Lycorma delicatula (White) (Hemiptera: Fulgoridae), the spotted lanternfly, is an invasive herbivorous insect first identified in Pennsylvania, USA, in Sep 2014 (Barringer et al. 2015). Since its original detection, this species has spread throughout the mid-Atlantic region, with current infestations in 14 eastern states, prompting quarantines in portions of Connecticut, Delaware, Maryland, New Jersey, Ohio, Pennsylvania, and Virginia, USA (NY IPM 2022; USDA-APHIS 2022). Long-distance spread has led to isolated detections in at least 2 additional eastern states (Vermont and New Hampshire, USA) (NY IPM 2022). The rapid range expansion and long-distance spread of L. delicatula are human assisted, because individuals can hitchhike on vehicles, and egg masses are easily moved on solid objects (Urban 2019). Modeling studies have identified broad areas of habitat suitable for establishment of L. delicatula populations, including much of eastern North America, portions of the west coast of the USA, broad areas of Europe, and portions of Australia, South America, and southern Africa (Jung et al. 2017; Wakie et al. 2019). This broad global habitat suitability and propensity for rapid spread have led to expectations of increased economic and ecological impacts of this species in the future.

Individuals of L. delicatula damage plants directly by feeding on phloem, leading to puncture wounds, loss of photosynthate, and sap exudation, with reports of direct damage contributing to plant mortality in grape vineyards (Vitis spp.) (Lee et al. 2019; Urban 2019). Feeding also leads to indirect damage, because honeydew and exuded sap both facilitate the growth of sooty mold, which blocks photosynthesis when it grows on green plant tissues (Dara et al. 2015; Urban 2019). Economic losses are currently of greatest concern for grapes (Lee et al. 2019; Urban 2019; Leach & Leach 2020a), although tree fruits, hops, nursery stock, and forestry also have been identified as potentially at risk for direct and indirect economic losses (Harper et al. 2019; USDA-APHIS 2019). Plant diseases are not known to be vectored by L. delicatula (Urban 2019; Brooks et al. 2020), and individuals do not bite humans or other animals, but L. delicatula is a notable nuisance pest in residential landscapes. Adult insects can be found in high densities on trees and shrubs in these landscapes; adults often fly in large aggregations in the autumn, and individuals produce copious amounts of honeydew (Urban 2019).

Lycorma delicatula is known as a feeding generalist, with observations on over 100 plant species across its native and invasive range, with particular affinity for tree of heaven, Ailanthus altissima (Mill.) Swingle (Simaroubaceae) (Dara et al. 2015; Liu 2019). Although lists of potential host plants assembled from the literature are available for L. delicatula (Barringer et al. 2015; Dara et al. 2015; Liu 2019; Avanesyan & Lamp 2020; Barringer & Ciafré 2020), quantitative information on host use remains incomplete and appears to depend in part on host availability (Liu 2019). When herbivores are introduced to new ranges, predicting host plant use can be challenging (Bertheau et al. 2010; Pearse & Altermatt 2013). Different assemblages of plants may be present in the native and invaded ranges, and preferences for plants shared across the global range of the species may not be consistent. Even for polyphagous species, use of plants typically is non-uniform and may change across ontogeny, may depend on the locally available plants, or may be a function of plant species abundance or geographic provenance (Mason et al. 2011; Martinson et al. 2016, 2020).

Although adults of L. delicatula exhibit strong preference for A. altissima (Barringer et al. 2015; Liu 2019), egg masses are oviposited on a variety of substrates, including live plants, landscape materials, and vehicles (Liu 2019; Urban 2019). Nymphs are highly mobile, move within and among plants, and likely feed on several host plant species (Kim et al. 2011; Jung et al. 2017; Urban 2019; Avanesyan & Lamp 2020; Nixon et al. 2020, 2022). Factors that determine female oviposition preference and the use of plants by nymphs remain unclear but are important in understanding risks and developing management strategies for L. delicatula (Urban 2019). In addition to plant species, plant size may play a role in oviposition and feeding, with larger plants expected to support higher abundances of insect herbivores because of the greater availability of resources (Marques et al. 2000; Whitfeld et al. 2012).

