Introduction

A long-standing goal in ecology has been to determine the underlying mechanisms that give rise to species coexistence in local communities, especially in assemblages with multiple competing species (MacArthur 1958; Hutchinson 1959). Numerous mechanisms have been proposed for maintaining species coexistence (Wright 2002; Silvertown 2004). Interspecific competitive trade-offs, whereby the superiority of a particular species in an environment or biotic condition is balanced by the inferiority of other species, can lead to segregation among species (Tilman 1994; Levine et al. 2004). These interspecific interactions are thought to lead to the long-term stable coexistence of ecologically similar species (Levins 1979; Holt et al. 1994; Chesson 2000; Bever 2003; Rudolf and Antonovics 2005), and may also be characterized by dominance hierarchies. Dominance hierarchies have been widely observed in a wide range of taxa, from vertebrates to invertebrates (Chase and Seitz 2011). Species can be ranked into dominant species (i.e. higher-ranked individuals) or subordinate species (i.e. lower-ranked individuals) on the basis of aggression or ritual displays (Drews 1993). Interspecific dominance hierarchies have been used to understand patterns of local species coexistence in ecological communities with as consequence that higher ranked individuals monopolize resources resulting in fitness benefits (Morse 1974; Schoener 1983).
In ant communities, dominance hierarchies have been used to examine interspecific tradeoffs to explain species coexistence patterns (Stuble et al. 2013). These trade-offs include the discovery-dominance trade-off, the discovery-thermal tolerance tradeoff, and the discovery-colonization trade-off (Cerdá et al. 1998a ; Stanton et al. 2002; Lebrun and Feener 2007; Stuble et al. 2013). In addition to testing interspecific trade-offs, dominance hierarchies have been used to understand the role of dominant species in structuring local communities and composition, such as partitioning dominant and subdominant species within guilds (Baccaro et al. 2010; Arnan et al. 2012). Dominant ant species play an important role in the structuring of local communities. For example, Formica species dominating a boreal ecosystem divert resources away from subdominant competitors (Savolainen and Vepsäläinen 1988). In Mediterranean ecosystems, subdominant species forage at nearly lethal environmental conditions while dominant species reduce mortality risk by foraging at more favorably temperatures (Cerdá et al. 1998c ). In tropical ecosystems, competing arboreal ants can be structured into a dominance hierarchy with higher ranked ant species having greater access to nesting sites and extrafloral nectaries. However, levels of uncertainty associated with outcomes of interspecific interactions are often not quantified (Stuble et al. 2017). Furthermore, its remains unclear how arboreal ants or tropical ants are structured at higher-order interactions, such as when interspecific interactions are viewed as a network.
In this study, we examine dominance hierarchies for a community of arboreal twig-nesting ants in a coffee agroecosystem. Both arboreal and ground-dwelling twig-nesting ants in coffee agroecosystems are nest-site limited in terms of number (Philpott and Foster 2005a ), diversity (Armbrecht et al. 2004; Gillette et al. 2015), and sizes (Jiménez‐Soto and Philpott 2015) of nesting resources. Many studies focused on interspecific dominance hierarchies lack a clear and consistent definition of dominance. For twig-nesting ants, nest takeovers are common and nest sites are often limiting, thus dominance in this system is defined as competition for nest sites (Brian 1952), and in at least one case has been experimentally demonstrated (Palmer et al. 2000a ). Although dominance studies use various methods to rank species, they don’t typically account for uncertainty in rankings, except for a few cases (Adler et al. 2007; Stuble et al. 2013). This present study aims to develop dominance hierarchies for twig-nesting ants living in hollow twigs in a Mexican coffee agricultural ecosystems due to competition for nest resources. We adopt methods from network analysis to infer dominance hierarchies from competitive interactions by estimating uncertainties and steepness in rankings (Pinter-Wollman et al. 2014; Shizuka and McDonald 2015; Sánchez‐Tójar et al. 2017). Furthermore, by viewing twig-nesting ants as a network we estimate the orderliness of the hierarchy within the community. Specifically, we tested the hypothesis that tropical, arboreal twig-nesting ants form a clear, dominance hierarchy for nesting sites in controlled environments.