Cardiocondyla (Hymenoptera: Formicidae) ants are small and inconspicuous species. All Cardiocondyla are native to the Old World, but 5 species have been broadly dispersed through human commerce, achieving widespread distributions in both the Old World and the New World. Here I examine the worldwide spread of Cardiocondyla minutior. I compiled published and unpublished C. minutior specimen records from > 300 sites. I documented the earliest known C. minutior records for 56 geographic areas (countries, island groups, major Caribbean islands, and US states), including several for which I found no previously published records: Austral Islands, Barbados, Cuba, Dominican Republic, Grenada, Honduras, Îles Éparses, Jamaica, Madagascar, and Mayotte. Cardiocondyla minutior appears to be originally from the Indo-Malayan region, where all its closest relatives are found. Surprisingly, there are few records of C. minutior from this region. Instead, most Old World records of C. minutior come from Pacific and Indian Ocean islands where there are few or no native ants, e.g., the Chagos Archipelago, Hawaii, Îles Éparses, Easter Island, the Marquesas Islands, the Seychelles, and the Society Islands. Cardiocondyla minutior is becoming widespread in tropical and subtropical parts of the New World. The success of Cardiocondyla species in exotic locales may be due, in part, to their ability to co-exist with dominant invasive ants, such as the Argentine ant, Linepithema humile.
Cardiocondyla (Hymenoptera: Formicidae) ants are small and inconspicuous species, which Deyrup et al. (2000) dubbed “sneaking ants.” All Cardiocondyla are native to the Old World, but 5 species have achieved widespread distributions in the New World as well, spread by human com merce: Cardiocondyla emeryi Forel, Cardiocondyla mauritanica Forel, Cardiocondyla minutior Forel, Cardiocondyla obscurior Wheeler, and Cardiocondyla wroughtonii Forel. In earlier reviews, Wetterer (2012a, b) examined the spread C. emeryi and C. mauritanica. Here I examine the worldwide geographic spread of C. minutior (Fig. 1).
Taxonomy, Identification, and Biology
Forel (1899) described Cardiocondyla nuda minutior (= C. minutior) from Hawaii. Wilson & Taylor (1967), however, designated C. minutior to be a junior synonym of C. nuda. Although Creighton (1950) tentatively treated C. minutior as a distinct subspecies of C. nuda, Creighton & Snelling (1974) wrote: “It now seems clear that the senior author was mistaken in treating Forel's variety minutior as a subspecies in 1950. At that time there were few long nest series of nuda available for study; hence it was not certain how the single nest series which had yielded workers of the typical nuda and others of the variety minutior ought to be handled. Subsequent studies have shown that the above situation is normally encountered in any long nest series of nuda. It follows that minutior must be treated as a synonym of nuda, as shown by Wilson and Taylor (1967).” Even after synonymy in 1967, some authors continued to use the name C. nuda minutior (e.g., Whitcomb et al. 1972). Heinze (1997, 1999), however, revived C. minutior as a full species. Seifert (2003) confirmed C. minutior as a valid species and determined Cardiocondyla tsukuyomi Terayama (described from Okinawa) to be a junior synonym of C. minutior. Seifert (2003) found genuine C. nuda records restricted to Australia, New Guinea, and Oceania. With the revival of C. minutior as a valid name, Fisher & Cover (2007) removed C. nuda from the list of ants known from North America.
One character which C. minutior shares with C. nuda is a flat dorsal mesosoma in profile with no distinct metanotal groove. This contrasts with the tramp species C. emeryi, C. obscurior, and C. wroughtonii, which have a distinct metanotal groove. It is relatively simple to distinguish C. nuda from C. minutior. In C. nuda, the postpetiolar sternite is convex and the height of petiole about equal to height of postpetiole. In C. minutior, the postpetiolar sternite is nearly flat and the maximum height of petiole greater than maximum height of postpetiole. Workers of C. minutior are generally smaller (total length = 1.5-1.7 mm) than those of C. nuda (total length = 1.6-2.0 mm), no doubt the basis of its Latin name “minutior,” meaning smaller or lesser.
