Survey of ants (Hymenoptera, Formicidae) in the city of Providence (Rhode Island, United States) and a new northern-most record for Brachyponera chinensis (Emery, 1895)

. We surveyed ants in Providence, Rhode Island, from 2015 to 2019. Methods including repeated pitfall trap sampling and manual searching were used to collect ants at Providence College and a rapid biological assessment was conducted at Roger Williams Park. A total of 36 species were identified based on morphology, including the first observations of a colony of Needle Ants ( Brachyponera chinensis Emery, 1895) in New England. Twenty-six species identified were new county records and seven species were new state records, representing a substantial update to the list of known ant species in Rhode Island, currently totaling 41 species in Providence and 69 species from six subfamilies across the state. These results are comparable with similarly scaled surveys conducted at parks and cities across the world, and they also offer a reminder that while urbanization can be associated with reductions in habitat availability for some fauna, cities can be accessible and ecologically important locations for exploring myrmecological biodiversity.


Introduction
Ants are among the most ecologically successful animals on the planet. Their social nature allows them to operate as complex adaptive systems, responding to and structuring ecological communities, providing critical ecosystem services, and with the potential to impact economic stability and agricultural productivity (Davidson 1997;Del Toro et al. 2012;Evans et al. 2011;King et al. 2013;McGlynn 1999;Ward 2006). Biodiversity data and the species distributions of many ant taxa have been widely studied, making them a key indicator species for identifying disturbed habitats and effects of climate change (Dunn et al. 2007;Jenkins et al. 2011). While the diversity of ants in many places has been relatively well sampled, this was not the case for Rhode Island, a state at the southern coastal boundary of New England where it may have a higher likelihood for biotic introductions and potential colonizations by ANNOTATED LIST OF SPECIES introduced species.
Rhode Island's geological history was strongly influenced by glaciation events 14,000 years ago and its diverse habitats now include maritime coastal and wetland systems, freshwater wetlands, forests, peatlands, lakes and ponds, salt marshes, pine barrens, farmland, islands, and urban and suburban residential and industrial areas (RIDEM 2015). Although it is a small state, it has the highest ratio of coastline to land area of any state, and it is the second most densely populated state in the country. The capital city, Providence, is a gateway to Narragansett Bay, providing shipping access for the state's primary export (scrap metal) and it is also home to many universities and College campuses.
In 1906, William Morton Wheeler documented 84 species of ants across New England but only 12 species of ants in Rhode Island (Wheeler 1906). More than a century later, Aaron Ellison and colleagues exhaustively compiled 28,205 ant specimen records from across New England and published a guide to the 132 described species of ants in this region, including 47 in Rhode Island and 13 from Providence (Ellison et al. 2012). A recent targeted survey of a small parcel near the southern-most extremity of the state added nine new species to the list from Rhode Island (Ellison and Farnsworth 2014). Since the most under-surveyed part of the state remained Providence County, we focused on surveying two urban sites in the city of Providence (Fig. 1A) which were accessible to students engaged in this project on campus at Providence College (PC) and at Roger Williams Park (RWP).

Study Area
Providence College was founded in Providence in 1917 and the College community is made up of about 5,000 students who, together with faculty, administrative staff, and Dominican Friars, are engaged in study on a campus of 0.43 km 2 located adjacent to the Elmhurst, Smith Hill, and Wanskuck neighborhoods within the city (Fig. 1B). The campus consists of about 50 buildings (including academic, dining, residential, religious, and athletics facilities) on heavily maintained grounds. A recent campus inventory counted more than 1,000 trees from 65 species. The largest public park in Providence, Roger Williams Park (1.7 km 2 ) is located on the south side of the city approximately 11 km from Providence College (Fig. 1C). The park is located on land that was a gift from the Narragansett people to Roger Williams in 1638. It was, for a while, used as farmland, and then gifted to the people of Providence in 1872. Roger Williams Park is now home to a zoo, museum, ponds, a boathouse, the Providence Police Department's Mounted Command center, sporting fields, a botanical garden, a concert venue, and many walking paths and roads supporting vehicular traffic. Like Providence College, Roger Williams Park is surrounded by a densely inhabited residential neighborhood with nearby commercial and industrial districts.