Table 1.

Locations of field surveys in Pennsylvania for Lycorma delicatula egg masses and nymphs.

In this study, we investigated egg mass locations and nymphal host plant use at 3 sites in eastern Pennsylvania. We assessed patterns of egg mass and nymphal abundance with respect to plant species and plant size, focusing on trees, shrubs, and vines, because woody plants constitute most known host plants (Barringer & Ciafré 2020). We then applied this information and information from the literature to predict threats to 2 currently uninvaded sites in central Maryland. In doing so, we hope to demonstrate that a combination of local tree inventories and information about L. delicatula host plant use can be used to assess future risks. As L. delicatula continues to spread in the mid-Atlantic region, understanding risks to locations likely to be infested will be a key to developing a priori management strategies.

Materials and Methods

RISK ASSESSMENT IN MANAGED AND SEMI-NATURAL SETTINGS

Using data from the Pennsylvania field surveys and published reports of host plant use, we then assessed the extent to which L. delicatula would pose a risk to the woody plants present in 2 settings in Carroll County, located in north-central Maryland. Recent establishment of L. delicatula populations in many of the counties in central and northern Maryland, along with climatic suitability throughout the mid-Atlantic region (Wakie et al. 2019), indicate that risk of L. delicatula to locations throughout Carroll County, Maryland, is high. We therefore assessed risk of L. delicatula impacts at the main campus of McDaniel College in Westminster, Maryland (39.586327°N, 77.000360°W), and the McDaniel Environmental Center, a 55-acre semi-natural property in nearby New Winsor, Maryland (39.527894°N, 77.042380°W). These 2 sites allowed us to work in locations similar to those we surveyed in Pennsylvania in that they both included managed and unmanaged areas and assemblages of native and introduced plants common in the region.

Fig. 1.

Sites of partial tree inventories for Lycorma delicatula risk assessment at (A) McDaniel College in Westminster, Maryland, and (B) the McDaniel Environmental Center in New Windsor, Maryland. At the college, trees were selected randomly and included ornamental plantings near academic and residential buildings and trees in semi-natural habitats around the golf course. At the McDaniel Environmental Center, trees were surveyed at forest edges.

A subset of trees on McDaniel College's campus and at the McDaniel Environmental Center were identified and mapped in Jun and Jul 2019. The college campus is managed and park-like, with ornamental plantings near academic and residential buildings and semi-natural habitats around a golf course (Fig. 1A). We conducted a partial tree inventory of the campus and included a random set of both ornamental and semi-natural regions to develop a robust risk assessment for L. delicatula. The McDaniel Environmental Center contains a variety of habitat types, including agriculture, old fields, forest patches, ponds, and small access roads (Fig. 1B). At the McDaniel Environmental Center, the tree inventory focused on trees within 1 m of forest edges. Though our survey at the McDaniel Environmental Center was constrained to edge habitats, forest edges represent an important habitat for a favored host plant, A. altissima, and surveys in vineyards in Pennsylvania have documented strong edge effects for L. delicatula (Leach & Leach 2020b).

For each species in the tree inventories, we determined risk in the following manner. Plant species described in the literature as being used for oviposition or as hosts for nymphs or adults were designated in our inventories as “known risk.” Plant species used for oviposition or as a nymphal host in our Pennsylvania field surveys but not yet reported elsewhere in the literature were designated as “of new concern.” Finally, plant species in the campus or McDaniel Environmental Center tree inventories but not yet evaluated or reported as hosts from the literature were designated as “unknown risk.” Literature used for assessing risk included Dara et al. (2015), Barringer and Ciafré (2020), and Liu et al. (2020).

