Introduction

Cities are major introduction sites for invasive species due to their role as hubs for global trade and travel (Gotzek et al. 2012; Padayachee et al. 2017; Wilson et al 2009). Once established, invasive species can spread outside urban centers and cause serious ecological and economic destruction (Diagne et al. 2021; Singh et al. 2021; Turbelin et al. 2023). By the time introduced species gain enough visibility to be discovered in cities, however, they have often already reached pest status. Even when species are noticed early, it takes time to make appropriate identification, and it is often too late to mitigate the impact of an invasive species or its spread. Despite our knowledge that cities are hotspots for species introductions, there are few, if any, policies in place to identify or monitor species before they reach pest status (Simberloff et al 2005). Though some monitoring efforts do exist, introduced species can still slip past undetected (Menchetti et al 2024; Schifani 2019; Wong et al 2023). This makes it difficult to catch problem species before they spread and become even more difficult to control.

In 2011, an unidentified ant was discovered in the heart of New York City, which made national headlines and produced the memorable nickname “ManhattAnt” (Karni 2012; Nuwer 2012). Despite initial media attention, there has been little to no follow up work on the ManhattAnt nor has it received an official identification. The species was noticeably absent during the first survey of New York City ants in 2009 (Pećarević et al. 2010), and its populations were relatively small and confined to forested parks when they were first collected in 2011 as part of an extensive ant community survey (Savage et al. 2015). Over the next 5 years, however, their populations expanded to become among the most common ants in both parks and highly urban habitats in the city (Penick et al. 2015). New York City has been the site of introduction for some of North America’s most damaging invasive pests—both chestnut blight and the Asian long-horned beetle were first identified in New York City, which are together responsible for the deaths of an estimated 4 billion trees (Hepting 1974; Jacobs et al. 2016; Nowak et al. 2001). Therefore, it is essential that emerging pests, such as the ManhattAnt, are correctly identified and monitored after they are first discovered.

Past research on invasive species has been hampered by uncertainty surrounding species identity and their source populations (Gotzek et al. 2012). It is estimated that the number of ant species introduced and established is much higher than those discovered, and even more so that those that have been accurately identified (Miravete et al. 2014). For ants and other insects, accurate identification can be difficult due to the diversity of insects globally and high resemblance among closely related species. This contributed to confusion around the identity of a recently discovered pest ant in Houston, TX, which was initially dubbed the “Rasberry Crazy Ant” after the exterminator who brought it to widespread attention, Tom Rasberry (Gotzek et al. 2012). Similar to the ManhattAnt, the Rasberry Crazy Ant’s rapid expansion, potential for ecological harm, and initial lack of identity gained it media notoriety. After several years of confusion surrounding its identity, researchers eventually combined morphological and genetic approaches to identify it as Nylanderia fulva, a species that had been documented in Texas decades before and had been known previously called the “Tawny Crazy Ant” (Gotzek et al. 2012; Schär et al. 2022).

Beyond clarifying the taxonomy of introduced species, it is also important to identify key characteristics that may contribute to their successful establishment. Invasive ants often share a number of common traits that contribute to their success, especially their ability to form supercolonies that display little to no aggression between neighboring nests (Chapman and Bourke 2001; Holway et al. 2002; Holway and Case 2000; Hölldobler and Wilson 1977; McGlynn 1999; Seifert 2018). Many of the most widespread invasive ant species exhibit supercoloniality within their invasive range, including Argentine ants (Linepithema humile) (Giraud et al. 2002; Heller 2004; Tsutsui and Case 2001), Asian needle ants (Brachyponera chinensis) (Warren et al. 2020), and the previously mentioned Rasberry Crazy ants (Nylanderia fulva) (Eyer et al 2018). This ability to exhibit supercoloniality allows introduced species to establish in high densities and dominate introduced habitats. Increased colony size and reduced interspecific aggression within the population can increase a supercolony’s ability to compete for local resources and displace native ant species (Holway and Case 2001). However, supercoloniality is often associated with limited natural dispersal abilities, which limits spread in highly fragmented landscapes (Schultner et al. 2016; Sundström et al. 2005). Therefore, identifying the colony structure and other traits that can facilitate invasion success (e.g., diet) is necessary to assess the potential of an introduced species to become an established pest.

