https://orcid.org/0000-0002-7067-0718
Abstract
Tracking tagged individuals is an emerging tool to locate invasive wasp nesting sites. Both tag size and transmitter/transceiver cost can limit the applicability of these technologies for eradication measures. This pilot study used a combination of lightweight, inexpensive harmonic radar tags fabricated with superelastic nitinol wire, and off-the-shelf transceivers designed for backcountry rescue. Larger tags weighing ~10 mg (~14 mg with adhesive) were used to track three vespid wasps, Delta esuriens (F.), Polistes aurifer Saussure, and Vespula pensylvanica (Saussure), while smaller tags ~5 mg (~8 mg with adhesive) were used only with V. pensylvanica. Wasps were successfully tagged in both laboratory and field settings. Tagged P. aurifer were shown to be flight capable in a large outdoor field cage. Subsequently, wasps were tracked in a macadamia nut field (D. esuriens and P. aurifer), a coastal habitat (D. esuriens), and in Hawaiʻi Volcanoes National Park (V. pensylvanica). Flight paths up to 6 steps (maximum 126.7 m) were recorded for D. esuriens (maximum single flight 76.6 m). Vespula pensylvanica were tracked to one nest; several other attempts failed to identify a nest location but did demonstrate the practicality and limitations of following wasps through dense vegetation. Additionally, we demonstrated that V. pensylvanica would carry off tags embedded in small pieces of meat although this technique did not lead to the discovery of any nests. This study demonstrates the feasibility of using an inexpensive method to track wasps, potentially allowing for a rapid and simplified method of locating invasive wasp nests.
Introduction
With the recent introductions of 3 large invasive wasps into the United States from Asia, the greater banded hornet, Vespa tropica (detected in Guam, 2018), the giant Asian hornet, Vespa mandarinia (detected in Washington State, 2019), the yellow-legged hornet, Vespa velutina (detected in Georgia, 2023), new methods to aid eradication of these pests are needed. One promising method used to eliminate invasive wasp nests is attaching a tracking device to a captured wasp and following the tagged individual back to the colony after it is released (Kennedy et al. 2018, Looney et al. 2023, Wang et al. 2024). This method is an example of a broader approach to finding invasive species that aggregate or are social, known as the Judas technique. The technique is based on the idea that pest animal conspecifics often evolve behaviors and/or superior sensory systems that aid in finding others of their species in a complex natural environment. The use of “Judas goats” in the eradication of feral goat populations is likely the earliest demonstration of this technique (Taylor and Katahira 1988, Campbell and Donlan 2005). While the technique has been most commonly employed to control mammals (Spencer et al. 2015), it has also been employed against fish (Bajer et al. 2011), birds (Woolnough et al. 2006), and recently insects, such as large beetles (Moore et al. 2017) and wasps (Kennedy et al. 2018, Looney et al. 2023). Until recently, application of the Judas technique to insects has been largely impractical due to the relatively high mass of available tracking devices. However, the development of small, lightweight tracking tags allows application of this technique with large bodied invasive insect species that aggregate in the wild.
A number of device technologies have been used to track insect movement including, most commonly, radio telemetry (RT) and harmonic radar (HR) (Batsleer et al. 2020). Both technologies have different advantages and tradeoffs. RT employs battery-powered tags that can uniquely identify individual insects. Tracking is often accomplished using a hand-held Yagi antenna and receiver with an effective range of up to several hundred meters. However, the added weight of the battery in the RT tags limits use to larger insects. Additionally, these tags are several orders of magnitude more expensive than HR tags. RT has been used to track a taxonomically diverse series of generally large insects (Kissling et al. 2014, Batsleer et al. 2020), including coleopterans (Moore et al. 2017), hemipterans (Hamer et al. 2018), and hymenopterans (Kennedy et al. 2018, Looney et al. 2023).
Likewise, HR, which was first used to study insect movement by Mascanzoni and Wallin (1986), has been used to track a number of hymenopterans (Batsleer et al. 2020). HR tracking employs 2 components: (i) a radar transceiver unit, which both emits a directional signal (TX component) and “listens” for a reflected signal at twice the broadcast frequency (RX component), and (ii) a diode tag that receives the original signal and reemits a frequency-doubled signal (Mascanzoni and Wallin 1986). HR units can be stationary ground-based (Maggiora et al. 2019) or mobile, which includes handheld units (Roland et al. 1996, Boiteau and Colpitts 2004). In addition to being small and lightweight, HR tags are less expensive and lack a battery, resulting in a relatively long shelf- and field-life. However, HR has several drawbacks compared to RT including a shorter detection range (particularly with hand-held transceiver units, max ~75 m) than with RT (max ~500 m), the inability to produce a uniquely identifiable signal, and substantial decreases in signal strength due to tag orientation relative to the transceiver and the influence of terrain and vegetation.
