Worldwide Feeding Host Plants of Spotted Lanternfly, With Significant Additions From North America

Abstract

The spotted lanternfly Lycorma delicatula (White) is an invasive insect spreading throughout southeast Asia and eastern North America. The rapid spread of this species is facilitated by the prevalence of its preferred host, tree of heaven (Ailanthus altissima (Mill.) Swingle), as well as its use of many other host plants. While the spotted lanternfly has been previously reported to use over 65 plant species, most of these reports are from Asia and may not be applicable in North America. Additionally, many of the known hosts have not been specified as feeding hosts or as egg laying substrates. To better understand the potential impacts of this invasive insect on natural and cultivated systems in North America, we reviewed records from published and unpublished results and observations of host plant use by spotted lanternfly. We aggregated 172 host plant records worldwide and found feeding behaviors associated with 103 plant taxa across 33 families and 17 orders, 20 of which were not previously known to be associated with SLF and 15 of which were not confirmed as feeding hosts. North American records account for 56 of these taxa which include native, cultivated, and nonnative species. As a result, the spotted lanternfly has the potential to impact a wide assortment of ecosystems throughout its potential range and its North American distribution may not be limited by the presence of tree of heaven.

host plant, polyphagous, invasive species

The spotted lanternfly (SLF), Lycorma delicatula (White), is a planthopper (Hemiptera: Fulgoridae) native to northern China (Chu 1930). Within the past 20 yr this species has spread through southeast Asia (Han et al. 2008, Kim et al. 2013) and, most recently, to the eastern United States (Barringer et al. 2015). Since its discovery in Pennsylvania in 2014, the North American range of SLF has expanded to include five other northeastern states (Gallagher 2019, Harper et al. 2019). Reports from southeast Asia document similarly rapid proliferation of SLF within Korea (Han et al. 2008, Lee et al. 2011) and Japan (Kim et al. 2013, Tomisawa et al. 2013). This rapid spread may be partly due to SLF’s apparent capacity for long-distance dispersal (Park et al. 2013) and potential to occupy a broad range of climactic conditions (Jung et al. 2017, Wakie et al. 2019).

The expansion of SLF is also likely facilitated by the presence and prevalence of its preferred host, tree of heaven (Ailanthus altissima (Mill.) Swingle [Sapindales: Simaroubaceae]), on all continents with the exception of Antarctica (Kowarik and Säumel 2007). Feeding on A. altissima may potentially allow for the uptake of quassinoid compounds by SLF as a chemical defense against avian predators (Song et al. 2018). SLF may not require A. altissima to complete its life cycle, however, as it is polyphagous feeder especially in early instars (Kim et al. 2011). As such it has been reported to use over 65 plant species in various capacities worldwide (Park et al. 2009, Dara et al. 2015, Lee et al. 2019) which likely contributes to its successful spread.

SLF feed exclusively on phloem tissues from nymph to adult instars (Chu 1930). Mass feeding, which often occurs later in the season, can result in oozing trunk wounds, wilting, branch death, sooty mold (Ding et al. 2006, Han et al. 2008). Grape (Vitis spp. L. [Vitales: Vitaceae]) crops have been particularly impacted by this damage which has decreased vine growth, lessened winter survivability, reduced crop yield by up to 90%, and killed vines outright (Song 2010, Urban 2020). While the potential economic impacts of damage on agriculture and forestry in Pennsylvania have already been assessed (Harper et al. 2019), the extent and effects of herbivory on other cultivated and wild plant taxa in North American are still little known. Here we review the published literature and aggregate unpublished results and observations to present an updated host list for this rapidly expanding invasive species.

Materials and Methods

SLF host records were divided into four categories by source: Literature records, laboratory studies, field studies, and field observations. Laboratory studies, field studies, and field observations were collected from May 2015 to November 2019. All taxa were compiled and separated by record origin, life history stage, and plant use (if specified). Growth habit for potential host taxa are also presented and follow those used in the online USDA PLANTS database where available (USDA 2020).

