An ongoing experiment now in its third decade at Cedar Creek has shown that even very low levels of nitrogen fertilization can reduce plant diversity. Cedar Creek is supporting a global experimental network (The Nutrient Network, or NutNet) testing the generality of these results at 70 sites on 5 continents. Established in 2005, this collaborative experimental network was established to understand the long-term consequences of global-scale changes in climate, multiple nutrient limitation, and food web structure for grassland diversity and ecosystem functioning.
The methods are relatively simple and labor and materiel costs are minimal. In 2013, over 90 scientists conducted a standard set of observations and manipulations at >70 herbaceous plant-dominated sites on six continents to develop a global perspective on the effects of multiple nutrients and herbivory on plant productivity and diversity. (Figure 1) Sites span a wide range of short-statured, primarily herbaceous vegetation types and environmental conditions and include seven LTER sites (CDR, SEV, SGS, KNZ, HJA, NWT, JRN). More than 50 sites are conducting standalone experiments within a randomized-block design with 5 m x 5 m plots and 3 replicate blocks per site. The Multiple-Nutrient Experiment is a factorial addition of three nutrient treatments (N, P, and K + micronutrients). The Consumer x Nutrient Experiment is a full factorial manipulation of nutrients (Control or All Nutrients) and consumers (Control or Fenced) using 180 cm fences designed to exclude aboveground mammalian vertebrates. Annual data collected in all plots includes aboveground live and dead plant biomass, available light, and species abundances. NutNet data are hosted on Minnesota Supercomputer Institute servers and the University of Minnesota – Cedar Creek provides a formal home for the data analysis, synthesis, and management of the ongoing NutNet experiment.
To date, NutNet data have been the basis for cross-site syntheses on the drivers of plant diversity and invader abundance, analyses that have demonstrated the insights that this unique dataset can provide. For example, NutNet data provided the strongest test to date of the hypothesis of consistent dependence of diversity on site or plot-scale productivity, finding no evidence of such a general pattern. Similarly, this powerful dataset revealed exotic species had similar abundances in their native and invaded ranges, suggesting that the same species traits or interactions controlled species abundance in invaded ranges as in home ranges. Because of NutNet’s global distribution, the dataset has been used to determine global climatic drivers of plot-scale decomposition. (O’Halloran et al PLoS ONE 2013) Finally the responses of individual species to the fertilization and fencing treatments have been shown to be constrained along a single trade-off axis, implying that one of the main hypothesized forces of diversity maintenance may be weaker than expected. (Lind et al Ecology Letters 2013)