Both RT (Kennedy et al. 2018, Looney et al. 2023) and HR (Milanesio et al. 2016, 2017, Maggiora et al. 2019, Lioy et al. 2021) have been employed to track invasive wasps in the context of applying the Judas technique. Kennedy et al. (2018) and Looney et al. (2023) used RT hand-held receivers to track tagged wasps back to previously undiscovered nests. In contrast, the HR tracking efforts used an off-the-shelf marine radar coupled with a purpose-built receiving module (Milanesio et al. 2016) mounted on a telescopic portable tower. Assessing these 2 approaches, RT employed more expensive (often several hundred dollars per tag), larger, and heavier tags, but used cheaper, mobile, off-the-shelf tracking equipment (Yagi antenna and receiver). By contrast, the HR tracking employed cheaper (several dollars per tag), smaller, and lighter tags while the transceiver used was moveable but relatively large, purpose-built, and more expensive. An alternative to the purpose-built TX/RX HR system is to use handheld HR transceivers manufactured for avalanche rescue by the RECCO corporation for insect tracking (Mascanzoni and Wallin 1986, Lövei et al. 1997, O’Neal et al. 2004). When RECCO transceivers are combined with HR tags fabricated using superelastic nitinol antennas, a third approach is possible that employs both off-the-shelf tracking equipment (RECCO transceivers) and ultra-lightweight (down to less than 1 mg), inexpensive, flexible tags (Miller et al. 2022, Siderhurst et al. 2023, Hurst et al. 2024). Thus, the approach we are using (i) employs smaller cheaper tags than have been used to locate wasp nests with RT methods and (ii) the tracking equipment (hand-held transceivers) are more accessible, both in terms of expense and ease of procurement, than HR methods previously used to locate wasp nests.
While none of the large invasive hornets mentioned above presently occur in Hawai’i, multiple smaller vespid wasps are present thereby providing model species to test tracking techniques. If tracking is possible with smaller vespids, larger wasps are also likely be able to accommodate nitinol wire HR tags. Although larger wasps may also make longer flights so the generalizability of the technique may not be straightforward. Regardless, the 3 nonnative vespid wasps tracking in this study are Delta esuriens (Fabricius), Polistes aurifer Saussure, and Vespula pensylvanica (Saussure) (Hymenoptera: Vespidae) (Carpenter 2008). Of these 3 vespid wasps, only V. pensylvanica is considered a pest or nuisance species. First noted on the island of Hawaiʻi 1976 (Nakahara 1980), V. pensylvanica both predates arthropods and forages for nectar. Invasive V. pensylvanica in Hawaiʻi have negative impacts on native invertebrates, birds, and plants (Gambino and Loope 1992, Banko et al. 2002, Wilson et al. 2009, Wilson and Holway 2010). Additionally, the aggressive behavior and stings of these wasps pose a risk to human health, especially for sensitive individuals (Golden et al. 2006). The impacts of yellowjackets are magnified by the lack of cold weather which allows the formation of perennial supercolonies (Jeanne 1980, Plunkett et al. 1989, Harris 1996, Visscher and Vetter 2003). The successful tracking of V. pensylvanica in the current study demonstrates the feasibility of using HR tracking to detect nesting sites, which may be an important tool in both controlling and further studying this invasive pest.
In this study, we aim to attach 2 sizes of HR tags to the 3 vespid wasp species mentioned above, test whether tagged wasps are flight-capable, and track tagged wasps in the field (ideally until they return to their nest). We also aim to trial a technique wherein wasps are induced to carry off tags embedded in meat, thereby eliminating the need to catch and tag wasps. Our objective is to provide a proof-of-concept for using off-the-shelf RECCO transceivers with inexpensive HR tags to enhance the availability of the Judas technique for organizations trying to quickly eradicate populations of invasive wasps.
