Plants harbor a diverse assemblage of microbial symbionts, including bacteria and fungi in leaves and roots.  The ecological consequences of many of these microbes remain unknown, particularly in natural ecosystems.  I investigated the costs and benefits of symbiosis between grasses and endophytic fungi that grow systemically in above-ground plant tissues.   Grass endophytes have been well studied in a few grass host species due to their economic and environmental impacts in forage and turf production.   These endophytes can protect hosts from damage by vertebrate and invertebrate herbivores, improve tolerance to drought, and enhance nutrient uptake, and their presence can have wide-ranging effects on the surrounding community.  However, the ecological function of endophyte symbiosis remains unresolved for the vast majority of grass species.  Here, I examined the costs and benefits of endophyte symbiosis for eight native grass species, spanning three tribes in the Poaceae (Aveneae, Poaeae, Triticeae).  Endophyte-infected seeds were experimentally disinfected using a heat treatment that eliminated the endophyte without reducing seed viability.  Plants were propagated clonally in the greenhouse then planted into field plots adjacent to the source population for the seeds.  I also subjected a subset of species to herbivory or drought in a controlled greenhouse environment. 

Results/Conclusions

Experiments revealed diverse benefits and costs of endophyte symbiosis in natural ecosystems.  For three species, symbiosis with the endophyte reduced insect herbivory by ~30% and increased plant growth by ~30%.  In contrast, symbiosis increased herbivory by small mammals and strongly reduced host plant survival in another host species.  Endophyte presence enhanced host growth under drought for two of three species tested.  In one case, symbiosis altered plant life history strategy, by increasing allocation to growth relative to reproduction.  In sum, results demonstrate that strong ecological consequences of endophyte symbiosis are not limited to managed, agronomic ecosystems.  Host grasses are often used in ecological experiments without knowledge of endophyte presence, despite the fact that endophytes may alter plant competitive hierarchies and the outcome of experiments.  Symbionts contribute an additional, but often overlooked, layer of diversity in ecosystems, and their inclusion in ecological research may help increase the realism of both experiments and theory.