Here we clarify the taxonomy of the ManhattAnt in New York City using a combination of genetic and morphological approaches. The ManhattAnt has been tentatively identified as the European ant Lasius emarginatus (Savage et al. 2015) due to its characteristic red and black coloration (Fig. 1). However, L. emarginatus is part of a species complex and shares morphological similarities with closely related species found across Europe and the Middle East (Seifert 2020). By clarifying its taxonomy, we identify the region from which the ManhattAnt was introduced as well as use information about its natural history to assess how this species could impact its newly invaded habitat. Because many successful invasive ants form supercolonies, we use aggression trials to test whether the ManhattAnt is supercolonial or whether they exhibit aggression between neighboring nests. Finally, we use published accounts and citizen science observations to build a distribution map of the ManhattAnt in New York City and surrounding areas, and we estimate its rate of spread.

Fig. 1
figure 1

Lasius emarginatus in profile (A) and top (B) view (Photo by April Nobile, specimen code CASENT0172762, from www.antweb.org) (AntWeb 2023)

Full size image

Methods

Field collection

We collected workers of the unidentified Lasius ant in Manhattan, New York City (NY, USA) during May 2022 from four separate locations covering an area of 1.6 km2 (Table S1). We stored specimens in 70% ethanol for genetic and morphological analyses to confirm the identity of this species. During field collection, we made targeted observations of Lasius colonies throughout the city to document nesting sites (i.e., underneath sidewalks) and foraging behavior. We also recorded any observation of interactions with co-occurring ant species.

Species identification

Accurate identification of new pest insects often requires a combination of genetic and morphological approaches. Genetic methods for identifying newly introduced species often rely on DNA barcoding of the COI gene (Gotzek et al. 2012; Rasool et al. 2020; Schär et al. 2022), but COI barcoding may be insufficient on its own to provide a positive identification. This is due, in part, to a lack of COI sequence data available for most of the world’s insect species. In cases where DNA barcoding fails to provide a sufficient match, morphological approaches can be useful to compare species to those in museum collections. One approach called numeric morphology-based alpha-taxonomy (NUMOBAT) has been successful to discriminate among closely related ant species that form cryptic species complexes (Centanni et al. 2022; Seifert 2009, 2019, 2020; Seifert and Galkowski 2016; Seifert 2009). NUMOBAT uses 15 unique morphological characters (linear or angular measurements and counts of elements) to compare specimens, and it can be used when COI data is unavailable or when DNA extraction is not possible. The benefit of using both approaches together is that, while both have their limitations, they can be used in combination to build confidence in species identifications and reduce misidentifications that can slow management decisions.

Genetic analysis

We amplified a portion of the mitochondrial cytochrome c oxidase subunit I gene (COI), which is the barcoding locus most often used to identify insects (Hebert et al. 2003). To extract DNA, we placed individual ants into separate plastic weigh boats and used a sterilized razor blade to break apart the tissue. We then extracted DNA using the QIAGEN DNeasy Blood and Tissue Kit (QIAGEN, Hilden, Germany) following manufacturer's instructions. To maximize compatibility with existing barcoding data, we used the primers LepF1 and LepR1 (Hebert et al 2004). We prepared reactions for two DNA extracts (ST009A and ST009B), and we tested two different annealing temperatures (45 °C and 52 °C) to optimize PCR conditions. We used Invitrogen Platinum SuperFi reagents (Thermo Fisher, Waltham, Massachusetts, USA), and each 25 µL reaction included: 5 µL 5 × buffer, 0.5 µL dNTPs, 0.25 µL polymerase, 13.75 µL nuclease-free water, 1.25 µL 10 µM LepF1, 1.25 µL 10 µM LepR1, and 3 µL extracted DNA. We used the following thermal cycler conditions: 98 °C for 30 s; then 40 cycles of 98 °C for 10 s, 45 °C or 52 °C for 10 s, and 72 °C for 20 s; then 72 °C for 5 min. We then purified PCR products with a 2:1 SpeedBead:DNA product volume ratio (Rohland and Reich 2012), validated them on a 1% agarose gel, and quantified them using a Qubit 4 Fluorometer (Thermo Fisher, Waltham, Massachusetts, USA). We then pooled approximately 250 ng of purified PCR product from one reaction (ST009B; 52 °C annealing temperature) with other amplicons from unrelated projects for library preparation.

