Herbivore‐Induced Species Replacement in Grasslands: Is it Driven by Herbivory Tolerance or Avoidance?

Herbivore‐induced shifts in species composition have been documented from grasslands throughout the world, but the mechanism(s) of species replacement remains largely unexplored. An experiment was conducted in a transplant garden, on the campus of Texas A&M University, to test the hypothesis that greater herbivory tolerance of mid‐seral species, rather than selective herbivory of a late‐seral dominant, is the dominant mechanism contributing to species replacement in a mesic grassland. Individual plants of a dominant, late‐seral, perennial grass, Schizachyrium scoparium, were grown in competition with conspecifics and two other perennial grasses (Bothriochloa saccharoides and Stipa leucotricha) known to increase in relative abundance in response to intensive grazing by domestic herbivores in the southern true prairie. Uniform defoliation of both S. scoparium and neighbors was imposed to test the relative expression of herbivory tolerance, while selective herbivory was imposed by defoliating S. scoparium plants, but not neighbors. After four successive defoliations, mean annual shoot production and cumulative shoot production of S. scoparium plants were suppressed to a greater extent by conspecific neighbors than by either of the two mid‐seral neighbors. Mean basal area of S. scoparium plants was not significantly suppressed by either intra‐ or interspecific neighbors; however, tiller number per plant was suppressed by B. saccharoides, but not by S. leucotricha or conspecific neighbors. Defoliation of S. scoparium plants, but not neighbors, negatively impacted the late‐seral plants. Selective defoliation of S. scoparium plants significantly reduced tiller variables of mean mass, leaf blade area, and leaf number, but did not significantly reduce plant variables including mean basal area, tiller number, or annual shoot production. Defoliation of both S. scoparium plants and neighbors increased annual shoot production, mean basal area per plant, mean tiller leaf area, leaf number, tiller mass, stomatal conductance to H₂O vapor, and plant xylem pressure potential in comparison with S. scoparium plants grown with comparable, nondefoliated neighbors. An increase in both plant and tiller variables in defoliated S. scoparium plants grown with uniformly defoliated neighbors establishes that replacement of a late‐seral dominant is not driven by a greater relative expression of herbivory tolerance of mid‐seral species. These results collectively suggest that the late‐seral dominant, S. scoparium, possesses a greater competitive ability and a comparable or greater degree of herbivory tolerance than the mid‐seral species that comprise the community. Therefore, the initial hypothesis was rejected. It can be inferred that the alternative mechanism, selective herbivory of the late‐seral dominant, is the dominant mechanism contributing to species replacement. Herbivore‐induced modifications of competitive interactions are most likely to drive species replacement in grasslands characterized by high and consistent resource availability. This may partially explain why condition and trend analysis was developed and initially implemented in the true and mixed prairie associations of North America and why it is widely used by rangeland managers in these grasslands.