The brown marmorated stink bug, Halyomorpha halys (Stål) (Hemiptera: Pentatomidae) is an invasive stink bug species native to China (Hoebeke & Carter 2003). Brown marmorated stink bug is an economically important agricultural pest (Leskey et al. 2012) as well as an urban nuisance (Inkley 2012). Presently, brown marmorated stink bug reproductive populations have colonized the Piedmont and are expanding into the Coastal Plain Region of Georgia and Alabama. The brown marmorated stink bug is highly polyphagous, ovipositing on a variety of hosts (Bakken et al. 2015; Bergmann et al. 2016), including sassafras (Sassafras albidum [Nutt.] Nees; Lauraceae), a non-crop host that is found in woodlands bordering agricultural fields in this region. In the southeastern US, populations of brown marmorated stink bug feed on and oviposit in fruit, vegetable, and row crops. Natural enemies (i.e., predators and parasitoids) may provide effective means of controlling brown marmorated stink bugs in this region, and sentinel egg masses often are deployed to assess potential suppression (Cornelius et al. 2016a, b; Herlihy et al. 2016; Morrison et al. 2016; Ogburn et al. 2016; Dieckhoff et al. 2017; Tillman et al. 2020). However, the duration of egg mass deployment may affect parasitism and predation rate. Therefore, it is crucial for assessing biological control services to understand the relative effects of sentinel egg deployment duration in a variety of hosts on the responses of native parasitoids and predators to this invasive pest.

Field surveys on biological control of brown marmorated stink bug eggs in various agroecosystems in the US have revealed that native natural enemies attack egg masses at varying rates of parasitism and predation. Presently, 20 species of Nearctic hymenopteran primary endoparasitoids in the genera Anastatus Motschulsky (Eupelmidae), Trissolcus Ashmead (Scelionidae), Telenomus Haliday (Scelionidae), Gryon Haliday (Scelionidae), and Ooencyrtus Ashmead (Encyrtidae) (all Hymenoptera) have been reported to parasitize brown marmorated stink bug eggs (Cornelius et al. 2016a, b; Herlihy et al. 2016; Morrison et al. 2016; Ogburn et al. 2016; Dieckhoff et al. 2017; Jones et al. 2017; Balusu et al. 2019a, b; Tillman et al. 2020). Adventive populations of Trissolcus japonicus (Ashmead) (Hymenoptera: Scelionidae), a parasitoid of brown marmorated stink bug in its native range, have been found in several US states but not the southeast (Talamas et al. 2015a; Milnes et al. 2016). A diverse complex of chewing and piercing-sucking predators attack brown marmorated stink bug eggs. Morrison et al. (2016) categorized predators of brown marmorated stink bug eggs into 4 characteristic patterns of predator feeding damage: complete chewing by Gryllidae and Tettigoniidae; incomplete chewing by Coccinellidae, Carabidae, and Dermaptera; stylet sucking by Anthocoridae and Pentatomidae; and punctured sucking by Salticidae. In addition, Tillman et al. (2020) described removal of whole eggs from egg masses by Formicidae, hole sucking by late instar Chrysoperla carnea (Stephens) (Neuroptera: Chrysopidae), and non-stylet sucking by Nabis spp. Latreille (Hemiptera: Nabidae).

The method by which brown marmorated stink bug egg mass predation and parasitism is measured may yield different patterns of activity. Higher rates of parasitism have been found in wild (naturally oviposited) brown marmorated stink bug eggs compared to fresh sentinel (laboratory-reared) eggs (Jones et al. 2014; Dieckhoff et al. 2017). Wild brown marmorated stink bug eggs may be attacked by a greater number of parasitoid species compared to sentinel egg masses (Jones et al. 2014), or the number of undeveloped parasitoids may differ in wild versus fresh sentinel eggs (Dieckhoff et al. 2017). Using wild egg masses to assess parasitism and predation is beneficial because the timing of assessment is linked to the phenology of both brown marmorated stink bug and host plants in the landscape. However, wild eggs can be challenging to find on host plants, especially in sufficient quantities for assessing rates of parasitism and predation. As such, deploying sentinel egg masses on host plants is necessary to assess patterns of parasitism and predation of brown marmorated stink bug egg masses across hosts and locations. Moreover, in recent stink bug sentinel egg studies, the number of d egg masses were deployed in fields ranged from 1 to 5 d (Cornelius et al. 2016a, b; Herlihy et al. 2016; Morrison et al. 2016; Ogburn et al. 2016; Dieckhoff et al. 2017; Jones et al. 2017; Tillman et al. 2020), which likely affects observed amounts of predation and parasitism. Therefore, determining the amount of time sentinel egg masses should be deployed for any given host is necessary to compare between and interpret patterns of brown marmorated stink bug egg predation and parasitism.

