Levees surround us. Yes, some hold back rivers that strain against their embankments. But others hold back diseases, which are ready to saturate and overwhelm the fragile walls of antibiotics we’ve erected. And sometimes levees fail. The metaphor extends beyond epidemiology. Nature ceaselessly advances, trespasses, embarrasses our every effort to keep it at bay, and ultimately bursts through. Its rivers will not be contained.

In A Natural History of the Future, the ecologist Rob Dunn sketches an arresting vision of this relentless natural world — a world that is in equal measures creative, unguided and extravagant. Fog a tree with pesticides and watch new beetle species tumble from the canopy by the hundreds, a “riot of unnamed life.” Chlorinate your water and, though you might wipe out most parasites, you’ll soon bedew your shower head with chlorine-resistant mycobacteria. Make a world fit for bedbugs, then try to kill them with chemicals, and you’ll end up — not in a world without bedbugs, but one in which they’ve “evolved resistance to half a dozen different pesticides.”

Life is not a passive force on the planet, and much as we might presume to sit in judgment of Creation — even sorting species by their economic value to us — we live on nature’s terms. The sooner we recognize this, Dunn argues, the better.

As humans retreat into more and more sanitized spaces, and our homes become spotless, Febrezed bunkers of sterility, we’ll increasingly find that we’ve not only failed to eradicate our microbial opponents, we’ve actually helped create new, more virulent forms of them.

Enter the terrifying “megaplate” experiment, carried out by researchers at Harvard. In it, E. coli is made to grow across what is essentially a giant petri dish, one partitioned into sections laced with increasingly lethal doses of antibiotics. In the final stage of the megaplate, the bacteria meet a concentration of antibiotics many thousands of times higher than that which can kill your garden-variety E. coli. Even so, at the rate of “one mutation per billion divisions,” the bug evolves, casually crossing the entire plate, even the almost impossibly lethal barrier at the end, slipping through the antibiotic revetments as surely as water through a crumbling dike. And in only 10 days.

“I show it in talks,” Dunn writes of a video of the experiment. “It makes people quiet. It is what Kant called the horrifying sublime.”

The experiment can be run again, and again it will take about 10 days. What goes for E. coli scales up to agricultural pests, only temporarily inconvenienced by new pesticides until they evolve around them. While ecology is sometimes regarded as one of the squishier sciences, these kinds of eventualities begin to point to something like a set of laws underlying it all. These laws — though lacking the bedrock status of the laws of physics — can sometimes be nearly as predictive. If we want to know what’s coming, then, we would be well advised to familiarize ourselves with them, Dunn argues. To that end, his book functions as a helpful crash course in ecology and, as the title implies, an augur of sorts.

The law of the niche, for instance, predicts that many hapless species will fail to track their habitats on our warming world and will go extinct (think snails marooned on quickly shrinking islands). Others, however, will be cast about the face of the earth as their climate niche expands, and will flourish. Most concerning, perhaps, Aedes is coming. Just as epidemiologists warned for years that a pandemic was not merely possible but inevitable, ecologists now warn us of the looming mosquito heyday as we warp the climate. Key to this warning is the choice of modal verb: It’s not that these tropical pests “could” establish themselves in much of the Southern United States if we’re not careful, but rather that they “will.” And they will carry with them “some complex mix of the dengue virus and the yellow fever virus, but also the viruses that cause chikungunya, Zika fever and Mayaro.”

While it might not surprise us to read that mosquitoes have a niche that affects their distribution on the planet, it might be more difficult to recognize that we humans do as well. We are animals after all, and can be studied as such by ecologists. Even with the spread of air-conditioning, and all the creature comforts afforded by burning fossil fuels by the gigaton, we still mostly inhabit the same shockingly narrow band of the globe that we have for millennia. But as we push the climate beyond the norms of the past three million years we will hit the hard limits of physiology. And as the familiar rhythms of the seasons grow more syncopated and strange, some swath of our range will be increasingly foreclosed, to God knows what geopolitical effect. Many of us will have to move.

Along this unsettling journey into the future, the mood is leavened here and there by oddities, which Dunn dusts off like the docent of a strange natural history museum. We learn that the Taung child, one of the earliest hominins known to science, was eaten by eagles. We learn that the yeasts that make beer come from the bodies of wasps. That when humans spread out into new landmasses our “face mites diverged.” The impression all this arcana leaves with the reader is that we live in a much weirder, more disorienting world than we tend to appreciate.

We simplify this chaos, this riot of life, at our peril. We hoist plants from their natural context, consign them to vast monocultures, then act surprised when the rest of nature conspires to tear them down. We panic when bees fail to submit to the rote demands of industrial agriculture. But if simplifying nature is the cause of so many modern ills, then Dunn’s policy prescription, conversely, comes down to one simple dictum: Diversify. Diversify the microbes in your intestines, the crops in your fields, the plants in your watershed, the research in your grant proposals. Recruit the forests to filter your water. Let a trillion microbial flowers bloom.

This strategy works because nature is cleverer than us. The science historian George Dyson once described evolution itself as a kind of computational process that solves problems like how to swim, and how to fly. But the new problems we’ve given it to solve are ill considered, and the solutions it produces often undesirable. We dare life to overtop the levees of pesticides, herbicides, antibiotics; to overrun the concrete outcrops of cities, insinuate itself in the cracks of human society and pick the locks of our immune system. If we wipe out charismatic megafauna, of the sort that graces the brochures of conservation nonprofits, fine, nature seems to say, a florescence of rats and crows it is. Want to live in modular outcrops of steel, glass and cement, fed by rivers of pavement spanning thousands of miles? Very well, this will be a migration corridor for mice, pigeons and disease. They represent life too, after all, and the planet gives not a whit if it’s inhabited by lions or cockroaches. There are now beetles that consume only grains, mosquitoes that live only in the London metro. “Evolution creates,” Dunn writes, “and acts of creation are never complete.”

Dunn’s account leaves an overwhelming impression of fecundity, growth, adaptation. But this isn’t a naïvely rosy vision of the future like some contrarian tracts on the resilience of nature in the Anthropocene. From a human perspective this will be an impoverished world, and many of Dunn’s warnings are concrete and sobering. . . .  The rivers are rising.