Methods
At Providence College, the primary survey method involved a repeated sampling protocol using pitfall traps. The traps were made from 50 mL plastic centrifuge vials filled with approximately 15 mL of soapy water, and they were placed in the ground so that the top of the vial was level with the surface. A total of 39 traps were spread throughout campus (Fig. 1D). Twelve locations were chosen at random, avoiding athletic fields and locations with impenetrable surfaces. Three pitfall traps were placed in a 10 m radius at each of these locations and an additional pitfall trap was placed at each of the three bioswale locations on campus which were designed with specific native vegetation to receive excess rainwater runoff. For each of the 10 weeks of the survey, a student deployed the 39 traps on one day of the week, retrieved them two days later (aiming to select days with minimal expected rainfall), and closed empty vials were left as placeholders in the ground between capture periods. Each week, the numbers of ants and other invertebrates were counted; ants averaged more than 80% of the specimens captured but their abundance varied across locations on campus and over the duration of the 10-week period (Figs. 2,3). Specimen sorting and identification of the 1,853 ants from the pitfall traps took approximately two years (Table 1). Additionally, baiting and manual collecting by students were conducted in subsequent years to expand on the results of the pitfall trap survey.
At Roger Williams Park, the Rhode Island Natural History Survey (https://rinhs.org) organized a BioBlitz rapid biological assessment event to catalog as many living things present over a 24-hour window, from 2 pm on May 31 to 2 pm on June 1, 2019. This event has been organized annually by the RINHS at different locations across Rhode Island for the last 20 years. Volunteer experts worked together with members of the general public, walking throughout the park, making observations, and returning collected specimens as necessary to an ad-hoc science center with resources for identification (including microscopes, reference books, insect pinning supplies, etc.). A total of 145 individuals actively engaged in the survey. Of these, 5-10 individuals were actively searching for ants, though many other participants donated ant specimens found among their samples.
Collected ants were preserved in ethanol, a subset of these were pinned, and specimens were identified using morphological characters and dichotomous keys (Ellison et al. 2012a). In the course of our work, specimens were examined under Motic and Wild stereomicroscopes (Martin Microscope Company, Easley, SC, USA) at 10-50×. Specimens were photographed using a Canon 6D, MP-E 65 mm 1-5× lens and with a commercially available focus-stacking system (Brecko et al. 2014). Specimen records were maintained in an online database and voucher specimens for each species we report new observations for were deposited with the  Cornell University Insect Collection (Ithaca, NY, USA).
in final volume of 200 µL of AE buffer. A section of the mitochondrial cytochrome c oxidase subunit 1 (COI) molecule was amplified in a PCR reaction using primers LCO1490 and HCO2198 (Folmer et al. 1994) at a final concentration of 5 µM for each primer. PCR amplification of the DNA began by heating the samples to 95 °C for 3 min followed by 35 cycles under the following conditions: 95 °C for 30 s, 50˚C for 30 s, and 72 °C for 45 s. PCR products were visualized and isolated using agarose gel electrophoresis (1.1% agarose in 1X TBE). Bands of appropriate size were excised from the gel and purified using a QIAGEN MinElute Gel Extraction Kit following the manufacturer's instructions. The concentration of the purified PCR products was measured using a nanospectrophotemeter. PCR product concentrations were standardized according to GeneWiz Sanger sequencing protocol, the sequencing reactions were then performed by GeneWiz (Azenta, Inc., Chelmsford, MA, USA). All four PCR products were sequenced in both directions. DNA sequences were aligned in MEGA v. 11 (Tamura et al. 2021). Similarity to known Brachyponera chinensis was tested using a GenBank BLAST search.

Results
Representatives of four subfamilies, 16 genera, and 36 species ( Fig. 4; Table 2) have been collected and identified with taxonomy information referenced from the online catalog and bibliography of the world's ants, AntCat (Bolton 2019).

