Geographic distribution of Leptogenys elongata ( Buckley ) and Leptogenys manni Wheeler ( Hymenoptera , Formicidae , Ponerinae )

Leptogenys elongata and Leptogenys manni (Hymenoptera, Formicidae, Ponerinae), the only described Leptogenys known from the continental US, were recently included on a list of exotic ants established in North America. To evaluate this possibility, I compiled and mapped published and unpublished specimen records of L. elongata and L. manni. Leptogenys elongata records have an essentially continuous distribution through central and eastern Texas (65 counties), southern Oklahoma (one county), and western Louisiana (eight parishes), south through much of Mexico (nine states), plus one questionable record from Colorado and one almost certainly erroneous record from the District of Columbia. Leptogenys manni records are known from much of Florida (22 counties), plus one isolated record from Georgia that needs corroboration. I found no credible evidence that L. elongata or L. manni have established populations anywhere beyond their native ranges.

Recently, Wittenborn and Jeschke (2011) investigated characteristics of exotic ant species by comparing species that they classified as native to North America with those they classified as exotic.Two species that Wittenborn and Jeschke (2011; supplementary material) considered exotic in North America were Leptogenys elongata (Buckley) and Leptogenys manni Wheeler.Of the 307 described Leptogenys species (Bolton 2015), L. elongata and L. manni are the only ones known from the continental US.Both L. elongata and L. manni appear to be specialist predators of terrestrial isopods (Wheeler 1904, Trager andJohnson 1988).Here, I examine the geographic ranges of L. elongata and L. manni and evaluate evidence concerning whether these species have any exotic populations in North America or elsewhere.
When a cosmopolitan ant species occurs in both the Old World and the New World, it is almost always clear that one of these ranges is entirely exotic.Within a species' native hemisphere, however, it can be difficult to determine what geographic area constitutes the native range and what area, if any, constitutes the exotic range.When evaluating the native and exotic ranges of a species, researchers consider a spectrum of distributional, historical, evolutionary, ecological, and genetic information (see Wetterer 2008).For example, evidence considered indicative of a species' native range includes 1) older records largely confined to a single continuous region, 2) occurrence in inland native communities, and 3) proximity to the ranges of closely related species.In contrast, evidence indicative of a species' exotic range includes 1) sudden appearance and spread of the species through an area discontinuous with other known populations, 2) occurrence exclusively in coastal and highly disturbed environments, and 3) geographic isolation from closely related species.

Materials and methods
Using published and unpublished records, I documented the known ranges of Leptogenys elongata and Leptogenys manni.I obtained unpublished site records from museum specimens in the collections of the Archbold Biological Station (ABS), the Museum of Comparative Zoology (MCZ), and the Smithsonian Institution (SI).In addition, I used on-line databases with collection information on specimens by Antweb (www.antweb.org).I obtained geo-coordinates for collection sites from published references, specimen labels, maps, or geography web sites (e.g., earth.google.comand www.tageo.com).
If a site record listed a geographic region rather than a "point locale," and I had no other record for this region, I used the coordinates of the largest town within the region or, in the case of small islands and natural areas, the center of the region.In the only exception, I mapped Deyrup et al.'s (1989) record of L. manni from Monroe County, Florida to a site in northeastern part of the mainland part of the county rather than the largest town (Key West) because Trager and Johnson (1988) and others wrote that L. manni is not known from the Florida Keys.
I was unable to map individually many records of L. elongata from caves because the site locations are kept secret to avoid vandalism.For example, Reddell and Cokendolpher (2001) and Cokendolpher et al. (2009) listed records of L. elongata from seven caves in Bexar County, Texas, including two caves on Camp Bullis Military Training Reservation; I included all these records as just a single site record, mapped near the center of Camp Bullis.

Results
In total, I mapped 139 site records of Leptogenys elongata (including 106 from Texas) and 27 site records of Leptogenys manni (all but one from Florida) (Fig. 1; Table 1).Vernon (Dash 2004, Lattke 2011, antweb) Two of the new species in the elongata group that Lattke (2011) described have distributions in Mexico that overlap with that of L. elongata (L.sianka in Veracruz and L. chamela in Jalisco).Lattke (2011) re-identified specimens reported as L. mexicana from Quintana Roo on the Yucatan Peninsula (Dejean et al. 1995;Dejean and Olmsted 1997) as L. sianka.Lattke (2011) confirmed species identification for 12 of the 22 Mexican L. elongata site records.It is possible that some Mexican specimens not examined by Lattke (2011) that are currently considered L. elongata are actually a different species in the elongata group.

