Supercolonial structure of invasive populations of the tawny crazy ant Nylanderia fulva in the US
Social insects are among the most serious invasive pests in the world, particularly successful at monopolizing environmental resources to outcompete native species and achieve ecological dominance. The invasive success of some social insects is enhanced by their unicolonial structure, under which the presence of numerous queens and the lack of aggression against non-nestmates allow high worker densities, colony growth, and survival while eliminating intra-specific competition. In this study, we investigated the population genetics, colony structure and levels of aggression in the tawny crazy ant, Nylanderia fulva, which was recently introduced into the United States from South America.
We found that this species experienced a genetic bottleneck during its invasion lowering its genetic diversity by 60%. Our results show that the introduction of N. fulva is associated with a shift in colony structure. This species exhibits a multicolonial organization in its native range, with colonies clearly separated from one another, whereas it displays a unicolonial system with no clear boundaries among nests in its invasive range. We uncovered an absence of genetic differentiation among populations across the entire invasive range, and a lack of aggressive behaviors towards conspecifics from different nests, even ones separated by several hundreds of kilometers.
Overall, these results suggest that across its entire invasive range in the U.S.A., this species forms a single supercolony spreading more than 2000 km. In each invasive nest, we found several, up to hundreds, of reproductive queens, each being mated with a single male. The many reproductive queens per nests, together with the free movement of individuals between nests, leads to a relatedness coefficient among nestmate workers close to zero in introduced populations, calling into question the stability of this unicolonial system in which indirect fitness benefits to workers is apparently absent.
KeywordsInvasive species Mating system Colony structure Supercolonies Social insects Ants
Cytochrome oxidase 1
Complementary sex-determining (locus)
Understanding the evolutionary factors affecting population structure and ecological assemblage is a central question in ecology. This question is especially complex in the context of biological invasions, as introductions usually prompt severe shifts in genetic structure and life history traits of invasive species, and profoundly disturb the ecological community of native species [1, 2]. Population bottlenecks associated with founder events following introductions reduce genetic diversity and may lead to inbreeding, while novel abiotic and biotic pressures in invaded environments require a rapid and efficient response to the new selective forces . Uncovering the mechanisms by which biological invasions induce post-introduction phenotypic changes in life history traits that allow invasive species to overcome the loss of genetic diversity and the reduced adaptive potential to successfully establish and achieve local dominance in a new environment remain important areas of study [4, 5, 6].
Social insects are among the most abundant and successful species at invading terrestrial environments. Despite the taxonomic diversity of invasive species, most of them share life history traits that may facilitate their introduction and dominance in new ecosystems . Among these, the invasion success of social insects is often associated with a unicolonial social system, under which the absence of aggressive behavior towards non-nestmates allows free mixing of individuals (workers, brood and queens) among geographically distant nests . Unicoloniality reduces intra-specific competition; this may allow high worker densities and an increased colony growth and survival due to the presence of many reproductive queens within nests. This social organization allows invasive populations to efficiently monopolize environmental resources and rapidly outcompete native species to achieve local dominance [2, 8].
Unicoloniality is a common trait in invasive ant species and several hypotheses have been proposed to account for its evolution in introduced populations [9, 10]. A first hypothesis posits that the loss of genetic diversity in bottlenecked populations lowers nestmate recognition reducing differentiation between colonies. If nestmate recognition is based on heritable cues [11, 12, 13], the overall loss of genetic diversity in introduced populations reduces the diversity at the recognition locus (or loci) and homogenizes recognition templates among colonies. Ultimately, if polymorphism at the recognition locus (or loci) is lost, unicoloniality could arise through the inability of workers to discriminate against non-nestmate conspecifics [2, 14, 15]. A second hypothesis suggests that high nest density in the introduced range has selected for reduced nestmate recognition to avoid recurrent fights with their neighbors [9, 16]. The relaxed environmental pressures in the introduced range often lead to high nest densities and increase the rates of encountering non-nestmate conspecifics . This could have selected for lower recognition cues if nonaggressive neighboring colonies attain higher worker number and outcompete aggressive ones [18, 19, 20]. A third hypothesis proposes that the species is already polydomous and polygynous, forming small supercolonies in its native range [9, 21]. It suggests that the absence of conspecific competition in the introduced range has enabled the invasive colony to grow extremely large . This scenario requires minimal evolutionary changes, since only the extent of the supercolony changes but not the behavior of workers. Whatever the evolutionary forces triggering unicoloniality [2, 9, 10], the loss of aggressive behavior toward non-nestmates results in the development of supercolonies, a social organization formed from a network of several interconnected nests without clear boundaries between them .