Literature records were compiled from a review of published materials including those that have been accepted for publishing at time of writing, including preprints. Review articles (EPPO 2016) were included only when the primary source for a host taxon was unable to be obtained or translated. In many cases life stage was not given for literature records. In these cases the reader may infer life stage based on the habit of the feeding host as it is unlikely that small herbaceous plants can support feeding by adult SLF (Song 2010). Some literature records may also be unreliable as they often did not differentiate resting, tasting, and feeding behaviors, or they implied that feeding occurred by use of ambiguous language (e.g., ‘depend on’ used in Tomisawa et al. 2013) or by reporting damage to the host (e.g., Park et al. 2009). The accuracy of host identification was also a potential issue. Sequencing of SLF gut contents by Avanesyan and Lamp (2020) yielded an 84–100% match with plant species in the NCBI GenBank database; in this case, we chose to conservatively include only species that they had confidently identified (i.e., 100% match).

Laboratory and field studies were compiled from ongoing or unpublished research projects either in controlled laboratory trials or in an outdoor setting under a variety of conditions. Studies ranged from unlimited to singular host availability. In conditions where host choice was limited, survival to at least one subsequent instar was required to demonstrate at least partial feeding host suitability for SLF. Finally, field observations were firsthand accounts made by those trained to recognize SLF feeding behavior. These records were primarily personal communications with researchers, regulatory inspection staff at the Pennsylvania Department of Agriculture (PDA), SLF survey and control personnel of PDA and the United States Department of Agriculture Animal and Plant Health Inspection Survey (USDA APHIS), and other collaborators. SLF is known to move among plants to find a suitable host both as a nymph and as an adult (Baker et al. 2019), so efforts to distinguish between resting and feeding were necessary. Anecdotal reports lacking observations or photographs of explicit feeding behavior (inserted mouthparts, feeding posture, and/or presence of honeydew) were excluded from consideration.

Results

In total, 172 plant host taxa in 100 genera (with 22 to genus only), 51 families, and 27 orders were compiled and organized by source type (Table 1). Field observations, lab observations, and lab studies contributed 46 feeding host taxa, 17 of which were not previously reported by the literature as a host plant for SLF in any capacity. Across all examined sources, feeding behavior was attributed to 103 host taxa worldwide in 33 families and 17 orders. Fifty-six of these records report feeding on these taxa in North America.

Table 1.

Worldwide plant usage list of Lycorma delicatula

Order, family, species, and common names are given for each host taxon. Usage of the taxon is divided by source (field observations, field studies, laboratory studies, and published literature) and further by life stage (egg, nymph, adult). N/A is used where life stage is not provided; nymph/adult is used when life stage is not given but it is clear that use was posthatching. Citations are given for each record in the published literature. The national origin of each record is also given except for field observations and studies; United States state abbreviations are provided instead for these as they occurred outdoors in situ. Abbreviations used are: NJ (New Jersey), PA (Pennsylvania), VA (Virginia). Feeding hosts are indicated in light gray (feeding already known from the literature) and dark gray (feeding not known from the literature); taxa in white have no specified support for feeding behavior at time of publishing. Bold text indicates a source (for literature records) or location (for field observations, field studies, and lab studies) indicating feeding use; those that are not bold were either associated with egg deposition or did not specify usage.

Worldwide, nymphs were reported feeding on 81 host taxa, only six of which were strictly herbaceous taxa: Angelica dahurica (Fisch.ex Hoffm.) (Apiales: Apiaceae), Aralia cordata Thunb. (Apiales: Araliaceae), Arctium lappa L. (Asterales: Asteraceae), Armoracia rusticana G. Gaertn, B.Mey. & Scherb. (Brassicales: Brassicaceae), Monarda sp. L. (Lamiales: Lamiaceae), and Ocimum basilicum L. (Lamiales: Lamiaceae). Adult feeding was reported for 47 host taxa, all of which were woody, nonherbaceous, or vining. Eight taxa were reported to have feeding behaviors associated with them but without a specified life history stage from any sources. Twenty-eight (60.0%) of the reported feeding host taxa were also documented as egg deposition substrates. Nineteen taxa were exclusively documented as egg deposition substrates. All taxa used as egg deposition substrates had shrubby or tree-like habits with the exception of Vitis labrusca L. (Vitales: Vitaceae) (a woody vine). Thirty of the 172 listed taxa were not associated with any specific SLF life history stage or behavior; these account for 17.4% of all of the reported taxa.

