https://orcid.org/0000-0001-7209-8295
Abstract
The spotted lanternfly, Lycorma delicatula (White), an invasive, phloem-feeding fulgorid generalist, was recently discovered in the United States. Current trapping methods include placing glue-covered sticky bands around trunks of host trees to exploit the lanternfly’s behavior of climbing up tree trunks. These bands are messy and need to be replaced often as they become covered in both target and nontarget insects and debris. Fourth instar nymphs and adults have also shown an ability to escape from traditional tree bands or avoid capture. A promising commercially available tree band (BugBarrier) design that faces inward to the trunk and targets larger developmental stages was tested. A modified pecan weevil trap (circle trunk trap) was also compared with tree bands. This design does not require the use of insect-trapping adhesive. Circle trunk traps caught more third and fourth instar and adult L. delicatula than BugBarrier bands. Flight intercept traps caught fewer adult L. delicatula than trunk-based tree bands. In a separate comparison, more spotted lanternflies were caught on adhesive-coated ‘tree mimicking’ traps placed along the edges of Ailanthus altissima Swingle (Sapindales: Simaroubaceae) stands than away from hosts in an open field. Circle trunk traps are recommended for their effectiveness at capturing L. delicatula as well as their relative ease-of-use and reusability.
The spotted lanternfly, Lycorma delicatula (White), is a phloem-feeding fulgorid generalist from China, which was recently discovered in the United States (Barringer et al. 2015, Dara et al. 2015). Prior to its introduction to North America, it was also found in Korea in the early- to mid-2000s (Han et al. 2008). With over 60 reported host plant species, it has the potential to be a serious pest of grapes Vitis vinifera L. (Vitales: Vitaceae; Lee et al. 2009, Park et al. 2009), stone fruits, and cultivated tree crops (Dara et al. 2015), and has been documented causing branch dieback on walnut, Juglans nigra L. (Fagales: Juglandaceae) and killing hops, Humulus lupulus L. (Rosales: Cannabaceae; Cooperband et al. 2018). The primary host in North America is tree of heaven, Ailanthus altissima Swingle (Sapindales: Simaroubaceae), an invasive tree from China (Barringer et al. 2015).
Following egg emergence, nymphs climb up the trunk of host trees in order to reach the canopy (Kim et al. 2011). This behavior may be interrupted by obstacles or wind, causing the nymphs to fall to the ground or jump to another surface. Once on the ground or a more suitable surface, L. delicatula will again climb upwards. As development continues, this behavior repeats itself, although according to Kim et al. (2011) the frequency of falling or jumping slows. Targeting this climbing behavior, trapping technology so far has been based primarily on glue-coated bands wrapped around tree trunks baited with a host volatile compound, methyl salicylate (Kim et al. 2011, Choi et al. 2012, Cooperband et al. 2019d). Sticky tree bands have shown limitations in their ability to capture L. delicatula as they walk up trees. First- and second-instar nymphs are captured, but later instars and adults avoid capture on sticky bands, and appear to pull themselves off of or avoid walking on the sticky surface (Cooperband et al. 2019d). Manufacturers can also change their glue formulations without notice, potentially causing possible changes in insect capture rates (JAF, MFC personal observation). More efficacious traps based on our increased understanding of their behavior and biology are needed for use in monitoring, detection, delimitation, and control of this invasive species.
In mark-release-recapture studies, nymphs at all stages were most often recaptured close (within 10 m) to their release points (Cooperband et al. 2019c). Similarly, following adult emergence, L. delicatula may walk or conduct short flights from one host plant to another to find suitable hosts, but tend not to travel far (Kim et al. 2011, Baker et al. 2019, Cooperband et al. 2019c, Myrick and Baker 2019, Wolfin et al. 2019). Adults begin to disperse during the early feeding stages (Cooperband et al. 2019c), and may do so in order to find more suitable maturation feeding and mating sites (Baker et al. 2019). Unmated females may conduct longer (10–50 m) flights during this time to find new trees in order to complete their maturation feeding (Wolfin et al. 2019), and once mated these females do not appear to be capable of flight greater than ~4m. Domingue and Baker (2019) showed that disturbed adults may fly away from woodlots toward sunlit, open fields. Females escaping from simulated predatory attacks also displayed a propensity to fly away from the disturbance and toward open fields. Light intensity and wavelength distribution can differ markedly between open day-lit and closed forested areas (Kelber 2006). Movement of L. delicatula toward open spaces and away from the inside of woodlots may correlate to greater amounts of preferred wavelengths of light in the open spaces (Jang et al. 2013, Domingue and Baker 2019). The placement of traps, vertically and horizontally in relation to woodlots, has been shown to be important to capturing forest insects (Su and Woods 2001, Lindhe et al. 2005, Wermelinger et al. 2007, Graham et al. 2012, Dodds 2014). For instance, spruce-feeding aphids, Elatobium abietinum (Hemiptera: Aphididae), were captured in the highest numbers on sticky traps placed in the upper canopy of host trees (Straw et al. 2011). Of 42 species of saproxylic beetles caught in traps along a horizontal gradient in relation to a woodlot, all but five of the species were caught in higher abundance along the edge or in a fully lit open area rather than in shaded, internal areas (Lindhe et al. 2005). Similarly, traps placed in an open field adjacent to an infested woodlot caught significantly more emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera: Buprestidae) than those placed inside the woodlot (Francese et al. 2008).