Table 1.

Sites, locations, and host plants where brown marmorated stink bug sentinel egg masses were deployed.

The objective of this study was to test the effect of deployment duration on predation and parasitism of brown marmorated stink bug sentinel egg predation and parasitism by native natural enemies in a variety of hosts. Over a 4-yr period, retrieved sentinel egg masses were evaluated for parasitism and predation at 2 to 5 d of exposure in plum, peach, blueberry, tomato, sassafras, corn, and soybean in Georgia and Alabama.

Materials and Methods

This study was conducted at 6 sites in Georgia and Alabama (Table 1). The presence of wild brown marmorated stink bug populations were confirmed at each site. Insecticides were not applied to plant species during the season. Brown marmorated stink bug sentinel egg masses were deployed in plum, peach, blueberry, tomato, sassafras, corn, and soybean.

Results

PREDATION

Across crops in 2017, percent predation of brown marmorated stink bug sentinel egg masses was predicted by the interaction between crop type and duration of deployment (F_{3, 155} = 3.14; P = 0.0267) of brown marmorated stink bug sentinel egg masses. Percent predation was significantly higher in plum exposed for 5 d compared to 2 d, whereas in soybean percent predation was similar regardless of exposure time (Table 2). Predation was significantly lower in plum at the 2 d deployment time than for soybean at either 2 d or 5 d of exposure.

In 2018 to 2020, percent of predation of brown marmorated stink bug sentinel egg masses was predicted by duration of deployment (F_{2, 699} = 4.09; P = 0.017) and crop type (F_{6, 699} = 2.74; P = 0.012). Regardless of crop type, percent predation was significantly higher after 5 d of exposure in the field compared to the 2 d and 3 d exposure times (Table 2). Percent predation levels were the lowest in tomato followed by blueberry, and highest in peach, corn, plum, and sassafras.

In plum, comparison of exposure time within rounds suggests that percent predation of brown marmorated stink bug sentinel egg masses varied by round and within the season. In 2017, significantly higher percent predation was detected for the 5 d deployment treatment compared to 2 d for round 3 (Table 3). Percent predation of brown marmorated stink bug egg masses was influenced significantly by the deployment duration in this fruit crop in 2018 for round 2 and in 2019 for both rounds (Table 4). In each case, percent predation was higher after a 5 d deployment time than after only 2 d (Tables 5, 6). For round 1 of 2019, the rate of predation was higher for the 3 d exposure time compared to 2 d. In general, percent predation increased in plum over the season through Sep (Tables 3, 5, 6). Early in the season, complete (Acrididae, Tettigoniidae, and Gryllidae) and incomplete (Coccinellidae larvae) chewing predators and removal of egg masses (Formicidae, especially Solenopsis invicta Buren [Hymenoptera: Formicidae]) were prevalent. However, punctured sucking predation (Salticidae) and piercing-sucking hole (Chrysopidae, mainly Chrysoperla carnea [Stephens] [Neuroptera: Chrisopidae]) predation also were detected. By mid-season, stylet piercing-sucking predation was detected, primarily due to cannibalism by brown marmorated stink bug nymphs.

Table 2.

Means (± SE) for percent predation and parasitism of brown marmorated stink bug sentinel eggs by crop and duration of deployment from 2017 to 2020.

Table 3.

Means (± SE) for percent predation and parasitism of brown marmorated stink bug sentinel eggs 2 and 5 d after exposure to plum in Byron, Georgia, USA, and soybean in Williamson, Georgia, USA, in 2017.

