinger and Chapman 1932, McGhee 1997). Additionally, because stink bugs migrate into orchards, their density and damage tend to be greatest at field edges (Tillman et al. 2009, Reeves et al. 2010). Therefore, sustainable integrated pest management practices must rely upon spatial and temporal precision in control tactics.

The increasing damage levels from the native stink bugs for the last 2 decades are no longer the only concern for establishing sustainable pentatomid management in the United States, and Washington in particular. The brown marmorated stink bug (Halyomorpha halys (Stål); Hemiptera: Pentatomidae), an invasive species from Asia, was first detected in the United States (Pennsylvania) in 1996 (Hoebeke and Carter 2003) and in the Pacific Northwest United States in Portland, Oregon in 2004 (Ingels 2014) and Washington in 2010 (Murray and Looney 2012). It has since spread to over 47 states, 4 Canadian provinces (stopbmsb.org), and through much of Europe (Haye et al. 2015). As a known pest of tree fruit and over 100 other host plants (Funayama 2002, 2004, Leskey et al. 2012), H. halys poses a major threat to Washington’s diverse agriculture (Beers et al. 2019). Insect pest management on high-value specialty fruit crops is difficult due to consumer demand for cosmetically impeccable fruit (Lye et al. 1988, Zalom et al. 1997). This results in reduced levels of tolerance for damage and reliance on broad-spectrum sprays, mainly pyrethroid and neonicotinoid insecticides, as soon as any stink bug damage is noticed in orchards (Krupke et al. 2001, Cullen and Zalom 2007, Kuhar et al. 2014). As with most invasive insects, H. halys has disrupted long-standing integrated pest management techniques by forcing growers to rely on chemical control alone due to the lack of effective alternative controls (Beers et al. 2019). For example, the heavy use of insecticide applications to control H. halys in the Mid-Atlantic United States causes severe outbreaks of secondary pests, necessitating the use of additional insecticides (Leskey et al. 2012).

Behaviorally based control techniques provide a promising alternative to broad-spectrum sprays. Techniques such as spraying alternate rows or orchard borders exploit aggregation and migration behavior while leaving a portion of the orchard untreated to conserve natural enemies (Bergh 2013, Blaauw et al. 2015). Attract-and-kill strategies, where only a pheromone lure-baited tree and the area around it are sprayed, further reduce the treated area (Morrison III et al. 2019). A refinement of this strategy is to replace the pesticide applications to the baited trees with insecticide-infused netting (IIN) either in orchards or near orchard borders (Krawczyk et al. 2019). The deltamethrin IIN is long-lasting and has been proven to be toxic to all stages of H. halys, with high mortality occurring after only a few seconds of contact (Kuhar et al. 2017, Sabbatini Peverieri et al. 2018).

Another recently proposed behavioral approach is the use of a physical barrier to prevent stink bug migration into crops. Physical barriers made of netting are being implemented in Washington State apple orchards to reduce fruit damage from wind, sunburn, and codling moth (Kalcsits et al. 2017, Mupambi et al. 2018, Marshall and Beers 2021, 2022a). It has been shown that completely enclosing trees with netting on all sides to form an exclusion cage effectively reduced H. halys ability to migrate and disperse within the enclosed area (Candian et al. 2018, Cottrell and Tillman 2019). Although the height of stink bug migration is not usually known, observation suggests that they migrate at the height of the canopy and not much above it (Tillman 2014). Further testing found that a single-wall barrier, instead of a complex enclosure, may be enough to reduce stink bug abundance and damage within a crop (Tillman et al. 2015). Similarly, stink bug damage in central Washington occurs along orchard borders with riparian or shrub-steppe habitat, which consists of plant species that rarely exceed 3 m in height (McGhee 1997, USDA 2014). Stink bugs migrating into orchards from the shrub-steppe are likely flying below the 3-m canopy of the natural habitats and have the potential to be excluded with physical barriers since migration and aggregation behavior are highly conserved across the family (McPherson and McPherson 2000, Panizzi et al. 2000, McPherson 2018). This study aimed to determine the height at which stink bugs migrate and the timing of dispersal into orchards through season-long trapping at the interface of the orchard and uncultivated vegetation. It also assessed the efficacy and nontarget effects of physical exclusion to prevent stink bug migration into the orchard by testing a net barrier at the orchard border with and without the addition of IIN.

