Invasive ants of Bermuda revisited

For 60+ years, two of the world’s most widespread and destructive invasive ant species, the African bigheaded ant (Pheidole megacephala) from tropical Africa and the Argentine ant (Linepithema humile) from subtropical South America, have been engaged in an epic battle on the islands of Bermuda. Both species are completely intolerant of the other and are also well-known for killing off native invertebrates, particularly other ants. Here I surveyed sites across Bermuda in 2016, including resurveys of the locations previously surveyed in 1963, 1966, 1973, 1986, and 2002, to provide an update on this conflict. The status of all other ant species present in the islands, including previous records from literature, is also provided. In addition, I surveyed ants nesting in red mangrove (Rhizophora mangle) trees to evaluate whether this arboreal habitat may serve as a refuge for previously overlooked ant species. In 2016, L. humile occurred at most surveyed sites in Bermuda, including all ten resurveyed sites. Pheidole megacephala was present at only two resurveyed sites, a lower proportion of sites than any of the five earlier surveys. Still, P. megacephala occupied substantial areas, particularly in and around Hamilton, the capital of Bermuda. This survey increased the number of ant species with verified records from Bermuda to 25, including four exotic species recorded for the first time: Cardiocondyla minutior, Pheidole navigans, Strumigenys emmae, and Strumigenys membranifera. I found five ant species nesting in mangroves: L. humile plus four Old World exotics, C. minutior, Cardiocondyla obscurior, Monomorium floricola, and Plagiolepis alluaudi. It appears that L. humile may be better suited to the subtropical climate of Bermuda than P. megacephala, except perhaps in warmer and sunnier habitats, such as plantings along urban streets and in open parks, where P. megacephala may hold the advantage. The dataset on the new ant records from 2016 is provided with the paper.


Introduction
For more than 60 years, two of the world's most widespread and destructive invasive ant species, the African big-headed ant (Pheidole megacephala) from tropical Africa and the Argentine ant (Linepithema humile) from subtropical South America, have been engaged in an epic battle on the Atlantic islands of Bermuda.Pheidole megacephala was first recorded in Bermuda in 1889 (Dahl 1892), and was the dominant ant species there in the early 20th century (Haskins 1939, Wetterer andWetterer 2004).When L. humile arrived in Bermuda in the 1940s, however, this new invader quickly took over much territory, displacing P. megacephala.Still, P. megacephala maintained many strongholds, and ever since, these two species have been contesting ever-shifting battlefronts between mutually exclusive territories that together occupy most of the land area of Bermuda (Haskins 1939;Haskins andHaskins 1965, 1988;Crowell 1968;Lieberburg et al. 1975;Wetterer and Wetterer 2004).
Both P. megacephala and L. humile are well-known for killing off native invertebrates, particularly ants (Erickson 1971;Human and Gordon 1996;Holway 1999;Wetterer et al. 2000Wetterer et al. , 2001;;Wetterer 2002).During their tenure in Bermuda, P. megacephala and L. humile have almost certainly had a devastating impact on native invertebrates.In addition to the two dominant ant species, Wetterer and Wetterer (2004) reported confirmed records of 18 other ant species from Bermuda.Subsequently, I found museum specimens of two more ant species from Bermuda: Nylanderia cf.fulva (reported as Paratrechina pubens; Wetterer 2007) and Monomorium floricola (see Wetterer 2010), bringing the total number of known species to 21 (plus one male specimen identified only to genus; see Results).It is unclear whether any ant species now living in Bermuda are native, i.e., predating human arrival.
In the present study, I surveyed sites across Bermuda to provide an update on this conflict and on the status of other ant species present.In addition, I surveyed ants nesting in red mangrove (Rhizophora mangle) to evaluate whether this arboreal habitat may serve as a refuge for some previously overlooked ant species.Red mangroves grow in and adjacent to shallow brackish water in subtropical and tropical estuaries around the world, providing a unique habitat, often completely isolated from any terrestrial habitat.

Methods
In 3-10 May 2016, I surveyed ants in Bermuda.I hand-collected ants using an aspirator, sifted litter using a Davis sifter, and collected ants from inside dead twigs and branches of red mangrove.
To examine changes in which ant species, P. megacephala or L. humile, dominated specific areas in Bermuda, Haskins and Haskins (1988) re-surveyed ten sites that had been surveyed earlier.Wetterer and Wetterer (2004) surveyed these same ten sites.
In 2016, I again surveyed ants at these ten sites using hand collecting and litter sifting (geo-coordinates and collection date in parentheses):

