Ecosystems aren't green; they are black and brown, at least in the colour palette favoured by Diana Wall. Wall, a soil ecologist at Colorado State University in Fort Collins, spends her days digging into the world's underground ecosystems. These beiges, ochres and charcoals reflect a three-dimensional mosaic of micro-environments, each with its unique set of inhabitants.


But very little is known about these inhabitants. Understanding soil is a matter of rising urgency. A July report from the US National Research Council listed soil quality as the biggest barrier to higher crop yields for farmers in sub-Saharan Africa and south Asia. And knowing what myriad organisms live in the soil, and how they interact, is crucial to creating a healthy ecosystem.

For those scientists who are willing to crouch down and dig, the diversity of soil denizens beats any above-ground system, even that of a tropical rainforest. A handful of soil from one spot may house a very different community from soil just a metre away, because of variations in the availability of water or nutrients. For example, the ground under a decaying plant or animal is a different environment from soil lacking such enrichment. And around plant roots, specialized organisms inhabit the rhizosphere, a thin layer where roots and soil organisms interact in myriad ways. Large animals such as moles contribute, changing and aerating the underground landscape by tunnelling. Even a small clump of soil has a gradient of oxygen from its edges to the centre, and each oxygen concentration may make the perfect habitat for different kinds of creatures. "It is the most incredible zoo," says Wall.

My dream is that you take a DNA sample from the soil, and then explain what species are there, and what benefits. Wim van der Putten

Take that view to a larger scale, and it is possible to appreciate just how complicated the world's soil ecosystems are. In one ongoing study, not yet published, Wall and her colleagues scooped soil cores from two sites in Alaska, one in the tundra and one in the taiga forest. Although the sites were only 400 kilometres apart, the species living there were radically different: only 18 invertebrate taxa out of an estimated 1,300 appeared in both locations. "That just blew me away," says Wall.

And that's just looking at invertebrates, not including microbes. "As far as I know, there is no environment on Earth that is more biologically diverse, per unit area, than soil," says Eric Triplett, a microbiologist at the University of Florida in Gainesville. Thanks to faster, cheaper DNA sequencing, scientists are now getting a grip on what is down there and what those organisms might be doing. That information, in turn, could help improve soil management for agriculture and forest management for conservation.

At this point, scientists don't even agree on how many creatures they are looking for. The first DNA-based estimate of soil microbial biodiversity, published in 1990, counted about 4,000 different bacterial genomes per gram of soil1. Since then, various studies and models have pushed the number up as high as 830,000 species per gram2, down to 2,000 (ref. 3), and back up again. Most recently, Triplett and his colleagues ran 139,000 individual sequences — more than other studies have used — and came up with an estimate of 10,000 to 50,000 species per gram of soil4. Complicating the matter is the fact that, because so few of these species have been described, researchers have to group similar organisms within 'operational taxonomic units', which correspond roughly but not precisely to species designations.

Valuable species


Quantifying such diversity illustrates just how much remains to be discovered, and soil scientists are teaming up to tackle the challenge. The Tropical Soil Biology and Fertility (TSBF) Institute, run by the International Center for Tropical Agriculture and headquartered in Nairobi, has united more than 300 scientists in seven countries to survey soil organisms. The project, which began in 2002, aims to identify living indicators for fertile or poor soil, and has already identified some novel organisms that could be useful to humans.

In the Veracruz rainforest, for instance, Mexican scientists have discovered Acaulospora, a mycorrhizal fungus that entwines with lily roots and provides water and mineral nutrients. Last spring the researchers injected Acaulospora into the soil of test lily plots in Benigno Mendoza, a community in Veracruz where lily bulbs are an important cash crop. As a result, this year's harvest consists of big, first-quality bulbs that match the yields gained through using inorganic fertilizer with none of the downsides of chemical treatments. Isabelle Barois, a soil ecologist at the Institute of Ecology in Xalapa and coordinator of the TSBF Mexican team, says that the fungus could eventually help replace the expensive nitrogen fertilizer and harsh agrochemicals that farmers apply to their land five or six times a year.