Materials and Methods

Two experiments were conducted in 5 commercial apple blocks located near Manson, Washington (N 47.9164°, W −120.1208°) from May to September 2018 and June to September 2019. Apple cultivars included ‘Honeycrisp,’ ‘Delicious,’ and ‘Golden Delicious.’ Each orchard block was bordered by uncultivated shrub-steppe vegetation, and the growers applied late-season pyrethroids (e.g., Lambda-cyhalothrin) to control stink bugs. The first experiment determined the height and timing of stink bug immigration and emigration using a 4 m × 2 m clear sticky barrier constructed at each orchard border using trellis poles, dimensional lumber, and 11 clear 0.30 m × 2 m double-sided sticky panels (AlphaScents, West Linn, OR, USA) (Fig. 1A). Traps were secured to the lumber so that their midpoint heights were at 0.23, 0.58, 0.94, 1.30, 1.65, 2.01, 2.36, 2.71, 3.07, 3.43, and 3.78 m from the ground on each barrier, with a 5 cm gap between traps to reduce shear wind load on the structure.

The second study evaluated the physical exclusion of stink bugs. At 3 of the apple blocks, a 69 m section of border was assigned 3 treatments: (i) a 4 m × 23 m net (20% pearl leno (white), 5 mm × 2 mm mesh size, Green-Tek West, Dinuba, CA, USA) barrier with deltamethrin (~3.85 mg a.i./g fiber) infused netting (mesh: 32 holes per cm², ZeroFly, Vestergaard-Frandsen, Lausanne, Switzerland) flaps (deltamethrin), (ii) a 4 m × 23 m exclusion net barrier with noninsecticide-infused flaps (plain), and (iii) a 23 m control (no net). Barrier frames were constructed with trellis poles, dimensional lumber, and metal wire secured between the poles (Fig. 1B). For Treatments 1 and 2, 20% pearl leno (white) shade netting with flaps sewn in at 0.5, 2, and 3 m from the top was attached to the posts with flaps facing the native vegetation. Another metal wire was secured in front of the net in an x-wise fashion between each post to provide support to the nets from wind gusts. The excess netting at the bottom of the barrier was buried in the ground to provide complete coverage and support. Deltamethrin-infused black netting was sewn into the flaps of Treatment 1. A 2 m × 4 m tarp was placed in the center of each treatment on the native vegetation side to evaluate the mortality of target and nontarget arthropods. The same netting and barriers were used for both years of the experiment.

Stink Bug Immigration and Emigration

Each sticky barrier was visually inspected for all motile life stages of stink bugs weekly from June to September (2018) and May to September (2019). Stink bugs were recorded by trap height, side (native vegetation vs. orchard), and species and then removed. While the side of trap capture is not known to conclusively indicate the direction of movement, since these traps were affixed to posts at each end to prevent twisting, the side of trap capture was used as a proxy for insect movement direction.

Stink Bug Physical Exclusion

Sampling of net exclusion barriers, the native vegetation, and the orchard occurred weekly from June to September for both years. Ten apple trees (5 adjacent trees in the nearest 2 border rows) of each treatment area were selected for beat tray sampling, which is the preferred method for sampling multiple species of stink bugs from tree fruits (McGhee 1997, Nielsen and Hamilton 2009). Two trees at the end of the rows were left as buffer trees and were not sampled. A beating tray consisted of a 45 cm × 45 cm white sheet secured to a metal frame with a 0.30 m section of metal pipe secured to the frame as a handle. To conduct a sample, the tray was held under the selected branch, and the branch was tapped 3 times using a 1 m section of stiff rubber hose. After tapping, all stink bugs on the tray were counted and recorded based on species and life stage. Each tree was beat tray sampled 4 times, one branch on the lower half and one on the upper half of each side of the tree. The total number of stink bug adults and nymphs was recorded for each species by orchard block and treatment. The native vegetation bordering the orchards including sage brush (Artemisia spp.), bitterbrush (Purshia tridentata), lupines (Lupinus spp.), balsamroots (Balsamorhiza spp.), and currants (Ribes spp.), was sampled in a similar fashion, except that a timed count (7 min) with a beat tray was used.