The overall goal of the studies described herein was to develop trapping and survey methods by attempting to exploit L. delicatula behaviors. Several trapping assays were conducted in 2018 to achieve this goal. Two approaches were taken: a) to compare tree bands or other trunk based traps to capture nymphs and adults walking on host trees, and b) to compare traps to capture flying, dispersing adults. As part of the first approach, two brands of tree bands were compared, including the standard Web-Cote. Several assays were also conducted to determine the efficacy of using pecan weevil circle traps (Mulder et al. 1997) modified to collect medium- to large-sized insects (Lance et al. 2013) as a potential monitoring tool for L. delicatula. Pecan weevil circle traps were originally designed to capture insects that walk up the trunk of host trees, so they provided a potential starting point for capturing this invasive pest. To achieve the second approach, flight intercept traps placed in the lower canopies of hosts were compared with tree bands to determine their effectiveness in capturing dispersing adults. Tall (~2.4 m) traps, with a similar structure to double decker traps (Poland et al. 2011) used for the emerald ash borer, were also compared along the edges of woodlots and in adjacent fields in order to exploit possible dispersing behavior as L. delicatula move from one woodlot to another.
Materials and Methods
Traps and Lures
Five trap types were compared in studies described below: Web-Cote ‘Sticky Tree Bands’ (referred to as Web-Cote; Web-Cote Industries, Hamburg, NJ), BugBarrier Tree Bands (BugBarrier; Environmetrics Systems USA, Inc., Victor, NY), modified pecan weevil traps (circle trunk trap), intercept panel traps (ChemTica USA, Durant, OK), and tall prism traps (tall prisms). For all of the studies described herein, traps were baited with a 53 mg/d methyl salicylate lure (AlphaScents, West Linn, OR) (as measured under laboratory conditions at 22–23°C) which had previously been shown to be attractive to all life stages of L. delicatula (Cooperband et al. 2019d). All traps attached to or hung from host trees were placed on A. altissima.
Web-Cote ‘Sticky Tree Bands’ (21.6-cm wide; Web-Cote Industries, Hamburg, NJ; Fig. 1a) has been shown to be the most effective tree band tested to date for trapping (Cooperband et al. 2019d). The Web-Cote sticky band was wrapped around an A. altissima host tree ~1.3 m above the ground. The band was attached to itself with the adhesive side facing out, away from the tree surface. Lures were placed near the top of the band, on the north side of the tree to limit sun exposure and prevent premature depletion.
Two types adhesive tree bands used in field assays: a) Web-Cote tree band wrapped around a host Ailanthus altissima tree, b) BugBarrier tree band wrapped around a host Ailanthus altissima.
BugBarrier tree bands (Fig. 1b) were attached to the tree using manufacturer instructions. Two layers of fiber material (7.6 cm wide) were placed around the host A. altissima tree. The BugBarrier adhesive film (12.7 cm wide) was then wrapped around the tree with the adhesive side facing inward toward the bark surface, with approximately 7.6 cm of adhesive material hanging downward from the fiber material. Later in the season (ca. 10 July), when L. delicatula reached fourth instar, a third layer of fiber material was added to the bark surface to prevent L. delicatula from climbing over the trap.