Seifert (2003) noted “intraspecific variability in C. minutior is rather low within the huge range of its distribution, extending over the Neotropic, Polynesian, Australasian, Indo-Australasian, and Oriental regions. Samples from central Sri Lanka have significantly smaller eyes, those from Okinawa, N. India, and Nepal shorter heads but all these deviating populations are in the vast majority of other characters consistent with the overall average.”
Seifert (2003) placed C. minutior in the ‘C. minutior species group' along with six other species: Cardiocondyla breviscapus Seifert from In- dia, Cardiocondyla carbonaria Forel from India, Cardiocondyla opaca Seifert from India, Cardiocondyla britteni Crawley found on butter beans in England (probably imported from India), Cardiocondyla goa Seifert from India, and Cardiocondyla tjibodana Karavajev known from Indonesia, Malaysia, the Mariana Islands, and Belize. Oettler et al. (2010) found that C. minutior and C. tjibodana are genetically almost identical, and both are quite distinct from C. nuda.
Cardiocondyla minutior is one of several Cardiocondyla species known to have dimorphic males (Heinze et al. 2004; Oettler et al. 2010). Winged males disperse, typically after mating within their natal colony. Wingless males typically do not disperse from their natal colony. Instead, they fight other wingless males, typically resulting in the death of all but one resident male.
Methods
Using published and unpublished records, I documented the worldwide range of C. minutior. I obtained unpublished site records from museum specimens in the collections of Archbold Biological Station (ABS, identified by M. Deyrup), the Museum of Comparative Zoology (MCZ, identified by S. Cover and M. Deyrup) and the Smithsonian Institution (SI, identified by M. Smith). W. P. Mackay sent me all Cardiocondyla specimens in his personal collection; M. Deyrup re-identified these specimens. In addition, I used on-line databases with collection information on specimens by Antweb.org, the Essig Museum of Entomology (essigdb.berkeley.edu), and the Global Biodiversity Information Facility ( www.gbif.org). I received unpublished C. minutior records from P. Ward (California, Colombia, Cuba, and Fiji). Finally, I collected C. minutior specimens on Pacific and West Indian islands.
I obtained geographic coordinates for collection sites from published references, specimen labels, maps, or geography web sites (e.g., earth.google. com, www.tageo.com, and www.fallingrain.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. I did not map records of C. minutior on boats, found in newly imported goods, or intercepted in transit by quarantine inspectors. Published records usually included collection dates. In a number of cases, publications did not include the collection dates for specimens, but I was able to determine the date based on information from museum specimens, on the collector's travel dates, or limit the date by the collector's date of death. Stefan Cover confirmed identification for all specimens in the MCZ.
Results
I compiled published and unpublished C. minutior specimen records from >300 sites (Fig. 2). I documented the earliest known C. minutior records for 56 geographic areas (countries, island groups, major Caribbean islands, and US states), including several for which I found no previously published records: Austral Islands, Barbados, Cuba, the Dominican Republic, Grenada, Honduras, Îles Éparses, Jamaica, Madagascar, and Mayotte (Tables 1 and 2).
Published records of C. minutior appear to be quite reliable, particularly in areas outside its purported native range, because authors reporting this relatively obscure taxon no doubt first determined the specimens were not the much more widely reported C. nuda. I found no cases where a specimen identified as C. minutior was later re-identified as a different species. Seifert (pers. comm.), however, wrote of his own identification: “You should know that all these determinations were achieved by an enormous measuring and analytical expense. Nobody (including me) can separate minutior and tjibodana by simple eye inspection. Both species possibly hybridize at localities where they were brought into contact. This is not verified but only suggested as the determinations are not very stable in some localities.” Seifert (pers. comm.) further warned that although C. tjibodana is only widespread in Indonesia and Malaysia, “findings of tjibodana from Belize, the Mariana Islands and Solomon Islands indicate that it has also some tramp species potential.” It is therefore possible that some of my mapped records of C. minutior are actually misidentified C. tjibodana. Alternatively, given their genetic closeness and possible inter-breeding, perhaps C. minutior and C. tjibodana are not actually distinct species.
