Taxonomy and Distribution of the Argentine Ant, Linepithema Humile (Hymenoptera: Formicidae)
Total Page:16
File Type:pdf, Size:1020Kb
SYSTEMATICS Taxonomy and Distribution of the Argentine Ant, Linepithema humile (Hymenoptera: Formicidae) ALEXANDER L. WILD Department of Entomology, University of California at Davis, Davis, CA 95616 Ann. Entomol. Soc. Am. 97(6): 1204Ð1215 (2004) ABSTRACT The taxonomy of an invasive pest species, the Argentine ant, is reviewed. Linepithema humile (Mayr) 1868 is conÞrmed as the valid name for the Argentine ant. Morphological variation and species boundaries of L. humile are examined, with emphasis on populations from the antÕs native range in southern South America. Diagnoses and illustrations are provided for male, queen, and worker castes. Collection records of L. humile in South America support the idea of a native distribution closely associated with major waterways in lowland areas of the Parana´ River drainage, with recent intro- ductions into parts of Argentina, Brazil, Chile, Colombia, Ecuador, and Peru. KEY WORDS Argentine ant, Linepithema humile, taxonomy, invasive species THE ARGENTINE ANT, Linepithema humile (Mayr) 1868, MCSN, MCZC, MHNG, MZSP, NHMB, NHMW, and is among the worldÕs most successful invasive species. USNM; see below for explanation of abbreviations). This native South American insect has become a cos- Taxonomic confusion over L. humile extends be- mopolitan pest, particularly in the Mediterranean cli- yond museum collections. At least one important mates of North America, Chile, South Africa, Austra- study, seeking to explain Argentine ant population lia, and southern Europe (Suarez et al. 2001). regulation in the native range through phorid para- Argentine ants have been implicated in the decline of sitism (Orr and Seike 1998), initially targeted the native arthropod (Cole et al. 1992) and vertebrate wrong Linepithema species (Orr et al. 2001). Errone- faunas (Suarez and Case 2002) and in the alteration of ous conclusions from that study were later perpetu- plant community structure (Christian 2001), and they ated in the invasion biology literature (e.g., Chapin et are also agricultural (Haney et al. 1987) and house al. 2000, Feener 2000). pests (Gordon et al. 2001). The genus Linepithema itself is a well-deÞned Several recent studies have been directed at the monophyletic group supported by several morpho- causes of the Argentine antÕs invasive success (Human logical synapomorphies (Shattuck 1992). IdentiÞca- and Gordon 1996, Orr and Seike 1998, Tsutsui et al. tion of specimens to genus can be accomplished with 2000, Giraud et al. 2002). Several of these hypotheses, the keys of Shattuck (1992) and Bolton (1994). How- particularly those of Orr and Seike (1998), Tsutsui et ever, species limits within Linepithema are poorly al. (2000), and Giraud et al. (2002) invoke a contrast known, and the only species-level identiÞcation key between some aspect of Argentine ant biology in its (Santschi 1929) is out of date and unusable. Since native range and Argentine ant biology in its intro- Mayr Þrst described Linepithema fuscum in 1866, 28 duced range. Such native/non-native comparisons re- species-level names have been assigned to Linepi- quire sufÞcient knowledge of the ant in both ranges. thema (Bolton 1995). There has been no effort to Unfortunately, the Argentine ant in its native range is synthesize these isolated descriptions into a coherent little understood compared with a wealth of knowl- taxonomy, and it remains a challenge to identify any edge accumulated from introduced populations. Linepithema species, including L. humile. Inadequate taxonomy for L. humile is at least partly The native distribution of the Argentine ant is some- responsible for hindering our understanding of the what better understood. Several lines of evidence Argentine ant in South America. Researchers have point to the Parana´ River basin in subtropical South displayed an unfortunate tendency to misdiagnose America as being the region of origin. First, Argentine other Linepithema species as L. humile, probably ow- ants from this area have high levels of genetic diversity ing to a combination of the high visibility of L. humile compared with populations elsewhere, and the ge- in the literature and a perplexing similarity in the netic diversity of introduced populations seems to be worker caste between Linepithema species. Examina- a subset of the Parana´ drainage diversity (Tsutsui et al. tion of major entomological collections reveals worker 2000, Tsutsui and Case 2001). Second, Argentine ants specimens of multiple Linepithema species stored un- are often found in relatively pristine natural areas der the humile label (personal observations; in LACM, within the Parana´ drainage coexisting with other ant 0013-8746/04/1204Ð1215$04.00/0 ᭧ 2004 Entomological