Efficacy of Kairomone Lures to Attract Parasitoids of Halyomorpha halys

Simple Summary The brown marmorated stink bug, Halyomorpha halys, is an invasive pest of agricultural crops, ornamentals, and human structures. In its native range, populations are suppressed primarily by parasitoid wasps that attack the egg stage. A promising adventive parasitoid, the samurai wasp, Trissolcus japonicus, has become established in the U.S., including Utah. According to ecological models, Utah is marginally suitable for the samurai wasp and poses unique challenges to its establishment from extreme climates and high elevation. Biological control enhancement efforts, such as deploying stink bug kairomones to attract parasitoids, may lead to the enhanced suppression of the brown marmorated stink bug. To evaluate the efficacy of this approach, experimental lures loaded with varying blends and release rates of stink bug kairomones were tested in field and mesocosm trials. This study found low parasitism in the field, while mesocosm trials demonstrated the efficacy of a single- and dual-compound blend at the 10 mg load rate for the attraction of the samurai wasp. These results support the validity of using rubber septa as a release device for kairomones of stink bugs and provide a baseline for future work on attracting samurai wasps with lures in a field environment. Abstract In its native range, Halyomorpha halys (Stål) is suppressed by parasitoids in the genus Trissolcus (Hymenoptera: Scelionidae). Trissolcus native to Utah have demonstrated low parasitism of H. halys, while adventive Trissolcus japonicus (Ashmead) have shown parasitism of up to 20%. Custom rubber septa lures containing stink bug kairomones, n-tridecane (attractant), and (E)-2-decenal (repellent), at 100%, 90%, and 80% levels of attractant (10 mg load rate), were placed adjacent to sentinel H. halys egg masses in northern Utah field trials. Egg masses were evaluated for the presence and intensity (proportion of parasitized eggs) of parasitism. Parasitism by T. japonicus and T. euschisti (Ashmead) was low; however, the 100% lure showed double the parasitism of the control and more than three times that of the 90% and 80%. Two-way choice mesocosm trials in the laboratory evaluated previous lures and a lower load rate of 5 mg—100% attractant treatment. Lures of 10 mg at 100% and 80% were more attractive to T. japonicus than the control, while 5 mg at 100% and 10 mg at 90% showed no significant attraction. Our results support a proof-of-concept of rubber septa as release devices for kairomones to attract T. japonicus and provide a baseline for future field-based studies.


