"Prior shorter-term studies, most about two years long, found that diversity increased productivity, but that having more than six or eight species in a plot gave no additional benefit," Peter Reich of the University of Minnesota, US, told environmentalresearchweb. "We found that over a 14-year time span, all 16 species in our most diverse plots contributed more and more each year to higher soil fertility and biomass production."

Reich and colleagues studied ecosystems at the Cedar Creek Biodiversity Experiment (BioDIV), which was planted in 1994–5, and the Biodiversity, CO2, and N Experiment (BioCON), set up in 1997. Both studies contained many of the same C3 and C4 grasses and nitrogen-fixing and nonfixing dicotyledonous herbs.

Over time, the number of planted species in each plot had a larger effect on total plant biomass. The team ascribed this to accumulating effects of diversity-dependent ecosystem feedbacks and interspecific complementarity. They reckoned this caused high-diversity species combinations that appeared functionally redundant during early years to become more functionally unique over time.

"The take-home message is that when we reduce diversity in the landscape – think of a cornfield or a pine plantation or a suburban lawn – we are failing to capitalize on the valuable natural services that biodiversity provides," said Reich. "These diversity effects are likely more important in the real world than prior studies showed, as our ˜14-year experiments are closer to the average age of most vegetation communities, which range from new following disturbance to hundreds of years old, than studies of communities when they are only two years old."

According to Reich, each species in a plot contributes to a function. But the level of diversity has been decreasing globally for decades due to factors such as invasive species, nitrogen pollution and elimination of natural disturbance regimes.

Reich says he and colleague David Tilman decided to begin long-term field experiments to test the effects of biodiversity on ecosystem functions directly because all other approaches, such as descriptive observations across productivity or diversity gradients, are highly problematic. "For example, other factors that vary across the environment – soils, fertility climate, topography, disturbance history etc – can influence both diversity and productivity, making it difficult to assess the direct impact of diversity on productivity," he said. "Only a direct experiment provides a clean test of this."

Now the team plans to continue the grassland experiments, and is beginning several studies in forest biodiversity. The researchers, who reported their work in Science, have planted three plots of forest in northern Minnesota and Quebec and are planning two new large experimental studies in Australia and central Minnesota.