Identification.
The petiole node appears to be missing a node, acidopore absent, characteristic odor of rotten coconuts or ripe bananas, horizontal slit at end of gaster.
Aphaenogaster picea Wheeler, 1908 Figure 4, Appendix Figure A26, Identification. Last four segments of the antennae are lighter in color than the rest, prominently depressed propodeum, long legs, dark color, ridge on top of mesonotum, propodeal spines short and point rearward, 2-segmented waist. Figure 4, Appendix Figure A27, Identification. Last four segments of the antennae are the same color as the rest prominently depressed propodeum, long legs, propodeal spines short and point upward, 2-segmented waist.

Identification.
Only one or two long erect hairs on each corner of the pronotum, 2-segmented waist, petiole attaches to top of heart-shaped gaster seen from a dorsal view.

Discussion
In our survey of Providence, we identified 36 species of ants, 24 at Providence College, 25 at Roger Williams Park, and 13 species found at both sites ( Fig. 4; Table 2). Of the 13 species of ants previously known to have been found in Providence, four species were not found in the current study, including Formica integra (Nylander, 1856), Lasius umbratus (Nylander, 1846), Monomorium emarginatum (DuBois, 1986), and Myrmica americana (Weber, 1939). If these species are still present in the city, and including an introduced population of Pheidole megacephala (Fabricius, 1793), we can count a total of 41 ant species in Providence. We found three introduced species including Tetramorium immigrans, Nylanderia flavipes, and Brachyponera chinensis. The Pavement Ant (T. immigrans) is originally from Europe and is regionally pervasive. The yellow-footed ant (N. flavipes) is Asian in origin, but as of only a few years ago was only found from a few locations in New England (Ellison et al. 2012b); now it appears to have spread abundantly. The Needle Ant (B. chinensis) is having an ecological impact as a competitive invasive species in the mid-Atlantic states, but previously it had not been observed to have spread as far north as New England (Guénard and Dunn 2010;Guénard et al. 2018); the closest recent observation was in New York State (Ellison et al. 2012b). We note that while the common name often used for this species often includes a regional identifier, we generally omit this part of the name since it is not necessary for uniqueness and including it can unintentionally promote negative associations and stereotypes. In our study, B. chinensis was at first identified from only a single specimen among thousands in the pitfall trap collection and manual search at first turned up no additional observations. Students at Providence College, incentivized by extra credit, searched the campus and found a colony of B. chinensis nesting around the perimeter of a dormitory (Fig. 5). In the years since we learned of its presence there, we have removed workers and queens for study, nevertheless it persists. The occurrence of individuals has expanded slightly to the perimeters of adjacent buildings, but to the best of our knowledge, it has not yet been found anywhere else on campus or more broadly across the region.
How many different ant species should there be in Providence? We found existing records and supplemented those with our own collections, but the final tally we counted (N = 41 species) does not answer the question of how many ants we may have been expected to find. Whether for a region, a city, a park, or a backyard, there is not a general answer to this kind of question. However, the literature can offer some context. At the broadest spatial scales, there are published totals for locations including 1,884 ant species (2,485 including unresolved infraspecific taxa) in Africa (Fisher and Bolton, 2016), 951 in China (Liu et al., 2015), and 237 from the Solomon Islands (Sarnat et al., 2013). For North America, we count almost 1,000 species (Fisher and Cover, 2007), including 94 in the Florida Keys (Moreau et al. 2014) and 143 species in New England (Ellison et al. 2012). How do we compare the findings from a relatively small site such as a park or campus to these larger inventories? On a spatial scale more comparable to this study, surveys of college campuses, parks, cities, and islands have found between eight and 164 species with a median count of 40 species and an average of 43 (Table 3). There is a lot of variation in species counts in different studies and using multiple methods can significantly increase species yield (Ellison and Farnsworth 2014;Ellison et al. 2007;Guénard et al. 2014), but without standardized survey methods or experimental approaches it is hard to attribute differences in diversity to ecologically meaningful factors. Even without standardized methods, however, studies such as this one, together with community science initiatives, can help raise awareness about local biodiversity and inform more ecologically oriented studies.