Circle trunk traps were constructed by modifying (Fig. 2a) a pecan weevil trunk trap (Great Lakes IPM, Vestaburg, MI) as described by Lance et al. (2013). The plastic cone opening to the trap was expanded from 0.7 cm to 1.5 cm, by removing a small portion of the tip of the cone, to accommodate the size of late instar and adult L. delicatula. The standard collection cup for the pecan weevil trap was replaced with a larger 1.89 liter, screw-top rectangular storage jar (Cornucopia Brands, Corbin, KY). An approximately 4.4 cm circular hole was cut into the lid of the jar, and the lid was attached using a hot-melt adhesive to the plastic section of the original trap opening near the base of the cone for the collection jar (Fig. 2b). Thus, the new larger collection jar could be unscrewed from its lid, allowing it to be easily removed and reattached. An 11.0 × 2.0 cm strip of Velcro (2.0-cm wide; Velcro Industries, Manchester, NH) was also glued to one of the long sides of the collection cup. The screen portion of the trap was attached around the host A. altissima tree per manufacturer recommendations with the following modifications. When attaching the trap, an approximately 5.0-cm section of Velcro was separated from the corresponding Velcro strip near the top of the collection jar (distal from the opening; Fig. 2c). This strip was then fastened to the tree using staples to prevent the opening of the trap from collapsing upon itself from the weight of the container and its eventual contents. The lure was then placed on the wooden structural strip facing in toward the tree (Fig. 2d). Two insecticidal strips (Vapona II 2,2-dichlorovinyl dimethyl phosphate (10%), Hercon Environmental, Emingsville, PA) were placed inside the collection jar. After 2 wk, jar contents were transferred into a plastic zipper bag for transport to the laboratory for sorting and counting.
a) A circle trunk trap (modified pecan weevil trap) used in field trapping assays, attached to an Ailanthus altissima tree. b) The trap is modified by increasing the opening to the collection cup and replacing the original collection cup with a 1.8 liter plastic storage container. A hole is cut into the lid of the new collection cup and glued to the original trap. c) The weight of the container is reinforced by gluing a Velcro strip to the container, and partially fastening the strip to the host tree. d) A lure is then placed on the wooden support of the trap facing inward toward the host to avoid impeding Lycorma delicatula walking up the trunk of the tree.
Intercept panel traps were precoated by the manufacturer with a solution of fluon, a fluoropolymer that increases the slipperiness of the trap surface and has been shown to increase trap catch in wood-boring beetles (Graham et al. 2010). They were hung from ropes in host trees ~5–8 m above the ground along the edge of woodlots. The collection cup at the bottom of each trap was filled with 300–400 ml of propylene glycol (Camco Easygoing-50, Camco, Greensboro, NC). Lures were hung from a hole in the center of the trap. Collection cup contents were poured through a medium mesh paper cone strainer that was then placed inside a plastic bag for transport to the laboratory for sorting and counting.
Tall prism traps (Fig. 3) were constructed from two 1.22 m × 1.22 m sheets (4.0 mm thick) of ‘brown’ corrugated plastic (Onyx Graphics, Hillsborough, NJ), as ‘brown’ sticky bands had been found previously to be more attractive than other colors (Choi et al. 2012). The plastic was scanned using a FieldSpec Pro (Analytical Spectral Devices Inc., Boulder, CO) contact probe spectrophotometer set to full range (350–2500 nm) scan mode to measure reflectance curves (methods described by Crook et al. 2009). The plastic used had peak reflectance of 7.2% and 8.2% at 360 nm and 620 nm, respectively (Fig. 4). Sheets were scored with a knife along the flutes and folded, to create a three sided trap surface where each side was 0.36 m wide × 1.22 m tall. Tabs, slots, and holes for holding the prism trap together and placing cable ties were cut in traps similarly to methods described by Francese et al. (2008), with modifications made for the increased size of the plastic used. Traps were coated with Tangletrap Insect trap coating (The Scotts Miracle-Gro Company, Marysville, OH) on all exterior surfaces. Similarly to Poland et al. (2011), a metal u-post (5.7 cm × 6.4 cm × 1.22 m) was driven into the ground to support the structure of the trap. A PVC pipe (8.9 cm outer diam, 7.8 cm inner diam. × 2.44 m) was then placed over the t-post, and attached to the base via a plastic cable tie. Holes were drilled into the pipe to accommodate attaching trap sections to the pipe and the pipe to the t-post. The bottom trap section was slid over the plastic and was attached to the PVC pipe at approximately 7.6 cm from the top and bottom of the trap. The top section was then placed over the bottom section so that there was no gap in between the two sections. A 2.5-cm strip of copper foil tape (XFasten, Miami, FL) was placed around the bottom of both the top and bottom sections of the trap to prevent slugs from climbing up the trap to scavenge dead L. delicatula captured on the trap surface. A lure was placed near the bottom of the top section by attaching it to the trap surface.
A tall prism trap placed along the edge of an Ailanthus altissima woodlot. Glue is being applied to the outside surface of the trap.
Reflectance spectra of brown corrugated plastic, used in the construction of tall prism traps.
