It is generally thought that ecosystems containing a greater diversity of species are more likely to thrive than those with fewer species. This has led to efforts to conserve or restore biodiversity in natural ecosystems. Diversity at the level of producers (such as photosynthetic algae) and decomposers (bacteria or fungi), in particular, may influence ecosystem productivity. On page 762 of this issue, Naeem et al.¹ report an elegant experimental study of the productivity -- the amount of biomass generated -- of a simple aquatic ecosystem. They find that productivity depends on the diversity of both producers and decomposers, but that the relationship between the two is not straightforward.

Naeem and colleagues' experiment involved independent manipulations of the number of algal and bacterial species in laboratory microcosms. The results were surprisingly complex, and do not support a simple relationship between diversity and productivity. Systems containing a single algal species and many bacterial species displayed a low combined productivity of algae and bacteria, and much of the total biomass consisted of bacteria. In contrast, systems with moderate to high algal diversity tended to have a higher combined productivity characterized by high algal biomass and relatively low bacterial biomass, regardless of bacterial diversity.

Naeem et al. also measured the ability of the bacteria in their microcosms to use different organic compounds as carbon sources. The found that bacteria from systems with high productivity could use a greater variety of carbon sources than those from systems with lower productivity. Interestingly, the ability of bacteria from a microcosm to use a variety of carbon sources did not depend solely on the number of bacterial species present, but also increased with increasing algal diversity. This may mean that ecosystem productivity is linked to the cycling of carbon or other nutrients.

The study also confirms a result previously described only for terrestrial ecosystems^2-4. Ecosystems containing more producer species tend to be more productive, but the pattern varies greatly with the number of bacterial species added to the system. Although they extend the generality of positive relations between diversity and ecosystem processes in terrestrial systems to the aquatic realm, Naeem et al. point out that such patterns may depend on whether researchers manipulate only producers (as is often the case) or other groups crucial to ecosystem function, such as predators. Most studies have manipulated the diversity of only producers^2,3 or mycorrhizal fungi⁵, or of entire food webs⁴, which makes it difficult to see whether the positive effects of diversity on ecosystem properties depend on other unspecified features of species composition.

Complex interactions between producers and decomposers potentially occur in all ecosystems, with the possible exception of systems such as deep-sea hydrothermal vent communities that are not directly powered by the capture of sunlight. Recent models have predicted the complexity of interactions between producers and decomposers^6,7. Positive interactions arise because photosynthetic producers create the organic compounds consumed by decomposers, whereas decomposers remineralize carbon and other nutrients required by producers.

Simpler models treat producers and decomposers as single species, an assumption that now seems suspect given Naeem and colleagues' results. Models that incorporate competition among species within nutrient cycles could better explain the complexity of interactions between producers and decomposers⁷. For instance, low productivity in systems with many bacterial species and few algae could be a consequence of the greater likelihood that those systems contain a bacterial species that tends to outcompete the algae for some essential nutrient. Greater productivity of systems containing a greater diversity of both algae and bacteria would be predicted by the increased likelihood that each group (producer or decomposer) would contain a species that is highly efficient in using resources⁷.

Although aquatic algae and bacteria may compete for limiting nutrients, there is ample evidence that they provide mutual benefits. Algae leak the sugars produced by photosynthesis, especially when nutrient-limited, providing a ready carbon source for bacteria. The total abundances of algae and bacteria are positively correlated in lakes⁸, so it seems that bacteria enjoy some sort of positive association with algae even though they may outcompete them for nitrogen and phosphorus.

In turn, bacteria may provide benefits for some algae beyond the remineralization of macronutrients. Not all producers can synthesize all of the compounds needed for cellular metabolism. Some, particularly photosynthetic flagellates such as Euglena, have an absolute dietary requirement for vitamin B₁₂ (ref. 9), which can be synthesized by bacteria. Although Euglena is only one of the several algal species used by Naeem et al., it and other species may benefit in similar ways from micronutrients synthesized by bacteria. Another possibility is that some photosynthetic flagellates (Euglena again) are mixotrophic -- that is, able both to photosynthesize and to consume other organisms, including bacteria. If so, this might explain why, in Naeem and colleagues' experiments, algae sometimes benefited from increased bacterial diversity, whereas bacteria did not appear to benefit from increased algal diversity.