Methods
Insects
Delta esuriens and Polistes aurifer wasps were captured at Hāpuna Beach State Recreation Area (19.9912217, -155.8246695) and Spencer Beach Park (20.0233422, -155.8221824) on 6 August 2024 (D. esuriens and P. aurifer), and again on 12 August 2024 (D. esuriens only). In both locations, wasps were collected around either showers or water taps with standing water. Wasps were captured directly in either small glass jars with screw top lids (with air holes) or slightly larger plastic containers with screen top lids. Captured wasps were given dental wicks wetted with both water and sugar solution. Captured wasps were transported to Hilo, HI, and held at 23 to 24 °C, 62 to 63 RH, except in the case of wasps that were immediately cooled and tagged in the field. Wasps in holding containers were mostly observed resting on one of the provided dental wicks or on the lids. Wasps did not appear to be agitated unless disturbed while being transported from Hāpuna or Spencer Beach Park.
Vespula pensylvanica were collected at 2 general locations (5 specific sites) in Hawaiʻi Volcanoes National Park (support/office build complex, 19.4194228, -155.2493979; hiking trail, 19.4216791, -155.2531211). Wasps were captured as described above and then held on ice. Since wasps were only ever held for a short period of time (never transported by vehicle), no water or sugar were provided.
Tag Fabrication and Attachment
Dipole HR tags were fabricated by attaching two 8.25 cm lengths of superelastic nitinol wire (Malin Co. Brookpark, OH) using a 2-part conductive epoxy (Electron Microscopy Sciences, Hatfield, PA) onto each of the leads of the Schottky diode (2 sizes were used, RECCO AB, Lindigo, Sweden, large diode shown in Fig. 1A). The wire lengths with a diameter of 0.076 mm were bonded to the diode leads by application of the epoxy directly to the wire-diode assembly. Diode wire assemblies were mounted to a 3-D printed jig with double-sided tape to facilitate rapid and uniform deposition of epoxy to tags (Supplementary Fig. S1). Once an initial layer of epoxy was applied to the wire and contact (with several minutes allowed for drying), a second coating was applied to increase the thickness of the conductive epoxy. This increased the conductivity of the attachment by maximizing surface contact between wire and lead, which contributed to the strength of the harmonic return signal. The working time of the conductive epoxy was found to be close to 0.5 h, providing ample opportunity to produce multiple tags and adjust the wire-diode assembly as needed. The only disadvantage of the long curing time (24 h at indoor temperature and humidity) was that completed tag assemblies were vulnerable to disruption. To accelerate curing, tags attached to jigs were placed onto a heated reflow station at 50 °C. After a full cure is achieved, the tags require no additional processing and are ready to apply to the insect of interest.
Large harmonic radar tag used (Schottky diode with nitinol wires affixed with conductive epoxy) (A). Tag range testing setup (B). Examples of tagged Delta esuriens (C) and Polistes aurifer (D) (the UV LED used to cure the attachment adhesive can be seen in (C)). Example landing location for the Delta esuriens tracking in the macadamia field ~5 m up in a tree (E).
Tags were attached to wasps either in the lab before transport to the release location or were attached in the field which shortened time in captivity. In the lab, the entire tagging procedure was carried out in a walk-in refrigeration unit (~3 to 4 °C). Wasps were prepared for tag attachment by chilling for 10 min or until cessation of movement. An individual wasp was then positioned so that the dorsal thorax was accessible. The tag was attached in a longitudinal orientation (Figs. 1C-D and 2A) to the scutum using a UV adhesive (Bondic, Niagara Falls, NY, USA was used for D. esuriens and P. aurifer; Dental Light Cure Dentin Enamel Bonding Resin, Prime-Dent, Chicago, IL, USA was used for V. pensylvanica) and cured with a UV light. Care was taken to avoid gluing the wings or head during tag attachment. Wasps were also carefully inspected for any damage after the tagging process and wasps showing any sign of damage were not released into the field for tracking. After tagging, wasps were placed in plastic holding containers (14 × 12 cm, H:W) covered with screen lids. Before release, containers holding both tagged and untagged captured wasps were placed in a large cooler (58 × 46 × 106 cm, H:W:L) to reduce wasp agitation and stress. Water and sugar solution were provided on dental wicks to wasps held for periods over 30 min.
