The molecular phylogenetics ofTrachymyrmexForel ants and their fungal cultivars provide insights into the origin and coevolutionary history of ‘higher‐attine’ ant agriculture

Abstract

The fungus‐growing ants and their fungal cultivars constitute a classic example of a mutualism that has led to complex coevolutionary dynamics spanning c. 55–65 Ma. Of the five agricultural systems practised by fungus‐growing ants, higher‐attine agriculture, of which leaf‐cutter agriculture is a derived subset, remains poorly understood despite its relevance to ecosystem function and human agriculture across the Neotropics and parts of North America. Among the ants practising higher‐attine agriculture, the genus Trachymyrmex Forel, as currently defined, shares most‐recent common ancestors with both the leaf‐cutter ants and the higher‐attine genera Sericomyrmex Mayr and Xerolitor Sosa‐Calvo et al. Although previous molecular‐phylogenetic studies have suggested that Trachymyrmex is a paraphyletic grade, until now insufficient taxon sampling has prevented a full investigation of the evolutionary history of this group and limited the possibility of resolving its taxonomy. Here we describe the results of phylogenetic analyses of 38 Trachymyrmex species, including 27 of the 49 described species and at least 11 new species, using four nuclear markers, as well as phylogenetic analyses of the fungi cultivated by 23 species of Trachymyrmex using two markers. We generated new genetic data for 112 ants (402 new gene sequences) and 95 fungi (153 new gene sequences). Our results corroborate previous findings that Trachymyrmex, as currently defined, is paraphyletic. We propose recognizing two new genera, Mycetomoellerius gen.n. and Paratrachymyrmex gen.n., and restricting the continued use of Trachymyrmex to the clade of nine largely North American species that contains the type species [Trachymyrmex septentrionalis (McCook)] and that is the sister group of the leaf‐cutting ants. Our fungal cultivar phylogeny generally corroborates previously observed broad patterns of ant–fungus association, but it also reveals further violations of those patterns. Higher‐attine fungi are divided into two groups: (i) the single species Leucoagaricus gongylophorus (Möller); and (ii) its sister clade, consisting of multiple species, recently referred to as Leucoagaricus Singer ‘clade B’. Our phylogeny indicates that, although most non‐leaf‐cutting higher‐attine ants typically cultivate species in clade B, some species cultivate L. gongylophorus, whereas still others cultivate fungi typically associated with lower‐attine agriculture. This indicates that the attine agricultural systems, which are currently defined by associations between ants and fungi, are not entirely congruent with ant and fungal phylogenies. They may, however, be correlated with as yet poorly understood biological traits of the ants and/or of their microbiomes.

Phylogenetic patterns of ant–fungus associations indicate that farming strategies, not only a superior fungal cultivar, explain the ecological success of leafcutter ants

“…striatus and with (iii) predominant Clade‐A cultivation but also some Clade‐B cultivation among all other leafcutter ants, any of the following three scenarios seem viable hypotheses for fungicultural specialization by the ancestral leafcutters: either an exclusive specialization on Clade‐A fungi, or exclusive specialization on Clade‐B fungi, or mixed Clade‐A & Clade‐B cultivation by the ancestral leafcutter ants. A comprehensive reconstruction of ant–fungus association throughout the phylogenetic history of higher‐attine ants, using a more representative sampling of fungal cultivars from all Trachymyrmex/Sericomyrmex lineages than in Figure , will be necessary to test these different hypotheses (Solomon et al., in preparation).…”

“…Because higher‐attine fungi originated from Clade‐1 fungi or from closely related free‐living leucocoprineaceous species (Figure ), this suggests the possibility that cultivation of Clade‐1 fungi by extant Trachymyrmex species could represent retention of the ancestral (plesiomorphic) state of Clade‐1 cultivation practiced by the ancestral ant lineage that gave rise to higher‐attine ants (Solomon et al., ). More comprehensive collection of Trachymyrmex ants and lower‐attine ants could further address this hypothesis, particularly more collections from South America (Solomon et al., in preparation). Ant–fungus associations reported so far for the well‐studied Panamanian Trachymyrmex/Sericomyrmex ants and leafcutter ants therefore represent only a subset of the far greater diversity of ant–fungus associations found in each of these same host clades elsewhere in the Neotropics.…”

“…We selected phylogenetically representative cultivar samples from the entire diversity of more than 2,500 samples of ethanol‐preserved garden collections stored frozen at −80°C in the Attine Collections of the Bacci Lab (UNESP Rio Claro) and the Mueller Lab (University of Texas at Austin). We selected samples to complement sequence information on attine cultivars generated in previous surveys of cultivars of lower‐attine ants (Kellner et al., ; Mehdiabadi et al., ; Mueller et al., ), nonleafcutter higher‐attine ants (Solomon et al., in preparation) and leafcutter ants (Mueller et al., ) and thus enable a global phylogenetic reconstruction of attine‐cultivated leucocoprineaceous fungi. We included representatives from all known lepiotaceous fungus clades cultivated by lower‐attine ants, such as Clade 1A & 1B and Clade 2 as defined by Kellner et al., ; plus representatives from the two known fungal clades and their known subclades cultivated by higher‐attine ant genera Atta , Acromyrmex , Trachymyrmex and Sericomyrmex ).…”

Development, characterization, and cross-amplification of polymorphic microsatellite markers for North American Trachymyrmex and Mycetomoellerius ants

“…This indicates that future analyses for additional individuals across a wider geographic region may be informative for understanding similar questions for T. arizonensis and T. pomonae and perhaps other closely related Trachymyrmex. It also seems likely that the markers developed here for M. turrifex should cross amplify in closely related Mycetomoellerius species, such as the highly studied M. zeteki [42,52,53], among others. While all markers across all four species showed polymorphism/allelic variation, several markers (particularly within T. septentrionalis) were hypervariable (over 20 alleles per locus; Table 1).…”

“…We also explored the potential for cross-amplification of the T. septentrionalis markers in two closely related species: T. arizonensis and T. pomonae. These three species belong to Trachymyrmex sensu stricto, unlike M. turrifex which belongs to a new genus, Mycetomoellerius [3,33,42]. These markers will provide a new tool to study patterns of attine intraspecific genetic variation and phylogeographic structure.…”