In soybean, comparison of exposure time within rounds suggests that percent predation of brown marmorated stink bug sentinel egg masses also varied by round in this row crop. Percent predation was significantly higher at 5 d than 2 d deployment for round 2 in 2017 (Table 3), and for round 1 in 2018 where predation was higher at 3 d deployment compared to 5 d (Tables 4, 5). Similar types of predation, mainly complete chewing (Acrididae, Tettigoniidae, and Gryllidae) and removal of egg masses (Formicidae, especially S. invicta) and some stylet sucking (Anthocoridae, primarily Orius insidiosus [Say] [Hemiptera: Anthocoridae], and Geocoridae, primarily Geocoris punctipes [Say] [Hemiptera: Geocoridae]) were detected in all soybean plots over time.

In contrast, in peach, blueberry, tomato, corn, and sassafras, deployment duration did not affect predation of sentinel egg masses by round (Tables 4, 5, 6). In peach, stylet piercing-sucking predation was the dominant type of predation detected upon examination of eggs in the laboratory. Chewing and stylet piercing-sucking predation, as well as some punctured sucking predation, were detected for egg masses deployed in blueberry. In tomato, complete chewing predation (Acrididae and Tettigoniidae) was the prevailing type of feeding damage on egg masses. In corn, complete chewing, egg removal, and stylet-sucking predation were detected for the brown marmorated stink bug sentinel egg masses. In sassafras, complete and incomplete chewing, punctured and stylet piercing-sucking, and removal of eggs from egg masses were the main types of predation on brown marmorated stink bug eggs.

PARASITISM

Across crops in 2017, percent parasitism of brown marmorated stink bug sentinel egg masses was predicted by a significant interaction between deployment duration and crop type (F_{3, 155} = 4.89; P = 0.004). Similar to patterns of predation, percent parasitism was significantly higher for 5 d of exposure compared to 2 d in plum and similar in soybean at each of these exposure times (Table 2). Parasitism was significantly higher for plum at the 5 d exposure time than for soybean at either deployment time.

Table 4.

Fixed effect estimates of the influence of exposure time on percent predation and parasitism of brown marmorated stink bug sentinel eggs in host plants in 2018, 2019, and 2020.

Across crops in 2018 to 2020, the percent of parasitism of brown marmorated stink bug sentinel egg masses was significantly influenced by duration of deployment (F_{2, 858} = 10.48; P < 0.0001) and crop type (F_{6, 858} = 5.58; P < 0.0001). Percent parasitism was significantly higher after 5 d of exposure in the field compared to a 3 d exposure time and higher after 3 d of exposure compared to 2 d (Table 2). Percent parasitism was highest in sassafras compared to all other host plants except tomato (Table 2). Parasitism was very low in peach and plum.

Table 5.

Least squares mean (± SE) for percent predation and parasitism of brown marmorated stink bug sentinel eggs 2, 3, and 5 d after exposure to host plants in 2018.

In plum in 2017, percent parasitism of brown marmorated stink bug sentinel egg masses was significantly higher when deployed for 5 d compared to 2 d for round 5 (Table 3). It was numerically higher at 5 d than 2 d in round 1 (only 6 egg masses) and 4; at the time of each of these rounds, predation was relatively low. Percent parasitism of brown marmorated stink bug egg masses was influenced significantly by the length of time the egg masses were deployed in this fruit crop in round 1 in 2018 (Table 4), and was higher when eggs were deployed for 5 d than for 2 d and 3 d (Table 5). Parasitism was numerically higher for the 5 d exposure time than for the shorter deployment times for round 1 in 2019 (Table 6). In 2018, the diversity of parasitoid species, Trissolcus euschisti (Ashmead) (Hymenoptera: Scelionidae), Trissolcus brochymenae (Ashmead) (Hymenoptera: Scelionidae), Trissolcus solocis Johnson (Hymenoptera: Scelionidae), Te. podisi, An. reduvii, and Ooencyrtus sp., that emerged from brown marmorated stink bug sentinel egg masses was higher than in 2017 and 2019 where only Tr. euschisti, An. reduvii, and Ooencyrtus sp. emerged from sentinel egg masses.

Table 6.

Least squares mean (± SE) for percent predation and parasitism of brown marmorated stink bug sentinel eggs 2, 3, and 5 d after exposure in host plants in 2019, and 3, 4, and 5 d after exposure to sassafras (Round 2) in 2020.

