Diversity-dependent stability under sowing and nutrient addition: evidence from non-weeded artificial grassland

Abstract

A growing body of evidence from diversity-manipulation and natural studies suggests that the stability of community productivity increases with biodiversity; however, few studies have researched this relationship in a non-weeded grassland. To clarify this issue, we established an artificial grassland in 2003 using three common species, Elymus nutans, Festuca sinensis and Festuca ovina, which included seven different community structures (three monocultures, three two-species mixtures and one three-species mixture based on sown species) and two nutrient addition treatments (non-nutrient addition and nutrient addition). Data was collected over a three-year period (2011–2013). Our results showed that the sowing species modified realized species richness (i.e. the number of total species we observed in a community) and species evenness, but had negligible influences on community- and population-level stability. Furthermore, all of these variables were reduced by nutrient addition. These dynamics did not alter the positive influence of realized species richness on community stability, but restricted the stable effect of evenness because this effect was only significant under nutrient addition condition. The potential mechanisms underlying these processes were statistical averaging and species asynchrony, rather than overyielding effect. Conversely, population stability decreased with realized species richness in non-nutrient addition treatments. We conclude that biodiversity contributed to community- and population-level stability even in non-weeded experiment. This process resulted from different mechanisms that observed in weeded experiments. Further studies in other ecosystems (e.g. aquatic ecosystem) are needed to find a more general conclusion.