Study Sites
Three field sites were used for the studies: HE, BA, and KU. Whereas KU traps were placed on the campus of Kutztown University, Kutztown, Berks County, PA, the other two sites, BA and HE, refer to privately owned land located in Barto, Berks County (roughly, 40°23′27.2″N 75°36′38.5″W) and Hereford Township, Berks County (roughly, 40°26′50.7″N 75°37′20.2″W), respectively. Because numerous L. delicatula egg masses per tree were found at each site prior to the field season at both KU and BA, they were considered medium to high population density sites. Conversely, at HE, almost no egg masses could be found on any host trees in early 2018. Due to overfeeding by L. delicatula, these trees were depleted, wilted and brown at the end of summer 2017. Based on these observations, HE was considered a low to medium density site for the beginning of the 2018 field season. All three sites had a high proportion of A. altissima. Due to limited space at sites, trapping studies were placed and taken down to make way for other studies throughout the season. At the beginning of the trapping season, pairs of trees were selected at each site to be used throughout the season and over multiple studies, with the exception of the edge versus field comparison (see below). Trees in individual pairs used at each site were adjacent (within 2 to 5 m) to each other, and were selected to have almost identical diameter at breast height (DBH) prior to trapping studies being conducted. Pairs were at least 20 m apart from each other. Dates when traps were placed and checked for each study are listed in the individual sections for each study.
Web-Cote Versus BugBarrier
The two adhesive tree band products, Web-Cote and BugBarrier, were compared in several field assays at the HE site during three developmental periods (Cooperband 2019a, 2019b) in 2018: a) early nymph (13 to 30 May), b) late nymph (11 to 23 July), and c) early adult (20 August to 5 September). For each trapping period, 7, 10, and 10 pairs were placed in the field, respectively. At the end of each trapping period, adhesive bands were removed from the trees and brought back to the laboratory for counting specimens.
BugBarrier Versus Circle Trunk: Late Nymph Period
Ten pairs of traps were placed in a field assay comparing BugBarrier bands with circle trunk traps in the field during the late instar nymph period on 10 July at KU (n = 6) and BA (n = 4) sites. Traps were then checked three times, on 25 July, 31 July, and 7 August and samples returned to the laboratory. During the 31 July trap check, traps in each pair of trees were rotated to the opposite tree in the pair to reduce potential individual tree effects on trap catch. In the laboratory, specimens were separated by life stage. Adults were also sexed, when possible. Catch from these last two trap checks were then summed.
BugBarrier Versus Circle Trunk: Adult Flight Period
Traps were placed in the field to compare trap catch between BugBarrier and Circle Trunk traps when set during the adult flight period. Ten pairs of traps were placed at HE on 20 September 2018. Traps were then checked twice, on 2 October and 17 October, and samples returned to the laboratory. During the 2 October trap check, traps in each pair were rotated to the opposite tree in the pair to reduce potential individual tree effects on trap catch. In the laboratory, specimens were separated by life stage. Adults were also sexed, when possible. Catch from these two trap checks were then summed.
BugBarrier Versus Intercept Panel Traps
BugBarrier tree bands were compared with intercept panel traps in a field assay conducted at KU and BA sites in 2018. Ten pairs (n = 6, n = 4, respectively) of traps paired on adjacent A. altissima trees were placed in the field on 21 and 22 August and checked every 2 wk with trap checks on 5 September, 19 September, 4 October, and 16 October. At each trap check, intercept panel collection cups were emptied and BugBarrier bands were removed for analysis in the laboratory. Trap catch was summed for the entire trapping period.
Edge Versus Field Comparison
Tall traps were placed at the KU site to compare catch on the traps at two locations in relation to an A. altissima woodlot, in 10 replicates. Two 2.4-m traps were placed in each replicate, with one trap placed along the edge of the woodlot, and the other placed 15 m outside the woodlot in an open field. Each side of both the top and bottom sections of each trap was labeled with large letters (~5–7 cm tall): ‘A’, ‘B’, or ‘C’. Due to the large size and unwieldy nature of this trap, photographs of each trap were taken so that the number of L. delicatula caught could be recorded during each trap check in the laboratory without having to move the trap. Traps were placed in the field on 21 and 22 August 2018 and checked every 2 wk with trap checks on 5 September, 19 September, 4 October, and 16 October.
