Advancing realism in biodiversity research.

Numerous experimental studies suggest that biodiversity loss is detrimental to ecosystem functions such as production and nutrient cycling. These experiments, however, have been criticized as unrealistic because they use combinations of species that do not resemble what is observed in nature. Bracken et al. take a novel approach, using species combinations observed in the field and computer simulations to explore many other combinations. Their approach represents a significant advance in making biodiversity research more realistic.