Introduction
The brown marmorated stink bug, Halyomorpha halys (Stål) (Hemiptera: Pentatomidae), is an invasive pest originating from East Asia. H. halys was first detected in Utah in 2012, with the first injury to tree fruit and vegetable crops reported in 2017 [1]. The rapid spread of H. halys can be attributed to its ability to utilize more than 300 different host plants including fruits, vegetables, field crops, and ornamentals [2][3][4] upon which it can cause economic damage [5]. In addition to causing mild to severe damage to agricultural crops, H. halys is an urban nuisance due to its overwintering congregation behavior on and within human structures [6].
In its native range, H. halys populations are suppressed primarily by parasitoid wasps that attack the egg stage [7,8]. The wasp genus Trissolcus (Hymenoptera: Scelionidae) contains species that attack H. halys eggs; however, those native to Utah have demonstrated only modest parasitism rates of H. halys to date [9,10]. Trissolcus japonicus (Ashmead), or the samurai wasp, has been identified as a key biocontrol agent with parasitism rates of H. halys eggs up to 60-90% in Asia [11]. Adventive populations of T. japonicus have been detected in the U.S. since 2014 [12,13], with the first detection in Utah in Salt Lake City in June 2019 [9]. Early surveys found H. halys egg masses parasitized by T. japonicus on Catalpa speciosa (Warder) Warder ex Engelm. (Lamiales: Bignoniaceae) trees [9]; thus, northern catalpa has been a focus for H. halys and T. japonicus surveys in northern Utah [14].
Utah is a unique geographic location and climate for both H. halys and T. japonicus with high elevation (>1200 m in northern Utah), substantial snowfall in winter, and hot, arid summers. For up to eight months of the year, average low temperatures fall below the minimum threshold for the development of T. japonicus (12.2 • C) [15,16], and current climatic modeling suggests only marginal suitability in the state [17]. A contributing factor may be insect behavioral aspects unaccounted for in laboratory studies and geographic models [17]. Early research in northern Utah indicates H. halys egg parasitism rates have increased in the two summers since T. japonicus was discovered, thus demonstrating the potential to provide biological control of H. halys [9]. Promoting and conserving Utah's adventive populations of T. japonicus may be a viable option for the sustainable management of H. halys.
Though lab-reared sentinel H. halys egg masses are a common way to assess wasp parasitism rates, many studies have shown that they underestimate parasitism rates and attract fewer wasps than wild egg masses [9,18,19]. It is thought that T. japonicus utilizes volatile cues associated with H. halys oviposition and feeding during host location. One study found that the bioactive volatile n-tridecane significantly attracted T. japonicus and reduced its host search time, while (E)-2-decenal acted as a repellent [20]. Subsequent research by Malek et al. [21] found that combining these kairomone compounds at a ratio of 4:1 n-tridecane to (E)-2-decenal, or 80% attractant, performed better than n-tridecane alone in a Y-tube olfactometer experiment. A more recent study used a ratio of 9:1 (90% attractant) during field trials, but with little success, likely due to inadequate release rates from filter paper sources [22]. The kairomone compounds have proven attractive in small-scale experiments, but more research is needed to understand how they perform in larger settings.
Traditionally, rubber septa are a standard chemical release device used in pest management (e.g., pheromone lures) [23][24][25]. Septa can be loaded with kairomone compounds to attract beneficial parasitoids and enhance target pest parasitism [26,27]. Our objective was to determine the viability of synthetic kairomone-loaded septa lures to increase T.japonicus parasitism rates of H. halys egg masses in field and laboratory mesocosm settings. This research will contribute to the identification of effective techniques to deploy stink bug kairomones to attract parasitoids and may contribute to increased accuracy in estimates of H. halys parasitism rates and enhance early detection of T. japonicus. Future research could lead to the creation of a highly attractive lure for T. japonicus with the potential for use in field sites to increase parasitism, thereby reducing H. halys populations and subsequent crop and nuisance damage.