Some taxa were reported to have been explicitly presented to SLF for feeding but not were fed upon despite tasting attempts and, therefore, were not included as hosts. These included Pyrus calleryana Decne. (Rosales: Rosaceae), Hibiscus syriacus L. (Malvales: Malvaceae), and Pinus densiflora Siebold & Zucc. (Pinales: Pinaceae) (Lee et al. 2009). Prunus persica (L.) (Rosales: Rosaceae) and Malus pumila Mill. (Rosales: Rosaceae) were also not fed upon following tasting in this study; however, other sources have reported feeding on these species.

Discussion

In the 6 yr since its discovery in North America, SLF has been observed feeding on both Eurasian and native species (Barringer et al. 2015). While 38 plant species were reported in the literature as host plants for SLF in the United States, the vast majority of these records are limited to egg deposition (Dara et al. 2015, Liu and Hartlieb 2020), and only 10 taxa were documented as feeding hosts (Barringer et al. 2015, Baker et al. 2019). Through field observations, ongoing research, and recent publications, SLF is now reported to feed on at least 56 taxa in North America, increasing the known worldwide feeding taxa to 103. The data we compiled contributed 20 plant taxa not previously known to be used by SLF in any capacity and an additional 15 taxa that had not yet been confirmed as feeding hosts.

Given the relatively few records of feeding on herbaceous taxa, SLF may prefer hosts with a particular structure. SLF are known to exhibit ontogenetic host partitioning in which nymphs are typically restricted to leaves, petioles, branches, and young stems while adults are often found closer to or on the trunk (Chu 1930, Tomisawa et al. 2013). This shift is supported by developmental changes in SLF tarsi and arolia that may allow adults to physically attach to a wider range of substrates (Kim et al. 2011, Avanesyan et al. 2019). The development of the mouthparts also likely plays a role in host partitioning as the stylets and labium of nymphs are much smaller than adults and may be too short to penetrate thick bark (Avanesyan et al. 2019). Host partitioning may also be indicative of structural preferences, potentially to hide from predators during early instars when they may lack aposematic coloration (Song et al. 2018, Domingue and Baker 2019), to prevent desiccation during the hottest parts of the day (Chu 1930, Lieu 1934), or to prevent physical displacement from the host by the wind (Kim et al. 2011). Branch structure may also play a role in the selection of sites for egg deposition as well (Liu 2019).

SLF may also likely prefer hosts with higher feeding quality such as hosts with greater available sap. Researchers attempting to cultivate SLF colonies have reported that small numbers of adult SLF can quickly wilt cuttings of woody materials (Song 2010). It is, therefore, unlikely that small herbaceous plants would support feeding by more than a few individuals, especially at later instars. Sap content also likely influences feeding preference in SLF. For example, SLF has been shown to prefer feeding on hosts with particular sugar content ratios and survived longer when fed sugar solutions similar to those produced by preferred hosts (A. altissima and V. vinifera L. [Vitales: Vitaceae]) (Lee et al. 2009).

SLF may also seek out feeding hosts that produce defensive chemicals. SLF is already known to sequester ailanthone, a bitter quassinoid produced by A. altissima to make it distasteful to predators (Song et al. 2018). Limonoids are bitter compounds related to quassinoids and are produced by plants in Meliaceae, Rutaceae, and Simaroubaceae (Roy and Saraf 2006). Other large, colorful Fulgorid planthopper species are known to feed on these families (Johnson and Foster 1986) potentially to sequester limonoids from them to make their bright coloration an honest aposematic signal. In addition to A. altissima (Simaroubaceae), SLF has been reported to have a particular affinity toward feeding on members of these three families including Pricrasma quassioides (D. Don) Benn. (Sapindales: Simaroubaceae), Melia azedarach L. (Sapindales: Meliaceae), Celedra fissilis Vell. (Sapindales: Meliaceae), Toona sinensis (A. Juss.) M.Roem. (Spindales: Meliaceae), Phellodendron amurense Rupr. (Sapindales: Rutaceae), Tetradium daniellii (Benn.) (Sapindales: Rutaceae), and Zanthoxylum simulans Hance (Sapindales: Rutaceae) (Park et al. 2009, Tomisawa et al. 2013). Many of these species are used in Chinese medicine, supporting the presence of limonoids or other toxic secondary metabolites that SLF may seek to utilize for defense (Kim et al. 2011) in addition to, or in the absence of, A. altissima.