Statistical Analysis
For all assays comparing treatments placed on trees, the difference in catch data within each pair was tested for normality. Among these studies, with one exception, the differences were normally distributed, and paired t-tests (P = 0.05) were conducted to determine whether there were significant differences between treatments in JMP (SAS Institute 2012). The differences in trap catch within pairs were not normally distributed in the BugBarrier versus Circle Trunk: Late nymph period assay. A Wilcoxon signed-rank test (P < 0.05) was performed to determine whether there were significant differences between treatments in this assay. The data set in the field versus edge comparison was also tested for normality, and was not normally distributed. In order to test the interaction effect between trap location and height of trap section in the field versus edge comparison, the trap catch was log-transformed (y + 0.05) to normalize the data. An ANOVA was conducted on the transformed data to compare treatment and interaction effects. Confidence intervals (95%) were calculated from the standard error of the transformed trap catch. Means and confidence intervals were then back-transformed for presentation in the Edge Versus Field Comparison section.
Results
Web-Cote Versus BugBarrier
Trap type played a significant role in trap catch during the late nymph (t = 3.54, df = 1, P = 0.0062) and early adult (t = 8.33, df = 9, P < 0.0001) trapping periods. Trap catch was higher on BugBarrier traps than on Web-Cote traps during these periods (Table 1). There was no significant difference in trap catch between the two band types during the early nymph period (t = 1.21, df = 6, P = 0.27) (Table 1).
Mean number (± SE) Lycorma delicatula caught per trap in a trapping study comparing two types of adhesive-coated ‘tree-wrap’ type traps, BugBarrier and Web-Cote, at one site in Berks County, PA
| Trap | 30 May to 13 June (early nymph) | 11 to 23 July (late nymph) | 20 Aug. to 5 Sept. (adult—early flight) |
|---|---|---|---|
| BugBarrier | 8.4 ± 2.4 | 3.3 ± 0.7a | 55.5 ± 8.6a |
| Web-Cote | 3.8 ± 3.2 | 0.6 ± 0.3b | 14.6 ± 4.2b |
| Trap | 30 May to 13 June (early nymph) | 11 to 23 July (late nymph) | 20 Aug. to 5 Sept. (adult—early flight) |
|---|---|---|---|
| BugBarrier | 8.4 ± 2.4 | 3.3 ± 0.7a | 55.5 ± 8.6a |
| Web-Cote | 3.8 ± 3.2 | 0.6 ± 0.3b | 14.6 ± 4.2b |
Three trapping periods were observed, early nymph following egg emergence (n = 7), late nymph coinciding with third and fourth instar development (n = 10) and early adult flight (n = 10). Traps were taken down between trapping periods to accommodate other studies being conducted at the trapping location. Letters within a column represent a significant difference in trap catch between the mean trap catch at that life stage between traps (paired t-tests; P < 0.05).
Mean number (± SE) Lycorma delicatula caught per trap in a trapping study comparing two types of adhesive-coated ‘tree-wrap’ type traps, BugBarrier and Web-Cote, at one site in Berks County, PA
| Trap | 30 May to 13 June (early nymph) | 11 to 23 July (late nymph) | 20 Aug. to 5 Sept. (adult—early flight) |
|---|---|---|---|
| BugBarrier | 8.4 ± 2.4 | 3.3 ± 0.7a | 55.5 ± 8.6a |
| Web-Cote | 3.8 ± 3.2 | 0.6 ± 0.3b | 14.6 ± 4.2b |
| Trap | 30 May to 13 June (early nymph) | 11 to 23 July (late nymph) | 20 Aug. to 5 Sept. (adult—early flight) |
|---|---|---|---|
| BugBarrier | 8.4 ± 2.4 | 3.3 ± 0.7a | 55.5 ± 8.6a |
| Web-Cote | 3.8 ± 3.2 | 0.6 ± 0.3b | 14.6 ± 4.2b |
Three trapping periods were observed, early nymph following egg emergence (n = 7), late nymph coinciding with third and fourth instar development (n = 10) and early adult flight (n = 10). Traps were taken down between trapping periods to accommodate other studies being conducted at the trapping location. Letters within a column represent a significant difference in trap catch between the mean trap catch at that life stage between traps (paired t-tests; P < 0.05).
Average DBH was 13.7 ± 0.7, 14.7 ± 1.1, and 15.0 ± 1.1 cm, for BugBarrier banded trees during the early nymph, late nymph, and early adult trapping periods, respectively. It was 14.0 ± 0.9, 13.9 ± 0.7, and 14.2 ± 0.83 for Web-Cote banded trees. The difference in DBH was not significantly different during the early nymph (t = 0.53, df = 6, P = 0.61), late nymph (t = 1.73, df = 9, P = 0.12), and early adult (t = 0.66, df = 9, P = 0.52), so trap catch as a function of DBH was not compared.