Field Trials
To assess the attractiveness of kairomone chemicals in a field setting, custom gray rubber septa lures loaded with 10 mg of test compounds were developed by Trécé, Inc., (Adair, OK, USA). This field study included four lure treatments with varying ratios of ntridecane to (E)-2-decenal: 100% n-tridecane, 90% n-tridecane, 80% n-tridecane, and hexane Insects 2023, 14, 125 3 of 10 (control). Lab-reared H. halys sentinel egg masses were deployed adjacent to kairomone lures as hosts for parasitoid wasps.
Trials (n = 6) were conducted from 24 June to 27 August 2021 in a strip of residential Catalpa speciosa trees in Salt Lake City, UT (40.772480, −111.854975). Catalpa trees were selected due to consistent and relatively high populations of H. halys observed on leaves and pods. Treatments were replicated in four trees (three in the final deployment due to a lack of egg masses) with a blank buffer tree between each treatment tree. Each replicate tree (3 m wide canopy) contained four H. halys egg masses and one of each lure treatment with one mass and lure placed approximately 2 m laterally from the trunk of the tree at each cardinal direction. Egg masses of H. halys were attached to small rectangles of white cardstock (2 cm by 3 cm) and clipped with a lure to the underside of tree leaves ( Figure 1). Treatments were placed at approx. a 2 m height above the ground with cardinal direction of treatments randomized for each tree in each deployment. To assess the attractiveness of kairomone chemicals in a field setting, custom gray rubber septa lures loaded with 10 mg of test compounds were developed by Trécé, Inc., (Adair, OK, USA). This field study included four lure treatments with varying ratios of ntridecane to (E)-2-decenal: 100% n-tridecane, 90% n-tridecane, 80% n-tridecane, and hexane (control). Lab-reared H. halys sentinel egg masses were deployed adjacent to kairomone lures as hosts for parasitoid wasps.
Trials (n = 6) were conducted from 24 June to 27 August 2021 in a strip of residential Catalpa speciosa trees in Salt Lake City, UT (40.772480, −111.854975). Catalpa trees were selected due to consistent and relatively high populations of H. halys observed on leaves and pods. Treatments were replicated in four trees (three in the final deployment due to a lack of egg masses) with a blank buffer tree between each treatment tree. Each replicate tree (3 m wide canopy) contained four H. halys egg masses and one of each lure treatment with one mass and lure placed approximately 2 m laterally from the trunk of the tree at each cardinal direction. Egg masses of H. halys were attached to small rectangles of white cardstock (2 cm by 3 cm) and clipped with a lure to the underside of tree leaves ( Figure 1). Treatments were placed at approx. a 2 m height above the ground with cardinal direction of treatments randomized for each tree in each deployment. Egg masses (n = 92) were ~48 h old and produced from an H. halys laboratory colony at the Oregon Department of Agriculture (Salem, OR, USA). Sentinel egg masses were deployed with kairomone treatment lures in the field for approximately 96 h and returned to the laboratory for evaluation. Any wasps found guarding egg masses were collected into a 9-dram plastic vial (Thornton Plastics, Salt Lake City, UT, USA) using a WHO (World Health Organization) in-line tube aspirator (Bioquip, Compton, CA) for later identification. After eggs were incubated at 25-27 °C for 14 days in the laboratory, they were evaluated for parasitism incidence and intensity (proportion of parasitized eggs per mass), and emerged wasps were identified to species using the key to Nearctic Trissolcus [28]. Egg masses were observed again 14 days later (4 weeks after collection) to identify late-emerging wasps or eggs with partially developed wasps or stink bugs. Individual egg fate was recorded as emerged or undeveloped parasitoid, hatched or unhatched H. halys nymph, predated, sunken, empty, or missing. Egg masses (n = 92) were~48 h old and produced from an H. halys laboratory colony at the Oregon Department of Agriculture (Salem, OR, USA). Sentinel egg masses were deployed with kairomone treatment lures in the field for approximately 96 h and returned to the laboratory for evaluation. Any wasps found guarding egg masses were collected into a 9-dram plastic vial (Thornton Plastics, Salt Lake City, UT, USA) using a WHO (World Health Organization) in-line tube aspirator (Bioquip, Compton, CA, USA) for later identification. After eggs were incubated at 25-27 • C for 14 days in the laboratory, they were evaluated for parasitism incidence and intensity (proportion of parasitized eggs per mass), and emerged wasps were identified to species using the key to Nearctic Trissolcus [28]. Egg masses were observed again 14 days later (4 weeks after collection) to identify late-emerging wasps or eggs with partially developed wasps or stink bugs. Individual egg fate was recorded as emerged or undeveloped parasitoid, hatched or unhatched H. halys nymph, predated, sunken, empty, or missing.

Mesocosm Trials
In order to mitigate multiple uncontrollable factors in a field setting, a second experiment was conducted in a mesocosm-scale lab-based experimental system in 2022. In a 0.  Four to five 1-20 day post-emergence (according to colony availability), honey-fed, female T. japonicus from the Utah State University T. japonicus colony (originating from females collected in 2019 from Salt Lake City, UT, USA) were introduced to the center, bottom of the cage via a small access hole (stoppered by a cork) and allowed up to 30 min to select between the control and treatment lure by landing on the adjacent clear panel sticky trap (2-way choice). Treatments consisted of four combinations of chemical attractant to repellent ratios and load rates: 5 mg-100% n-tridecane, 10 mg-100% n-tridecane, 10 mg-90% n-tridecane, and 10 mg-80% n-tridecane.
Each treatment was tested against a hexane control lure and trials were replicated until each treatment had a minimum of 32 individuals make a choice. Wasps that did not make a choice within the 30 min experimental period were not included in the analyses. The observation cage was cleaned with laboratory detergent and allowed to air dry between lure types and treatment vs. control lures were rotated every other trial to account for side bias. Trials were conducted in a temperature-controlled rearing room at 21-27 • C and 30-50% RH.

Statistical Analysis
To analyze field trial results, data were pooled across sample dates and each egg fate category was analyzed using the Kruskal-Wallis test to compare differences across treatment lure types. In mesocosm trials, the null hypothesis that T. japonicus showed no preference between the control and treatment lure (an overall choice of 50:50) was analyzed with a Chi-square goodness of fit test. All statistical comparisons were run using R software [29] and were considered significant at p < 0.05.