Supercolonies may extend across large geographic distances, with populations ultimately consisting of a single, vast supercolony with no aggression towards colony-mates [2, 23, 24, 25, 26, 27]. In these populations, the combination of free exchange of individuals between nests and the occurrence of hundreds or even thousands of reproductive queens per nest results in extremely low relatedness between colony-mates that often approaches zero [28, 29, 30, 31, 32, 33]. However, studies on a broader geographical scale reveal genetic differentiation may exist such that the entire population comprises several supercolonies [9, 21, 31, 34, 35, 36, 37]. The presence of several supercolonies or limited dispersal abilities within a supercolony may reduce gene flow between nests , and restore relatedness among colony-mates or lead to hot spots of locally elevated relatedness between nestmates within supercolonies .
The tawny crazy ant, Nylanderia fulva, is native to South America from Brazil to Argentina along the border of Uruguay and Paraguay . This species has been introduced into Peru, Colombia and the Caribbean [39, 40], and recently was documented in the U.S.A., rapidly spreading across Florida, southern Mississippi, southern Louisiana and Texas [41, 42]. From the 1950’s to the 2000’s, this species was only reported in Florida (under the synonym Paratrechina pubens). It was later uncovered in Texas during a sudden outbreak in 2002; and more recently in Mississippi in 2009 . Field observations in the U.S.A. introduced range revealed that populations consist of dense networks of polygynous nests (0 to 5 reproductive queens, ), with ants freely moving among them without any aggression between non-nestmates [44, 45]. No nuptial flights have been observed in the invasive range, suggesting that the invasion front advances by nest fission (or budding), where queens establish new nests with the help of workers within walking distance of the natal nest [41, 46]. However, these studies only used field observations with limited behavioral tests; genetic studies are lacking to clearly determine the population genetics, colony structure and aggression patterns of the tawny crazy ant within its introduced range, and no information is available from the species’ native range.
In this study, we conducted large-scale genetic and behavioral analyses of the invasive tawny crazy ant. We first investigated patterns of population genetic structure within its native and introduced ranges to estimate the extent of the genetic diversity loss stemming from the founder effect following its introduction in the U.S.A. Second, we investigated the reproductive system of this species in its introduced range, assessing the number of queens per nest, the number of matings per queen, the possibility that queens reproduced through thelytokous parthenogenesis, and the relatedness among nestmate workers. A comparison of colony genetic structure in the native and introduced ranges allowed us to determine whether the recent introduction of N. fulva induced a shift in its social system, from multicoloniality to unicoloniality. Lastly, we performed behavioral assays testing whether workers from different colonies within the invasive range recognize each other as colony mates in order to define the number and the extent of the supercolonies observed in the U.S.A.
A total of nine mitochondrial haplotypes were found. All haplotypes were found in the native range, while only two were uncovered in the introduced range. In the native range, the mean genetic distance within group was 0.034 (20.04 bp difference on average), but only 0.002 (1.45 bp difference in average) in the introduced range.
Population and colony structure
Analysis of molecular variance (AMOVA) at different hierarchical levels for both native (Nat.) and introduced (Int.) populations of N. fulva
Source of variation
Sum of squares
Among nests within localities
Reproductive system and genetic relatedness
Colonies of N. fulva in the introduced range were spatially expansive, making a complete excavation and a precise count of queen number impossible. Despite this, 0 to 20 queens were typically found in most of the nests sampled, with up to 300 queens discovered from a single nest. Moreover, the assignment of worker genotypes was not compatible with the occurrence of a single queen for any of the nests sampled. This suggests that multiple queens shared reproduction and/or that workers freely moved between nests in the U.S.A.
A total of seven queens were successfully isolated with a group of workers into subcolonies and produced enough progeny to reliably infer whether polyandry occurs through mother-offspring comparisons. All the genotypes of worker pupae were unequivocally assigned to the genotype of the putative mother queen, and were consistent with a single mating of the queen in all seven subcolonies analyzed. Despite the low level of genetic diversity in the introduced range, the probability of non-detection of two males as a result of carrying the same alleles at all loci is very low (P non-detection = 2.99 × 10− 6).