Results
In 2016, I collected a total of 19 ant species in Bermuda, including four recorded for the first time: Cardiocondyla minutior, Pheidole navigans, Strumigenys emmae, and Strumigenys membranifera (Table 1, See Suppl.material 1.These records bring the total number of confirmed ant species records from Bermuda to 25. Bermuda Department of Agriculture has previously intercepted P. navigans (recorded as Pheidole moerens) on in-coming goods (Wetterer and Wetterer 2004).Wetterer and Wetterer (2004) reported a male dacetine ant (Tribe Dacetini) that could not be identified to species level.It seems likely that this dacetine was one of the two newly reported Strumigenys species.
In their 2002 survey, Wetterer and Wetterer (2004) found L. humile in large numbers at all ten sites studied by Haskins and Haskins (1988).At four of the sites, they also found P. megacephala (Table 2).In 2016, I again found L. humile at all ten longterm survey sites, but found P. megacephala at only two (Table 2).At Spittal Pond, I found L. humile at the park entrance and parking lot where P. megacephala occurred in 2002.At Harrington 100s, I found L. humile north of Knapton Hill Road as before.South of Knapton Hill Road, rather than P. megacephala, I found high numbers of P. navigans and T. simillimum.
At two long-term survey sites, I found both L. humile and P. megacephala: on Ireland Island and the Newstead Hotel (now Newstead Belmont Hills Golf Resort and Spa).On Ireland Island, I found P. megacephala along the North Breakwater and by the Maritime Museum (now the National Museum) as before.In addition, I collected P. megacephala in front of the Clocktower Mall and to the south end of the Glassworks Mall, two places occupied by L. humile 14 years earlier, indicating a modest expansion of the P. megacephala population on North Ireland Island.At the Newstead Hotel, I found the boundary between L. humile and P. megacephala territory, near the western edge of the property, essentially identical as 14 years earlier.At the Newstead Hotel, I collected in the same vial L. humile and P. megacephala workers from only a few meters apart; the ants immediately locked in battle, confirming their mutual intolerance.
I found five ant species nesting in red mangrove in Bermuda: L. humile plus four Old World exotics that Wetterer and Wetterer (2004) did not collect in 2002: Cardiocondyla minutior, Cardiocondyla obscurior, Monomorium floricola, and Plagiolepis alluaudi (Table 3).At nine of the 14 sites, I found ants nesting inside twigs and branches: Hungary Bay (four samples with M. floricola, six with P. alluaudi), Crow  At many red mangrove sites, I saw streams of L. humile on the branches of red mangrove growing in shallow water.In some cases, L. humile workers I found inside twigs could have belonged to land-based colonies, where they could access the man- grove growing in water via a continuous canopy.However, one sample from Pilchard Bay that had L. humile inside twigs came from a small mangrove island consisting of two mangrove trees isolated by water from all other mangroves.This L. humile colony had to be completely arboreal.
All three samples that people sent me from their houses included L. humile.One also included P. megacephala and Brachymyrmex cf.heeri, and one included P. navigans.

Discussion
In 2016, I found that L. humile dominated most sites I surveyed in Bermuda.Still, P. megacephala occupied substantial areas, particularly in and around the capital of Hamilton.Pheidole megacephala is a dominant species in tropical sites around the world, whereas Linepithema humile tends to dominate in subtropical areas with "Mediterranean" climates.It appears that L. humile may be better suited to the subtropical climate of Bermuda, except perhaps in warmer and sunnier habitats, such as the plantings along urban streets and in open parks, where P. megacephala may have the advantage.
One ant species that is possibly native to Bermuda is the trap-jaw ant Odontomachus ruginodis.This species was once common in Bermuda, but now appears to be quite rare (Haskins and Haskins 1965).Odontomachus ruginodis is common in South Florida and the West Indies.Although it is not known as a tramp species, Deyrup (1992) considered it to be probably exotic to Florida.Genetic analyses could help determine whether Odontomachus ruginodis (or any other ant species) is native or exotic to Bermuda.
Bermuda has the northernmost populations of mangroves in the world (Spalding et al. 2010).I surveyed ants in the red mangroves of Bermuda in the hope of finding some rare ants that use this unique habitat as a refuge and thus escaped extermination by P. megacephala and L. humile.However, because L. humile also nests in red mangrove, this habitat did not provide a refuge for ant species incapable of co-existing with this dominant species.Except for L. humile, all ant species I found nesting in the red mangroves of Bermuda were Old World exotics.Thus, red mangroves may serve as a refuge from the dominant invasive ants, but the only ants taking advantage of this refuge are other exotic species.In the absence of invader-free refuges, invasive ants can drive native species to extinction.For example, in 2003, I surveyed ants on Monte Gordo, the highest mountain of the Atlantic island of São Nicolau, Cape Verde, and found only P. megacephala at all sites sampled except for sites within 100 m of the mountain's peak (JKW, unpublished data).Here, where P. megacephala had not yet spread, a small pocket of other ant species survived, including an endemic species known only from this mountain, Monomorium boltoni.If P. megacephala continues its spread up the mountain, it seems unlikely that M. boltoni will survive once P. megacephala overruns what may be its last refuge.Similarly, it may be that all ant species originally native to Bermuda are now extinct.
I was surprised to collect P. navigans at five sites scattered across Bermuda (and I was sent specimens from a sixth site), given that this species had never before been collected in Bermuda, and that I surveyed three of these same locales in 2002.This suggests that P. navigans may be a recent arrival on Bermuda and is spreading quickly.Curiously, at four of the five sites, P. navigans was coexisting with L. humile.On Ordnance Island, I even found them nesting together under the same piece of concrete.It would be interesting to determine whether or not P. megacephala can tolerate P. navigans.Pheidole navigans, a big-headed ant native to the Neotropics, is known as an exotic in the southeastern US (formerly misidentified as P. moerens; see Sarnat et al. 2015).If P. navigans battles with P. megacephala, but not with L. humile, the presence of P. navigans could affect the competitive balance between P. megacephala and L. humile.
Pheidole megacephala and L. humile began their worldwide dispersal in the 19th century and have already spread through much of their potential range (Wetterer et al. 2009;Wetterer 2012).However, several other highly destructive ants have thus far come to occupy only a small portion of their potential ranges, most notably the red imported fire ant (Solenopsis invicta) (Wetterer 2013).Great vigilance is needed to prevent S. invicta and other invasive ant species from reaching Bermuda where it would likely thrive (Morrison et al. 2004) and cause substantial ecological and economic damage.

Table 1 .
Ants of Bermuda, arranged in descending order of the number of collection sites in 2002 (Wetterer and Wetterer 2004) plus the number of collection sites in 2016.+ = recorded for the first time in 2016.NX = New World exotic.OX = Old World exotic.

Table 3 .
Ants nesting in dead twig and branch samples from red mangrove, arranged in descending order of number of collection sites.