Field Trials
Of 92 total egg masses containing 2472 eggs deployed in the field, only 13 masses had evidence of parasitism (14.1%; guarding female present or eclosed parasitoids) of which 6 masses supported successful emergence of adult parasitoids (6.5%). The lure treatments in order of highest proportion of eggs with successfully emerged parasitoids were 100% ntridecane (10.6%), hexane control (5.2%), 80% n-tridecane (2.9%), and 90% n-tridecane (0.0%) (Figure 2). Though the 100% lure showed double the parasitism of the control and more than three times that of the 90% and 80%, the analysis of all egg fate categories did not support statistical differences (p > 0.05; Supplementary Table S1).
T. japonicus attacked 5.4% of deployed egg masses. This included one egg mass in each of the control and 80% attractant treatments and three in the 100% attractant treatment. Of the 129 H. halys eggs T. japonicus parasitized, there was a high rate of successful emergence (78.3%). T. euschisti attacked 6.5% of egg masses deployed with two egg masses in each of the control, 100%, and 90% attractant treatments and none in the 80% attractant treatment. Successful emergence of T. euschisti only occurred in one of the control treatment egg masses and the rate of overall emergence from attacked egg masses (136 eggs) was 1.8%. Two egg masses in the control treatment were parasitized by an unknown Trissolcus sp(p) and demonstrated no successful emergence (Table 1). In one instance, a T. euschisti female was found guarding an egg mass which later produced 26 T. japonicus individuals (92.9% emergence rate) and 2 inviable eggs.
Undeveloped parasitoid rates ranged from 1.6% to 5.9% across treatments, were highest in the 80% attractant treatment, and lowest in the 100% attractant treatment. The rate of unhatched H. halys was also variable across treatments, making up the largest percentage of egg fate for the control treatment (26.5%), and was noticeably lower in the 100% attractant treatment compared to others. There was less than a 2% difference in the predation between lure treatments. Sunken rates were relatively high (17.7-26.0%), and the combination of missing and empty on average across lure treatments made up <10% of egg fate. Rates of hatched H. halys nymphs were similar across lure treatments containing stink T. japonicus attacked 5.4% of deployed egg masses. This included one egg mass in each of the control and 80% attractant treatments and three in the 100% attractant treatment. Of the 129 H. halys eggs T. japonicus parasitized, there was a high rate of successful emergence (78.3%). T. euschisti attacked 6.5% of egg masses deployed with two egg masses in each of the control, 100%, and 90% attractant treatments and none in the 80% attractant treatment. Successful emergence of T. euschisti only occurred in one of the control treatment egg masses and the rate of overall emergence from attacked egg masses (136 eggs) was 1.8%. Two egg masses in the control treatment were parasitized by an unknown Trissolcus sp(p) and demonstrated no successful emergence (Table 1). In one instance, a T. euschisti female was found guarding an egg mass which later produced 26 T. japonicus individuals (92.9% emergence rate) and 2 inviable eggs.

Mesocosm Trials
In the mesocosm system, T. japonicus females showed a significant preference for the kairomone blend over the control for two of the four treatments tested (Figure 3). The 10 mg-80% n-tridecane showed the most significant response (p = 0.003) with 73% of individuals choosing the kairomone lure. There was also a significant difference between the control and 10 mg-100% n-tridecane treatment lure (p = 0.04) with 60% of wasps choosing the kairomone-treated side. For the 10 mg-90% n-tridecane treatment, 55% of Insects 2023, 14, 125 6 of 10 wasps chose the kairomone lure; however, there was no significant deviation from the expected response (p = 0.55). The choice between the low load rate treatment of 5 mg-100% n-tridecane and the control also showed no significance (p = 1); an even proportion of wasps chose each side. The proportion of no-choice wasps per treatment varied from 20% to 32% of wasps tested and was not associated with treatments.

Mesocosm Trials
In the mesocosm system, T. japonicus females showed a significant preference for the kairomone blend over the control for two of the four treatments tested (Figure 3). The 10 mg-80% n-tridecane showed the most significant response (p = 0.003) with 73% of individuals choosing the kairomone lure. There was also a significant difference between the control and 10 mg-100% n-tridecane treatment lure (p = 0.04) with 60% of wasps choosing the kairomone-treated side. For the 10 mg-90% n-tridecane treatment, 55% of wasps chose the kairomone lure; however, there was no significant deviation from the expected response (p = 0.55). The choice between the low load rate treatment of 5 mg-100% ntridecane and the control also showed no significance (p = 1); an even proportion of wasps chose each side. The proportion of no-choice wasps per treatment varied from 20 to 32% of wasps tested and was not associated with treatments.