The above findings indicate that introduced populations of N. fulva have highly polygynous nests containing up to hundreds of queens, each of them being singly mated. As a result, together with its supercolonial structure, the mean relatedness among nestmate workers is close to zero in the introduced range (r w-wa ± SD = 0.04 ± 0.05). This finding is in sharp contrast with the high relatedness among nestmates observed in the native range (r w-w ± SD = 0.57 ± 0.19) (Fig. 5). Interestingly, the relatedness in the introduced range differs from zero (r w-w = 0.16) when the global population is taken as a reference.
Our large-scale genetic and behavioral analyses of the invasive tawny crazy ant provide several new insights into the biological invasion and social system of this ant in its native and invasive (U.S.A.) ranges. Genetic data suggest this ant species experienced a genetic bottleneck following its introduction in the U.S.A. that led to a significant reduction (60%) in genetic diversity. Population genetic analyses show that N. fulva exhibits a multicolonial organization in its native range, with colonies genetically distinct from each another. In contrast, we found that this species displays a unicolonial system with no clear boundaries between nests in its invasive range. This latter finding is supported by the lack of genetic differentiation among nests within populations as well as between geographic populations, and relatedness coefficients among nestmate workers close to zero in introduced populations. Each invasive nest was headed by several, up to hundreds of singly-mated reproductive queens. Behavioral tests reveal no aggressive behaviors toward conspecifics from different nests, even ones separated by several hundred kilometers. Overall, these results suggest that the entire U.S.A. range of the species forms a single large supercolony spreading more than 2000 km.
Population bottleneck and inbreeding
In our study, we uncovered a loss of genetic diversity between native and introduced populations. Such reduction may be particularly costly for hymenopteran species because of their sex determination system. In these species, the sex of an individual is controlled by a single complementary sex-determining locus (multi-locus CSD is known but rare ). Heterozygous individuals at this locus develop into females while homozygous individuals develop into males. Females are diploid heterozygous individuals usually produced by sexual reproduction, whereas males arise from unfertilized eggs through arrhenotokous parthenogenesis, and are therefore haploid (i.e., thus homozygous) individuals . However, diploid individuals, homozygous at this locus, which can result from mating between individuals carrying the same sex allele (matched mating), develop as diploid males. Production of diploid males represents a cost for colonies because they are effectively sterile in most hymenopteran species [48, 49, 50, 51]. Loss of allelic diversity at the sex locus as a result of a population bottleneck significantly increases the chances of matched matings . As one example, in the introduced populations of the red fire ant Solenopsis invicta, colonies produce a higher proportion of diploid males than those from native populations [53, 54].
Some ant species have however evolved unorthodox reproductive modes, which may facilitate invasiveness by acting as pre-adaptations against the genetic loss due to bottlenecks during invasions . In some populations of four invasive ant species, Wasmannia auropunctata, Vollenhovia emeryi, Anoplolepis gracilipes and Paratrechina longicornis, queens are clones of their mothers and males are clones of their fathers, whereas workers arise from classical sexual reproduction [35, 55, 56, 57]. Male and female gene pools are completely segregated, even those produced by the same mother queen [58, 59]. In these species, a single-mated queen may thus establish an introduced population, producing 100% heterozygous workers. This queen may also produce new queens and males able to mate together inside the nest; yet still maintain heterozygosity in their worker offspring. Clonality was recently recorded in the native range of W. auropunctata in southern South America [60, 61]. This strategy thus circumvents the costs of inbreeding after an introduction event over an unlimited number of generations [55, 59] and act as a pre-adaptive trait to invasion. In our study, the level of heterozygosity is not significantly different between workers and queens of N. fulva, indicating that they are both produced through classic sexual reproduction.
Formation of supercolonies seems to be a common trait of invasive social insects [22, 62], allowing a rapid and efficient monopolization of resources to achieve local dominance, mainly in the introduced range where the competitive pressure exerted by the members of local ant community is lower than in the native range [2, 8, 63, 64]. Supercolonies have been reported numerous times in various invasive ant species, such as Linepithema humile , Monomorium pharaonis , Pheidole megacephala , Anoplolepis gracilipes , and Lasius neglectus , and has also been reported in invasive populations of the termite Reticulitermes urbis . Interestingly, the sizes of supercolonies apparently varies considerably among species, and occasionally even within species. For example, the invasive big-headed ant, P. megacephala, forms a single large supercolony covering up to 3000 km across northeastern Australia . In contrast, the yellow crazy ant, A. gracilipes, inhabiting a small geographic area in northeastern Borneo, comprises at least six supercolonies . In L. humile, two supercolonies are reported in the invasive range in southern Europe; one supercolony is 6000 km long, while the other is only a few km long . In this same species, the invasive area of California comprises at least five supercolonies ranging in areas from 1 to 1000 km , while four supercolonies were uncovered in Japan [27, 70], and several in the southeastern U.S.A. . Overall, these results suggest supercoloniality is a common trait in social insects, but the number and the size of their supercolonies can differ greatly among and within species.
