The designers of Biosphere 2 knew they would lose species. They stocked the building with something over three thousand on purpose, more than any of its small ecosystems could hold, on the theory the project called species-packing: that when a species died out, as some would, another would be there to take its place in the food web. What they could not predict was which. The two years were, among other things, the first large experiment in what happens when pieces of five biomes are cut out, shrunk to the size of a garden each, put under one roof, and left to sort themselves out.

What grew

Almost everything grew faster than expected. The rainforest’s biomass had increased by half before the doors even closed, between the first survey in late 1990 and the resurvey the following summer, and it kept going. The nurse trees, a fast-growing Leucaena planted to shade the tender species in their first years, reached twelve to sixteen metres and had to be felled in the first transition so that the trees they were meant to protect could grow. The “gingerbelt” at the rainforest’s edge and the bamboo belt that shielded it from salt spray off the ocean grew prolifically.

Vines did what vines do: morning glory and passionfruit climbed over the rainforest and savannah canopies faster than the crew could cut them, and had to be cut, or they would have shaded out everything beneath. Fast grasses invaded the savannah and the desert. The biospherians became, in their own phrase, deliberate keystone predators, pruning and culling to keep the hierarchy of dominance from collapsing into a monoculture, and they managed the vegetation for the air as well: savannah grass and ginger were cut and the biomass stored dry to slow its decay, so that the regrowth would draw down carbon dioxide in the dark months.

The desert changed character. Designed as a coastal fog desert of cactus and succulents on the Baja model, it became within two years a shrubland of annuals and grasses, more like a coastal scrub. Condensation dripping off the glass in winter, humidity higher than any real desert’s, and a deliberate strategy of keeping it green and active to help manage carbon dioxide all pushed it the same way. It was still alive, and arguably more productive, and it was no longer a desert.

What died

The pollinators went first. The hummingbirds died, and the honeybees, and with them nearly every insect the designers had brought in to carry pollen; flying insects as a class declined through the two years. Two bird species were lost. Of the small vertebrates stocked at closure, the count that has circulated since the mid-1990s is that nineteen of twenty-five species were gone by the end; it is repeated everywhere and traces to the review committee of that period rather than to a published census, which is a caution rather than a refutation. A galago, the small nocturnal primate called a bushbaby, conceived and gave birth inside, one of the few vertebrate successes on record.

What thrived in the vacuum was what thrives in any disturbed place. A tramp ant, Paratrechina longicornis, the crazy ant of ports and greenhouses the world over, came to dominate every other ant in the building and took the crickets and grasshoppers with them; a study after the mission found it had become the ecosystem’s ruling insect. Cockroaches exploded, and so did mites and mealybugs on the farm. These were not failures of the design so much as the design finding its level: a small, warm, humid, enclosed world with no winter and no predators is the ideal habitat of exactly those animals, and no amount of species-packing changes that.

The ocean

The ocean was expected to be the most fragile system in the building, and it was the one that held. Some two and a half thousand cubic metres of seawater, a reef of corals collected in the Caribbean and off Yucatán, a lagoon and a beach, with a vacuum wave machine, pumps circulating 2,500 litres a minute, protein skimmers, and algae turf scrubbers under artificial light to strip the nutrients that would otherwise have turned it green. White band disease took some of the brain corals. The crew buffered the water’s pH with chemicals against the high carbon dioxide in the air, and built extra skimmers themselves. At the end of two years the coral losses were less than anticipated and there was evidence that the corals had reproduced.

When the ecologists Joel Cohen and David Tilman reviewed the biodiversity results in Science in 1996 their verdict on the whole was sober: “Dismembering major biomes into small pieces, a consequence of widespread human activities, must be regarded with caution,” and no one yet knew how to engineer the services that natural ecosystems provide for free. The reef was the exception, and it turned out to be the beginning of the project’s real scientific afterlife.

What the reef taught

When Columbia University took the building over at the end of 1995, it opened the seal and turned the place into something it was better suited to be: a set of very large, very controllable experimental ecosystems. The ocean became the most important of them. Chris Langdon and colleagues used it for nearly four years to do something no one could do to a real reef: change the chemistry of the seawater on purpose and watch what the corals did.

By adding sodium bicarbonate, sodium carbonate or calcium chloride they moved the water’s saturation state with respect to the carbonate mineral corals build from, and measured the whole community’s rate of calcification as it changed. The rate tracked the saturation state as a straight line, over days and over years alike. As the water’s carbonate fell, the reef built less skeleton. The paper’s projection was that a doubling of atmospheric carbon dioxide would cut the surface ocean’s carbonate by about 30 percent from its pre-industrial level, and the reef’s building of skeleton with it.

They published in Global Biogeochemical Cycles in 2000, and it is one of the papers on which the science of ocean acidification stands: the first demonstration at the scale of a whole ecosystem, rather than a coral in a beaker, that what carbon dioxide does to seawater reaches the reef. Wallace Broecker, who had found the missing oxygen, went to the Bahama Banks the next year to check the finding against a natural one.

So the reef that was built to prove that humans could copy a living system ended up measuring how humans were changing the original. That is a strange legacy for a building sold as a prototype for Mars, and it is the one that has lasted.

Sources

• Mark Nelson and William F. Dempster, “Living in space: results from Biosphere 2’s initial closure,” in Strategies for Mars, 1996. Species-packing, the biomes and their origins, the rainforest growth, the nurse trees, the vines, the desert’s shift, the keystone-predator culling, the reef’s management and reproduction, the galago, the flying insects, the lost birds, the ants and cockroaches.
• Joel E. Cohen and David Tilman, “Biosphere 2 and biodiversity: the lessons so far,” Science 274, 1150–1151, November 15, 1996.
• James K. Wetterer, Diane S. Miller, Donald E. Wheeler and colleagues, “Ecological dominance by Paratrechina longicornis (Hymenoptera: Formicidae), an invasive tramp ant, in Biosphere 2,” Florida Entomologist 82 (3), 381–388, 1999.
• Chris Langdon, Taro Takahashi, Colm Sweeney, Dave Chipman, John Goddard, Francesca Marubini, Heather Aceves, Heidi Barnett and Marlin J. Atkinson, “Effect of calcium carbonate saturation state on the calcification rate of an experimental coral reef,” Global Biogeochemical Cycles 14 (2), 639–654, 2000. doi:10.1029/1999GB001195
• Wallace S. Broecker and colleagues, “Factors controlling the rate of CaCO3 precipitation on Great Bahama Bank,” Global Biogeochemical Cycles 15 (3), 589–596, 2001.
• “Biosphere II lives on,” Mongabay, June 2007, for the nineteen-of-twenty-five count; Carl Zimmer, “The Lost History of One of the World’s Strangest Science Experiments,” New York Times, March 29, 2019, for the hummingbirds and the bees. mongabay.com, carlzimmer.com