Society of America November 2004 WILD:TAXONOMY AND DISTRIBUTION OF L. humile 1205 species (Holway and Suarez 2004; personal observa- Buenos Aires, Argentina; MCSN, Museo Civico de tion), even though they also inhabit areas of human Historia Natural ÔGiacomo DoriaÕ, Genoa, Italy; disturbance. Finally, subtropical South America holds MCZC, Museum of Comparative Zoology, Cambridge, the highest diversity of Linepithema species (Shattuck MA; MHNG, Muse´um dÕHistoire Naturelle, Geneva, 1992) and seems to be a region of radiation for this Switzerland; MZSP, Museu de Zoologia da Univer- group. sidade de Sa˜o Paulo, Sa˜o Paulo, Brazil; NHMB, Despite the evidence favoring a native Parana´ Naturhistorisches Museum, Basel, Switzerland; drainage distribution, reliable records of Argentine NHMW, Naturhistorisches Museum Wien, Vienna, ants from this region are sparse. The genetic work of Austria; PSWC, Philip S. Ward personal collection, Tsutsui et al. (2000) focused on the southern Parana´ Davis, CA; QCAZ, Museo de Zoologõ´a de la PontiÞcia drainage, and beyond their collections few museum or Universidad Cato´lica del Ecuador, Quito, Ecuador; literature records have been veriÞed by genetic study UCDC, R. M. Bohart Museum of Entomology, Davis, or by trained morphologists. Details about the distri- CA; USNM, National Museum of Natural History, bution of this ant within the region are not well Washington, DC; and WPMC, William P. MacKay known, particularly in the northern Parana´ drainage, personal collection, El Paso, TX. whereas literature records based on dubious identiÞ- Morphological Analysis. Most observations were cations of L. humile from outside the Parana´ drainage made at 50ϫ on a Wild stereomicroscope. I conducted cloud our understanding of the true limits of the native morphometric measurements on a subset of male (n ϭ distribution. 75), queen (n ϭ 43), and worker specimens (n ϭ 364). Here, I clarify the taxonomy and the native distri- A majority of measurements were taken using a dual- bution of the Argentine ant. SpeciÞcally, this study axis Nikon stage micrometer with a precision of 0.001 examines the status of the scientiÞc name L. humile mm, but measurements at IFML and MZSP used an (Mayr) 1868 as it applies to the Argentine ant, deÞnes ocular micrometer with a precision of 0.01 mm. I the morphological limits of the species, creates a re- report measurements here to 0.01 mm. I repeated liable morphological diagnosis, and compiles veriÞed measurements on several specimens by using both collection records into the most complete data set yet optical and stage micrometers to conÞrm that mea- of this speciesÕ South American distribution. The goal surements were consistent between systems. of this study is to provide researchers with an unam- I used a number of standard morphometric char- biguous method for identifying Argentine ants and a acters. Head measurements are given with the head in baseline of knowledge about their native range dis- full-face view, with the anterior clypeal margin and tribution. the posterior border of the head in the same focal plane. I consider ant heads to be prognathous, such Materials and Methods that the clypeus is anterior and the frontal area is Specimens. I examined 6,540 worker, 249 queen, dorsal. and 366 male Linepithema specimens collected across Head Length (HL). In full-face view, the midline the global distribution of the genus. Particular atten- distance from the level of the maximum posterior tion was paid to ants collected in South America, projection of the posterior margin of the head to the including specimens from the Orr and Seike (1998) level of the most anterior projection of the anterior study that reported phorid ßy parasitism of L. humile clypeal margin. In males, I consider the posterior mar- in Brazil. Specimens were examined during visits to gin of the head as the vertex between, and not includ- several entomological museums and through institu- ing, the ocelli. tional and personal loans. Additionally, I observed and Head Width (HW). In full-face view, the maximum collected Linepithema in the Þeld in Argentina, Par- width of the head posterior to the compound eyes. aguay, Ecuador, Guatemala, the Dominican Republic, Minimum Frontal Carinal Width (MFC). In full- Jamaica, Puerto Rico, South Africa, and the western face view, the minimum distance between the frontal United States on several occasions from 1996 to 2002, carinae. in expeditions ranging in length from a few days to Antennal Scape Length (SL). Measured from the several months. These collections consisted mainly of apex of the Þrst antennal segment to the base, exclu- extensive visual searches targeting Linepithema nests sive of the radicle. and foraging ants, sometimes augmented with honey Profemur Length (FL). In posterior view, measured baits, Berlese funnels,