Example Vespula pensylvanica with (smaller) harmonic radar tag attached (A). Trail in Hawaiʻi Volcanoes National Park where V. pensylvanica were tracked to a nest site (B). Harmonic radar tag embedded in small piece of Vienna sausage and hypodermic needle used for insertion (C). Wasp lifting large piece of Vienna sausage (D).
In the field, wasps were immobilized by placing the capture container, with the wasp inside, into a large cooler filled with ice. After 10 min of chilling, the immobilized wasp was transferred to a 10 cm dia. petri dish which was then placed in a second larger glass petri dish (15 cm diameter) that was filled with ice (Supplementary Fig. S2). Tags were attached inside a front-opening insect cage (BugDorm-2F120 Insect Rearing Cage, Taichung, Taiwan) draped with a dark cloth to exclude UV light (Supplementary Fig. S2).
HR Tag Detection Range Testing
HR tags were secured longitudinally to a thick wooden dowel (~1.5 m) with double-sided tape (Fig. 1B). The maximum distance that tag signals could be detected was then determined using a RECCO transceiver (model R9, transmitting at 902.85 MHz, receiving at 1805.7 MHz). Testing was conducted in a large, quiet open field (19.70197, -155.12519).
Field Cage
Initial flight tests of tagged P. aurifer were conducted in a large outdoor field cage (2.4 × 6.2 × 15.0 m, H:W:L) at the Pacific Basin Agricultural Research Center in Hilo, HI. Tagged P. aurifer were used because these wasps were the most readily available at the time of testing. Wasp releases were conducted during a warm (28 °C) midafternoon on 7 August 2024. Wasps were observed to be actively moving inside holding containers just before release. After being released from holding containers, tagged wasps were located with the aid of the transceiver unit after they had taken flight. Tagged wasps were recaptured upon location.
Field Tracking
To test tracking feasibility, 2 vespid wasps, 1 D. esuriens and 1 P. aurifer, were released and tracked in a macadamia nut farm just outside Keaʻau, HI (release point: 19.61039821, -155.0771426). Tagged wasps were visually inspected in the field before release to ensure that tags were still attached, and the transceiver was used to test that each tag was functional. The release location was an open area between two rows of trees, well within the stand of trees, and 70 m from the access road. The tagged wasps were released during the midafternoon (a time when the wasps were previous observed to be active in the field) of 8 August 2024 with temperatures approximately 21 to 22 °C. Tracking a released wasp was accomplished with 4 spotters, each having a RECCO transceiver. When possible, an attempt was made to follow the wasp visually once it took flight. Once a tentative direction of flight was determined, vegetation along the trajectory of the flight path was swept with a RECCO transceiver. Care was taken to space out the searchers to avoid false-positive signals from nearby transceivers. Wasps were generally located within 5 min of taking flight. At most locations, the tagged wasp was visually located within a tree and photographed using a camera equipped with a telephoto lens (EOS Rebel T6i camera, Cannon 55-250 mm macro lens) (Fig. 1E). This helped to document whether each wasp still had an attached tag or whether the tag was lodged in foliage after becoming detached (Supplementary Fig. S3C–D). After being given 5 min to rest after landing, wasps were disturbed to induce flight by shaking the surrounding vegetation with a long extendable pole. No attempt was made to recover released wasps or dropped tags. Wasps were tracked until tag detachment was observed or until no signal was detected after 15 min of searching. Wasps were not marked to aid identification due to the small number of wasps released. Wasp releases were spaced out in time so as to allow the tracking team to concentrate search efforts on one individual at a time and to lower the chance of confusing 2 tagged wasps. GPS coordinates were recorded for each landing location.
To further test tracking feasibility, 5 wasps, all D. esuriens, were released and tracked at Spencer Beach Park (release point: 20.02302931, -155.82222956). The release location was an open shower area on the beach where wasps had originally been collected. The tagged wasps were released during the midafternoon of 13 August 2024 with temperatures approximately 28 to 29 °C. During the time of releases there were strong onshore winds blowing between 10 and 20 km/h. Wasp tracking and location were accomplished in a similar manner as described above. In this instance, most wasps flew inland so several spotters were positioned inland of the release point. Four of the 5 released wasps were tagged in the lab and transported to the release site with one additional wasp captured, tagged, and released in the field.
