Endophyte-grass-herbivore interactions:: the case of Neotyphodium endophytes in Arizona fescue populations

Neotyphodium endophytes in introduced agronomic grasses are well known to increase resistance to herbivores, but little is known of interactions between Neotyphodiuun endophytes and herbivores in native grass populations. We investigated whether endophytes mediate plant-herbivore interactions in a native grass species, Festuca arizonica in the southwestern United States, in two ways. First, to test the prediction that the presence and frequency of endophyte-infected (E+) plants should increase with increasing herbivory, we determined endophyte frequencies over a 4-year period in six natural Arizona fescue populations. We compared Neotyphodium frequency among plants growing inside and outside long-term vertebrate grazing exclosures. Second, we experimentally tested the effects of Neotyphodium infection, plant clone, and soil nutrients on plant resistance to the native grasshopper Xanthippus corallipes. Contrary to predictions based upon the hypothesis that endophytes increase herbivore resistance, levels of infection did not increase in plants subjected to grazing outside of exclosures relative to ungrazed plants within exclosures. Instead, endophyte frequencies tended to be greater inside the exclosures, where long-term vertebrate grazing was reduced. The grasshopper bioassay experiment corroborated these long-term patterns. Survival of grasshoppers did not differ between infected (E+) and uninfected (E-) plants. Instead, mean relative growth rate of grasshoppers was higher on Ec grasses than on E- ones. Growth performance of newly hatched grasshopper nymphs varied among host plant clones, although two of six clones accounted for most of this variation. Our results suggest that Neotyphodium-grass-herbivore interactions may be much more variable in natural communities than predicted by studies of agronomically important Neotyphodium-grass associations, and herbivory is not always the driving selective force in endophyte-grass ecology and evolution. Thus, alternative hypotheses are necessary to explain the wide distribution and variable frequencies of endophytes in natural plant populations.