In peach, although deployment duration did not significantly affect parasitism of sentinel egg masses in early Jul 2019 (round 1) (Table 4), during round 2 exposure time did significantly affect parasitism (Tables 4, 6). The rate of parasitism was higher when the exposure time was extended to 5 d (Table 6). Mainly Ooencyrtus sp. emerged from egg masses in peach, but Te. podisi and An. reduvii also parasitized eggs in this host crop.

In tomato, percent parasitism was influenced by exposure time in Jun 2019 (Table 4). Unlike peach, percent parasitism was highest for the 5 d exposure than the other 2 exposure times, and 3 d was higher than 2 d exposure (Table 6). Eggs were parasitized primarily by Gyron obesum Masner (Hymenoptera: Scelionidae), but also Tr. brochymenae and Ooencyrtus sp. in tomato.

In sassafras, percent parasitism was affected by exposure time in 2019 and 2020 (Table 4). In 2019, percent parasitism was higher at 5 d deployment compared to the 2 d and 3 d treatments (Table 6). In 2020, 5 d deployment resulted in higher percent parasitism compared to the 3 d treatment, but percentages were similar between 4 and 5 d of exposure. Also, 16.7% of egg masses were 100% parasitized at 4 and 5 d whereas 11.1% of those parasitized on d 2 of exposure were completely parasitized. Interestingly, approximately 1 to 4 h after deployment in sassafras in 2020, 8 of the egg masses were observed with an ovipositing T. euschisti female, and 5 of them were observed with an ovipositing An. reduvii female. Over both yr, retrieved egg masses were parasitized by Tr. euschisti, An. reduvii, Ooencyrtus sp., Tr. brochymenae, and Te. podisi.

In blueberry, deployment duration did not affect parasitism of sentinel egg masses in 2018 and 2019 even though parasitism was numerically higher at 5 d than at the 2 d and 3 d exposure times in 2018 (Table 4, 5, 6). The only parasitoid species emerging from brown marmorated stink bug egg masses was An. reduvii and Tr. brochymenae in 2018 and 2019, respectively.

In soybean and corn, duration of deployment did not significantly affect percent parasitism (Tables 3, 4, 5, 6). The diversity of parasitoid species emerging from brown marmorated stink bug eggs in soybean was relatively low, consisting primarily of Ooencyrtus sp. and few Te. podisi, with the exception of Tr. basalis that parasitized brown marmorated stink bug eggs when the southern green stink bug, Nezara viridula (L.) (Hemiptera: Pentatomidae), also was present in soybean. In corn, brown marmorated stink bug eggs were parasitized by Ooencyrtus sp., Te. podisi, and Tr. basalis.

Discussion

Across crops, overall patterns regarding the influence of duration of exposure on percent predation and parasitism of brown marmorated stink bug egg masses in the field were detected. Indeed, deployment at 5 d increased the rate of predation beyond that for the 2 d and 3 d deployment time. Unlike percent predation, the rate of parasitism also was higher for 3 d compared to 2 d. Thus, perhaps 5 d is the optimal deployment time to assess percent predation and parasitism.

For the individual within-crop rounds, the effects of time of exposure were tested in different yr, sampling dates, and different crops with variable results. Although deployment time significantly increased predation and parasitism rates for some rounds, in others it had no significant effect. For example, for the 2 crops in 2017, exposure time significantly affected predation in 2 out of 8 rounds and parasitism in 1 out of 8 rounds. Similarly, across the 7 crops in 2018 to 2020, exposure time significantly increased predation in 4 out of 18 rounds and parasitism in 5 out of 18 rounds. In plum, peach, tomato, and sassafras, when significant differences were detected for parasitism, egg masses deployed for 5 d resulted in higher percent parasitism compared to 2 d. In tomato, 3 d of exposure led to a higher percent parasitism than the 2 d treatment. Higher parasitism with 3 d compared to the 2 d treatment may have been due to a combination of moderate predation and smaller plant size, but this needs further evaluation. In sassafras, 4 d of exposure was similar to the 5 d treatment. In plum and soybean, when percent predation increased over the season, higher predation was detected 5 d after deployment compared to 2 d.

