Associate Professor Inner Mongolia University, China (People's Republic)
Abstract: Human activities can dramatically change plant diversity in global grasslands. The random-loss hypothesis predicts that species with low abundance (subdominants) are more likely to go extinct than those with high abundance (dominants) but we also need to understand which species persist or colonize such areas. Species-specific characteristics such as plant origin, life history, and growth form can affect competitive ability and thus influence composition. Furthermore, communities may diverge as treatment conditions filter the species pool and favor a subset of species.
Our goals were to quantify the responses of individual plant species to multiple years of nutrient addition and herbivore exclusion, and to evaluate whether species characteristics are associated with similar responses.
We addressed these goals by applying Indicator Species Analysis (ISA) to species-level compositional data from the Nutrient Network (NutNet), focusing on the effects of nutrient addition (unfertilized vs. NPK fertilized) and herbivore exclusion (fenced vs. unfenced). ISA accounts for the frequency of occurrence and the abundance within treatments of each species. We compiled cover data for 2,206 species in 90 grasslands on six continents over 12 years of treatment. Each site-year-factor combination was analyzed separately, providing an Indicator Value for each species in each treatment. Statistical significance was assessed via permutations. We tallied the number of significant indicators of one treatment over the other. Finally, we tested whether the likelihood that species of a given characteristic were significant indicators differed from their expected prevalence based on the relative abundance of that characteristic in the community.
Here, we show strongly increasing patterns with respect to nutrient addition but relatively few indicators of herbivore exclusion. Subdominant species were much more likely to be indicators of unfertilized plots, suggesting that nutrient addition resulted in their extinction from fertilized plots. The number of significant indicators that were exotics or annuals did not differ between fertilization treatments, but species with these characteristics were more likely than expected to be indicators of fertilized plots. Species with all other characteristics were more often indicators of unfertilized than fertilized plots, particularly after 5+ years of treatment. Legumes were more likely than expected to be indicators of unfertilized plots.
In conclusion, this study highlights the value of considering the dynamics of individual species from a long-term perspective when examining how grassland communities respond to different types of human activities. This information provides necessary nuance for grassland management and biodiversity conservation.