A nest location trial (NL1) was attempted using the release of approximately a dozen tagged V. pensylvanica along a hiking trail in Hawaiʻi Volcanoes National Park (19.4216791, -155.2531211). The area along the trail was largely ʻōhiʻa lehua, Metrosideros polymorpha Gaudich., forest with fairly dense understory vegetation (Fig. 2B). Two release points were used each ~40 m along the trail from the known location of the nest, which was ~2 m from the edge of the trail. Wasps were collected and released (after tagging) from the same location to minimize disorientation. Tagging was performed in a light-excluding enclosure (Supplementary Fig. S2) in a truck bed located at the trailhead several hundred meters from the nest. Both collection/release locations were initially used but the northern site had a much higher density of foraging V. pensylvanica. This led to only one wasp being released from the south location while all others were released from the north location. Capture, tagging, and tracking took place during the morning of 15 October 2024 with temperatures approximately 18 to 23 °C. Wasp tracking and location were accomplished in a similar manner as described above. In this instance, most wasps flew along the trail toward the nest, so a spotter was positioned along the trail to observe. If tagged wasps flew into the grasses along the trails, they were recaptured and placed in open tree foliage (Fig. 2A) to facilitate further flights. Further tracking was conducted on 29 October 2024 using 2 collection/release sites, one forest location (NL2: 19.4184522, -155.2493474) and one open grassy area with picnic tables (NL3: 19.4194414, -155.2493988). Approximately 20 tagged wasps were released between the 2 locations on this date.
Trials to investigate whether V. pensylvanica could be induced to carry off HR tags embedded in small pieces of meat (Vienna sausage, Fig. 2C) were conducted at various locations within and around a support/office campus located in Hawaiʻi Volcanoes National Park (19.4194228, -155.2493979). Large tags embedded in Libby’s Vienna sausage (made with chicken, beef and pork in chicken broth, Chicago, IL, USA) weighed ~30 to 45 mg while embedded small tags weighed ~15 to 25 mg. Trials which included the largest numbers of wasps and baits were conducted on 19 September (18 to 19 °C) and 3 October 2024 (19 to 23 °C) during which ~10 and ~30 wasps were observed taking off carrying meat-embedded tags, respectively. Initial trials also included tags embedded in small pieces of sardine meat or gummy bear candy. During initial trials, small food items were also presented without embedded tags to assess wasp preferences (Fig. 2D). Tags were embedded in food items using a hypodermic needle (Fig. 2C). The needle was first passed through the small food item, one tag antenna was then threaded through the needle, and finally, the tag was pulled back through the hole made by the needle so that the diode was concealed within the food item and the antenna wires protruded on each side (Fig. 2C). To lessen the weight contributed by the Vienna sausage, excess meat was shaved off using a scalpel after the tag insertion. Wasp tracking and location were accomplished in a similar manner as described above. In this instance, wasps were observed flying relatively short distances, often high into nearby trees. A single observer was used at each food presentation site (no additional spotter was used). Two food presentation sites were used on 19 September and 3 sites were used on 3 October. Food items were initially presented on plastic lids or directly on the ground. Later trials used a plastic container (5 × 16 × 24 cm, H:W:L, Rubbermaid, Atlanta, GA, USA) with a screen lid to display the tagged Vienna sausage pieces. Several grams of meat were placed inside the screened container allowing the release of attractive odors (meat inaccessible to wasps). The tagged Vienna sausage pieces were then placed on top of the screen and were available to visiting V. pensylvanica. The final iteration of this method also included laying the tagged Vienna sausage pieces on a moist towelette (still above the screen) to prevent rapid desiccation on the meat.
Wasp and Tag Masses
Masses were recorded using a DeltaRange XP105 balance (Mettler Toledo, Columbus, OH, USA). Tags were massed after assembly. Wasps were brought back to the laboratory after tracking activities, frozen in a −80 °C freezer, and massed with and without tags attached. Tags with attached adhesive were removed from wasps with a pair of forceps (UV adhesives remained attached to the tag in all cases).
Results
The mean mass for the larger tags was 10.74 ± 0.05 mg and 5.80 ± 0.03 mg for the smaller tags. The mean mass of the tag plus adhesive was 14.2 ± 0.4 mg for the large tags and 6.34 ± 0.05 mg for small tags. The mean mass of captured wasps was 105 ± 8 mg for D. esuriens, 100 ± 11 mg for P. aurifer, and 52 ± 2 mg for V. pensylvanica. Mean mass with tag (including adhesive) to body mass ratios were 13.5% for D. esuriens (large tag), 14.2% for P. aurifer (large tag), and 12.1% for V. pensylvanica (small tag). The maximum detection range for the large tags was 64 ± 1 m and 54 ± 1 m for the small tags.