Discussion
This study is the first to assess and provide a proof-of-concept for kairomone-infused rubber septa lures to attract T. japonicus and other potential H. halys parasitoids in field and mesocosm settings. Though the kairomone chemicals tested herein have been previously evaluated in a small-scale lab setting (Y-tube olfactometer) [20], implementation in the field may provide different results due to the influence of external factors on the shape, concentration, longevity, and spatial extent of the kairomone odor plume. In addition, the

Discussion
This study is the first to assess and provide a proof-of-concept for kairomone-infused rubber septa lures to attract T. japonicus and other potential H. halys parasitoids in field and mesocosm settings. Though the kairomone chemicals tested herein have been previously evaluated in a small-scale lab setting (Y-tube olfactometer) [20], implementation in the field may provide different results due to the influence of external factors on the shape, concentration, longevity, and spatial extent of the kairomone odor plume. In addition, the interaction of the lure plume with nearby plant surfaces and mixing with other volatiles may give rise to plume masking or plume amplification, of which little is known in a parasitoid searching context [30].
Our field results contrast those of Malek et al. [21] where the combination of the attractant and repellent was preferred over the attractant alone. In our study, not only did the 100% n-tridecane attractant lure have a numerically (not statistically) higher parasitism rate than the control, the lures containing the previously identified repellent (E)-2-decenal had less parasitism than those without the repellent, including the hexane control. These results may suggest that under field conditions a combination of the attractant and repellent is less effective at increasing parasitism by Trissolcus spp., at least with the specific load rate tested and placement in near proximity to the target host egg masses.
The lures used in the field study contained a load rate of 10 mg per lure, high in comparison to standard pheromone lures (1 mg), in the hopes of ensuring the attraction of parasitoids (Trécé, Inc., Adair, OK, USA); however, this relatively high load rate may have been counterproductive to our objective. In studies of pheromone lure load rates, it has been found that attraction often plateaus and can even become repellent to the insect at high release rates [23][24][25]. In addition, our lures were placed directly adjacent to the H. halys egg masses (Figure 1), which may deter parasitoid attraction as the host female stink bug does not typically remain near the eggs following oviposition.
The effect of lure load rate on parasitoid attraction was explored by examining both a 5 and 10 mg treatment of only the attractant n-tridecane in the mesocosm trials. The results of these trials demonstrate that the higher load rate of 10 mg was a more viable option for attracting T. japonicus and did not have a repellent effect. In contrast, the 5 mg treatment demonstrated no difference in attraction as compared with the hexane control lure. Based on these findings, the load rate may not have been detrimentally high in the field trials; however, we did not assess if a load rate greater than 10 mg could increase parasitoid attraction in a field setting.
Given the low number of egg masses tested in field trials and a lack of statistically significant results, the parameters of this field study may have been unsuitable to discern differences in the attractiveness of the different kairomone lure treatments. We observed low parasitism rates across all egg mass deployment periods in this study. In previous surveys, the site selected for the field study had the highest abundance of T. japonicus observed in northern Utah with concomitant high parasitism rates of H. halys [9]. However, compared to T. japonicus populations in its native geographic regions and other adventive populations in the U.S., the abundance of T. japonicus in Utah is relatively low [13,31]. In addition, Utah suffered from significant drought over the time frame of this study (summer 2021), with 99.94% of the state in "extreme" or "exceptional" drought categories [15], which may have had a negative effect on host abundance, parasitoid wasp populations, and H. halys egg parasitism rates [6].
Other researchers have observed that lab-reared egg masses perform inferiorly to wild egg masses in terms of their attractiveness to parasitoids [9,18,19]. The lures tested here were an initial attempt to solve this issue and increase the accuracy of parasitism rates detected in deployed egg mass surveys. Parasitism was observed in wild H. halys egg masses near deployed, unparasitized egg masses during the field study. Without an in-depth analysis of the kairomone plume release from lures, it is unknown if the plumes for each lure treatment may have overlapped within the tree block (lures were separated by approximately 3 m of tree canopy) and potentially interfered with one another. In addition, this may suggest the lure-dispersed kairomones had unknown interactions with the surrounding environment or may have lacked the necessary kairomone load rate and/or ratio necessary to match the high attractiveness of wild egg masses with their natural kairomones intact.
