Development of an Attract-and-Kill Strategy for Drosophila suzukii (Diptera: Drosophilidae): Evaluation of Attracticidal Spheres Under Laboratory and Field Conditions

Drosophila suzukii is an invasive fruit fly that was accidentally introduced to the United States from Asia (Bolda et al. 2010). First detected in California in 2008, D. suzukii rapidly spread across the North American continent, including Canada and Mexico, and is currently reported in 45 states in the USA (Asplen et al. 2015). Additional introductions have established populations throughout Europe and South and Central America (Asplen et al. 2015), suggesting that D. suzukii will emerge as a cosmopolitan invasive species. D. suzukii attack healthy intact blueberries, caneberries, strawberries, and cherries, resulting in severe economic damage (Lee et al. 2011, Walsh et al. 2011, Asplen et al. 2015, Lee et al. 2015).

Unlike most Drosophilid spp. that are restricted to oviposition in softer overripe fruit, D. suzukii females have serrated ovipositors, allowing them to lay eggs in firm, intact ripe fruit. Fruit infested with D. suzukii are unmarketable, with larval feeding resulting in fruit collapse, which, in turn, leads to increased pathogen infections (Cini et al. 2012, Hamby et al. 2012, Asplen et al. 2015, Ioriatti et al. 2015). Annual losses from D. suzukii in the United States have been estimated at $718 million (Walton 2013). Drosophila suzukii have few effective natural enemies in introduced regions (Stacconi et al. 2013, 2015), a rapid generation time (∼11 d from egg to adult at 26°C), (Tochen et al. 2014) and a high reproductive rate (∼400 per female) (Cini et al. 2012), which allow populations to quickly reach outbreak levels. Currently, trapping systems available for D. suzukii can be used to provide information regarding fly presence, but not necessarily guide management decisions. Since the arrival of D. suzukii, small fruit growers have switched to calendar-based weekly insecticide applications (Haviland and Beers 2012, Van Timmeren and Isaacs 2013). Increased pesticide applications can negatively affect pollinators and reduce natural enemy abundance, resulting in secondary pest outbreaks. Obviously, this is not a sustainable management practice, and new IPM-based options are needed.

Behaviorally based management strategies, such as attract-and-kill, provide pest control while reducing insecticide applications (Lanier 1990, El-Sayed et al. 2009). Attract-and-kill methods lure the pest species to a concentrated area where they can be annihilated with a killing agent, thereby reducing the total area subjected to pesticide applications. In apple orchards, attracticidal spheres have been used as attract-and-kill devices to manage apple maggot, Rhagoletis pomonella Walsh (Wright et al. 2012, Morrison et al. 2016). Apple maggot are attracted to the visual stimulus afforded by the red sphere, with olfactory lures used to enhance their attractiveness. Unlike typical insecticide applications, attracticidal spheres combine a toxicant with a feeding stimulus (sucrose) to promote oral uptake. Adult flies alight on spheres, beginning feeding on the mixture of sucrose and toxicant, ingest a lethal dose, and die (Wright et al. 2012). In commercial apple orchards, attracticidal spheres reduced apple maggot damage to that achieved with standard insecticide applications (Morrison et al. 2016). Drosophila suzukii is also visually attracted to red spherical objects (Rice et al. 2016) and to flat, red, disc-shaped stimuli (Kirkpatrick et al. 2016), indicating that the attracticidal sphere system originally developed for apple maggot may provide alternative management options for this destructive invasive pest species.

We conducted laboratory bioassays where we compared the efficacy of 10 candidate insecticides as potential toxicants for inclusion in attracticidal spheres against D. suzukii adults. On the basis of the results from laboratory trials, we deployed attracticidal spheres in raspberry plots and compared D. suzukii infestation rates in plots protected by attracticidal spheres alone, weekly insecticide applications, the combination of spheres and insecticide applications, and untreated control plots.