Secure tag attachments were achieved both in the lab and in the field. In general, tag attachments were found to be much more durable with D. esuriens than with P. aurifer (when using the Bondic UV adhesive) and this led to D. esuriens becoming the focus wasp for the initial field releases. In one instance, a P. aurifer was observed losing the attached tag when one of the antenna wires became tangled in foliage (Supplementary Fig. S3C–D). Similar tag attachment issues were noted with V. pensylvanica, which prompted the switch to the Prime-Dent UV adhesive after which no tag detachments were observed. The Prime-Dent UV adhesive was also less sensitive to low level UV light (ambient outdoor light), which made tagging in the field easier.
Vespid wasps in this study were observed to be flight capable with attached tags in field cage testing (P. aurifer), field releases (all species) and V. pensylvanica were also observed to be able to carry off (fly with) tags embedded in small pieces of meat. Likewise, tracking was successful in a macadamia nut field (D. esuriens and P. aurifer) (Fig. 3A and Supplementary Fig. S3), at Spencer Beach Park (D. esuriens) (Fig. 3B and Supplementary Fig. S4), and at Hawaiʻi Volcanoes National Park (V. pensylvanica). Wasps with attached tags generally few within 1 to 2 min of being released. The mean step distance for tagged D. esuriens tracked at Spencer Beach Park was 21 ± 5 m, max 76.6 m, with a mean path length of 68 ± 17 m, max 126.7 m.
Harmonic radar tracking of tagged wasps in a macadamia nut field (A) and at Spencer Beach Park (B). Species is indicated by line color. Tagging location is indicated by a solid or dashed line. Map data: Google, Airbus.
For the successful nest location trial (NL1) with V. pensylvanica in Hawaiʻi Volcanoes National Park (along a hiking path), all released wasps flew toward the nest (except one that went slightly backwards into tall trees). In general, tagged V. pensylvanica flew up into trees when released or “fell” down into the grasses along the path. This resulted in many wasps only being tracked for one step. Three tagged V. pensylvanica were observed either at the nest entrance or going into the nest. There was no detectable signal once the wasp entered the nest (in the ground). One wasp was located (by tag signal, no visual sight) on the back side of the nest (away from the trail) but the nest was active enough to make closer visual inspection problematic. Wasps tracked to the nest took 3 to 5 observable steps. That is, several landing locations were noted before the tagged wasp got to the nest. Subsequent unsuccessful nest location trials resulted in tracking tagged wasps for up to ~20 m (NL2) and ~30 m (NL3) but did not identify the nest locations. For releases in the forested area (NL2), all wasps flew in a southerly direction with headings falling within a 25° arch. At ~20 m out from the release site, dense vegetation prevented expedient tracking of tagged wasps beyond this point. Tagged wasps released near the picnic area (NL3) flew either NNW into a forested area or SSW with the prevailing wind. Wasp landing locations identified with the NL2 and NL3 releases were again generally in trees overhead with a handful of cases in which tagged wasps landed in the grass.
Trials with tags embedded in small pieces of food showed that V. pensylvanica are able and willing to carry off these parcels. In total, ~40 wasps were observed flying off with meat-embedded tags. Wasps were often observed to take short initial flights to nearby vegetation before either initiating longer flights (presumably toward the nest) or dropping the meat-embedded tag. At some locations, wasps were observed to chew down the meat and in one instance a forager was observed methodically removing the embedded tags before carrying off a smaller piece of meat. Roughly a quarter of the tags were observed to have been dropped (or the tags were removed from the meat), and it is unknown how many wasps, if any, carried tags back to the nest. The maximum distance a meat-embedded tag was located after being carried off was roughly 20 m. However, some wasps did initiate long flights on first takeoff with the meat-embedded tag and were not subsequently located. Presumably, these flights were over 50 m as the tags were not located by searching. Some wasps flew high into trees (beyond visual location) which also complicated subsequent tracking. It was also noted that wasp activity toward the tag/meat combination was quite variable. At one location, during an ~20 min period, over a dozen tags were carried off, while on another day at the same location, no tags were taken. This is despite the fact that similar numbers of wasps were observed on both days and the weather was very similar. Why wasps were more interested in meat items at some times and places than others was not discernable at the time of tracking.