Despite the lack of aggressiveness within supercolonies, genetic identities of two adjoining supercolonies can be maintained because workers display strong aggression towards allocolonial sexuals and workers at colony boundaries, as observed in the Argentine ant [69, 72, 73]. This aggression towards allocolonial sexuals strongly reduces potential for mating between partners from distinct supercolonies. Thus, gene flow is reduced between abutting supercolonies, resulting in maintenance of genetic distinctiveness even after prolonged contact with one another [9, 21, 69, 74]. Supercolony differentiation has been suggested to come about one of two ways. On the one hand, supercolonial structure may stem from an initial colony differentiation, in which different supercolonies came from multiple introductions. These distinct introductions from genetically and chemically differentiated source populations are more likely to result in distinct supercolonies in the invasive range [75, 76]. For example, the worldwide supercolonies of the Argentine ant originated from at least seven founding events out of the native area in Argentina . The dominant supercolonies of Europe, Japan and California probably arose from the same primary introduction  and consist of a single supercolony that globally expanded through secondary introductions, since these populations are not aggressive toward each other  and have similar hydrocarbon profiles . On the other hand, supercolony differentiation may occur through divergence after introduction. Queen recruitment, intranidal mating and female dispersal through budding may lead to a reduction of gene flow between geographically separated fragments of the same initial supercolony. Over time, the accumulation of genetic and cuticular hydrocarbon (CHC) differentiation may result in mutual aggression between fragments . In the introduced population of L. humile in Corsica, the clear reduction of gene flow between the island and the mainland supercolonies has led to noticeable chemical and behavioral differentiation . A similar pattern has been reported in a single population in the Californian invasive range, yet coming from the introduction event that gave rise to the other supercolonies , and in A. gracilipes in Borneo, in which spatial separation has enhanced genetic and CHC differentiation over time . Eventually, these cases may result in allopatric fragmentation if enough differentiation occurs before both fragments come into contact again. However, no pattern of isolation by distance has been found within supercolonies of other invasive ant species [9, 34, 71, 74], suggesting that gene flow is high enough in these supercolonies to prevent differentiation among geographically distant areas within supercolonies.
In N. fulva, no abrupt genetic transition was discovered across all introduced N. fulva populations studied, suggesting that this species forms a single large supercolony from Texas to Florida. We uncovered a weak pattern of isolation by distance in the introduced range, several orders of magnitude lower than that in the native range. This result may stem from the absence of nuptial flights in N. fulva and its invasion front expansion through budding. These features usually lead to a genetic population viscosity, which may, over time, result in population differentiation. The invasion of N. fulva in the U.S.A. is recent, and may not have had sufficient time to induce genetic differentiation between localities and split the invasive range into distinct supercolonies. Although we cannot exclude that other supercolonies of N. fulva are present but were not sampled, our results suggest that introduced populations in the U.S.A. may come from a single introduction from South America, which then spread through human mediated jump dispersal. The hypothesis of a single introduction is also supported by the positive relatedness (r w-w = 0.16) observed when the global population is taken as reference.