Credit: A. SYRED/SPL

Global soils contain a bounty of unusual and potentially useful organisms such as Acaulospora — more, theoretically, than they should. Although some species are common, there are also countless taxa found in vanishingly small numbers. Many species also seem to be redundant, eating the same foods and fulfilling the same ecosystem jobs, so scientists don't quite understand why they're there at all. "There is some debate about how many species need to be present in the soil to make an ecosystem," says Wim van der Putten, an ecologist at the Netherlands Institute of Ecology in Heteren.

Heikki Setälä, an ecologist at the University of Helsinki, took on this question with experiments in which he controlled the number of animal or microbial species in artificial ecosystems. In one study5, he set up soil microcosms in glass jars and added fungal species: only one in some jars, and up to 43 in others. Diverse systems decomposed more organic matter — demonstrated by higher carbon dioxide production — and produced more nitrogen compounds in the soil. But that relationship held true only at the lower end of the spectrum. Six species were better than one, but 43 weren't any better than six. "It was kind of a bummer," Setälä says. "It would be nice to tell the audience that we need all the species to make the planet green and sustain it."

The explanation for the wealth of soil biodiversity, then, remains an open question. Maybe the multitudinous creatures are simply adapted for niches that humans don't yet understand. Alternatively, they could literally be waiting for a rainy day; some organisms spring into action after a storm, fire or other disturbance, and so make the ecosystem more resilient. Or perhaps those organisms are truly redundant. "We know virtually nothing about what controls the diversity of soil communities," says soil ecologist Richard Bardgett of Lancaster University, UK.


Triplett disagrees. "I don't think it's a vast unknown," he says. "I think there are some dominant genera out there that we could learn about pretty fast." In a follow-up to his soil biodiversity survey6, Triplett and his colleagues found that up to around 65% of the DNA samples from soil microbes fell into known genera, which makes those genera prime candidates for further study. For example, Chitinophaga was prevalent in the four distinct soils tested, from Canada, Illinois, Florida and Brazil. But a PubMed search for the genus finds only ten papers on the genus (and one of those is Triplett's), highlighting the lack of work that has been done in this area.

"My dream for the future would be that you would just take a DNA sample from the soil, and then explain what species are there, and what benefits," van der Putten says. But this kind of quick DNA test is years in the future.

Setting microbes to work

For some scientists, just defining the diversity isn't enough. Triplett, for instance, wants to alter it. He envisions harnessing the nitrogen-fixing power of bacteria that form nodules on the roots of some plants, such as legumes, and convert nitrogen from the air into a form the plants can use. He thinks he could insert some of the nitrogen-fixing (nif) genes from the bacteria into agricultural crops — which could then collect their own nitrogen from the atmosphere and eliminate the use of artificial nitrogen fertilizer. It has already been shown that some nif genes can function in plants7. A nitrogen-fixing plant would require at least ten new genes, making the task difficult, Triplett says, but not impossible.

Policy-makers are slowly starting to pay attention to the problem of soils. In 2006 the European Union agreed that soils need protection from erosion, landslides and salinisation, but has not yet finalized the laws that would ensure this happens. Some countries, including France, would prefer to see individual countries regulate soil. "I'm pretty confident that the politicians will swallow the hook sooner or later," Setälä says.


Avoiding that hook comes with a price tag: one estimate valued the free services provided by the world's soil biota at US$1.5 trillion or more each year8. Soils are also important as a carbon sink; soil stockpiles 1,500 gigatonnes of organic carbon, more than Earth's atmosphere and all the plants on the planet, according to the United Nations Food and Agriculture Organization. If soils remain degraded and their many denizens disappear, the world might lose access to organisms that improve crop yields, degrade toxins, or make useful by-products such as drugs — before they're even discovered.