First field validation of Ontsira mellipes (Hymenoptera: Braconidae) as a potential biological control agent for Anoplophora glabripennis (Coleoptera: Cerambycidae) in South Carolina

1 Introduction

The Asian longhorned beetle, Anoplophora glabripennis (Motschulsky) (Coleoptera: Cerambycidae), is an invasive woodboring pest native to China and the Korean peninsula (Figure 1). A. glabripennis likely first arrived in North America in wood packing material and was first discovered in the U.S. in 1996 (Haack et al. 1996), and since then, infestations have been detected in multiple locations. While highly polyphagous, A. glabripennis appears to prefer maple trees, Acer spp. L. (Sapindales: Sapindaceae) as hosts in North America (Coyle et al. 2021; Dodds et al. 2013; Turgeon et al. 2022). Management of A. glabripennis has been dominated by eradication efforts via surveys and removal of infested trees, which has resulted in successful eradication of invasive A. glabripennis populations from 10 quarantine zones in North America (Turgeon et al. 2022; USDA APHIS (United States Department of Agriculture 2024)). However, in South Carolina, where A. glabripennis was detected in 2020, the Asian longhorned beetle eradication program faces serious challenges. The preferred host of A. glabripennis, A cer rubrum L. (Coyle et al. 2021), is a highly adaptable species that thrives in a wide range of conditions including swamps and wetlands (Figure 2). The common distribution of Acer rubrum in swamps and wetlands in South Carolina has made tree removal difficult as the heavy machinery required to do large scale tree removal can degrade this sensitive habitat and often access is not possible at many sites. Additionally, detection of A. glabripennis is inherently challenging, as visual inspections are estimated to be only 33–60 % effective (Golec et al. 2016). As such, biological control, the promotion or introduction of natural enemies to reduce the ecological and economic impacts of a pestiferous species (Van Driesche et al. 2010), may help manage A. glabripennis populations in these areas as self-dispersing natural enemies can reduce pest impacts over landscapes rather than being restricted to individual point-locations.

Figure 1:

Anoplophora glabripennis life cycle and damage: (A) adult; (B) adults feed on bark of twigs (noted by arrow) and females chew oviposition pits in the bark (circled); after larvae hatch they feed on phloem, which causes Acer rubrum to ooze dark colored sap (C); older larvae move deeper into the host to feed on wood, which results in wood shavings being pushed out of the tree (D); larval feeding creates large holes in the stem and branches (E, noted by arrows), weakening the structural integrity of the tree and often leading to breakage; mature larvae (F) will pupate inside the wood (G) before emerging as adults. All photos by D. Coyle, Clemson University.

Figure 2:

Wet, swampy conditions common in the South Carolina Anoplophora glabripennis infestation zone. Arrows point to infested Acer rubrum that require removal.

Classical biological control programs involve using natural enemies from the target pest’s native range to reduce pest populations in its invaded region. Extensive efforts for the exploration of natural enemies have occurred in A. glabripennis’ native (e.g., Golec et al. 2018; Hu et al. 2009; Li et al. 2020; Liu et al. 2016; Wang et al. 1999) and introduced ranges (Hérard et al. 2013; Lupi et al. 2017), and these efforts have led to the discovery of several hymenopteran parasitoid species (Johnson et al. 2024). Unfortunately, none of these Asiatic parasitoids appear to be suitable for potential introduction against A. glabripennis in North America, largely because of their broad host ranges (Rim et al. 2018). Further, the use of generalist parasitoids for invasive forest pest management can result in unforeseen impacts, for example, Compsilura concinnata Meigen (Diptera: Tachinidae) was found to have widespread impacts on native species in addition to its target host, Lymantria dispar L. (Lepidoptera: Erebidae) (Elkinton and Boettner 2012). While many classical biological control programs now focus on specialists rather than generalists, new regulations have imposed a great cost and difficulty to the introduction of insect natural enemies for biological control of invasive pests (Ward 2016).

Classical biological control is one type of biological control method (Roderick et al. 2012), and increasingly the use of natural enemies native to the invaded region to reduce the impacts of invasive pest species in that region is being promoted through a practice known as conservation biological control (see Tscharntke et al. 2007). A rich cerambycid fauna existed in North America prior to the arrival of A. glabripennis, and as with other invasive woodboring pests (e.g., Sirex noctilio [F.] [Hymenoptera: Siricidae]), there is also a diverse group of natural enemies already present that use these groups of insects as hosts (Coyle and Gandhi 2012). Native parasitoids can develop a new association with the invasive target pest (e.g., Broadley et al. 2019), which will not only eliminate the ecological risk associated with the introduction of novel species but may also be more effective in suppressing the newly acquired hosts. In fact, in some instances, new associations can prove to be up to 75 % more effective control than introduced parasitoids (Hokkanen and Pimentel 1984).