Re-Identification
Between 1967 and 1997, C. minutior was considered a junior synonym of C. nuda. Since the revival of C. minutior, some authors have re-identified published records of ‘C. nuda' as actually C. minutior. For example, Herrera & Roque-Albedo (2007) reported C. nuda from many islands in the Galapagos and no C. minutior, but Herrera & Heraty (2011) reported no C. nuda from the same islands, replacing all these records with C. minutior.
Before synonymy in 1967, there were several published records of C. nuda minutior from the New World (e.g., Wheeler 1932, Smith 1933, 1944, 1967, Vivar 1957), but none of C. nuda. It appears that subsequent published records of C. nuda from the New World may actually all be C. minutior. I confirmed with M. Deyrup, J. King, Z. Prusak, and J. A. Rodríguez (pers. comm.) that they now consider all specimens they reported as ‘C. nuda' from Florida, the Bahamas, and Mexico (e.g., Deyrup et al. 1988, 1998, 2000; Deyrup & Trager 1986; Deyrup 1994; Ferster & Prusak 1994; King 2007; Rodríguez 2008) as actually C. minutior. Torres & Snelling (1997) listed many records of C. nuda from Puerto Rico, but Snelling & Torres, in an unpublished book manuscript, changed all records of C. nuda to C. minutior. Oddly, although Fernández & Palacio (1995) listed C. nuda in Colombia, Fernández & Sendoya (2004) listed neither C. nuda nor C. minutior from anywhere in the Neotropics.
Mark Deyrup (pers. comm.) examined New World Cardiocondyla specimens in W. P. Mackay's collection and found specimens from 11 sites to be C. minutior: from Columbia (7 sites), Florida (1), Georgia (1), and Nicaragua (2), including 8 that Mackay (1995) previously listed as C. nuda. Deyrup re-identified 1 specimen that Mackay (1995) listed as C. nuda (from New Iberia, Louisiana) as Cardiocondyla venustula Wheeler. Finally, Deyrup was uncertain of the identity of 1 queen that Mackay (1995) listed as C. nuda (from Dauphin Island, Alabama).
I have mapped 11 unconfirmed records of ‘C. nuda' from the New World as C. minutior records (“+” in Fig. 2), i.e., Alabama (one record: Mackay 1995), Columbia (5 records: Mackay 1995, Armbrecht et al. 2001, and gbif), Georgia (one record: Smith 1979), Louisiana (2 records: Dash 2004), South Carolina (one record: Davis 2009), and Texas (Creighton & Snelling 1974). Only 2 New World areas with published records ‘C. nuda' lack any confirmed C. minutior records: Alabama and South Carolina (Table 2).
Wholesale replacement of the name C. minutior for ‘C. nuda' in the New World cannot be repeated in the Asia and Oceania, where true C. nuda occurs. In addition, several other species in the Old World have been misidentified as C. nuda. For example, many published records of ‘C. nuda' from the Palearctic are actually C. mauritanica (Seifert 2003, Wetterer 2012b). Starr et al. (2007) wrote: “In Hawaii, specimens of this species were previously called C. nuda in error, so all previous records of C. nuda are now considered C. minutior (Krushelnycky pers. comm.).” Seifert (2003), however, noted that some specimens from Pacific region identified previously as ‘C. nuda' are actually Cardiocondyla kagutsuchi Terayama, a species from East Asia and Oceania that has been recently reported from Hawaii (e.g., Seifert 2003; Buczkowski & Krushelnycky 2012). Jürgen Heinze (pers. comm.) wrote: “in my opinion, C. kagutsuchi is the most widespread Cardiocondyla inland at least in Hawai'i and Kauai. What has been referred to as C. nuda or C. venustula is mostly C. kagutsuchi. They occur even in na- tional parks far from the coast at elevations of 1000m and more and in some areas are the only ants around. C. venustula appears much rarer.” I therefore only mapped Hawaiian records specifically identified as C. minutior.
Table 1.
Earliest known records for Cardiocondyla minutior from 29 old world areas. Unpublished records include collector, museum or web source, and site. mcz = Museum of Comparative Zoology. ucb = Essig Museum of the University of California, Berkeley. + = No known published record.
