Diversity Of Species Triumphs

Their early results were controversial, however. Some scientists have asked whether the association between diversity and productivity might be a "sampling effect"; for example, plots with more species have a higher chance of containing plants that naturally produce well. Such questions led to a major debate about biodiversity and whether the effects of biodiversity loss were relevant to public policy. In work to be published in the Oct. 26 issue of Science, Tilman and his colleagues present evidence that such "sampling effects" are not significant, and diverse ecosystems likely do better because differences among species allow such ecosystems to utilize resources better.

The researchers found that no plant species growing by itself could do at all as well as some mixtures of several species. This disproved the "sampling effect" conjecture, which said that effects of diversity came solely because there would be a species which, although growing alone, could perform as well as all diverse mixtures of species, Tilman said.

"Some thought that if diversity was having its effects for simple reasons, like the sampling effects, then a manager could do just as well with a single species," he said. "In fact, however, when many species are together, they do better as a system than any species does alone. As the number of species goes up, more and more plots do better than the single species that does best in monoculture."

These results suggest that managed systems like hayfields, pastures, forests and even fisheries might benefit from this principle, said Tilman. To get highest yields, it isn't enough to have one or two high-producing plants (or animals). Rather, what's required is an appropriate mix of many organisms that live their lives differently. For example, mixes of plants with deep roots and plants with shallow roots, plants that thrive in cool and those that thrive in warmer weather, or other such combinations ensure a greater range of traits that keeps the ecosystem functioning as conditions change.

In 1994 and 1995, Tilman and his colleagues planted 168 plots, each 9 meters on a side, with 1, 2, 4, 8 or 16 species chosen at random from an assortment of grassland plants with different traits. The researchers then measured the productivity of the plots by weighing the biomass produced each year in each one. In the current study, they used their long-term data set to test among the different hypotheses that had been proposed in the biodiversity debate. A variety of tests all rejected the sampling effect.

Results showed that as the number of species in a plot rose, so did the likelihood that the plot would outperform the top-performing monoculture. Specifically, the average productivity of 16-species plots was higher than the productivity of the single species that did best in monoculture. Further, this effect became more pronounced with time; in data from 2000, 16-species plots did better than in 1999, 1998 or 1997. In 2000, the average productivity of 16-species plots outperformed the best monoculture by 40 percent. When the average productivity of all single-species plots (monocultures) was compared to the average productivity of all 16-species plots, the diverse plots produced 170 percent more biomass.

The researchers also tested whether plots with only a few species did poorly because, by the luck of the draw, they had been stuck with the least productive plants in the pool. The researchers threw out data from plots with any of the least productive species in them, regardless of the total number of species in the plots. The overall pattern still held. They then threw out data from all plots whose production ranked in about the lowest fifth of all plots. Again, the pattern held. No evidence was found that the diversity effect depended solely on plots having the good luck to contain a productive species or the bad luck to contain only poor performers.

"By having the experiment run so long, we've had the chance for plots to come to equilibrium," said Tilman. "It's still not a long time in terms of nature, but we're seeing the real, sustainable outcomes of the interactions among these organisms."

Tilman emphasized that a diversity of species led to higher productivity because it brought something new into the picture: better utilization of resources. No single species can do everything.

"There's no magic effect of diversity--it's just that you get a greater range of traits with more species," said Tilman. "Each species is like a profession in human societies. We do better with lots of professions."

Tilman worked with colleagues in the department of ecology, evolution and behavior and department of forest resources at the University of Minnesota and the School of Biological Sciences and School of Natural Resource Sciences at the University of Nebraska. The work was supported by the National Science Foundation and the Andrew W. Mellon Foundation.