Except for 2 new records, Tr. solocis and Tr. basalis (Balusu et al. 2019a, b), each native parasitoid species that emerged from brown marmorated stink bug eggs in our study has been reported to parasitize eggs of this pest in other regions of the US (Cornelius et al. 2016a, b; Herlihy et al. 2016; Ogburn et al. 2016; Dieckhoff et al. 2017; Jones et al. 2017). Prevalence of parasitoid species that emerged from brown marmorated stink bug eggs was primarily habitat specific, as previously reported for brown marmorated stink bug and native stink bug species (Okuda & Yeargan 1988; Cornelius et al. 2016a; Herlihy et al. 2016; Tillman 2016; Jones et al. 2017). Anastatus reduvii, Tr. brochymenae, and Tr. euschisti were prevalent in woody habitats (i.e., plum, peach, and sassafras). Ooencyrtus sp. was the primary species emerging from brown marmorated stink bug eggs in tomato, soybean, and corn. Gryon obesum and Tr. basalis and Te. podisi parasitized brown marmorated stink bug eggs in these crops when egg masses of N. viridula and Euschistus servus (Say) (Hemiptera: Pentatomidae), respectively, were attacked by these parasitoids.

The type of predation damage detected for brown marmorated stink bug sentinel egg masses varied by crop. Complete and incomplete chewing, punctured sucking, and hole sucking predation were detected for eggs deployed in plum and sassafras. As brown marmorated stink bug nymphs develop in plum and peach over the season, they cannibalize sentinel eggs resulting in an increase in prevalence of stylet sucking feeding on eggs (Tillman et al. 2020). Nymphs likely led to the increase in predation of brown marmorated stink bug egg masses over the season in plum in the current study and have the potential to increase predation in other plants. Indeed, development of brown marmorated stink bug nymphal populations and subsequent stylet-sucking predation of sentinel egg masses by nymphs has been observed in sassafras (G. Tillman, unpublished data). In contrast to plum, predation of brown marmorated stink bug egg masses in soybean and corn potentially could reach high levels, but also was consistent over the season and across treatments. Thus, parasitism remained low over time as reported in Tillman et al. (2020). A complex of predator species prey on stink bugs, including their nymphs, in corn and soybean (Tillman 2011; Tillman et al. 2015; Morrison et al. 2016; Ogburn et al. 2016). Complete chewing, removal of egg masses, and stylet sucking represented the major predator groups in these crops. In a previous report, predation was higher than parasitism in 2 row crops, cotton and soybean, and chewing predation and egg removal were the prevalent types of predation (Tillman et al. 2020).

In 2017 and 2018, we discovered Acroclisoides sinicus (Huang & Liao) (Hymenoptera: Pteromalidae), a hyperparasitoid of primary scelionid egg parasitoids, emerging from naturally occurring stink bug eggs masses, including a parasitized brown marmorated stink bug egg mass (i.e., eggs black) in 2017 in pecan, 3 Chinavia hilaris (Say) (Hemiptera: Pentatomidae) egg masses collected in mimosa 7 d after detected as fresh egg masses in 2018, and 2 N. viridula egg masses collected in cotton 7 d after being found as fresh egg masses (Sabbatini Peverieri et al. 2019; Giovannini et al. 2021). This hyperparasitoid also has emerged from Euschistus sp. and Brochymenae sp. egg masses in Maryland, USA (Sabbatini Peverieri et al. 2019). A preliminary exposure time test using brown marmorated stink bug sentinel egg masses in sassafras trees in a woodland habitat indicated that the duration of exposure influenced the likelihood of parasitization of scelionids by A. sinicus (Giovannini et al. 2021). Adult A. sinicus did not emerge from egg masses exposed for 3 and 4 d, 1 A. sinicus male emerged from egg masses exposed for 5 d, and females and males emerged from egg masses exposed for 6 d. In a laboratory study, A. sinicus emergence was obtained when exposing 5 d and 7 d old primary parasitized egg masses for 72 h to females of A. sinicus (Giovannini et al. 2021). Thus, unless sentinel egg masses are deployed in the field for at least 5 d, this hyperparasitoid species, and perhaps others, will not be detected.