To prepare amplicons for sequencing, we used a Ligation Sequencing Kit (SQK-LSK112) from Oxford Nanopore (Oxford, UK) and followed the manufacturer’s protocol. We quantified the final libraries and sequenced them on a MinION Flow Cell v. r10 (FLO-MIN112). We used the live, fast basecaller in Guppy v6.3.8 through MinKNOW v22.10.7 (Oxford Nanopore Technologies, Oxford, UK) and saved the first 804,000 reads for downstream analysis. We then used medaka v1.7.3 (Oxford Nanopore Technologies, Oxford, UK) to align reads to a COI sequence from L. umbratus (median depth =  > 8000 reads) and to create and polish a consensus sequence from these aligned reads. Our raw sequencing data are available through the NCBI SRA (PRJNA977684), and our consensus sequence is available through NCBI GenBank (OR063821).

We used two approaches to identify the New York Lasius sp. First, we used an NCBI BLAST query (Altschul et al. 1990) for highly similar sequences using default parameters. Second, we used the “Animal Identification” tool to search the “Public Record Barcode Database” on the Barcode of Life Database (BOLD; Ratnasingham and Hebert 2007). For both approaches, we downloaded the 100 top matches from each search. Separately for each dataset, we added the consensus sequence from our New York Lasius sp. and a sequence from Lasius umbratus (GenBank accession = MZ610655) for use as an outgroup. We then aligned sequences using MUSCLE v3.8.31 (Edgar 2004) within AliView v1.28 (Larsson 2014). We then used RAxML-NG v1.1.0 (Kozlov et al. 2019) to estimate maximum-likelihood phylogenies for each alignment. We used a GTR + GAMMA substitution model, 10 random + 10 parsimony-based starting trees, and used 1000 bootstrap replicates to assess confidence in relationships.

Morphometric analysis

We measured a total of 15 Numeric Morphology-Based Alpha-Taxonomy (NUMOBAT) characters (Table 1) (Seifert 2020) of three worker specimens from three nest samples collected in Manhattan in 2022 to compare with four closely related species: Lasius emarginatus (Olivier 1792) from Europe; L. illyricus Zimmermann 1935 from the Balkans, Asia Minor and Caucasus; Lasius tebessae Seifert 1992 from Morocco and Algeria; and Lasius maltaeus Seifert 2020 from Malta. To measure NUMOBAT characters of Lasius specimens collected in Manhattan, each ant was point-mounted and observed under a high-performance stereomicroscope (Leica M165C) fitted with an a 2.0 × planapochromatic objective (resolution 1050 lines/mm) at 120–360 × magnification. We then compared measurements to 292 Lasius specimens representing the four species described above, including the type specimens for all species (holotype, neotype, lectotype, and paratypes; Table S2). We deposited vouchers from all morphometrically examined Lasius workers collected from New York City in the Senckenberg Museum für Naturkunde Görlitz, Germany, and we deposited vouchers from nests used for molecular analyses in the Auburn Museum of Natural History at Auburn University, AL, USA.