Other host plants presented here may also have the potential to provide other secondary metabolites for sequestration by SLF. For example, SLF appears to feed on many members of Juglandaeae (Carya Nutt. [Fagales: Juglandaceae], Juglans L. [Fagales: Juglandaceae] and Pterocarya spp. Nutt. ex Moq. [Fagales: Juglandaceae]), a family known for its production of juglone (5-hydroxy-l,4-naphthoquinone) (Rietveld 1983). Juglone and other naphthoquinones are feeding deterrents which are toxic to many insects, though some taxa appear to be tolerant to moderate concentrations of these compounds (Piskorski et al. 2011). Furthermore, there is some evidence for the use of secreted juglone by Gastrolina depressa Baly (Coleoptera: Chrysomelidae) to repel ants and other small predators (Matsuda and Sugawara 1980). Other SLF host taxa also show such potential for defensive compound sequestration, such as those in Betulaceae (Betula L. [Fagales: Betulaceae], Corylus L. (Fagales: Betulaceae), and Ostrya spp. Scop. [Fagales: Betulaceae]) and Salicaceae (Populus L. [Malpighiales: Salicaceae] and Salix spp. L. [Malpighiales: Salicaceae]) (Matsuda and Sugawara 1980, Opitz and Müller 2009).

While care was taken to ensure feeding behaviors were associated with the field observations presented here, there is still some question as to whether those taxa are substantive feeding hosts. SLF is known to insert mouthparts into host xylem for up to 30 min without subsequently feeding on phloem (Lee et al. 2009). If these tasting behaviors can last this long, observations of feeding behaviors are likely not sufficient to fully designate a plant as a feeding host. We can only confirm that true feeding occurred in hosts recognized in the presented field and lab studies because they required survival to successive instar to qualify for reporting in those categories.

To address this issue, researchers recently developed methods to identify genetic sequences found in SLF gut contents (Avanesyan and Lamp 2020). While molecular sequencing has the potential to be a powerful tool in confirming feeding hosts in SLF, species-level identifications may not be sufficiently accurate to identify new ones at this time. Some species tentatively identified by Avanesyan and Lamp, both from SLF guts and collected plant material, are unlikely to occur in the northeastern United States where their study took place. These sequences had as low as an 84% match to those in the NCBI GenBank Database suggesting either geographic or taxonomic gaps present in the current data. Until these gaps are addressed, their method could be used to confirm suspected feeding hosts by comparing gut content sequences to local plant material that has been morphologically identified by a qualified botanist.

SLF has the potential to cause short- and long-term damage to both natural and cultivated systems. In addition to the projected economic impacts reported by other publications, the damage it causes by its heavy feeding will likely be an additional stressor to ecosystems already threatened by climate change, land use changes, and introduced species. As SLF is continuing to encounter new feeding hosts as it spreads across North America, the full host range of this species remains unknown. Many of the taxa fed upon in Asia also occur in North America where feeding has not yet been observed. It is becoming clear that the potential feeding host range of SLF is extremely broad. Moving forward, differentiating feeding, tasting, and feeding behaviors and identifying preferred characteristics of hosts will be critical toward estimating the larger impacts of SLF. As such, research and mitigation techniques will have to adapt and evolve to deal with this rapidly expanding invasive species.

Acknowledgments

We thank the multiple groups and researchers who provided first-hand accounts, reports, and preliminary data to the host list. These individuals and groups, in no particular order, include PDA plant inspectors, the State of Virginia, Dr. Charles Bartlett, Paul Kurtz, Trilby Libhart, Dr. Alberto Marcon, Dr. Kelly Hoover, PDA SLF field staff, Dr. Dennis Calvin, Dr. Miriam Cooperband, Dr. Chris Tipping, Longwood Gardens, and Dr. Houping Liu. Thanks are also owed to Jay Losiewicz for revisions and to the reviewers of this manuscript.