Material and Methods

D. suzukii Colony

Drosophila suzukii were maintained in a laboratory colony at the USDA-ARS Appalachian Fruit Research Station in Kearneysville, WV. Adults were housed in 50-ml plastic drosophila vials (Flystuff.com, San Diego, CA) in groups of 20–30 flies, and fed 2.5 g of artificial drosophila diet with 3–5 grains of yeast (Formula 4-24 instant drosophila medium, Carolina Biological Supply, Burlington, NC). The colony was maintained in an environmentally controlled room at 25 ± 2°C, 50 ± 10% RH and a photoperiod of 16:8 (L:D) h. All D. suzukii used in experiments were 7- to 14-d-old and sexually mature (Asplen et al. 2015). Before laboratory trials, flies were transferred to empty vials that contained a cotton wick with 20% sugar water, and stored at 20 °C for 15–18 h.

Sphere Production

Laboratory Trials

Each cap was formulated with 0.1, 0.5, or 1.0% insecticide, fitted to a flat-topped sphere base and then misted with water for activation. The following compounds were examined in this study: control (no insecticide added to sphere), dinotefuran, imidacloprid, spinetoram, spinosad, boric acid, acephate, lambda-cyhalothrin capsule suspension (CS), lambda-cyhalothrin water dispersible granule (WG), Chromobacterium subtsugae Martin, lambda-cyhalothrin (WG) + spinetoram, and dinotefuran + spinetoram (see Table 1 for rates). As trials were conducted over a period of time, control spheres were evaluated continuously. Therefore, based on the trial date, results from control spheres and spheres containing specific compounds were compared (see Table 3 for groupings).

Colony-reared 7- to 14-d-old, healthy D. suzukii (20 males and 20 females) were exposed individually to spheres of each insecticide treatment, and each treatment was replicated five times. For each trial, a single D. suzukii was gently placed at or near the equator of the sphere base and cap and allowed to forage freely for up to 5 min (Fig. 1B). Total feeding time was recorded for each D. suzukii. Afterward, flies were then held in individual containers with a cotton wick containing 20% sugar water. Fly condition (alive, moribund, or dead) was assessed immediately after the trial was completed (time = 0), and then at 24 and 48 h. After 24 h, only 0.7% of D. suzukii were moribund, and therefore for statistical analysis, moribund were combined with the dead category. Drosophilasuzukii mortality was compared between each insecticide treatment and control spheres using multivariate logistic regression using mortality as the response variable and sex and feeding time as explanatory variables for each observation interval (SAS Institute 2004, version 9.1).

Field Trials

All field trials were conducted at the Appalachian Fruit Research Station in Kearneysville, WV, in 2013 and 2014. When standard sugar–yeast traps indicated that D. suzukii populations were present, trials were initiated. Traps were constructed from 750-ml clear plastic containers with 0.64-cm-diameter side entry holes and were placed in control plots at a height of 149 1.5 m. Traps were baited with 250-ml sugar–yeast water solution (40 g: 7 g: 0.95L) and were replaced each week.

Spinetoram Spheres

During 2013, potted raspberry plants (cv. Joan J) were established from root cuttings (Nourse Farms, Deerfield, MA) in February and maintained in a greenhouse until May. After ripe fruit were present, plants were transferred to an open field to create four semi-field plots. Each plot consisted of three rows of plants with five plants per row (total = 15 plants). Plants were supported with trellis wire and spaced 1 m apart with 2-m row spacing. Plots were separated by 45 m. Each plot was assigned to one of the following treatments: 1) weekly insecticide application (see Table 2 for details) applied with a backpack sprayer; 2) attracticidal spheres containing spinetoram at 1.0% a.i. and hung on the trellis wire at a height of ∼1.5 m and next to each plant; 3) weekly insecticide applications and attracticidal spheres (as described above); and 4) control plots that received no insecticide applications or attracticidal spheres. All attracticidal spheres were hand-misted directly after deployment to initiate release of toxicant-laden feeding stimulant residue. Each week, from July 3 through September 3, all ripe fruit were removed from each plant and placed into individual 473-ml paper cups with a mesh lid in groups of ≤10, then held in the laboratory at 25 ± 2°C, 50 ± 10% RH and a photoperiod of 16:8 (L:D) h for 9–10 d to allow flies to pupate and/or emerge as adults. Each treatment was replicated three times, and the total number of D. suzukii emerging from each treatment was compared using Wilcoxon–Mann–Whitney with a Dwass–Steel–Critchlow–Fligner mean separation test because of violation of normality.