Table 1 Description of 15 Numeric morphology-based alpha-taxonomy (NUMOBAT) characters measured for each Lasius specimen
Full size table

Distribution mapping

To map the distribution of the ManhattAnt in New York City and surrounding areas, we combined previous data from ant surveys in New York City (Penick et al. 2015; Savage et al. 2015) with community reported data on iNaturalist (GBFI.org 2023; iNaturalist 2023). The introduced Lasius sp. is relatively easy to identify in photographs due to its size (3–5 mm) and its characteristic black and red coloration, which is not found on other co-occurring ants. We individually verified research-grade photos reported to iNaturalist as L. emarginatus. In addition, we launched an iNaturalist citizen science project, Project ManhattAnt, to map the continued expansion of the ManhattAnt in the northeastern United States.

Aggression trials and colony structure

To determine colony structure and assess whether New York City Lasius sp. exhibited supercoloniality, we conducted aggression tests between workers collected from the same or different nests. We collected workers from 19 individual nests from sidewalks and street trees in New York City (Table S1) and stored them in 1.5 mL vials until the start of aggression trials. We randomly selected an individual worker from two nests (or two from a single nest for control trials) and placed them each into a plastic 180 mL vial (5.08 cm × 10.77 cm) and observed them for 15 min to quantify aggression between the two workers. We quantified aggression as the behaviors of biting one another or formic acid spraying. In addition, we quantified rapid antennation and mandible gaping as potentially non-aggressive behaviors as they are typically scored low in aggression trials (Giraud et al. 2002; Jelley and Moreau 2023; Suarez et al. 1999) We conducted trials blind so that the observer did not know whether a trial was between ants taken from the same or different nests (Wilgenburg and Elgar 2013) and we ran each trial within 24-h of ant collection.

Statistical analyses

For our morphometric analysis, we used a linear discriminant analysis (LDA) to test the probability that the Lasius specimens collected from New York City were members of a known species using the software package SPSS 15.0. We ran the New York samples as “wild-cards” (i.e. without imposing a species hypothesis) against the background of Lasius emarginatus, L. illyricus Zimmermann, L. tebessae Seifert 1992 and L. maltaeus. The LDA was run with the morphological characters corrected for allometric variance according to the algorithms given in Seifert (2020). For aggression trials, we used Fisher’s exact test using R version 4.3.2 (R Core Team 2023) to compare the percentage of trials during which any aggressive encounter was observed (biting or formic acid spraying) to test whether aggression was significantly more common in between-nest trials compared to within-nest control trials. If colonies displayed supercoloniality, we predicted that we would observe little to no aggression among trials and that we would not observe a significant difference in aggression for between-nest or within-nest trials. If colonies were not supercolonial, then we predicted that we would observe significantly higher aggression in between-nest trials compared to within nest-trials.

Results

Species identification

Our genetic and morphological analyses supported identification of the New York ManhattAnt as Lasius emarginatus, a European species. We amplified a portion of the mitochondrial cytochrome c oxidase subunit I gene (COI), the barcoding locus used most often to identify insects. In the phylogeny that included samples from GenBank, we found strong support (bootstrap support = 88%) for the placement of the New York specimens in a clade that included only L. emarginatus and two specimens identified only as L. sp. In the phylogeny that included samples from the Barcode of Life Data System (BOLD) we found similarly strong support (bootstrap support = 90%) for the placement of New York Lasius specimens in a clade that included only L. emarginatus, one sample identified only as L. sp., and one sample identified as L. brunneus (Fig. 2). Genetic data were not available for the other three members of the L. emarginatus species complex.

Fig. 2
figure 2

Maximum-likelihood phylogenies inferred from cytochrome c oxidase subunit I (COI) barcoding sequences, with pie charts at nodes showing bootstrap support for edges. A Phylogeny inferred from a COI sequence from a New York City Lasius sp., its 100 closest matches from an NCBI BLAST search, and an outgroup of Lasius umbratus. B Phylogeny inferred from a COI sequence from a New York City Lasius sp., its 100 closest matches from a Barcode of Life Database search, and an outgroup of Lasius umbratus. The inset in each panel shows the smaller clade in which the New York City specimens are nested