Results

PENNSYLVANIA FIELD SURVEYS

We investigated the presence and abundance of L. delicatula egg masses and nymphs on a total of 15 species of trees, shrubs, and vines in our Pennsylvania surveys (Table 2). The surveys included 55 unique plants, most of which (69.1%) were visited twice, for a total of 93 timed surveys. We found a total of 174 L. delicatula egg masses and 460 nymphs in our 179 min of field surveys. Egg masses were present on 10 of the 15 plant species we surveyed, and nymphs were present on 12 plant species (Table 2). Notably, our observations of egg masses on eastern redbud, Cercis canadensis L. (Fabaceae) and sweetgum, Liquidambar styraciflua L. (Altingiaceae), and egg masses and nymphs on green hawthorn, Crataegus viridis L. (Rosaceae) are, to the best of our knowledge, newly recorded host plants for L. delicatula (Table 2).

Sites differed substantially in the composition of plants and the use of these plants by L. delicatula. Far more egg masses were observed at the residential site (160 egg masses) than the arboretum (14 egg masses), and no egg masses were observed at the roadside site (Fig. 2). Nymphal abundance also was highest at the residential site (214 nymphs), followed by the arboretum (159 nymphs) and the roadside (87 nymphs (Fig. 2). Overall, the highest proportion of egg masses across all the surveys was observed on sugar maple, Acer saccharum Marshall (Sapindaceae) at the residential site, and the highest proportions of nymphs across all the surveys were found on black walnut, Juglans nigra L. (Juglandaceae) and ornamental rose, Rosa sp. (Rosaceae) at the residential site; on bee-bee tree, Tetradium daniellii (Bennett) T. G. Hartley (Rutaceae) at the arboretum, and A. altissima at the roadside site (Fig. 2).

Table 2.

Presence and total counts of egg masses (EM) and nymphs (Nym) of Lycorma delicatula among surveyed plants, with the total number of each species surveyed (N). Literature accounts of plants used for feeding and oviposition also are indicated, as are new records of host plant use from this study.

The most parsimonious model for egg mass abundance included only the effect of plant size (square root transformed maximum egg mass abundance per plant; Likelihood ratio test, LRT χ² = 15.388; df = 1; P < 0.0001), and egg mass abundance increased as a function of plant size (Fig. 3A). In contrast, nymphal abundance tended to be lower on larger trees, though the relationship between nymphal abundance and plant size was not retained for the final model (square root transformed nymphal abundance; LRT χ² = 2.522; df = 1; P = 0.112) (Fig. 3B).

Egg mass abundance varied considerably among surveyed plants, but the effect of plant species was not retained in the final model (LRT χ² = 13.578; df = 9; P = 0.138) (Fig. 4A). Again, the pattern for nymphs differed from that of egg masses, and nymphal abundance varied significantly among plant species (LRT χ² = 49.352; df = 9; P < 0.0001) (Fig. 4B), with the highest average abundances on A. altissima (AIL), Rosa sp. (ROS), J. nigra (JUG), and T. daniellii (TET). The final model for nymphal abundance included only the effect of plant species; the effect of date was non-significant (LRT χ² = 1.956; df = 3; P = 0.582) and was removed during model simplification.

RISK ASSESSMENT IN MANAGED AND SEMI-NATURAL SETTINGS

We surveyed a total of 307 trees at the McDaniel College campus and the McDaniel Environmental Center representing a range of ornamental plantings and trees in semi-natural habitats. The partial tree inventory of the college campus included a greater number of tree species (15 species) compared to the McDaniel Environmental Center (6 species) (Table 3). The campus included 12 species known to be used by L. delicatula for oviposition and feeding, making up 86% of the trees surveyed. We identified 1 species on the campus that may be of new concern as a host for L. delicatula, the commonly planted native sweet gum tree (L. styraciflua), based on our Pennsylvania field surveys. The diverse tree community on the campus also included 2 species (Quercus palustris Munchh. and Gleditsia tricanthos L. [both Fagaceae]) that previously have not been assessed or reported for L. delicatula host plant status. In contrast, the semi-natural habitats surveyed at the McDaniel Environmental Center contained far more trees of species known to be host plants or oviposition substrates for L. delicatula, including A. altissima and J. nigra; all together, trees of known risk made up 98.6% of the trees surveyed at the McDaniel Environmental Center.