Discussion
This study provides proof-of-concept for using off-the-shelf RECCO transceivers with inexpensive HR tags for tracking invasive vespid wasps. Successful pre-tracking aspects of the study included attaching tags in both the lab and the field and determining maximum detection distances. Problems with tag detachments were encountered with all 3 wasp species when using the Bondic UV adhesive. This adhesive has performed well when used to attach HR tags in previous studies with tephritid fruit flies (Miller et al. 2022, Hurst et al. 2024) and is widely available. No tag detachments were observed when using the dental UV adhesive (used only with V. pensylvanica), demonstrating more secure tag attachments, but this material is less readily available in the United States due to restrictions on purchasing dental supplies. Maximum detection distances for the larger tags were roughly 10 m further than for the smaller tags. This distance was determined under ideal conditions and the working distance in the field was found to be closer to 10 to 15 m for both tag sizes due to interference from vegetation and misalignment of the transceiver and tag antenna. As both tag sizes used the same antenna lengths, the shorter detection range of the smaller tags was likely due to the fact that the smaller diode had less surface area, which may reduce contact with the antenna wire and lead to reduced conductivity at the diode terminals.
Wasps of all 3 species were observed to be able to fly with HR tags attached. Once in flight, wasps appeared to fly naturally with no detectable modifications in flight behavior. However, tagged wasps were observed to struggle with the nitinol wire antennas when in some types of vegetation. Flat smooth surfaces did not present obstacles to movement or take-off, but grass and smaller leaves presented greater surface complexity and opportunities for the antennas to become entangled or struck. In the context of testing the Judas technique (with an objective of following wasps to a nest, not observing natural behavior), we found it advisable to reposition wasps found in the grass to a smooth surface higher in the vegetation. This appeared to result in longer more direct flights. In addition to antenna entanglement, the flight ability of tagged wasps is also likely affected by the tag mass. All 3 vespid wasps used in this study had tag-to-body mass ratios above the often quoted “5% rule,” which was largely derived for vertebrates (Batsleer et al. 2020). However, vespid wasps tend to be strong fliers, and all 3 species are adapted to carrying food items back to nesting sites. The most readily observed examples of impeded flight were when V. pensylvanica carried off chunks of meat, which were sometimes roughly equal to the wasp’s own body mass. Therefore, while attached tags likely negatively impact wasp flight capacity, the wasps are still capable of flights exceeding 100 m (D. esuriens) and presumably further.
All 3 wasp species were successfully tracked in a variety of habitats. The tracking distances achieved are likely to be sufficient for nest location efforts, although tagged wasps were only observed returning to a nest in one instance. Most landing locations were in trees which can make a positive (visual) identification difficult. A number of recent studies have begun to demonstrate the feasibility of tracking insects, either using RT or HR, with the aid of drones (Shearwood et al. 2020, Lavrenko et al. 2021, Ju and Son 2022, Kim et al. 2022, Pak et al. 2024). These approaches have the potential to improve tracking results by both allowing searching areas further above the ground (such as high in trees) or across larger areas (presumably drones will be able to move faster and won’t be encumbered by difficult terrain). Of the 3 tracked species, P. aurifer has the most exposed nest (single layer, umbrella-shaped), D. esuriens nests are likely to be above ground but with thicker walls that might attenuate HR signals, and V. pensylvanica are the most problematic being underground. Pinpointing the nest location for V. pensylvanica was likely made harder by the fact that once wasps moved into the underground nest, tag signals were lost. In general, the tags and transceivers used in this study seem to be appropriate for the application of the Judas technique for finding invasive wasp nests although a longer detection range would make tracking easier.
Previous wasp tracking studies using HR (Milanesio et al. 2016, 2017, Maggiora et al. 2019, Lioy et al. 2021) used larger, more expensive transceivers and heavier tags but these systems also had greater detection ranges. Furthermore, the transceivers used in these previous studies were purpose-built and mounted on movable towers which operate from a fixed point. By contrast, the transceivers used in the current study were less expensive, handheld, and readily available from a commercial source (RECCO). To our knowledge, this study constitutes the first instance of tracking wasps with handheld HR transceivers. Additionally, the tags used were lighter, cheap, and easy to fabricate. Together, off-the-shelf transceivers and small, inexpensive tags mean this tracking technique should be readily accessible to action agencies trying to respond to invasive wasp incursions.