Full size image

Due to a lack of COI data from other closely related Lasius species, we could not resolve the species identity using genetic data alone. Therefore, we conducted a morphological analysis using 15 allometrically corrected NUMOBAT characters that we compared with morphological measurements taken from L. emarginatus and three other closely related species (L. illyricus, L. maltaeus, and L. tebessae). We ran a linear discriminant analysis (LDA) comparing three New York specimens against the four members of the L. emarginatus species complex with the New York specimens run as “wild-cards” (i.e., without imposing a species hypothesis). There was over a 93% posterior probability that the New York specimens were L. emarginatus compared to less than 7% match with the other three species in the L. emarginatus complex (Table 2). We then ran a cluster analysis with L. malteus excluded due to a zero percent probability match, which showed that the New York specimens clustered within the European L. emarginatus and showed that they could not be allocated to the northwest African L. tebessae nor to the L. illyricus from the southern Balkans, Asia Minor, or the Middle East (Fig. 3).

Table 2 Posterior probabilities for allocation of New York City Lasius sp. to four closely related species based on 15 NUMOBAT characters using a linear discriminant analysis (LDA) with New York specimens run as wild-cards
Full size table
Fig. 3
figure 3

Linear discriminant analysis (LDA) means based on 15 NUMOBAT characters measured for European Lasius emarginatus (white rhombs, n = 60), L. illyricus (black triangles, n = 39), L. tebessae (black dots, n = 6), and Lasius sp. collected from New York City (red rhombs, n = 3). New York City specimens are nested within European L. emarginatus

Full size image

Distribution and spread

L. emarginatus was initially absent from a survey of ant diversity conducted in New York City in 2006 (Pećarević et al. 2010) but appeared in 33% of park sites and 10% of traffic island sites during a sample in 2011 (Savage et al. 2015). The percent of sites in which L. emarginatus were collected in studies of New York City ants increased to 42% in park sites and 42% in traffic island sites by 2013 (Penick et al. 2015). In addition to formal sampling efforts, there have been over 570 sightings reported to iNaturalist from 2014 to 2023 (GBFI.org 2023; iNaturalist 2023). From the initial discovery of L. emarginatus in Midtown and upper Manhattan, they have expanded to lower Manhattan and across the Hudson River into New Jersey (Fig. 4). When comparing the maximum range of L. emarginatus each year from 2011 to 2023, we found that their populations are expanding at a rate of 2 km per year, which is consistent with expansion through typical mating flights, though we cannot rule out anthropogenic dispersal.

Fig. 4
figure 4

Distribution of Lasius emarginatus in New York City, NY, USA. Polygons illustrate maximum range at 2-year intervals from 2017–2023 (A–D). Note, the first L. emarginatus specimens were collected in New York City in 2011. Map made using ArcGIS Pro (Version 3.0)

Full size image

Colony structure

Many introduced ants exhibit supercoloniality and display little aggression among neighboring colonies, but we found no evidence of supercoloniality in L. emarginatus in New York City. In dyadic trials between workers collected from the same or neighboring nests, we found a significant difference in aggression, quantified as performance of biting or formic acid spraying (Fisher’s exact test, Nsame = 10, Nneighboring = 10 p = 0.023, Fig. 5). Workers from neighboring nests displayed aggression in 80% of trials, while workers collected from the same nests displayed aggression in only 20% of trials. In addition to clearly aggressive behaviors, we also quantified incidences of mandible gaping and rapid antennation, which were performed in over 80% of trials regardless of treatment and did not significantly differ between workers paired from the same or neighboring nests (Fisher’s exact test, Nsame = 10, Nneighboring = 10 p = 0.47).