BugBarrier Versus Circle Trunk: Late Nymph Period
During the trapping period between 10 July and 7 August, there was a significant difference in catch between trap types among the total number of nymphs (χ 2 = 6.04, df = 1, P = 0.0140) and fourth-instar nymphs (χ 2 = 7.61, df = 1, P = 0.0058) with more L. delicatula caught in circle trunk traps (Table 2). Second-instar L. delicatula were collected in 5 of 10 Circle Trunk traps and on 6 of 10 BugBarrier bands. Third- and Fourth-instar nymphs were collected on 8 of 10 and 9 of 10 traps each. There was no significant difference in trap catch between BugBarrier and Circle Trunk traps among second-instar nymphs (χ 2 = 2.44, df = 1, P = 0.12), third-instar nymphs (χ 2 = 0.75, df = 1, P = 0.38) or adult L. delicatula (χ 2 = 1.41, df = 1, P = 0.23) (Table 2). Trap type did not play a role in the proportion of males (χ 2 = 2.89, df = 1, P = 0.09) to the total number of L. delicatula caught during this trapping period.
Mean number (±SE) Lycorma delicatula caught per trap in a trapping assay comparing two trap types (n = 10), BugBarrier and Modified Pecan Weevil (Circle Trunk) at two sites in Berks County, PA
| Trap type | Nymph total | Second instar | Third instar | Fourth instar | Adult total | Proportion of males caught |
|---|---|---|---|---|---|---|
| BugBarrier | 190.9 ± 92.8b | 9.1 ± 5.4 | 66.0 ± 32.5 | 115.8 ± 55.5b | 11.6 ± 6.1 | 0.53 ± 0.10 |
| Circle Trunk | 271.4 ± 75.5a | 2.2 ± 0.9 | 61.2 ± 18.0 | 208.0 ± 57.0a | 6.6 ± 2.7 | 0.59 ± 0.06 |
| Trap type | Nymph total | Second instar | Third instar | Fourth instar | Adult total | Proportion of males caught |
|---|---|---|---|---|---|---|
| BugBarrier | 190.9 ± 92.8b | 9.1 ± 5.4 | 66.0 ± 32.5 | 115.8 ± 55.5b | 11.6 ± 6.1 | 0.53 ± 0.10 |
| Circle Trunk | 271.4 ± 75.5a | 2.2 ± 0.9 | 61.2 ± 18.0 | 208.0 ± 57.0a | 6.6 ± 2.7 | 0.59 ± 0.06 |
The trapping period for this study (10 July 2018 to 7 August 2018) mainly coincided with late nymph stages of development for the insect. The proportion of males to the total number of adults captured during this period is also represented. Letters within a column represent a significant difference in trap catch between the mean trap catch at that life stage between traps (Wilcoxon signed-rank test; P < 0.05).
Mean number (±SE) Lycorma delicatula caught per trap in a trapping assay comparing two trap types (n = 10), BugBarrier and Modified Pecan Weevil (Circle Trunk) at two sites in Berks County, PA
| Trap type | Nymph total | Second instar | Third instar | Fourth instar | Adult total | Proportion of males caught |
|---|---|---|---|---|---|---|
| BugBarrier | 190.9 ± 92.8b | 9.1 ± 5.4 | 66.0 ± 32.5 | 115.8 ± 55.5b | 11.6 ± 6.1 | 0.53 ± 0.10 |
| Circle Trunk | 271.4 ± 75.5a | 2.2 ± 0.9 | 61.2 ± 18.0 | 208.0 ± 57.0a | 6.6 ± 2.7 | 0.59 ± 0.06 |
| Trap type | Nymph total | Second instar | Third instar | Fourth instar | Adult total | Proportion of males caught |
|---|---|---|---|---|---|---|
| BugBarrier | 190.9 ± 92.8b | 9.1 ± 5.4 | 66.0 ± 32.5 | 115.8 ± 55.5b | 11.6 ± 6.1 | 0.53 ± 0.10 |
| Circle Trunk | 271.4 ± 75.5a | 2.2 ± 0.9 | 61.2 ± 18.0 | 208.0 ± 57.0a | 6.6 ± 2.7 | 0.59 ± 0.06 |
The trapping period for this study (10 July 2018 to 7 August 2018) mainly coincided with late nymph stages of development for the insect. The proportion of males to the total number of adults captured during this period is also represented. Letters within a column represent a significant difference in trap catch between the mean trap catch at that life stage between traps (Wilcoxon signed-rank test; P < 0.05).