Several native North American braconid parasitoids including Ontsira mellipes Ashmead, Rhoptrocentrus piceus Marshall, Spathius laflammei Provancher, Heterospilus spp. Haliday, and Atanycolus spp. Förster (Hymenoptera: Braconidae), have successfully attacked young A. glabripennis larvae under controlled conditions and produced both male and female progenies (Duan et al. 2016). Of these parasitoids, O. mellipes has successfully adapted to utilize A. glabripennis larvae through continuous rearing and has produced female biased progeny (6:1 female to male ratio) at each generation with increased parasitism of A. glabripennis (Golec et al. 2019). In the laboratory, O. mellipes parasitized 70% A. glabripennis in a no choice test using host material naturally infested with A. glabripennis (Duan et al. 2016). It was also the most abundant parasitoid captured in field surveys in the northeastern U.S. and has a known distribution throughout much of eastern North America (Golec et al. 2020). Rearing protocols for O. mellipes are established (Wang et al. 2020) and mass production for use in controlling A. glabripennis populations is possible and currently being developed (J. Duan, personal communication). Given this promise, our goal was to evaluate O. mellipes as a biological control agent of A. glabripennis to supplement the current eradication efforts. Thus, we conducted the first field releases of laboratory-produced O. mellipes adults on A. glabripennis-infested trees in South Carolina and subsequently sampled A. glabripennis larvae for parasitism by the released O. mellipes.

2 Materials and methods

2.3 O. mellipes releases

Newly emerging O. mellipes adults (∼1 wk old) from parasitized A. glabripennis larvae were shipped overnight from the USDA Agricultural Research Service (ARS) Beneficial Insects Introduction Research Unit in Newark, Delaware, U.S. The wasps were hosted in vented vials with honey streaked on the ventilated screens for shipping and arrived in good condition. Thus, no additional supplemental water or honey were given to the wasps before release. A total of 1,106 adults were released (Table 1) by securing the vials sideways at breast height on the tree boles using zip ties (Figure 3). Wasps were released between 2:16 PM and 7:57 PM on 10 July 2023 under mostly sunny skies, a starting temperature of 35.6 °C, and 45 % relative humidity with 3 mm of precipitation that day. On average, it took 5 min and 33 s for 80 % of the adult wasps to leave the vial.

Table 1:

Ontsira mellipes release trial for biological control of Anoplophora glabripennis on 10 July 2023 in South Carolina United States Department of Agriculture Asian longhorned beetle Stono Facility and on Royal Harbor Road). Trees 1–9 received O. mellipes releases while trees 10–13 were untreated controls.

Tree	Location	Total O. mellipes released	Live A. glabripennis larvae retrieved	A. glabripennis parasitized by O. mellipes
1	Stono Facility	91	5	0
2	Stono Facility	130	1	0
3	Royal Harbor Rd.	136	38	0
4	Royal Harbor Rd.	140	5	2
5	Royal Harbor Rd.	134	1	0
6	Royal Harbor Rd.	143	3	0
7	Royal Harbor Rd.	115	2	0
8	Royal Harbor Rd.	107	1	0
9	Royal Harbor Rd.	110	0	0
10	Stono Facility	0	7	0
11	Stono Facility	0	16	0
12	Royal Harbor Rd.	0	2	0
13	Royal Harbor Rd.	0	4	0
Total		1,106	95	2

Figure 3:

Vials containing Ontsira mellipes were secured to Acer rubrum boles using flagging tape and zip ties (A) which allowed adult wasps (B) to escape on their own. Photos by M. Lupu, Clemson University.

3 Results

Anoplophora glabripennis larvae were recovered from 12 out of 13 trees. Two out of 56 field-collected A. glabripennis larvae (3 %) were parasitized by O. mellipes (Table 1); both were found on tree #4. One larva on a 76 mm diameter branch had seven O. mellipes larvae attached to it (Figure 4). One O. mellipes larva was found in an A. glabripennis egg site, also on a 76 mm diameter branch, but the A. glabripennis larva was missing and presumed consumed by the O. mellipes larva. No larvae from control trees showed evidence of O. mellipes parasitism.

Figure 4:

Several Ontsira mellipes larvae on one Anoplophora glabripennis larva from tree #4, on a 76 mm diameter branch. Photo by M. Lupu, Clemson University.

After DNA primers were aligned and trimmed, a 464 base pair fragment of COI was amplified from three O. mellipes voucher samples. A multisequence alignment indicated that there were no differences (100 % similarity) between the voucher samples and the unidentified larval parasitoids. All generated sequences from this study have been uploaded to GenBank (Accession Numbers PV765558–PV765562) and to the NCBI BioProject (Accession Number PRJNA1274380).