Fig. 5
figure 5

Bar graph showing the percent of trials in which aggression was observed (i.e., biting and/or formic acid spraying) between workers collected from either different nests or the same nest of L. emarginatus in New York City (Fisher’s exact test, Nsame = 10, Nneighboring = 10, p = 0.023)

Full size image

Natural history observations

During field work in New York City, we recorded observations of L. emarginatus nesting sites, foraging behavior, and interactions with co-occurring ants. We observed workers walking into and out of nesting holes in the ground, which is consistent with observations in Europe that indicate L. emarginatus is a ground-nesting species (Seifert 2018). We also observed what appeared to be satellite nests at the base of trees and in holes in trees more than 1 m off the ground in park and sidewalk habitats. More recently, we observed L. emarginatus workers on the balcony of a sixth-floor apartment in the Lower East Side, and there have been growing reports of L. emarginatus workers foraging inside buildings at ground level and above (Stewart 2022). One colony of L. emarginatus was even found nesting in a small flower planter in the middle of Times Square far away from street trees or large ornamental plants that would typically serve as nesting habitat. Often, the only visual non-humans in Times Square are humans dressed as Disney animals and yet, hidden in this flower pot was a whole society of animal beings.

In our observations of foraging, we saw L. emargaintus workers interacting with aphids on vegetation in city parks, and we observed them carrying small herbivorous insects, including scale insects, while foraging on street trees. We observed several occasions of L. emarginatus workers surrounding and dismembering workers of the second most common urban ant in New York City, Tetramorium immigrans (Penick 2021), another introduced species. We also observed L. emarginatus workers carrying some of their sister workers up and down trees. We found that L. emarginatus colonies foraged on street trees all hours of the night, even during and after rain events. They continued to forage on dry portions of smooth barked London plane (Platanus sp.) trees and seemed less deterred by the rain on coarser barked locust trees (Robinia sp.).

For New York City populations, L. emarginatus mating flights were observed flying in the evening 24 July 2020 through the morning of 25 July 2020 (Ivan Lacroix, personal communication). Other Lasius species were also observed flying that night, including L. murphyi. According to NOAA weather station data (NOAA 2023b) for July 2020, there was rain on the 22nd (3.6 cm, evening), 23rd (trace amount, sporadic), and 24th (0.5 cm, morning), with relatively dry conditions for the week preceding these dates and the week following. L. emarginatus queens were observed and reported to iNaturalist on 21 June 2020, 23 July 2022, 27 July 2022, and 4 August 2022. No rain events occurred on or immediately before 21 June 2020 or 4 August 2022, but there were trace amounts of rain on 22 July 2022 and 27 July 2022 (2.5 cm, afternoon, trace amount in the evening).

Discussion

Using genetic and morphological evidence, we have identified the “ManhattAnt” in New York City, NY (USA) as Lasius emarginatus, a species native to central and southern Europe. Genetic sequencing of the COI barcode gene placed the ManhattAnt within the genus Lasius and tentatively identified the species as L. emarginatus. Because COI data were lacking for other closely related Lasius species, we performed a morphological analysis using measurements from additional Lasius species, which confirmed the identity as L. emarginatus. While the ManhattAnt was first reported in New York City in 2011, this is the first confirmation of its identity, and it is the first documented population of L. emarginatus outside of its native range. Over the last decade, L. emarginatus populations have been expanding within and outside of New York City at a rate of 2 km per year, particularly within urban habitats where they have been reported as an emerging pest (Stewart 2022). This expansion rate is consistent with natural expansion following mating flights. Initial collections of L. emarginatus occurred mainly in parks, with occurrences in 33% of park sites and 10% of traffic islands along Broadway (Savage et al. 2015). A subsequent study conducted in 2013 found an expansion of L. emarginatus in Manhattan, particularly in highly urban areas, with occurrences in 42% of park sites and 42% of traffic islands along Broadway (Penick et al. 2015). Unlike other common invasive ants, we found no evidence that L. emarginatus exhibits supercoloniality, though their expanding range suggests they could have negative impacts on previously established ant species. In addition, they form mutualisms with honeydew-producing tree pests, which could increase stress on urban trees.

