1 | INTRODUCTION
Invasive ants often reach extremely high densities, outcompete and prey upon native species, disrupt mutualisms, and lower ecosystem biodiversity (Berman, Andersen, Hély, & Gaucherel, 2013; Holway, 1998; Holway, Lach, Suarez, Tsutsui, & Case, 2002; LeBrun, Abbott, & Gilbert, 2013; McGlynn, 1999; Porter & Savignano, 1990). Understanding the factors that promote the success of invasive ants is critical to discerning and managing their ecological impacts. Characteristics that are thought to play a role in the success of many invasive ant species include reduced nestmate recognition and an absence of boundaries between unrelated nests (Eyer & Vargo, 2021; Holway et al. 2002; Passera 1994). By avoiding the costs of colony defense and competition against neighboring conspecifics, these invasive ant species can reach higher densities (Giraud, Pedersen, & Keller, 2002; Porter, Fowler, & Mackay, 1992) and achieve greater ecological dominance by more effectively outcompeting other species (Holway et al., 2002; Holway & Suarez, 2004; LeBrun et al., 2013). For example, the number of Argentine ant workers (Linepithema humile ) was approximately 50-fold higher in sites where nests were interconnected compared with sites where nests defended distinct boundaries and competed with conspecifics (Holway & Suarez, 2004). Consequently, interconnected nests of Argentine ants more effectively outcompeted native ants, as native ant species richness was reduced by over 50% compared with sites where nests competed with each other (Holway & Suarez, 2004). Reduced intraspecific competition has also been implicated in the success and ecological impacts of other invasive and non-invasive social insects (Hanna et al., 2014; Korb & Foster, 2010; Perdereau et al., 2015, Wilson, Mullen, & Holway, 2009).
Colony boundaries and nestmate recognition are typically delimited using worker aggression assays, but there is increasing evidence that workers can discriminate nestmates from non-nestmates without an aggressive response (Breed 2003). For example, although Argentine ant workers do not aggressively attack non-nestmates from within the same supercolony (Giraud et al., 2002; Tsutsui, Suarez, Holway, & Case, 2000), they spend more time antennating non-nestmates than nestmates (Björkman-Chiswell, Van Wilgenburg, Thomas, Swearer, & Elgar, 2008), indicating nestmate recognition despite a lack of aggression. Perhaps as a consequence of nestmate recognition, sharing between Argentine ant nests was consistently limited to distinct clusters of nests within a single supercolony over a three-year-period (Heller, Ingram, & Gordon, 2008). By preferentially sharing food resources with nestmates over non-nestmates, workers may increase their inclusive fitness, particularly if nestmates are more related to them (Hamilton, 1964; Helanterä et al., 2009). Because aggression bioassays do not always reliably indicate nestmate recognition and colony boundaries, it is important to use alternative methods to assess intraspecific interactions in the field, such as genetic analyses and direct quantification of resource exchange between nests (Ellis, Procter, Buckham-Bonnett, & Robinson, 2017). Assessing some of the more subtle interactions between and within nests improves our understanding of the factors enhancing ant invasions and has important implications for invasive ant management.
We examined worker interactions between nestmates and non-nestmates in red imported fire ants (Solenopsis invicta ; hereafter fire ants). Fire ants occur in two social forms: the polygyne form (i.e., colonies with multiple egg-laying queens) and the monogyne form (i.e., colonies with only a single egg-laying queen; Gotzek, Shoemaker, & Ross, 2007; Ross, 1993; Ross, Vargo, & Keller, 1996; Tschinkel, 2006). These two social forms are under the control of an inversion-based supergene, which spans over 13 Mb of a “social chromosome” (Muers 2013; Wang et al., 2013). This social chromosome contains over 400 protein-coding genes (including Gp-9, which has been used as a marker to estimate the social form of the colonies), and experiences greatly reduced recombination (Arsenault et al., 2020; Yan et al., 2020). Interestingly, the fire ant supergene (and all the complex traits associated with it) exhibits two haplotypes (SB and Sb), which are passed on via Mendelian inheritance (Arsenault et al., 2020; Keller & Ross, 1998; Ross & Shoemaker, 2018; Wang et al., 2018). Colonies bearing theSb supergene haplotype express the polygyne phenotype (i.e., colonies accept multiple SB/Sb queens and reject any SB/SBqueens); whereas colonies bearing exclusively the SB haplotype express the monogyne phenotype (i.e., colonies accept only oneSB/SB queen and reject all SB/Sb queens; Arsenault et al., 2020; Gotzek & Ross, 2008, 2009; Ross & Keller, 2002). Moreover, supergene control appears to be complete, as social organization is independent of genetic diversity within the colony (Gotzek & Ross, 2008), as well as non-genetic factors such as environmental odors, queen reproductive status, and prior social experience of the workers (Gotzek & Ross, 2007; Ross & Keller, 2002).
