Approximately one third of herbaceous forest plant species in eastern North America rely on ants for seed dispersal. In this mutualistic interaction (myrmecochory), foraging ants transport seeds to their nest, where seed coat-derived elaiosomes are removed and consumed. Seeds either remain in the nest or are redispersed. Herein, plants obtain dispersal related benefits, including escape from distance- and density-dependent mortality agents and a nutrient-rich germination site. However, chemical byproduct benefits (i.e. benefits to one partner through the selfish actions of the other partner) to ant-dispersed plants have been understudied in this system. Ants, including a keystone seed-dispersing genus Aphaenogaster, produce chemical secretions from their glands. These secretions are used for a variety of purposes, including antiseptic cleaning behaviors. During their tenure in the ant nest, seeds likely come into contact with these antimicrobial chemicals, which could alter the seed coat microbial community, including plant pathogens. We investigated the effects of Aphaenogaster picea and A. rudis handling on the seed coat microbial community of three ant-dispersed plants in the southeastern U.S.: Sanguinaria canadensis, Asarum canadense, and Jeffersonia diphylla. Using seed coat washes, we assessed changes in the overall bacterial and fungal communities and the plant pathogenic fungal community with and without ant handling via amplicon sequencing of the 16S rRNA gene and the ITS2 gene region.
Results/Conclusions
We found that ant handling changes the bacterial and fungal seed coat communities. While seeds of the three species differed in their initial bacterial and fungal communities, distance-based redundancy analyses (db-RDAs) indicate that the seed coat microbial communities of all species differed significantly after ant handling from seed coat microbial communities of unaltered seeds and from seeds with manually-removed elaiosomes. Furthermore, ant handling caused convergence of bacterial and fungal communities. In most instances, the alpha diversity of microbial communities decreased after ant handling. A similar pattern was observed for the FUNGuild-identified fungal plant pathogen community: plant pathogen relative abundance decreased for two seed species. These results support the possible existence of a byproduct benefit in ant seed dispersal; we hypothesize that the mechanism by which ant handling alters seed-coat microbiomes—specifically plant pathogens—is through glandular chemicals that are produced for ant cleaning behaviors. This change in the seed coat-associated microbial community could carry important ramifications for plants, including the reduction of seed susceptibility to pathogens. This work illustrates how chemical byproducts in mutualisms may operate on selective agents that have been previously overlooked.