Successful management of emerging pests requires accurate and timely identification of newly introduced species, which can require both genetic and morphological approaches. We first used genetic sequencing of the COI barcode to place the ManhattAnt within the genus Lasius, and more specifically within a clade that included L. emarginatus. However, we acknowledge that this approach is limited by at least two issues: (1) mtDNA barcoding is unreliable as a primary method of species delimitation—a problem that is exacerbated by both biological (e.g., introgression) and methodological (e.g., misidentification) processes (Seifert 2018); and (2) the value of mtDNA barcoding depends upon the completeness of a reference database, and the barcoding data available for the L. emarginatus species complex are relatively sparse. In the future, as the amount of COI data available in public databases grows, this latter problem may be resolved. However, additional genomic resources (e.g., full-genome sequencing) will be necessary to fully explain evolutionary history of the L. emarginatus species complex, delimit species within it, and facilitate a more precise source of the introduced population in New York City. To account for the lack of barcoding data available for other members of the L. emarginatus species complex, we used morphological traits to distinguish among four closely related Lasius species found in Europe, the Middle East, and northern Africa and provide an identification of the ManhattAnt. Based on our analysis of 15 NUMOBAT characters, we were able to rule out other closely related species and identify the ManhattAnt as L. emarginatus with 93% certainty in the least clearly classified nest sample. In this case, genetic sequencing allowed us to quickly place the ManhattAnt within the L. emarginatus species complex, but detailed morphological measurements were needed to increase confidence in our identification.

The current distribution of L. emarginatus in Europe suggests their populations could find suitable habitat across a wide swath of the eastern United States. The range of L. emarginatus in Europe extends from 52.6° N in the north in southern England and the Netherlands to 37° N in the south across Iberia, Apennine, and the Balkans (Seifert 2018), which covers a span in mean annual temperature (MAT) from 9.2 to 18.8 °C (NOAA 2023a). Translating this to the eastern United States, the introduced range of L. emarginatus could extend as far north as Portland, ME (9.2 °C MAT) and as far south as Atlanta, GA (18.5 °C MAT) (NOAA 2023a). However, we caution that we do not yet know what other biotic or abiotic factors may limit expansion. While populations of L. emarginatus in southeastern Europe are primarily found in closed-canopy broadleaf forests, populations in central Europe are more common in open forests as well as urban and suburban habitats. In parts of southeastern and northeastern Germany, in particular, L. emarginatus lives in close contact with humans and can commonly be found in homes and churches. There have also been observations of L. emarginatus workers in Europe foraging inside buildings and feeding on sugary human food sources, though research on ant community diets in New York City suggests that introduced populations of L. emarginatus populations do not heavily feed on human foods (Penick et al. 2015). Many of the most destructive recent invasive ant species display supercoloniality with little aggression between neighboring colonies (Holoway et al. 2002), but we did not find evidence of supercoloniality in L. emarginatus. We observed aggression between ants collected from neighboring colonies in over 80% of trials, which suggests that colonies are independent. This differs from another invasive Lasius species found in Europe, L. neglectus, which forms massive supercolonies within its introduced range (Espadaler et al. 2007; Gippet et al. 2022; Ugelvig et al 2008)—one L. neglectus supercolony in Budapest was estimated to have more than 1010 workers and over 30 million queens, and it expanded in size at a rate of 89 m/yr over 17 years (Espadaler et al. 2007). Though supercoloniality is a common trait among successful invasives, there are successful invasive species that do not form supercolonies, such as monogynous fire ants, Solenopsis invicta, which are abundant throughout the southeastern United States, and pavement ants, Tetramorium immigrans, which are abundant in the northeastern United States, including New York City (Jelley and Moreau 2023).

Although supercoloniality reduces intraspecies competition and promotes larger colony sizes, there is a trade-off between supercoloniality and natural dispersal ability because supercolonial species often do not perform nuptial flights (Schultner et al. 2016; Sundström et al. 2005). In contrast, species that engage in aerial mating flights might successful dispersing within highly fragmented urban landscapes. In contrast, supercolonial species may depend on more passive dispersal through human activities to facilitate their spread (Schultner et al. 2016; Sundström et al. 2005). In the case of Tapinoma sessile, a species native to and extremely common in New York City, supercoloniality is a trait observed in urban environments and these colonies thrive in densely populated areas (Blumenfeld et al. 2022; Penick 2021). However, T. sessile can perform nuptial flights, and because they are native to the area may have maintained populations as the city developed around them. Supercoloniality is therefore not the only trait that should be considered when assessing invasion success in urban habitats, and there are likely other features that influence success in urban habitats.

