Comment arising from a paper by Wittmer et al.: hypothesis testing for top-down and bottom-up effects in woodland caribou population dynamics

Asexual fungal endophytes, which do not produce reproductive structures, spread in host populations only vertically via the propagules of their hosts. With such a close relationship between the fitness of the asexual endophyte and that of the host, the relationship is traditionally thought to be mutualistic. Here we present data showing that the positive effect of asexual endophytes may concern only a short period of the host’s life and that its lifetime reproductive success may be reduced. We followed 180 individuals of a perennial grass, Puccinellia distans, for the first 3 years of their growth both in the field and in a common garden experiment. In the first 2 years, infected individuals produced more generative shoots with longer inflorescences. Three-year-old individuals produced significantly fewer generative shoots with shorter inflorescences when infected with the fungus Epichloë typhina. Moreover, the dry mass of above- and belowground parts after three seasons was significantly lower in infected individuals. We suggest that if the endophyte retains control over plant reproduction and if shorter generation time is more crucial for the fungus than for the perennial host, then the fungus should stimulate plant reproduction early in life even though infection incurs a future cost. This reasoning suggests that discovering an endophyte’s beneficial effect on its perennial host in one season provides insufficient grounds for concluding that a mutualistic relationship exists.