Fig. 2.

Plant composition and the distribution of Lycorma delicatula egg masses and nymphs among plants varied among residential (top), arboretum (middle), and roadside sites (bottom) in the Pennsylvania field surveys. Plotted are the abundances of egg masses as the proportion of all egg masses observed in the study (orange circles) and the abundances of nymphs as the proportion of all nymphs observed in the study (blue diamonds), with symbol size scaled to proportional abundance.

Discussion

Lycorma delicatula is a major new threat to natural environments and agriculture, and is a significant nuisance pest in its invaded ranges (Dara et al. 2015; Lee et al. 2019). Since its first detection in Pennsylvania in 2014 (Barringer et al. 2015), the list of plants used as oviposition substrates or for feeding by nymphs and adults has continued to grow and now includes over 100 species from various geographic provenances (Barringer & Ciafré 2020). As populations continue to spread and these insects move into new sites with varying environmental conditions and varying plant communities, the list of plants susceptible to oviposition, feeding, and damage is likely to expand. Continued investigations into host plant use and oviposition substrates certainly will be a key to developing effective monitoring and management strategies throughout the invaded range of L. delicatula.

Fig. 3.

Plant size (as diam at breast height, DBH, in cm) differentially affected the abundance of Lycorma delicatula egg masses and nymphs in the Pennsylvania field surveys. (A) Egg mass abundance was an increasing function of plant size. Points are maximum egg mass counts per plant, and the line depicts the back-transformed model-estimated effect of plant size. (B) Nymphal abundance (plotted as average nymphal abundance per plant but analyzed as raw count data) tended to decrease with plant size.

Though limited in temporal and geographic scope, our study found similar patterns in favored oviposition and host plant use compared with field studies in other locations in Pennsylvania. Specifically, we found that nymphal abundance in our timed surveys differed significantly among potential host plants, with highest nymphal abundances on A. altissima, Rosa sp., J. nigra, and T. daniellii (Fig. 4B). These top nymphal hosts correspond well with favored host plants from the recent literature (Dara et al. 2015; Lee et al. 2019; Liu 2020). Additional host plants used by nymphs in our study (Acer saccharum Marshall [Sapindaceae], Amelanchier sp. Medik. [Rosaceae], Fagus grandifolia Ehrh. [Fagaceae], Liriodendron tulipifera L. [Magnoliaceae], Pinus strobus L. [Pinaceae], Prunus avium L. [Rosaceae], Rubus spp. L. [Rosaceae], and Vitis sp. Bunge [Vitaceae]) also correspond with literature reports and support the conclusion of a broad host range for nymphs (Kim et al. 2011; Song et al. 2018; Lee et al. 2019). We also report early instar nymphs on green hawthorn (Crataegus viridis L. [Rosaceae]), which is a new nymphal record for L. delicatula. We suggest further investigation into whether these common, native understory trees are true feeding hosts, potentially by using DNA gut analysis (Avanesyan & Lamp 2020) and by assessing the frequency of nymphal feeding on these trees at other sites.

Table 3.

Risk assessment and counts of trees in partial tree inventories at the McDaniel College campus in Westminster, Maryland, and the McDaniel Environmental Center (MEC) in New Windsor, Maryland. Risk categories are based on literature reports (Known risk: Dara et al. 2015; Barringer & Ciafré 2020; Liu et al. 2020), new data from the current study (Of new concern), or no known reports (Unknown risk) for Lycorma delicatula egg masses and nymphs.

Fig. 4.