4 Discussion

This study represents the first record of O. mellipes successfully parasitizing A. glabripennis in the field and provides a proof of concept for O. mellipes as a viable new-association biological control agent that may be used to supplement the current eradication effort against this invasive pest. While this is an exciting and promising development, more work is needed to determine optimal release time, number and interval of releases, and density of wasps sufficient to result in a measurable reduction of A. glabripennis populations. Further, as O. mellipes is known to occur naturally in South Carolina and likely (based on known records) much of the eastern U.S. (Ashmead 1888; Golec et al. 2020; Kula and Marsh 2011; Marsh 1966), the possibility that this particular parasitization occurred naturally cannot be discounted.

All O. mellipes larvae found in this study were collected on smaller diameter (i.e., 74–78 mm) branches. While no conclusion can be made about preferences with such a relatively small sample size, this may be due to ease of ovipositor insertion through thinner bark. Like many trees, Acer rubrum exhibits thinner bark on younger, smaller diameter branches compared to the thicker plates developing as the stems age. Thicker bark in the main stem may be an obstacle for the wasps in late-stage infestations, as O. mellipes has a short ovipositor (3.91 ± 0.05 mm in length) (Wang et al. 2019), and Acer rubrum bark can be up to 20 mm thick (Schafer et al. 2015). However, A. glabripennis prefers to oviposit on smaller diameter material (Bean 2022; Turgeon et al. 2024), which typically has thinner bark (Bova and Dickinson 2005). Our recent laboratory observation indicates that O. mellipes may also attack large A. glabripennis larvae during the time when the large larvae need to come out of xylem to push the frass out of their galleries (JJD and EMA, unpublished data).

Weather on the days of release appeared to have a drastic effect on the behavior of the wasps. Wasps appeared lethargic and inactive during the duration of the release, compared to the time of their arrival 2 days prior (ML, personal observation). The temperature was 36 °C on the day of wasp release, which is outside of the range of optimal temperatures for this wasp (Golec et al. 2017). Wasps may have experienced shock coming from a cooler laboratory setting to a hot and humid field setting, and it is undetermined how long an adjustment period may be needed. When vials were opened, the wasps attempted mating, likely due to the new stimuli. The likelihood of O. mellipes failing to locate A. glabripennis larvae in the field is undoubtedly higher than in the laboratory due to prevailing winds, rain and other adverse weather, and distance. This presents another challenge in determining the ideal time and date of release. It is possible that success might be greater if weather at the time for release consisted of no rain, no storms in the next 24 h, no severe winds, and releases took place in the morning to avoid releasing into temperatures above 32 °C. This may allow the wasps to better acclimate to the new conditions in the field.

The environmental temperature has a positive relationship with the rate of development of O. mellipes from egg to adult; however, after reaching 30 °C, survival is nearly halved, and fecundity is significantly reduced. The ideal temperature for fastest development and survival is 25 °C (Golec et al. 2017). This might provide some insight into the low parasitism in this trial, which occurred at 35.5 °C. Since 1,107 individuals were released in this trial, based on the findings in the laboratory study we can speculate that at least half of eggs laid did survive to adulthood, likely due to the temperature limitations (Golec et al. 2017). Future studies should attempt to repeat releases during weeks which do not exceed 30 °C. This suggestion aligns with recommendations in the northern states to release the wasps during early spring and late fall.

O. mellipes is being reared for potential use to manage A. glabripennis (Duan et al. 2016) but has the potential to target other cerambycids including Elaphidion mucronatum Say, Monochamus caroliniensis Olivier, and M. notatus Drury (Wang et al. 2019). Moreover, there may be additional cerambycids in the southeastern U.S. that have not been evaluated as hosts, as original studies were done in the mid-Atlantic and northeastern U.S. (Wang et al. 2019). If an A. glabripennis biological control program using O. mellipes will be investigated further, potential cerambycids in the South Carolina quarantine area should be included in the tests. Moreover, understanding how cerambycid populations will impact effectiveness of management using O. mellipes is crucial to this project. A future cerambycid survey may be useful in determining these populations in areas where A. glabripennis has infested.

In previous studies, O. mellipes showed no significant difference in preference for M. caroliniensis (a cerambycid native to North America) over A. glabripennis. Thus, it may be beneficial to survey for M. caroliniensis in South Carolina to sustain a population of O. mellipes where A. glabripennis populations are dwindling but control is still necessary (Wang et al. 2019). Of all species tested on A. glabripennis, R. piceus and O. mellipes had up to 100 % parasitism in no choice laboratory assays, however R. piceus has been difficult to rear in the laboratory because of its diapausing behavior (Golec et al. 2016; JJD and EMA, unpublished data). The potential of other new association biological control agents (e.g., R. piceus) to control A. glabripennis should be further investigated in the context of current eradication-based management programs.