Problematic records
There are three problematic reports of L. elongata (from Colorado, the District of Columbia, and Maryland).Creighton (1950) Mayr's (1886) record of Pergande's specimens reportedly from the District of Columbia.Wheeler (1923) wrote that L. elongata "occurs in the Gulf States from Texas to Florida and north to Colorado and the District of Columbia."Records of L. elongata from Florida were actually L. manni (see above), and it seems likely that published records of L. elongata from Colorado and the District of Columbia did not come from outdoor populations in these areas.Both southernmost Colorado (37.0°N) and the District of Columbia (38.9°N), have considerably higher latitudes (and much cooler climate) than the next highest latitude records for L. elongata (33.9°N, see above).
The single record of L. elongata from Colorado was based on specimens that Ezra T. Cresson collected (Emery 1894 as L. septentrionalis).Cresson made substantial collections of Hymenoptera in the Western US for the Academy of Natural Sciences of Philadelphia.Cresson collected in Colorado, but he also worked extensively in Texas (see Cresson 1872).Even though the validity of this record is questionable, I mapped it to Campo in southeastern Colorado.
The single record of L. elongata from the District of Columbia was based on specimens that Theodore Pergande collected (Mayr 1886 as L. septentrionalis).Pergande worked for the Federal Bureau of Entomology in Washington DC.All other ant specimen records I have found for Pergande from Washington DC were from inside houses and greenhouses (e.g., Monomorium pharaonis, Paratrechina longicornis, and Tapinoma melanocephalum in the Smithsonian collection).It is possible that the specimens Mayr (1886) reported came from a greenhouse.It seems more likely, however, that the specimens were mislabeled or their provenance was misunderstood.Pergande (1893Pergande ( , 1894Pergande ( , 1896) ) collected ants extensively in Mexico and Texas, which are more likely sources of his specimens.Lattke (2011) similarly concluded that this record was "an obvious error as the District of Columbia lies far beyond the range of L. elongata."Smith (1947) wrote that he saw specimens of elongata labeled "Stone Mt., Decatur, Georgia."These specimens are at the Smithsonian and Lattke (2011) identified them as L. manni.Lattke (2011) wrote: "The specimens labeled as from Georgia constitute a single series and lack a date, though the state of the label suggests the early 1900's.L. manni is now known to range into northern Florida, but over 300 kilometers separate Decatur from the nearest L. manni collection sites in Florida.The Georgia locality is a possibility but needs corroboration."Trager and Johnson (1988) wrote that Leptogenys elongata occurs in "western Louisiana, Texas, and northeastern Mexico."However Lattke's (2011) designation of Leptogenys mexicana as a junior synonym of L. elongata extends the reported range of this species south into central Mexico (Fig 1).Despite the paucity of records, L. elongata appears to show an uninterrupted range across much of Texas (65 counties) into southernmost Oklahoma (one county), east to central Louisiana (eight parishes), and south through much of Mexico (nine states) (Fig. 1).This continuous range gives no indication that L. elongata is exotic to any part of this region.The one questionable record from Colorado could be an isolated northern extension of this native range, but seems more likely to be an error, and the one record of L. elongata from the District of Columbia is almost certainly an error.The relatively recent first records of L. elongata from some states in Mexico (Table 1) are likely to be due to the scarcity of specimens and the difficulty of distinguishing L. elongata from the 16 other Leptogenys species known from Mexico (Antweb 2015), including five other members of the elongata species group (L.chamela, L. oaxaca, L. peninsularis, L. sianka, and L. sonora), rather than recent expansion into parts of Mexico.Trager and Johnson (1988) listed Leptogenys manni only from Florida and wrote that L. manni "must be added to the growing list of Florida's endemic ants."Leptogenys manni records are now known from much of Florida (22 counties), plus one isolated record from a natural area in central Georgia that could be based on a labeling error and needs confirmation.

Discussion
Leptogenys elongata and L. manni are large, distinctive ants that have been rarely collected, perhaps due to largely subterranean habits.In fact, Reddell and Cokendolpher (2001) and Cokendolpher et al. (2009) found L. elongata relatively often when surveying caves in Texas.Trager and Johnson (1988) speculated: "in south Florida, L. manni may be entirely subterranean."Virtually all records of L. elongata and L. manni come from relatively undisturbed natural areas, a character not normally indicative of an exotic species.In fact, I found no credible evidence that L. elongata or L. manni have established exotic populations beyond their native ranges.I conclude that Wittenborn and Jeschke (2011) were wrong in classifying L. elongata and L. manni as exotic to North America.
In documenting the ranges of other species on Wittenborn and Jeschke's (2011) list of North American exotics, I have found that numerous other species also appear to be misclassified.For example, Wittenborn and Jeschke (2011) categorized Gnamptogenys hartmani (Wheeler), Labidus coecus (Latreille), and Pachycondyla harpax (Fabricius) as exotics in North America, but all three species have apparently continuous ranges from South America, through Central America, and into Mexico and the southern US and give no indications of being exotic in any part of their ranges (Wetterer 2014, in press a, Wetterer andSnelling 2015).Similarly, Cephalotes varians (Smith), Tapinoma litorale Wheeler, Temnothorax allardycei (Mann), and Trachymyrmex jamaicensis (André), also on the list of exotics, appear to be native throughout their seemingly continuous ranges in the West Indies and southern Florida (Wetterer in press b, in prep.).It is unfortunate that native North American ant species wound up on Wittenborn and Jeschke's (2011) list of exotics.Exotic species are an important ecological problem around the world.In order to protect native species and minimize the negative impacts of exotic species, it is essential to distinguish which species are native and which are exotic.

Table 1 .
(Deyrup et al. 1989rds for Leptogenys elongata and Leptogenys manni in the US and Mexican states.*=probablyerroneous.Site information given for unpublished records.Volusia, and Wakulla(Deyrup et al. 1989, Lattke 2011, antweb, ABS collection).The northernmost Florida record came from Tall Timbers Research Station in Leon County (30.7°N;ABS).There is a single, isolated record of L. manni from Georgia (33.8°N; see below).
wrote: "Records for elongata have been reported from Colorado, the District of Columbia and Maryland.It is unlikely, but not impossible, that elongata occurs in southeastern Colorado.But records from the District of Columbia and Maryland seem plainly impossible."Wheeler (1908) reported L. elongata from Texas, Florida, Colorado, and Maryland, though Wheeler (1908) cited Pergande and Mayr for the Maryland record and therefore this must refer to