We reviewed the ant species records on the popular BugGuide and iNaturalist platforms for comparison with our data and previously curated species records. BugGuide (https://bugguide.net) has been operational as a website since 2003 and iNaturalist (https://www. inaturalist.org) has been available as a website and mobile application since 2008. As of July 2019, the BugGuide website contained records for 119 ant observations in Rhode Island, the vast majority of which were from Block Island. State-wide, there were 21 species identified and just three from Providence. Two    Table 3. Ant species counts for local surveys. The studies listed were selected based on surveys that were conducted in parks, cities, and small islands, though the spatial scale is varied and some are more broadly regional or habitat-specific. Other varying factors include survey intensity and duration, methods applied, and the degree of urbanization as this list includes habitats ranging from conservation wilderness to one of the most densely populated urban areas on the planet (Macau). was not found using the described methods of our study which focused on collecting outdoors, after being alerted about an unknown ant inside a rainforest exhibit at the zoo, we collected individuals, verified their identification, and have shared pinned specimens. One major question raised in interpreting our results is whether or not the urbanization of areas contributes to the loss or gain of myrmecological biodiversity. Some studies have shown a general trend of urbanization associated with a decrease in overall diversity, although perhaps mitigated among arthropods by an increase in abundance (Faeth et al. 2011). The results of our survey cast doubt on the assumption that cities are not diverse places and others have concluded similarly based on surveys for ants in Raleigh (Menke et al. 2010) or Macau (Leong et al. 2017) and for bee diversity in New York City (Matteson et al. 2008) and Vancouver (Tommasi et al. 2004). Cities may have more asphalt and concrete than rural areas, but they also have a high flux of potential resources mediated by human activity ranging from invasive plant transport to food waste (Penick et al. 2015). The urban heat island effect offers a refuge against lower critical thermal limits (Stringer et al. 2009). Especially at the small scale of an individual ant, cities offer highly heterogeneous, spatially compartmentalized, and highly variable thermal micro-habitats (Pincebourde et al. 2016). Cities may also be less likely to be sprayed with large amounts of pesticides as might be the case in more agriculturally developed regions.
Our study only focused on two sites, and both were on institutionally maintained grounds. It is possible that the number of species we identified might not be found in the residential and commercial districts throughout the city-that the ant diversity is relatively concentrated in the urban parks-but this remains to be determined. If a diversity of ant species may be found at a number of parks scattered across and within the city, would they not also be found under stones, on trees, and within houses more generally throughout the city? In Taichung City (Taiwan), there was not a significant change in ant species diversity across the city with respect to the distance from urban parks, though there were associations with park size, soil moisture, and the number of trash bins (Liu et al. 2019). Gradients for urban insect diversity have been mapped out in other cities including Phoenix and Los Angeles, but while a combination of microhabitat temperatures, humidity, surface permeability, and plant drought-tolerance have been identified as important factors, they can have variable impacts in the different cities and for different taxa sampled (Adams et al. 2020;McGlynn et al. 2019).
As E.O. Wilson and others have implored (Pimm et al. 2014;Saunders 2019;Tschinkel and Wilson 2014;Wilson 2017), we have only scratched the surface of identifying the biodiversity on the planet, the smallest habitats are likely the most threatened, and there is an imminent need to identify and conserve the wildlife all around us before it disappears. There is also value to highlighting the biodiversity found within urban ecosystems, as this is the nature many people will encounter most frequently and can inspire future conservation efforts more broadly (Dunn et al. 2006). There is a great opportunity for community science initiatives such as the successful School of Ants project (Lucky et al. 2014) and local BioBlitz events to continue to address these questions across a broader range of localities and spatial scales. whose ongoing generosity has helped support our students along with our fieldwork and DNA sequencing efforts. Collecting permits were granted by the Rhode Island Department of Environmental Management.