Dinotefuran Spheres

During 2014, 12 raspberry plots (cv. Joan J) were established from root cuttings (Nourse Farms, Deerfield, MA). Plots were spaced 7.6 m apart. Each plot consisted of five rows with 10 plants per row. Plants were spaced 0.5 m apart with 3 m between rows. In each plot, the inner three rows were assigned to one of four management treatments: 1) a weekly insecticide application (see Table 2 for details); 2) attracticidal spheres formulated with 1.0% dinotefuran; 3) weekly insecticide application and attracticidal spheres formulated with 1.0% dinotefuran; and 4) no insecticide application or attracticidal spheres. Insecticides were applied by an airblast sprayer. One sphere was hung from a trellis wire in the upper one-third of each raspberry plant and positioned to be visually apparent to foraging flies within and near the upper canopy. Spheres were misted with water to activate the release of sugar and toxicant from the cap. The two edge rows of each stand served as buffers between treatments. The fruit sampling protocol for this trial included harvesting 10 ripe fruit weekly from each of five randomly selected plants within each replicate from August 19 through September 9. Unharvested, ripe fruit were not removed from the treated plots or the buffer rows. Fruit from each plant were placed in individual cups, transferred to the laboratory and held as previously described, and the number of D. suzukii adults emerging from each treatment was compared using Wilcoxon–Mann–Whitney with a Dwass–Steel–Critchlow–Fligner mean separation test because of violation of normality.

Dinotefuran Spheres and D. suzukii Lures

From September 26 to October 9 2014, semiochemical baits were combined with attracticidal spheres to determine if olfactory attractants increased sphere efficacy and reduced D. suzukii infestations using the field plots described above. Each plot was assigned one of the following treatments: 1) a weekly insecticide application applied with an airblast sprayer (see Table 2 for details); 2) attracticidal spheres formulated with 1.0% dinotefuran; 3) attracticidal spheres formulated with 1.0% dinotefuran in combination with commercially available D. suzukii lures; and 4) a control (no insecticide applications, attracticidal spheres, or D. suzukii lures). Spheres were deployed as described above, and D. suzukii lures (Trece Inc, Adair, OK) were suspended directly above each sphere in plots combining stimuli. Ten ripe fruit were harvested weekly from each of five randomly selected plants within each replicate. Each treatment was replicated three times and compared using Wilcoxon–Mann–Whitney with a Dwass–Steel–Critchlow–Fligner mean separation test because of violation of normality.

Results

Laboratory Trials

D rosophilasuzukii spent significantly less time feeding on attracticidal spheres containing dinotefuran (0.1%), acephate (0.1, 0.5, and 1.0%), lambda-cyhalothrin (WG) + spinetoram (1.0%), lambda-cyhalothrin (CS) (1.0%), and permethrin (1.0%) than control spheres (Table 3).

No significant difference in mortality between male and female D. suzukii in any insecticide treatment was observed (see Tables 3–4). After 5-min exposure to attracticidal spheres, 13 insecticides resulted in significantly higher D. suzukii mortality than exposure to control spheres (Table 3). Spheres containing dinotefuran (0.5 and 1.0%) killed 100% of D. suzukii within 5 min. Spheres with dinotefuran (0.1%), acephate (0.1%), lambda-cyhalothrin (WG) (1.0%) and spinosad (1.0%) resulted in 90%, 85%, 85%, and 80% mortality, respectively. Drosophilasuzukii feeding on spheres containing spinetoram (1%), boric acid (10.0%), boric acid (0.1%), and C.subtsugae (0.1%), experienced no mortality after 5 min.