Because colonies require only a small number (10-15%) of workers of theSb haplotype to express the polygyne phenotype, workers appear to regulate social organization (Gotzek & Ross, 2008). One possible mechanism for worker control includes habituating the colony to an odor unique to b -carrying adults (Gotzek & Ross, 2008). This hypothesis is supported by a lack of nestmate recognition between polygyne workers in the field (Vander Meer, Obin, & Morel, 1990) and a supposed absence of colony boundaries within polygyne populations throughout their invaded range in the USA (Bhatkar & Vinson, 1987). As a consequence, North American polygyne fire ant populations are often referred to as unicolonial (e.g., Greenberg, Vinson, & Ellison, 1992; Holway et al., 2002; Morel, Vander Meer, & Lofgren, 1990; Plowes, Dunn, & Gilbert, 2007; Porter et al., 1992; Vander Meer, Obin, & Morel, 1990). The exchange of workers and resources between nests in polygyne fire ants is thought to correspond with a greater abundance compared with the monogyne form due to reduced intraspecific competition (Porter, Bhatkar, Mulder, Vinson, & Clair, 1991). For example, polygyne mounds were over twice as abundant on average compared with monogyne mounds in Texas (mean ± SE: 680 ± 475 polygyne mounds/ha vs. 295 ± 240 monogyne mounds/ha; Porter et al., 1991). The greater abundance of the polygyne form may increase the likelihood of ants interacting with and preying upon native species, thereby increasing their ecological impact (Allen, Epperson, & Garmestani, 2004; Porter & Savignano, 1990).
Despite the assumption that polygyne nests are highly interconnected (see Bhatkar & Vinson, 1987), the physical exchange of workers and resources between polygyne nests in the field is poorly documented. Moreover, although polygyne workers from different nests do not aggressively attack each other in bioassays (Vander Meer et al., 1990), their interactions within the nest are relatively unknown, particularly in the case of worker-brood interactions. Prior research on within-colony interactions has focused almost exclusively on worker-queen interactions (DeHeer & Ross, 1997; Gotzek & Ross, 2008; Ross & Keller, 2002), but worker-brood interactions are also critical to colony dynamics and can differ from worker-queen interactions. For example, although Formica argentea workers in polygyne colonies show no preference towards related or unrelated queens, they preferentially care for brood that are more closely related to them (Snyder, 1993). Within-colony relatedness between polygyne fire ant workers is often near zero throughout their invaded range in the USA (DeHeer & Ross, 1997; Goodisman, Sankovich, & Kovacs, 2007; Ross, 1993; Ross & Fletcher, 1985; Ross et al., 1996), but workers may increase their inclusive fitness by preferentially caring for more related brood.
Our study tests fundamental assumptions about inter- and intracolonial interactions in introduced populations of fire ants. First, we compared colony boundaries between the two social forms in the field. To delimit boundaries between colonies, we quantified the exchange of a15N-glycine tracer dissolved in a sucrose solution and correlated this exchange with colony genetic structure. Using a labeled resource in combination with genetic data allows for two different ways to define colony boundaries (Ellis et al., 2017). We also examined polygyne brood-tending behaviors towards nestmates and non-nestmates in the laboratory. By studying interactions between and within colonies ofS. invicta , we further elucidate the primary factors influencing the ecology and success of this invasive species.