To assess target and nontarget effects of the net barrier treatments, all dead arthropods found on the tarps were collected and placed in glass vials filled with 70% ethanol. Vials were labeled with treatment, rep, and date. Specimens were then identified to the following taxonomic categories: Orthoptera, Neuroptera, Coccinellidae, other Coleoptera (excluding coccinellids), Pentatomidae, Bees, Hymenoptera (excluding bees), and Diptera.

Fruit damage was assessed on 1 September for all the sites. Fruit from the same trees selected for beating tray samples were visually inspected in situ, scanning only half of the fruit that was visible. At each site, up to 2,000 fruit (2018) or 3,000 fruit (2019) were inspected from each treatment area. Fruit suspected of having stink bug damage from the visual assessment was removed and cut to confirm the type of damage. The total apples examined were recorded to calculate the percentage of stink bug damage.

Data Analysis

Means on each sample date were calculated for the entire data set. An index of seasonal populations was calculated for the in situ beat tray samples (cumulative insect days, or CIDs), similar to Ruppel (1983) equation, such that CID = $∑[(Yi+1+Yi)/2](Ti+1−Ti)$where Y_i+1 and Y_i are consecutive insect counts, and T_i+1 and T_i are the corresponding sampling dates. For the tarp counts, seasonal captures were the sums of the weekly captures (cumulative tarp days, or CTDs). Treatment differences between CIDs and CTDs were analyzed with a generalized mixed linear model (PROC GLIMMIX; Statistical Analysis Institute 2019) using a normal distribution, except for the nontarget counts. The nontarget CTD counts were log-transformed, and treatment differences were analyzed across both years with a mixed model (PROC MIXED; Statistical Analysis Institute 2019), including the treatment-by-year effect. Proportional data (damaged fruits per number of fruits evaluated) were analyzed using a generalized mixed linear model with a binomial distribution and logit link (PROC GLIMMIX; Statistical Analysis Institute 2019). For all analyses, the Tukey’s adjustment was used to control experiment-wise error for the mean separation test (P ≤ 0.05).

Results

Stink Bug Immigration and Emigration

Seasonal stink bug movement from the native vegetation to the orchard (immigration) began in early June, peaked between late July (2018) and early August (2019), and tapered off by the end of August for both years. Movement from the orchard to native vegetation (emigration) began in early July (2018) and early August (2019) and ended in late August for both years. Peak movement occurred at the same time for both directions each year (Fig. 2A and B). There was a significant difference in the height of immigration into the orchard for both years (Fig. 3). In 2018, the highest count was at 1.30 m and the lowest at 3.78 m, and in 2019, the counts were highest at 2.36 m and lowest at 0.23, 3.43, and 3.78 m. Each year, the majority of stink bugs moving into the orchard are trapped between 1 m and 3 m (76%, 2018; 84%, 2019). There was no difference in the height of emigration out of the orchard for either year (Supplementary Fig. 1). Both years, there were nearly twice as many stink bugs immigrating into the orchard compared to emigrating out of the orchard (Fig. 2A and B, Y-axis Fig. 3 vs. Y-axis Supplementary Fig. 1).

Stink Bug Physical Exclusion

The cumulative number of stink bugs caught in beat tray samples in the native vegetation was significantly lower in the control section than in the plain netting section (2018) (Fig. 4A). However, this trend was not present the following year and there were no treatment differences between the native vegetation samples (2019) or orchard samples for both years (Fig. 4A and B). The cumulative stink bug capture in the native vegetation was highly variable between plots compared to the orchard capture for both years. In 2018, the barrier treatments showed a consistent decrease in stink bug counts between the native vegetation and orchard, while the counts in the control plots had similar amounts between the native vegetation and orchard (Fig. 4A; Supplementary Fig. 2A). This trend was not consistent in 2019, with a decrease between the orchard and native vegetation found in all treatments and the smallest reduction occurred in the deltamethrin treatment (Supplementary Fig. 2A). The majority of stink bugs caught in the native vegetation for both years belonged to the genus Chlorochroa (62%, 2018; 56%, 2019), followed by Euschistus conspersus Uhler (28%, 2018; 29%, 2019), and then the genus Thyanta (10%, 2018; 15%, 2019) (Supplementary Fig. 3A and B). Conversely, the majority of stink bugs caught in the orchard for both years were E. conspersus (62%, 2018; 41%, 2019), and the only nymphs caught in the orchard were E. conspersus in 2018, with no nymphs found in 2019 (Supplementary Fig. 4A and B).