BugBarrier Versus Circle Trunk: Adult Flight Period
During the trapping period between 20 September and 17 October, no nymphs were caught in traps. Trap type played a significant role in adult L. delicatula trap catch (t = 4.15, df = 9, P = 0.0025) with more L. delicatula caught in Circle Trunk traps than on BugBarrier bands (Table 3). Trap type did not play a role in the proportion of males (t = 0.92, df = 9, P = 0.38) caught during this trapping period. Adult L. delicatula were captured on every trap in the assay.
Mean number (± SE) Lycorma delicatula adults caught per trap in a trapping study comparing two trap types (n = 10), BugBarrier and Modified Pecan Weevil (Circle Trunk) at two sites in Berks County, PA
| Trap type | Mean adults | Proportion of males caught |
|---|---|---|
| BugBarrier | 278.9 ± 22.1b | 0.20 ± 0.02 |
| Circle Trunk | 461.1 ± 79.9a | 0.18 ± 0.03 |
| Trap type | Mean adults | Proportion of males caught |
|---|---|---|
| BugBarrier | 278.9 ± 22.1b | 0.20 ± 0.02 |
| Circle Trunk | 461.1 ± 79.9a | 0.18 ± 0.03 |
The trapping period (20 September to 17 October 2018) for this study coincided with the mid to late adult flight for the insect. The proportion of males to the total number of adults captured during this period is also represented. Letters within a column represent a significant difference in trap catch between the mean trap catch at that life stage between traps (paired t-tests; P < 0.05).
Mean number (± SE) Lycorma delicatula adults caught per trap in a trapping study comparing two trap types (n = 10), BugBarrier and Modified Pecan Weevil (Circle Trunk) at two sites in Berks County, PA
| Trap type | Mean adults | Proportion of males caught |
|---|---|---|
| BugBarrier | 278.9 ± 22.1b | 0.20 ± 0.02 |
| Circle Trunk | 461.1 ± 79.9a | 0.18 ± 0.03 |
| Trap type | Mean adults | Proportion of males caught |
|---|---|---|
| BugBarrier | 278.9 ± 22.1b | 0.20 ± 0.02 |
| Circle Trunk | 461.1 ± 79.9a | 0.18 ± 0.03 |
The trapping period (20 September to 17 October 2018) for this study coincided with the mid to late adult flight for the insect. The proportion of males to the total number of adults captured during this period is also represented. Letters within a column represent a significant difference in trap catch between the mean trap catch at that life stage between traps (paired t-tests; P < 0.05).
BugBarrier Versus Intercept Panel Traps
Significantly more L. delicatula were caught on BugBarrier (384.3 ± 81.1) adhesive band traps than in intercept panel traps (87.2 ± 17.5) (t = 3.41, df = 9, P = 0.0077) over the course of the trapping season.
Edge Versus Field Comparison
Trap location played a significant role in L. delicatula catch (F = 33.07, df = 1, P < 0.0001) with more mean L. delicatula caught along the edge of woodlots [62.7 (44.6–88.0 95% confidence interval (CI)] than in the open field [18.3 (12.0–28.0 CI)]. There was no significant difference in catch between the upper [28.0 (20.0–40.0 CI)] and lower [41.0 (25.7–64.9 CI)] 1.2 m sections (F = 3.07, df = 1, P = 0.091) or the interaction between trap height and location in relation to a woodlot (F = 2.76, df = 1, P = 0.108). Replicate also played a significant role in L. delicatula caught (F = 23.50, df = 9, P < 0.0001).
Discussion
In order to improve upon existing trapping capabilities, we evaluated five types of traps for efficacy against L. delicatula, two of which represent significant improvements to the previously best available trap. BugBarrier bands were more effective at catching L. delicatula on the trunks of hosts than Web-Cote bands. The BugBarrier bands may have had several advantages to the Web-Cote bands. With the trap surface facing inward toward the tree surface, the accumulation of dust and debris on BugBarrier bands was less than on the outward facing Web-Cote, which also resulted in a reduction in nontarget captures. While walking up the tree surface, L. delicatula tend to avoid obstacles or changes in the surface of the host tree (Kim et al. 2011), and later instar nymphs and adults were shown to walk over or avoid the surface of sticky band traps (Cooperband et al. 2019d). The inward facing surface of the BugBarrier bands may deter this behavior, as the lanternflies may not seek an alternate route until after they have been funneled under the sticky band to the fiber material where their only options are to either fall or walk downward, or jump, fly, or walk onto the sticky trap surface where they become captured. This may be demonstrated by the fact that many of the L. delicatula captured on BugBarrier traps were stuck to the trap surface by their wings or dorsal body surface (EGB, SMD pers. obs.).