After 24 h, exposure to spheres containing insecticides killed significantly more D. suzukii than controls (Table 4) with spinetoram, spinosad, permethrin, lambda-cyhalothrin (CS), and lambda-cyhalothrin (WG) (all 1.0%) killing 100% of flies. Spheres containing acephate (0.5 and 1.0%) and imidacloprid (0.1%) resulted in 95%, 90%, and 80% mortality, respectively. At 24 h, D. suzukii feeding on spheres containing dinotefuran (0.5%) experienced 92.5% mortality, suggesting a slight recovery rate from the 100% mortality observed at 5 min, perhaps resulting from the combined moribund and dead categories for analysis. Spheres containing C.subtsugae (0.1%) and boric acid (10.0%) produced low mortality rates. After 48 h, 17 candidate insecticides produced significantly higher mortality than controls (Table 5).

Treatments with two insecticides within a single attracticidal sphere killed more D. suzukii than controls (Table 1). After 5 min of exposure, lambda-cyhalothrin (WG) (0.1%) + spinetoram (0.1%) killed 17.5% of D. suzukii, a lower mortality rate than lambda-cyhalothrin (WG) (1.0%) alone. Dinotefuran (0.1%) + spinetoram (0.1%) resulted in 95% mortality after 5 min. At 24 h, both combinations produced 92.5% mortality. Lambda-cyhalothrin (0.1%) + spinetoram (0.1%) remained at 92.5%, whereas dinotefuran (0.1%) and spinetoram (0.1%) produced 97.5% mortality.

Field Trials

Discussion

In laboratory trials, attracticidal spheres containing 1.0% a.i. dinotefuran killed 100% of D. suzukii within 5 min, and spinetoram (1.0%), spinosad (1.0%), permethrin (1.0%), lambda-cyhalothrin (1.0%) killed 100% of flies within 24 h. Fast-acting compounds are likely preferable because they immediately remove pests from the population (El-Sayed et al. 2009), mitigating potential economic damage. Attracticidal spheres may provide faster control than traditional insecticide applications because they contain feeding stimulants (sucrose), so insects receive a lethal dose of toxicants by both ingestion and direct contact. The neonicotinoid compound, dinotefuran, provided the most rapid effect in laboratory trials, although pyrethroids, an organophosphate, and spinosyns all killed 100% of exposed flies within 24 h. Potential sublethal effects need to be investigated for these compounds, particularly if females are able to continue oviposition. In general, insects consuming a sublethal dose of insecticides may have reduced reproduction and feeding, but increased vulnerability to natural enemy attacks (Haynes 1988, Suckling and Brockerhoff 1999).

In field trials, attracticidal spheres used as a stand-alone control measure decrease D. suzukii infestations compared with untreated controls in plots with ripe raspberries in all trials conducted in 2013 and 2014. This is the first evidence that attracticidal spheres used to effectively manage apple maggot in apple orchards in New England (Wright et al. 2012, Morrison et al. 2016), appear to hold promise against D. suzukii. Attracticidal spheres are based on D. suzukii attraction to red, spherical objects in the laboratory (Kirkpatrick et al. 2016, Rice et al. 2016) and field (Rice et al. 2016), and these visual stimuli are combined with a feeding stimulant (sucrose) and toxicants to attract and remove foraging flies from the population.

In trials conducted in 2013, plots with spheres had infestation rates comparable to full insecticide application fields, suggesting attracticidal spheres, manage D. suzukii as effectively as full field spray applications. However, during 2014, plots with attracticidal spheres experienced significantly greater infestation rates compared with insecticide applications. However, all ripe fruit were removed from each plant weekly in 2013, whereas only 10 fruit from five plants were removed each week during 2014, to sample D. suzukii infestation rates. This difference in harvesting and sampling tactics meant that plants had much higher ripe fruit densities throughout the season in 2014 compared with 2013. Fruit density is often positively correlated with pest density (Root 1973, Jackson and Lee 1985, Prokopy et al. 1987, Dalby-Ball and Meats 2000), and behaviorally based management methods are often less effective under high pest pressure (El-Sayed et al. 2009). For example, attract-and-kill methods that successfully controlled codling moth, Cydia pomonella, in apple and Egyptian cotton leafworm, Spodoptera littoralis, in cotton at low pest densities, failed under high pest pressure (Carde and Minks 1995, Downham et al. 1995, Trematerra et al. 1999, Angeli et al. 2000, Charmillot et al. 2000, Cocco et al. 2013). Increasing the number of attract-and-kill units per acre when pest abundance is high is one mechanism for reducing competition imposed by higher pest densities (Lösel et al. 2000). Another approach can be to reduce the resource itself. In the case of our trials, higher densities of ripe and overripe fruit in 2014 likely led to a higher population of D. suzukii and thereby, higher infestation rates in plots protected by attracticidal spheres. In 2013, weekly harvest of all ripe fruit resulted in infestation rates that were comparable among plots protected by spheres. Indeed, current management recommendations recommend frequent harvest of ripe fruit and good sanitation practices to reduce pressure from D. suzukii (Walsh et al. 2011). Thus, reducing the ripe fruit load in plots frequently appears to increase the efficacy of attracticidal spheres.