Unicoloniality results in several ecological advantages in terms of colony growth, nest density, productivity and survival, and may favor invasive success outcompeting native species through resource monopolization . But on the other side of the coin, unicoloniality reduces to zero the relatedness between nestmate workers, and, thus, the workers’ indirect benefits from helping. In this context, selfish behaviors are expected to disrupt social cohesion within colonies [81, 82]. For this reason, unicoloniality might represent an evolutionary dead-end; an idea supported by the fact that there is no unicolonial species but only unicolonial populations, and by its scattered distribution along the ant phylogeny . In N. fulva, the relatedness between nestmate workers did not differ from zero in the introduced range, while it varied from 0.29 to 0.86 in the native range. A similar loss of relatedness has also been reported in several supercolonial populations of the species L. humile, L. neglectus, P. megacephala, and S. invicta [2, 32, 83, 84]. However, most of these species were comprised of several genetically distinct supercolonies, with members of the same colony more related to each other than to members of other supercolonies. In this case, it is important to measure relatedness with respect to the local competing population rather than to the global population . Taking the two supercolonies of L. humile in the Southern European range as an example, it is unlikely that two workers separated by thousands of kilometers still compete with each other. Therefore, in most parts of the supercolony, workers most likely compete with colonymates; while selection for altruism should only take place at colony boundaries . In our study, the whole introduced range seems to comprise a single supercolony, even if workers within the introduced range are more related to each other than to workers from native populations, introduced workers do not compete with native workers, making introduced relatedness equivalent to zero. In contrast, several supercolonies of A. gracilipes inhabit the island of Borneo . In this limited area, workers of a given supercolony are more likely to compete against workers from other supercolonies. The relatedness coefficients observed in this species are quite high, making social cohesion sustainable when supercolony size is reduced . Actually, supercolonies of smaller size are uncovered in native populations of the Argentine ant [9, 21, 34, 85, 86] and the little fire ant [61, 87, 88]. In noninvasive species, the turnover of supercolonies suggests the occurrence of local competition . Overall, these outcomes suggest that unicoloniality is not only a derived trait in invasive populations, but might represent a sustainable social strategy when the reduction of relatedness outweighs its ecological advantages.
Overall, this study shows that like several unrelated ant species, introduced populations of N. fulva developed a unicolonial organization, giving another example of the independent evolution of this social structure in ants . Yet, the scattered distribution of unicoloniality along the ant phylogeny casts doubt on the long-term stability of this system, in which one might expect a complete breakdown of co-operation due to the absence of relatedness among nestmates. This study reports another ant species exhibiting plasticity in reproductive strategy and behavior that allows it to take advantage of the loss of genetic diversity in the invasive range. Further studies investigating whether native populations of this species consist of small localized colonies or smaller supercolonies should shed further light on whether the large supercolonies formed in the U.S.A. are due to a post-introduction shift in social structure or whether it is related to pre-adapted traits present in the native population.
A total of 36 populations of N. fulva were mainly collected between 2015 and 2017 (Fig. 1). Sampling comprised of 14 populations in its introduced range from Texas to Georgia, U.S.A. and 22 populations in its native range in South America (Additional file 2: Table S1). For each population, 1 to 8 nests were sampled (X ± SD = 1.7 ± 1.6; N = 65). Colonies from Texas populations were brought back alive to the laboratory, where they were maintained under standard conditions (28 ± 2 °C, 12:12 h photoperiod, and fed sugar water and cockroaches) for behavioral and breeding system analyses. A subset of individuals from the Texas colonies were then removed and stored in 95% ethanol. All samples from other U.S.A. localities and from South America were immediately stored in 95% ethanol for subsequent genetic analyses.
For each individual, total genomic DNA was extracted following a modified Gentra Puregene extraction method (Gentra Systems, Inc. Minneapolis, MN, USA). Thirteen new microsatellite markers (Additional file 3: Table S2) were developed for N. fulva based on the transcriptome generated by Valles et al. (2012; ). Amplicons were labelled with 6-FAM, VIC, PET or NED dye to facilitate multiplexing. PCR conditions and multiplexing arrangements are given in the online supplementary material (Additional file 3: Table S2). PCR were run on a Bio-Rad thermocycler T100 (Bio-Rad, Pleasanton, CA, U.S.A.). PCR products were sized against LIZ500 internal standard on an ABI 3500 capillary sequencer (Applied Biosystems, Foster City, CA, U.S.A.). Allele scoring was performed using Geneious v.9.1 . A fragment of the COI mitochondrial gene was also sequenced using the Jerry and Pat primer pair previously developed for Apis mellifera . PCR products were purified with EXOSAP-it PCR purification kit (Affymetrix), and sequenced using the ABI BigDye Terminator v.3.1 Cycle Sequencing Kit on an ABI 3500 Genetic Analyzer (Applied Biosystems). Base calling and sequence reconciliation were performed using CodonCode Aligner (CodonCode Corporation, Dedham, MA, U.S.A.).