Interestingly, the results of the mesocosm environment did not fully align with the field trial results, suggesting lures containing (E)-2-decenel may be repellent, but were better aligned with those of Malek et al. [21] where the 80% n-tridecane ratio was more attractive than n-tridecane alone. The result of wasp attraction to the 100% and 80% lures and unexpected lack of attraction to the 90% treatment occurred both in the field and in the mesocosm trials. Research has suggested that the variable secretions by different sexes and life stages of H. halys may be linked to their specific functions at said life stages and physiological states [21,32,33]. Perhaps the 100% and 80% n-tridecane treatments in this study are more closely associated with the gravid/ovipositing adult female or egg life stages that can be exploited by T. japonicus, while the 90% treatment is associated with other life stages. Further investigation is necessary to understand these counterintuitive results.
The mesocosm trials presented the opportunity to test the response of a much larger sample size and demonstrate significant differences between the control and treatments, verifying the validity of these lures. While these trials supported the attractiveness of certain lures in a controlled environment, they also did not include many of the external factors at play in an authentic environment such as plant volatiles and competing parasitoids and predators. In addition, our mesocosm trials did not include the presence of host eggs which may have altered the T. japonicus response.
In contrast, the field trials provide essential information about the interaction between hosts and parasitoids in a novel geographic and climatic environment. The results further verify the preliminary research in Utah on the effectiveness of the exotic T. japonicus and the common native T. euschisti. In this study, as in Holthouse et al. [9], T. euschisti demonstrated the ability to parasitize H. halys in a similar proportion to the natural parasitoid of H. halys, T. japonicus. However, T. euschisti seems an inviable option for successful biological control due to very low adult wasp emergence and successful stink bug nymph development likely associated with the failure of T. euschisti eggs to hatch or early death of larvae within H. halys eggs [34]. Additionally, this poses an evolutionary trap for T. euschisti and other native Trissolcus species that accept H. halys eggs as ovipositional sites despite their poor reproductive investment [35]; although, there is evidence that the recent arrival of T. japonicus may have implications for the success of native Trissolcus on H. halys egg masses.
Research has demonstrated that Trissolcus spp. are able to parasitize host eggs that have previously been parasitized. In the case of T. japonicus and T. mitsukurri (Ashmead), each species outperformed the other when it was the first to oviposit, though T. mitsukurri was more aggressive in chasing off T. japonicus when present concurrently [36]. The similar timing of the competition and outperformance may have been the case in our observation of T. euschisti guarding an egg mass (presumable parasitized by T. japonicus prior to T. euschisti oviposition) that only produced T. japonicus offspring. Conversely, research by Konopka et al. [37] demonstrated that parasitism by the exotic T. japonicus can provide facultative parasitism opportunities for native Trissolcus, such as T. cultratus (Mayr), to successfully develop in H. halys eggs when they would otherwise fail. Regardless of their reproductive success, native Trissolcus species can reduce the developmental success of H. halys embryos providing low levels of control for the pest [38]. The attraction of T. eushisti and other native Trissolcus species to kairomone lures containing n-tridecane and (E)-2-decenal were not explored in a mesocosm environment in this study. However, the use of these kairomone lures in the field did result in attacks from T. euschisti in addition to T. japonicus. Consequently, further research into the physiological and behavioral interactions between the exotic T. japonicus and North American native Trissolcus requires investigation to fully evaluate the potential efficacy of biological control programs against H. halys.
While much investigation is needed into the complex system of parasitoids using semiochemicals to locate their host, the mesocosm results presented support the validity of using rubber septa as a release device for the kairomones of stink bugs, and this research provides not only preliminary results but also a baseline for future work with experimental lures for H. halys parasitoids in a field environment.

Conclusions
As biological control continues to be a preferred approach for managing the invasive H. halys, it is important to explore all avenues for attracting and retaining effective parasitoids. Here, we provide support for the validity of infusing rubber septa lures with H. halys kairomones to attract T. japonicus. This novel strategy has the potential to increase parasitism and, therefore, suppression of H. halys in agricultural and urban settings and deserves further investigation.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/insects14020125/s1, Table S1: Statistical results from Kruskal-Wallis tests to compare each egg fate category differences across treatment lure types.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.