In addition, higher ripe fruit density may increase visual competition with attracticidal spheres, so that flies are more likely to alight upon fruit instead of the attracticidal spheres, as was found for apple maggot in apple orchards (Rull and Prokopy 2003, 2004). For instance, visual traps targeting Lygus lineolaris experience reduced captures after leaf development because of increased competition (Prokopy et al. 1979). Thus, frequent harvesting of ripe fruit appears to potentially reduce visual competition with attracticidal spheres in raspberry plots and decrease fly pressure.

During 2013, attracticidal spheres coupled with insecticide treatments reduced damage more than all other treatments. Combined attract-and-kill with conventional insecticide treatments managed Mediterranean fruit fly, Ceratitis capitata (Wiedemann), more effectively than either method alone, reducing crop loss by fivefold compared with conventional sprays (Rahman and Broughton 2016). Likewise, mass trapping combined with spray applications reduced Mediterranean fruit fly damage more than either methods alone in citrus orchards. For D. suzukii, attracticidal spheres may not serve as a stand-alone treatment, particularly under high pressure, but they do provide a mechanism for continuously killing flies between insecticide applications. Considering the rapid generation time of D. suzukii and available wild host reservoirs (Mitsui et al. 2010, Cini et al. 2012, Wiman et al. 2014, Lee et al. 2015), continuous killing stations such as attracticidal spheres could be useful in managing this invasive species.

When commercial D. suzukii lures were deployed in association with spheres, D. suzukii infestations in fruit increased compared with unbaited sphere treatments alone. Baits available to D. suzukii do not appear to be point source attractants. Chemotaxis by other taxa, particularly Lepidoptera to sex pheromones, involves directly contacting the emitting point source of pheromone (Roelofs and Cardé 1977). In the case of D. suzukii, adults may be attracted toward the bait plume but alight on adjacent fruit. Similar trends have been observed in other baited traps targeting agricultural pests. For instance, brown marmorated stink bug, Halyomorpha halys (Stål), is attracted to traps baited with pheromone in combination with the pheromone synergist (Morrison et al. 2016) or with the pheromone synergist alone (Sargent et al. 2014), but H. halys presence (Morrison et al. 2016) and feeding injury to surrounding crop (Sargent et al. 2014, Morrison et al. 2016) generally extend 1-2 m beyond the location of the baited trap itself. Similarly, when traps containing Dendroctonus pseudotsugae aggregation pheromone were placed in forest stands, adjacent trees experienced greater attack rates (Laidlaw et al. 2003).

Attracticidal spheres may provide a novel IPM-based management tool for D. suzukii. Because male and female D. suzukii respond strongly to red spheres (Rice et al. 2016), attracticidal spheres have the capacity to kill both sexes. Ultimately, more studies regarding the deployment strategy of attracticidal spheres (location within the plant, sphere density, etc.) and compatible horticultural practices (e.g., spray programs and harvesting intervals) will be needed to establish how to maximize the effectiveness of this attract-and-kill technique. Certainly, this technique holds merit, particularly because it has the potential to reduce the need for frequent insecticide application and decrease nontarget effects on natural enemies and pollinators, while reducing secondary pest outbreaks.

Acknowledgments

We thank Cesar Rodriguez-Saona for providing initial D. suzukii colony. This research was funded, in part, by the USDA-ARS and USDA NIFA CPPM # 2015- 70006-24152 and OREI # 2015-51300-24154 grants.

References Cited

Author notes