Population and colony structure
For the mitochondrial dataset, we sequenced 1–3 workers for each population, which overall included 26 workers in 14 native locations and 14 workers in 13 introduced locations in the U.S.A. We included two samples of Nylanderia terricola from one location in Texas as an outgroup. The conservation of some samples in alcohol was not optimal, especially those from the native range, often resulting in poor quality DNA, and some samples could not be sequenced successfully. Haplotype network was used to visualize phylogeographic relationships between mitochondrial haplotypes. Networks were produced by the median-joining method  implemented in the program NETWORK v.22.214.171.124 (available at http://www.fluxus-engineering.com/). Nucleotide diversity and genetic distance were compared within and between populations using MEGA v. 5.0 .
For microsatellite analyses, 2–20 individuals per nest were genotyped at 13 microsatellite markers in each locality (X ± SD = 15.11 ± 5.22; n = 937; N = 63 nests). The number of alleles, allele frequencies, measures of observed and expected heterozygosity, and F-statistics were determined using FSTAT . We looked for evidence of a recent bottleneck by testing for an excess of heterozygotes with Bottleneck 1.2 . The loss of rare alleles after a bottleneck is expected to lead to excess heterozygosity compared with expectations under mutation-drift equilibrium . We used the two-phase model (TMP) to generate expected heterozygosity in Bottleneck 1.2.
For both native and introduced ranges, the hierarchical partitioning of the genetic diversity among localities, among nests within localities, and within nests was assessed using analysis of molecular variance (AMOVA) implemented in Arlequin . We assessed the level of genetic differentiation between localities by estimating genetic differentiation FST, and tested its statistical significance by a permutations test using FSTAT . We investigated population structure and isolation-by-distance by plotting [FST/(1 – FST)] coefficients between pairs of nests against the ln of their geographical distance (Slatkin 1993). The significance of the correlation was tested using Mantel tests implemented in GENEPOP ON THE WEB . We visualized population structure by plotting individuals on a Principal Component Analysis (PCoA) using FactoMineR R package . We tested for the presence of genetic structure within and among populations inferring the number of genetic clusters (K) in our samples using the Bayesian clustering method implemented in STRUCTURE v.2.3 . Simulations were run separately for all individuals with K ranging from 1 to 36, for individuals from the native range only (K from 1 to 22), and for individuals from the introduced range only (K from 1 to 14). The simulations were replicated 10 times for each number of K. The analyses were run using a combination of a correlated-allele frequencies and an admixture model. Each run comprised a first step of a 5 × 104 burn-in period and 1 × 105 iterations of the MCMC. The log-likelihood value and the ΔK method  implemented in Structure Harvester v.0.6.8  were used to estimate the most likely number of clusters. Finally, whether different nests within populations belonged to the same colony was determined by comparing genotypic frequencies at all loci with a log-likelihood (G)-based test of differentiation using GENEPOP ON THE WEB . The overall significance across loci was determined using a Fisher’s combined probability test after Bonferroni correction for multiple comparisons (α after Bonferroni correction = 0.00006).
Reproductive system and genetic relatedness
Relatedness coefficients (r) among nests were estimated using COANCESTRY v.1.0 , following the algorithm described by Queller & Goodnight (1989; ). Relatedness coefficients were calculated separately for the introduced and the native range to account for the differences in allele frequencies between populations. We also calculated relatedness within the introduced range using the global population as a reference for allelic diversity. Finally, we assessed the possibility that queens produce new queens via thelytokous parthenogenesis by comparing heterozygosity level and relatedness between castes in the introduced range using an extensive sampling of 16 nests from four localities in Texas (27, 30, 31 and 33). Thelytokous parthenogenesis through automixis decreases homozygosity over time [107, 108]. Parthenogenetic production of queens would lead to a difference in observed heterozygosity between queen and worker castes due to a decline of heterozygosity and increased relatedness among queens compared to the sexually produced workers.