In the Pennsylvania field surveys, egg mass abundance varied greatly (A) but did not significantly depend on plant species, whereas nymphal abundance did (B). Plotted are mean (± 1 SE) maximum egg mass abundance per plant and mean nymphal abundance per plant, as in Figure 3. Abbreviations: ACE (Acer saccharum), AIL (Ailanthus altissima), AME (Amelanchier sp.), CER (Cercis canadensis), CRA (Crataegus viridis), FAG (Fagus grandifolia), JUG (Juglans nigra), LIR (Liriodendron tulipifera), PIN (Pinus strobus), ROS (Rosa sp.), RUB (Rubus spp.), TET (Tetradium daniellii), VIT (Vitis sp.).

Egg mass abundance also varied considerably among individual plants in our study (Fig. 4A). In contrast to the clear differences in nymphal abundances among plant species, egg mass abundances did not vary significantly among plant species. This lack of influence of host plant species on oviposition patterns is perhaps unsurprising given the incorporation of both living and non-living substrates for oviposition (Urban 2019). In addition, egg mass abundances may be influenced by factors we did not assess directly in this study. Notably, sites varied in both plant composition, abundance, and invasion history. Additionally, females may oviposit on plants or structures based on proximity to adult feeding host plants (Liu 2020). We do, however, report here the use of 3 plant species not previously known as oviposition substrates: C. canadensis, C. viridis, and L. styraciflua (Table 2). As with the nymphal assessment above, we suggest these 3 species, all which are common in eastern forests and regularly used as valued native ornamental species in landscape and street tree plantings, be examined at additional sites to evaluate whether their use as oviposition substrates is widespread in the invaded US range.

As more information about oviposition and host plant use for L. delicatula becomes available, more effective strategies for surveying and management may become clear. Our work reinforces that targeted investigations on potentially important nymphal hosts during the summer and A. altissima and other key aggregation hosts for adults in the fall will be important (Lee et al. 2019; Liu 2020). We also suggest that plant size be taken into consideration for oviposition surveys in the future. In our study, egg mass abundance was an increasing function of plant size (Fig. 3A), with the cumulative use of large trees over time much higher than the use of smaller trees, shrubs, and vines. In habitats with trees of varying sizes, large trees would be clear targets for initial survey efforts. This strong, positive relationship between egg mass abundance and tree size also underscores that herbivorous insects may respond to plant size and other traits associated with resource availability, despite variation in phylogenetic relatedness and secondary chemistry among available plants (Marques et al. 2000; Whitfeld et al. 2012; Castellis et al. 2017).

Based on information from the literature and our Pennsylvania field surveys, we then assessed the extent to which plant communities in nearby, currently uninvaded sites may be at risk for use by L. delicatula. At the campus site, our inventory included a diverse set of native and non-native ornamental and wild trees. Similar to studies of risk in natural habitats (Dara et al. 2015), most of these managed trees can be used by L. delicatula for feeding or oviposition. Unfortunately, though diverse plantings in campuses and other urban forest settings can help mitigate risks caused by any 1 new pest or pathogen (Raupp et al. 2006), the highly generalized feeding and oviposition habits of L. delicatula represent formidable problems for management in both natural and managed settings.

Notably, our tree inventory at the semi-natural site revealed the near ubiquity of at-risk plants, largely because of the dominance of A. altissima, J. nigra, and A. saccharum (Table 3). It should be noted that our tree inventory at the semi-natural site was a partial inventory emphasizing the composition of the forest edge habitats. Thus, the dominance of edge-associated species like A. altissima is higher than would be representative of the forest overall. Nevertheless, the high abundance of A. altissima in forest edge habitats in our local site and in the mid-Atlantic region overall (McAvoy et al. 2012; Slodonja et al. 2015) indicates that such habitats may be hit hard by L. delicatula as it spreads through the region, especially with recent documentation of edge effects in the abundance of L. delicatula (Leach & Leach 2020b). Management strategies should include development of tree inventories, monitoring dense stands of A. altissima in areas adjacent to established L. delicatula populations, and removal of A. altissima where possible (Urban 2019).