The Serengeti plain of East Africa is one of the world’s great wild lands — teeming with lions, leopards and migrating wildebeest. But is it ecologically intact, a rare fragment of the earth unaltered by the hand of humanity? Or is it, as many researchers argue, a human-created landscape, nurtured by generations of Maasai cattle herders?

And should we care? In the Anthropocene, should conservation be about protecting iconic species, ecological intactness, nature’s resilience, or human custody of landscapes —whether in the Serengeti or other famed wild landscapes such as the rainforests of the Congo basin or the vast tundras of Siberia and Canada?

These questions have all been addressed in three new research papers, all published this month, which reach very different conclusions on the nature we have, how to conserve it, and the best way to fulfil the UN call to make the 2020s a Decade on Ecosystem Restoration.

The issue of which of the world’s wild lands are truly pristine has long bugged ecologists. Conventional estimates put the amount of “intact” wilderness on the planet’s land surface outside Antarctica at between 20 and 40 percent. But Andrew Plumptre, who heads the Key Biodiversity Areas Secretariat, a coalition of NGOs, in Cambridge, England, says this is a false assumption. It is based on satellite assessments of human impact that pay little regard to what species are down below.

In the first of the three papers, published in Frontiers in Forests and Global Change, Plumtre sets a new, higher bar for intactness. It requires both untouched habitat and the presence of all the species known from the historical record to have occurred there, in numbers “sufficiently abundant to play their ecological roles.” By this measure, he estimates that 43 percent of the land has intact habitat — higher than some estimates — but once species loss is taken into account, the proportion falls dramatically, to just 2.9 percent.

In practice, Plumptre and his international team of co-authors count mainly large mammals such as gorillas, chimpanzees, forest elephants, jaguars, buffalo, bears, and orangutans; and they set their baseline from which decline is measured at AD 1500, the moment when Europeans showed up in the Americas. An earlier date might have produced an even lower number.

Many apparently pristine ecosystems, Plumptre says, should not be considered intact because they are empty of key large mammal species, often because of hunting. That loss may fundamentally change how their ecosystems work. For instance, the disappearance of elephants and other large herbivores will reduce nutrient cycling and seed dispersal in those animals’ dung. And the loss of predators such as lions or wolves will lead to an excess of herbivores, to the detriment of the vegetation they eat.

In a separate commentary, Plumptre names some of the areas that still meet his strict measure of intactness. They include the Nouabale-Ndoki National Park in the rainforest of the Republic of the Congo; the mosaic of forests, fjords, lakes, and wetlands that make up the islands of the Kawesqar National Park in Patagonia in southern Chile; and the unfenced megafauna-rich grasslands of the Serengeti. “These are very rare and special places that should be conserved,” he says.

His diagnosis sounds bad. Especially because he finds that only 11 percent of these intact ecosystems are within government protected areas. But the good news, Plumptre says, is that we could restore intactness to up to 20 percent of the land surface by reintroducing five or fewer species into areas otherwise intact, and that this could be done during the UN Decade on Ecosystem Restoration.

The decade was agreed upon at the UN General Assembly in 2019 and runs from this year to 2030. It has so far concentrated on restoring habitats, Plumptre says, but should do much more to restore species. Candidates should include putting forest elephants back into the Congo, and returning buffalo, giraffes, and zebra to African grasslands where they have been hunted out.

Plumptre’s focus on species contrasts strongly with a paper published in Science three days later by Fernando Blanco of the Leibniz Institute for Evolution and Biodiversity Science in Berlin. In it, he and a team of Spanish and German co-authors argue that we care too much about species. However well-liked our great iconic species may be, these scientists note, they are transient. They come and go. What should matter to conservationists interested in protecting nature is preserving its resilience, which lies in ecosystems that persist over millions of years.

“Actions carried out to preserve ecosystem functioning will last longer than actions oriented towards protecting individual species,” says co-author Juan Cantalapiedra of the University of Alcala.