Many factors can impact the rate of parasitism in the field, including timing and duration of stink bug oviposition in a host plant, parasitoid host searching behavior, parasitoid habitat preferences, activity within and between nearby crop and non-crop host plants, availability of floral resources, management practices, environmental conditions, and host and parasitoid species density and diversity. The first step a foraging female takes to parasitize a stink bug egg mass is to find suitable habitat and search for a host (Bin et al. 1993; Ponzio et al. 2016). Once a female locates a host egg mass, she assesses host suitability (Laumman et al. 2009), and if suitable, she oviposits an egg into the host egg (Wilson 1961; Weber et al. 1996). During this process, oviposition by a female can be influenced by many factors such as her age and nutritional status, the age of the stink bug eggs, presence of natural enemies and competitors, and environmental conditions such as temperature and humidity (Ohno 1987; Weber et al. 1996; Takeshi & Numata 2000; Kivan & Kilic 2004; Cusumano et al. 2012; Abram et al. 2015). Unfortunately, much of the information regarding factors that lead to variation in oviposition remains unknown when sentinel egg masses are deployed. At most, we may understand the phenology of both brown marmorated stink bugs and suitable hosts in the landscape, host age of sentinel eggs, and environmental conditions. The nutritional status of parasitoids may be unknown even if flowering plants are present within agricultural farmscapes. One possibility for enhancing parasitism rates could be to provide nectar-producing flowers in within the agricultural farmscape (Tillman & Carpenter 2014). One suggestion to determine the presence and species of parasitoid adults and their competitors and level of predation is to conduct a sentinel egg pre-test by deploying and watching a small number of egg masses over a short period of time. Another suggestion would be to deploy yellow sticky cards to sample presence or absence of parasitoids. Also, using field cages to exclude predators and release parasitoid females of known age and mating status could result in maximizing parasitism in experiments with sentinel egg masses (Herlihy et al. 2016).

A longer deployment period in the field may provide more time for females to search for and find suitable host plants and brown marmorated stink bug sentinel egg masses, as well as provide more time to parasitize an egg mass. Currently, limited information is available on specific search times of native parasitoid species in the southeastern US. The test conducted in sassafras in 2020 revealed that although some parasitoids began to parasitize egg masses on the day of deployment, parasitism was still higher at the 5 d deployment treatment compared to 2 d. Limited information is available on oviposition behavior of egg parasitoids in the laboratory. After finding an egg mass, a Tr. basalis or Trissolcus utahensis (Ashmead) (Hemiptera: Pentatomidae) female examines several eggs by antennal palpation (Wilson 1961; Weber et al. 1996). Once she selects an egg for oviposition, she deposits her egg into the host egg. Afterwards, the female leaves a host marking pheromone on the egg that deters her and other females from ovipositing into the egg (Wilson 1961; Bin et al. 1993; Colazza et al. 1996). Oviposition continues until all eggs within the mass are parasitized. The duration of the process of oviposition for Trissolcus spp. ranges from 2 to 15 min (Wilson 1961; Ohno 1987; Weber et al. 1996). Host marking takes only 14 to 25 s for Tr. basalis (Wilson 1961). The time spent by a Tr. basalis female to select the next host for oviposition takes up to 2 min. At the end of parasitization of the egg mass, the female can search for unmarked eggs for 15 or more min before abandoning the search. The majority of brown marmorated stink bug egg masses have 28 eggs (Nielsen & Hamilton 2009). Based on timing of oviposition in the laboratory, the duration of oviposition could range from 2.2 to 7.3 h for a brown marmorated stink bug egg mass in the field. In addition, during oviposition, competition by conspecific females can occur, and driving a female from the egg mass may occupy about 10 min (Wilson 1961). In the current study, individual egg masses rarely were parasitized completely, indicating that parasitoids were competing with other natural enemies for egg masses. In plum and sassafras, percent parasitism per egg mass was higher for the longer duration of deployment in the field, suggesting that the increase in deployment time may have provided parasitoids more time to find and parasitize individual egg masses, but this needs further investigation.

Overall, our results strongly suggest that 5 d is the optimal deployment time to assess percent predation and parasitism of brown marmorated stink bug egg masses in the field. Even though significant differences in deployment times were not detected for each round in crops, the trend over all crops showed that percent predation and parasitism was higher at 5 d deployment times compared to 2 d and 3 d exposure times. Also, the hyperparasitoid A. sinicus cannot be detected at deployment times lower than 5 d. Furthermore, naturally occurring egg masses are exposed to primary parasitoids and predators for approximately 5.5 to 6 d (G. Tillman, unpublished data). Perhaps exposure duration for sentinel egg masses should attempt to mimic naturally occurring egg masses, especially given the variability of factors that affect parasitism and predation under field conditions.