Beyond HR, wasps have also been tracked to nest locations using RT (Kennedy et al. 2018, Wang et al. 2024). The main advantages of RT over HR are the greater detection range (~800 m; Kennedy et al. (2018)) and unique identification. Downsides are higher tag mass (200 to 300 mg; Kennedy et al. (2018)), larger tag size (which may inhibit movement or cause tactile irritation), increased tag costs, and shelf-life issues (battery failure). For larger wasps that can accommodate the heavier tags, RT may be the best option as added detection range and signal strength are likely key factors for successfully locating nests. However, for smaller wasp species, HR is a viable option, especially if wasps forage close to nests, limiting the need for longer-range tracking. To our knowledge, V. pensylvanica is the smallest wasp that has been followed in the field using a transmission tracking system (in this case HR).
Finally, the observations of V. pensylvanica carrying off tags embedded in pieces of meat demonstrates a novel method of tracking wasps. Ideally, having wasps simply pick up tags makes the tracking process easier as there is no need to capture, immobilize, tag, and then release individuals that are subsequently followed. In practice with V. pensylvanica, the periodic lack of interest in the presented meat and high drop rates made this tracking technique less consistent, less efficient (in terms of wasps tracked per time) and more time-consuming. The observed behavior of initially flying to a nearby location where the meat was chewed is consistent with previous description of yellow jacket forager behavior (Akre et al. 1976, Ross 1983). Akre et al. (1976) suggests that yellow jackets investigate potential prey objects smaller than 5 mm and Gambino (1992) consider prey items over 4.5 mm to be “large.” Thus, the meat-embedded tags are roughly the right size to be recognized by V. pensylvanica foragers as potential prey. Whether the meat-embedded tags are substantially heavier than normal V. pensylvanica prey items is not readily obvious. Gambino (1992) has reported on prey taken by V. pensylvanica in Hawaii Volcanoes National Park, and the predominate items are spiders, orthopterans, hemipterans, dipterans, and lepidopterans (similar results were observed by Akre et al. (1976)). It seems likely that the optimal meat-embedded tag size is somewhat smaller than what was used in this study and continued effort to reduce the size of the package seem warranted.
Under given favorable conditions, eg many active wasps consistently carrying off tags, this technique using meat-embedded tags could potentially realize improved productivity over attaching tags to wasps. Several improvements in tags/meat presentation were made during the course of the trials. These include having extra, inaccessible meat, which helped to attract more wasps, and keeping the meat-embedded tags moist. For the latter, desiccation seemed to be a problem when working in the sun on a hotter day, with dried-out pieces of meat becoming less attractive over time. Given the moderate success of using meat-embedded tags, this method deserves further investigation. It is possible that this method could work more efficiently with the addition of an attractant, such as heptyl butyrate, which has been shown to increase visitation and removal of chicken by V. pensylvanica (Hanna et al. 2012).
Funding
This research was supported by the U.S. Department of Agriculture, Agricultural Research Service appropriated project “Development of New and Improved Surveillance, Detection, Control, and Management Technologies for Fruit Flies and Invasive Pests of Tropical and Subtropical Crops” (2040-22430-027-000-D) and by the Florida Department of Agriculture & Consumer Services.
Acknowledgments
We would like to thank Hawaiʻi Volcanoes National Park for allowing us to conduct experiments with Vespula pensylvanica (Permit Number: HAVO-2024-SCI-0041) and Jonathan Makaike for his help in locating nest sites. We also extend our appreciation to Island Princess Mac Nut Inc. for allowing us to conduct experiments on their property. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture (USDA). The authors declare no competing interest. The USDA is an equal opportunity provider and employer.
Author contributions
Matthew Siderhurst (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal], Project administration [equal], Resources [equal], Supervision [equal], Visualization [equal], Writing—original draft [equal]), Katherine Fairbanks (Conceptualization [equal], Investigation [equal], Methodology [equal], Writing—review & editing [equal]), Nicolas C. Ladizinsky (Investigation [equal], Methodology [equal], Writing—review & editing [equal]), James Snyder (Conceptualization [equal], Investigation [equal], Methodology [equal], Writing—review & editing [equal]), and Anika Hurst (Investigation [equal], Methodology [equal], Writing—review & editing [equal])
Conflicts of interest. None declared.