Within a week of collection, standardized aggression tests were conducted by placing two workers in a 5 cm petri dish arena for 5 min. Workers were not starved before the beginning of the experiment. The arena floor was covered with filter paper to prevent odor transfer between replicates, and the sides were coated with Fluon to prevent escapes. Interactions were scored on a 5-level scale: levels 1 (ants still or huddled together) and 2 (antennation, allogrooming or trophallaxis) were considered non-aggressive behaviors, whereas levels 3 (biting and quickly releasing), 4 (prolonged biting > 3 s) and 5 (balling, fighting, spraying formic acid) were considered as agonistic. Aggression tests were conducted with three nests per locality for four localities separated by at least 30 m in the introduced range in Texas (localities 26, 27, 28 and 29; Fig. 1). Interactions were measured between nestmates, then between non-nestmates, either from the same or different localities. We also tested the interactions between one worker from each locality and a red imported fire ant worker (Solenopsis invicta), to control for the ability of N. fulva to be aggressive. Each combination was replicated three times, yielding a total of 36 encounter type assays between nestmates, 36 between non-nestmates from the same locality, 18 between non-nestmates from a different locality, and 12 against a fire ant. Aggression levels were compared using ANOVA tests between groups using R software . All figures were made using the free software Inkscape v.0.92 (available at http://www.inkscape.org/).
We thank E. Lebrun for providing samples from the native range.
Funding was provided by a grant from the Texas Invasive Ant Research and Management Seed Project to ELV and RP and by the Urban Entomology Endowed Chair fund at Texas A&M University. Most collections in southern South America were funded with grants from the Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT) and ARS-USDA to LAC. The funding agencies did not participate in the design of the study, the collection of the samples, the analysis, the interpretation of data, or in writing the manuscript.
Availability of data and materials
The datasets generated during the current study are available in the GenBank repository, accession numbers MH973808 to MH973849.
ELV and RTP designed the study. BM, MBF, LAC and PAE collected samples. DWS developed the microsatellites. PAE, BM, LNLJ, MBF performed the genetic analyses. PAE wrote the paper with contributions of LNLJ, LAC, RTP, DWS and ELV. All authors have read and approved the manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 10.Steiner FM, Schlick-Steiner BC, Moder K, Stauffer C, Arthofer W, Buschinger A, Espadaler X, Christian E, Einfinger K, Lorbeer E, et al. Abandoning aggression but maintaining self-nonself discrimination as a first stage in ant supercolony formation. Curr Biol. 2007;17(21):1903–7.PubMedCrossRefPubMedCentralGoogle Scholar
- 33.Pamilo P, Zhu D, Fortelius W, Rosengren R, Seppa P, Sundström L. Genetic patchwork of network-building wood ant populations. Ann Zool Fenn. 2005;42(3):179–87.Google Scholar
- 41.Meyers J, Gold R. Identification of an exotic pest ant, Paratrechina sp.nr. pubens (Hymenoptera: Formicidae), in Texas. Sociobiology. 2008;52:589–604.Google Scholar
- 42.MacGown J, Layton B. The invasive rasberry crazy ant, Nylanderia sp. near pubens (Hymenoptera: Formicidae), reported from Mississippi. Midsouth Entomologist. 2010;3:44–7.Google Scholar
- 43.Zenner-Polania I. Biological aspects of the “hormiga loca”, Paratrechina (Nylanderia) fulva (Mayr), in Colombia. In: Meer RKV, Jaffe K, Cedeno A, Boulder CO, editors. Applied Myrmecology: A World Perspective. USA: Westview Press; 1990. p. 290–7.Google Scholar
- 57.Ohkawara K, Nakayama M, Satoh A, Trindl A, Heinze Jr: Clonal reproduction and genetic caste differences in a queen-polymorphic ant, Vollenhovia emeryi. Biology Letters 2006, 2(3):359–363.Google Scholar
- 60.Rey O, Estoup A, Vonshak M, Loiseau A, Blanchet S, Calcaterra L, Chifflet L, Rossi JP, Kergoat GJ, Foucaud J, et al. Where do adaptive shifts occur during invasion? A multidisciplinary approach to unravelling cold adaptation in a tropical ant species invading the Mediterranean area. Ecol Lett. 2012;15(11):1266–75.PubMedCrossRefPubMedCentralGoogle Scholar
- 61.Chifflet L, Rodriguero MS, Calcaterra LA, Rey O, Dinghi PA, Baccaro FB, Souza JLP, Follett P, Confalonieri VA. Evolutionary history of the little fire ant Wasmannia auropunctata before global invasion: inferring dispersal patterns, niche requirements and past and present distribution within its native range. J Evol Biol. 2016;29(4):790–809.PubMedCrossRefPubMedCentralGoogle Scholar
- 91.Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper A, Markowitz S, Duran C, et al. Geneious basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012;28(12):1647–9.PubMedPubMedCentralCrossRefGoogle Scholar
- 109.Development Core Team R. Language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2016.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.