Seifert (2003) listed site records of C. minutior from the Marquesas Islands (Kopaafaa, Hiva Oa; 1927; collected by A. M. Adamson & E. P. Mumford) and the Society Islands (Hitiaa, Tahiti; 1928; collected by A. M. Adamson) from the Museum of Comparative Zoology that match reports by Wilson & Taylor (1967) of C. emeryi, suggesting that these are C. minutior specimens previously misidentified as C. emeryi. Eguchi et al. (2011) reported that the specimens that Eguchi et al. (2005) listed as C. emeryi were actually C. minutior.
Forel (1903) reported C. nuda minutior (= C. minutior) from Kondul Island in the Nicobar Islands, but Mohanraj et al. (2010) incorrectly reported Forel's (1903) record as C. nuda.
Discussion
Most Old World records of C. minutior have come from oceanic islands with few or no native ants, e.g., Hawaii, Easter Island, the Marquesas Islands, the Society Islands, the Seychelles, and the Chagos Archipelago. Cardiocondyla minutior, however, appears to be originally from the Indo- Malayan region, where all its closest relatives are found (Seifert 2003). Cardiocondyla minutior specimens from Sri Lanka, Okinawa, India, and Nepal are morphologically more variable than those collected in the rest of the world (Seifert 2003), providing further evidence that C. minutior is native to this region. There are surprisingly few records of C. minutior from the Indo-Malayan region (Fig. 2), though this may simply be a testament to how inconspicuous and difficult to identify these ants are. Many, if not most, published records of C. nuda from the Indo-Malayan region (e.g., Thailand; Jaitrong & Nabhitabhata 2005) may be based on misidentifications of C. minutior. In this case, the dates of the earliest known records of C. minutior in the Old World appear to provide little guidance concerning where this species originated.
Table 2.
Earliest known records for Cardiocondyla minutior from 27 new world areas. * = earliest record unconfirmed. ** = sole record unconfirmed. abs = Archbold Biological Station. wpmc = W.P. Mackay Collection. All other abbreviations as in table 1.
The success of Cardiocondyla species in exotic locales may be due, in part, to their ability of coexist with dominant invasive ants (Heinze et al. 2006). Ito (1953) reported C. minutior occurring in an area occupied by the Argentine ant, Linepithema humile (Mayr, 1868), in Opaeula, Hawaii and speculated that C. minutior “was perhaps tolerated or overlooked by the Argentine ant because of its minute size.” Wilson & Taylor (1967) noted that L. humile “excludes other larger ant species, including the formidable Pheidole megacephala. One species found to be compatible with it on Hawaii is the diminutive Cardiocondyla nuda” (possibly referring to C. minutior, which they had newly synonymized with C. nuda). Ward (2005) reported that both C. mauritanica and C. minutior “are able to survive in sites invaded by the Argentine ant,” L. humile.
Five species of Cardiocondyla have become cosmopolitan tramps and have achieved widespread distributions in both the Old World and the New World: C. emeryi, C. mauritanica, C. minutior, C. obscurior, and C. wroughtonii (Seifert 2003). Several additional Cardiocondyla species have begun to spread outside their native range, including C. kagutsuchi, C. tjibodana, and C. venustula. None of these species, however, are known to have significant ecological impacts, and it seems unlikely that any of these inconspicuous sneaking ants will ever become significant pests as they continue to spread, largely unnoticed, around the world.
Acknowledgments
I thank M. Wetterer, B. Seifert, and J. Heinze for comments on this manuscript; S. Cover (MCZ) and M. Deyrup (ABS) for help, encouragement, and ant identification; P. Ward, J. Chaul, and D. McDonald for providing unpublished records; M. Deyrup, J. King, Z. Prusak, and J. A. Rodríguez for additional information; B. Guénard for sending overlooked references; W. O'Brien for GIS help; D. P. Wojcik and S. D. Porter for compiling their valuable FORMIS bibliography; R. Pasos and W. Howerton of the FA U library for processing so many interlibrary loans; Florida Atlantic University and the National Science Foundation (DES-0515648) for financial support.