The circle trunk trap was significantly more effective at capturing L. delicatula adults and fourth-instar nymphs than the BugBarrier bands, with no significant difference in catch between the two trap types for the earlier nymph stages which were less prevalent during that timeframe. There was also no difference in the detection rate (defined as finding at least a single lanternfly on a trap in a replicate) for the two traps at the elevated population densities. All of the circle trunk traps catching L. delicatula when their paired BugBarrier counterpart caught them, with the exception of one replicate.
Although there was no difference between traps in the sex ratio of adults captured, we did see a seasonal shift in sex ratio from 53–59% males in the early adult flight stage (Table 2) to 18–20% males in the mid to late adult stage. This is similar to a seasonal sex ratio shift on tree trunks seen in 2016 by others (MFC, pers. obs., Domingue et al. 2019). The shift occurred prior to observed mating, suggesting that males may be aggregating in another part of the tree or forest to attract or intercept, court, and mate with receptive females. Studies are underway to better understand this process.
Compared with the BugBarrier tree bands, intercept panel traps were relatively ineffective at catching L. delicatula as 81.5% of all of the lanternflies caught in the study were captured on tree bands. The flight traps and the tall traps placed outside the woodlots may not have provided a large enough or attractive enough (both visually and chemically) target to pull L. delicatula adults dispersing to or from woodlots or from tree to tree, whereas the tree band was more effective at targeting the walking behaviors and short dispersal range of early instar nymphs (Kim et al. 2011, Cooperband et al. 2019c) and later nondispersing adults (Cooperband et al. 2019c, Wolfin et al. 2019). When tall prism traps were compared at locations along the edge and the field of A. altissima woodlots, the edge traps caught more than three times as many L. delicatula than their field counterparts. This tendency to stay close to the host tree or tree line appears to be similar to earlier findings that most L. delicatula do not disperse far, except when feeding following emergence (Baker et al. 2019, Cooperband et al. 2019c; Wolfin et al. 2019). These traps rely on a small single-component kairomone lure to attract L. delicatula away from the trunks of attractive host trees, and so may not be able to overcome the attraction of the complex volatile profile of the host tree itself. Because of this, at this time on-trunk traps are probably the most attractive traps to survey for L. delicatula.
Aside from the improved trap catch, there were some notable benefits in handling of circle trunk traps over sticky bands. With sticky bands, debris such as leaves, dust, and nontarget insects gradually become stuck to the surface rendering it less sticky over time. The entire sticky portion must be changed regularly in order to refresh the dirty or saturated trapping surface with a clean, new one allowing the target insects to be trapped. Additionally, under the surface of adhesive tree bands, lanternflies may find a suitable shelter to safely feed and oviposit (KMM, SLC pers. obs.). As seen with multiple traps used for a variety of insects, adhesive-coated traps can also pose a problem for nontarget vertebrates that may inadvertently become stuck to a glued trap surface (JAF, MFC unpublished data). Replacing glue-coated tree bands can be messy and time consuming, especially in a large-scale survey. Once the trap has been checked for the presence or absence of L. delicatula, the glue-coated material must then be discarded. Most municipalities will not accept these materials for recycling, because of the glue (if the initial trap material is even recyclable). The circle trunk trap allows the surveyor to take a sample from the trap without disturbing the trap. Care in trap set-up and take-down allows them to be reused over multiple field seasons. In 2019, USDA APHIS PPQ provided this trap to multiple states for use in L. delicatula surveys. Studies being conducted in 2019 include comparing survey tools at varying population densities and earlier developmental stages, as well as improving the circle trunk trap to increase its efficiency and ease of use.
Disclaimer: Mention of trade names or commercial products in this publication is solely to provide specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture.
Acknowledgments
Leslie Abreu and Shannon Adams of the USDA APHIS PPQ Otis Laboratory provided field work assistance. Isaiah Canlas, Samuel Stella, and Christopher Smith assisted in the processing of trap samples both during and at the conclusion of the field season. Damon Crook (Otis Lab) provided technical laboratory assistance. Daniel Carrillo (University of Florida) and Ann Ray (Xavier University) provided scientific critique and helpful discussions. This work was funded by USDA APHIS PPQ Methods Development, cooperative agreements from USDA APHIS PPQ PPA 7721 to cooperating institutions: AP18PPQS&T00C017 (East Stroudsburg University), and AP18PPQS&T00C042, AP17PPQS&T00C119 and AP18PPQS&T00C169 (Xavier University).