In addition to greater natural dispersal ability of L. emarginatus, their success in highly urban habitats could also be related to adaptations to foraging in an open environment and their generalist diet supplemented by hemipteran-produced honeydew. Foragers of L. emarginatus have the fastest walking speed and the best visual system of any Lasius species occurring in western Palaearctic cities (Seifert 1986; Seifert 1992), which may facilitate their success in urban habitats with smooth stone or concrete surfaces. Based on studies within their native range, L. emarginatus tends to perform best in relatively open habitats with smooth surfaces with low spatial resistance, such as a mixture of concrete and stony surfaces that are common in cities (Seifert 2018). They typically forage long distances (up to 40 m) with faster recruitment rates compared to co-occurring Lasius (Seifert 2018). They also have a relatively broad diet, which may contribute to their success in urban habitats where generalists tend to perform better than those with more specialized diets (Penick et al. 2015). L. emarginatus exploits a wide range of resources, which includes hunting or scavenging dead insects as well as consuming plant sap, nectar, and elaiosomes found on seeds (Seifert 2018). Within New York in particular, L. emarginatus forms mutualisms with honeydew producing insects, such as aphids and scale insects, which are common in urban street trees and may provide an abundant a reliable food source (Meineke et al. 2013; Seifert 2018).

The impact of L. emarginatus on local ecosystems is not known, but they display a number of characteristics that could make them serious pests. First, they have reached high abundances in New York City and have been increasingly reported as pests inside human dwellings. In 2022, local hardware stores in New York City reported selling out of ant baits when New Yorkers found L. emarginatus colonies infesting their apartments (Stewart 2022). Second, their tendency to form mutualisms with tree pests, such as scale insects and aphids, could have negative impacts on urban trees. New York City, alone, is home to 7 million trees that are estimated to remove 51,000 tons of carbon per year, 1100 tons of air pollution, and to reduce annual residential energy costs by $17.1 million USD (Nowak et al. 2018). The total cost of replacing New York City trees is estimated at $5.2 billion USD, with insect pests identified as a primary concern (Nowak et al. 2018). We observed that L. emarginatus workers foraged in trees throughout the entire day and night, as well as during rain events where the bark was dry. These foraging behaviors are similar to foraging behavior observed in another introduced species, the Argentine ant (Linepithema humile) (Burford et al., 2018). Third, L. emarginatus could have a negative impact on local insect diversity, particularly other ant species. We observed L. emarginatus workers surrounding and dismembering pavement ant workers (T. immigrans) in sidewalk habitats where both species are abundant. In parks, which have a higher diversity of native ant species, we observed L. emarginatus less frequently, so their negative impacts on these communities may be reduced. Further monitoring of L. emarginatus populations is warranted to determine whether they are simply outcompeting other introduced ants in urban habitats (e.g., T. immigrans), or whether they will have negative impacts on native ant communities in less disturbed habitats as their populations expand.

Because L. emarginatus was detected early after its introduction, it may be possible to assess their impacts on native insect populations by resampling diversity before and after their expansion into previously sampled habitats. We have established a community science project through iNaturalist to monitor and document L. emarginatus expansion. In addition, further research on their interactions with honeydew-producing insects and subsequent impacts on urban trees could also assess any potential negative impacts on tree health. Finally, it remains unclear which traits of L. emarginatus have facilitated their success as urban pests. We showed that they do not form supercolonies in New York City, but they are likely to possess other traits that have contributed to their success. Continued monitoring of their introduced population and studies of their interactions with co-occurring species will be necessary to assess their potential impacts on urban ecosystems as their populations expand.