The fruit damage evaluation resulted in extremely low levels of damage (<0.5%), which were not significantly different between treatments for both years (Table 1). Mortality of stink bugs migrating from the native vegetation to the orchard (as measured by the ground tarps) was highest in the deltamethrin-flap treatment, followed by the plain netting and the control treatments (Fig. 5). The nontarget effects were also the highest for the deltamethrin net flaps for each category (Orthopetera, other beetles, other Hemiptera, and beneficial insects) followed by the plain netting, and only a single dead arthropod (orthoptera) was found in control (Fig. 6).

Discussion

The height of stink bug dispersal between crops has been suggested through indirect evidence to be dependent on the canopy height of the crop (Tillman 2014). Recent studies have demonstrated that stink bugs do indeed stratify vertically on their plant host, with some species, including H. halys, being more abundant in the upper canopy (Quinn et al. 2018, Cottrell et al. 2023). The results from the clear sticky barrier trial support the hypothesis that stink bugs dispersing from the uncultivated vegetation into the orchard was limited to <3 m from the ground, which is similar to the predicted height of most plant species within the shrub steppe (McGhee 1997, USDA 2014). This limitation was not evident for stink bugs flying from the orchard to the native vegetation, which is expected since apple trees, although mostly on dwarfing rootstocks and training systems, are above 3 m in height (Looney and Lane 1983, Lawson 1994). These patterns are likely due to the general negative geotaxis that stink bugs exhibit, where they climb to the top of objects they encounter (Tillman and Cottrell 2016). In areas with tall native vegetation, such as the large deciduous trees in the eastern United States where stink bug capture is highest >9 m (Quinn et al. 2018, Cottrell et al. 2023), a net barrier may not be feasible. However, the relatively low height of eastern Washington’s native shrub-steppe vegetation is more suitable for a physical barrier that could provide season-long protection from stink bug damage.

Protective net enclosures and physical barriers have been shown as potential nonchemical techniques for preventing stink bug dispersal into crops; however, our net barrier study only partially supports the existing literature (Tillman et al. 2015, Candian et al. 2018, Cottrell and Tillman 2019). Our trial found that stink bug orchard border densities and damage near the net barrier did not differ from a border without a physical barrier, even with the addition of insecticide netting into the flaps. This lack of difference in stink bug densities within the orchard could be due to several causes. Our experiments were conducted in commercial orchards treated with late-season pyrethroids (e.g., Lambda-cyhalothrin), which likely caused the low capture and damage for all treatments within the orchard for both years. It is also possible that the scale of the barriers (width of border covered) was not sufficient to prevent movement into the orchard around the sides. Although stink bug densities and damage were not different, the deltamethrin-infused netting treatment successfully killed larger numbers of stink bugs than nontreated netting, indicating that stink bugs were being intercepted by the barrier and killed by the insecticide-treated flaps.

Physical barriers may not be selective between pests and natural enemies and may present an equal obstacle to the immigration of beneficial insects (DeBach and Huffaker 1971, Gontijo et al. 2015, Marshall and Beers 2022b). Uncultivated areas, such as the riparian and shrub-steppe habitats near agriculture, provide refuge for many beneficial insects, which provide valuable ecosystem services (Letourneau 1998, Horton and Lewis 2000, Miliczky and Horton 2005, Isaacs et al. 2009). Through our nontarget assessments, we determined that the net barrier does have deleterious effects on insect movement between the orchard and native vegetation, and the addition of deltamethrin-infused netting exacerbated this impact. Some of the main groups recorded were parasitic hymenopterans, native bees, neuropterans, and coccinellids. Building a physical barrier that completely encloses an orchard border may exclude these beneficial groups of parasitoids, pollinators, and predatory insects and lead to the outbreak of secondary pests, which the reduced chemical tactics are trying to prevent. However, the deleterious effects would only be seen for natural enemies immigrating into the orchard from the extra-orchard habitat. Species that complete their entire life cycle within the orchard should be less affected by net barriers, as shown for earwigs and Aphelinus mali (Marshall and Beers 2022b), especially compared to multiple broad-spectrum sprays.

A recent study documented that H. halys has established populations in the tree fruit-growing regions of eastern Washington and can complete development in the native shrub-steppe habitat, although it was suboptimal for survivorship and fecundity (Hepler et al. 2023). Aside from disrupting established integrated pest management techniques, another interesting aspect of invasive species is their influence on local insect communities. H. halys is bivoltine, while most, if not all, of the native stink bug species in Washington are univoltine (Borden et al. 1952, McPherson and McPherson 2000, Rice et al. 2014). It is possible that H. halys may outcompete the native stink bugs and change the species composition as it spreads throughout the area. Thus, it is essential to evaluate the composition of native species before, during, and after the establishment of the novel species to determine the impact of the invasive species on local communities. Our study provides this opportunity as H. halys has not yet been found in these sampling sites but has been detected near them (Beers 2019). Currently, there are over 50 species of stink bugs recorded in Washington State (Zack et al. 2012), and of those, E. conspersus (Uhler) is considered the most abundant stink bug pest in central Washington tree fruit crops (McGhee 1997, Krupke 2004). Through our orchard sampling, we found that E. conspersus is still the most abundant stink bug in orchard borders in central Washington; however, it composed less than one-third of the stink bugs found in the vegetation bordering the orchards. Interestingly, the reverse was true of stink bugs belonging to the genus Chlorochroa for both years. If H. halys begins to establish, future sampling of these sites can document the effects the invasive species has on the community of native stink bugs.

Conclusion

Evaluating novel reduced or nonchemical control tactics is essential to developing a sustainable control program as stink bug damage continues to increase, and H. halys spreads across the United States and throughout Washington. Currently, it is thought that most stink bug pressure in crops occurs when their native hosts begin to senesce, and control has primarily consisted of late-season broad-spectrum sprays to target this increase (Toscano and Stern 1976, McGhee 1997, Panizzi et al. 2000). Using clear sticky barriers, we were able to determine stink bug dispersal behavior between the uncultivated shrub-steppe and orchard borders of central Washington. Our study does not support the existing notion that a single migration occurs at native vegetation senescence. Instead, stink bugs began to disperse into orchards in June and moved between the shrub steppe and orchard multiple times throughout the growing season. The presence of stink bugs in orchard borders early in the year could explain the lack of control experienced from relying on mid-August sprays. Most stink bugs immigrated into the orchard below 4 m, and physical barriers of this height intercepted stink bugs and nontarget species; the addition of deltamethrin-infused netting strips increased mortality for all groups. This study indicates that stink bug control strategies should begin earlier in the growing season, and barriers at an orchard border can provide a season-long method of slowing stink bug movement into the orchard.

Supplementary material

Supplementary material is available at Environmental Entomology online.

Acknowledgments

The authors would like to acknowledge Bruce Greenfield, Peter Smytheman, Chris Sater, James Hepler, Thomas Smytheman, Andy Terrazas, Ali Everhart, Allie Wright, CJ Squires, and Daria Gausman for their technical assistance.

Funding

This work was supported by the Washington Tree Fruit Research Commission [grants #CP-16-101 and #CP-19-105], the National Institute of Food and Agriculture Specialty Crop Research Initiative [grant #2016-51181-25409], the United States Department of Agriculture National Institute of Food and Agriculture Hatch Act [project #1016563], and the Western Sustainable Agriculture Research and Education Graduate Student Research and Education [grant #GW17-025]. The use or mention of trade names and products herein does not represent an endorsem*nt by the USDA. USDA is an equal opportunity provider and employer.

Author Contributions

Adrian Marshall (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Investigation [lead], Methodology [equal], Project administration [equal], Software [equal], Validation [equal], Visualization [equal], Writing—original draft [lead], Writing—review & editing [equal]), and Elizabeth Beers (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Funding acquisition [equal], Investigation [supporting], Methodology [equal], Project administration [equal], Resources [lead], Software [equal], Supervision [equal], Validation [equal], Visualization [equal], Writing—review & editing [equal])

References

Published by Oxford University Press on behalf of Entomological Society of America 2024.

Issue Section:

Behavioral Ecology

Email alerts

Citing articles via

More from Oxford Academic