They base their case on evidence from newly assembled and extensive fossil records of large mammals that lived over the past 21 million years in the Iberian Peninsula of Spain and Portugal. They identified 396 mammalian species living there during the period, and categorized them all according to their diet, body size, and method of locomotion to infer their role within their ecosystems.

Blanco and his colleagues found massive shifts through time in the specific species present within the ecosystems. Extinctions were frequent. A typical survival time for a species before its extinction was a little under a million years. Yet the ecosystems themselves persisted much longer. Even a drying up of the Mediterranean Sea and the recurring ice ages of the past 2 million years failed to shift the ecological equilibrium. Each iteration of species carried out essentially the same jobs as their predecessors. So, the paper’s authors say, making efforts to save them is of limited value if what we care about is protecting ecosystems and nature’s resilience.

“In the last 20 million years, only two environmental changes significantly affected this ecological structure,” says co-author Manuel Hernandez Fernandez of the Complutense University of Madrid. Those changes on the Iberian Peninsula occurred 14 and 9 million years ago, and both were associated with climate change, when rainfall patterns shifted worldwide. As it got drier on the peninsula, browsers of shrubs and trees gave way to grazers of grasslands. But which browsers, and which grazers, seemed to matter little to the ecosystems themselves.

These “deep-time ecological dynamics” suggest that what matters is not preserving today’s inheritors of the grazing mantle but ensuring that there is a role in the wider ecosystem for such grazers, says Blanco.

This conclusion, based on fossil evidence, gives weight to a controversial study in 2019 that found a constant churning of species within today’s ecosystems, in which new species often replaced others doing similar jobs. The study, covering 50,000 sites across the world, found that individual places swapped a staggering 28 percent of their species every decade. It found the churn occurred in apparently pristine ecosystems as well as those obviously disturbed by human activity.

“Of course, we know that conservation efforts are often guided by politics as well as science,” Blanco agrees. “But it is important to think more explicitly about our long-term priorities.”

The two papers offer very different perspectives on what ecological intactness means, and what conservationists should be protecting. But both look essentially at ecosystems without humans. That is a big mistake, says geographer Erle Ellis of the University of Maryland Baltimore County, in a paper in PNAS co-authored with archaeologists, ecologists, and geographers from 10 other institutions in the United States, Netherlands, China, Australia, Argentina, and elsewhere. They argue that “archaeological data demand new approaches to biodiversity conservation.”

Their study, published a week after the other two, maps ecosystems and biodiversity over the past 12,000 years, since the end of the last ice age, and compares them with known locations of human activity. It contends that the popular presumption — not least in reporting of Plumptre’s paper — that most of the world’s ecosystems were all but untouched until 500 years ago is nonsense. Plumptre says in his paper that species composition changed earlier due to human influence, but says he chose AD 1500 because it is the baseline for assessing species extinctions used in the IUCN Red List of Threatened Species.

“Even 12,000 years ago, nearly three-quarters of terrestrial nature was inhabited, used and shaped by people,” says Ellis. “Areas untouched by people were almost as rare then as they are today.”

“Hunter-gathers, early farmers and herders transformed wildlands into human biomes” through burning, shifting cultivation, hunting, and animal domestication, he says. Often, far from wrecking ecosystems, they enhanced the number of plant species, spreading seeds and improving soils by burying household waste.

The problem today “is not human use per se,” says co-author Nicole Boivin of the Max Planck Institute for the Science of Human History in Jena, Germany. “The problem is the kind of land use we see in industrialized societies,” with monoculture plantations taking over former forests and grasslands fenced and plowed.

Ellis says that most of the areas that Plumptre defines as intact “actually have a land-use history of thousands of years.” But the impact that land use had on ecosystems was very much less back then. By and large, he says, the ancients “sustained the vast majority of Earth’s biodiversity.” And the good guys then remain the good guys today. The study finds “a close correlation between areas of high biodiversity today and areas long occupied by Indigenous and traditional peoples.”

He says this insight has often been ignored by modern conservationists, who work from the presumption that the areas most worth protecting are “intact” places with no human influence. The real truth, he says, is that nothing is intact, but many ecosystems remain in good working order, often thanks to human stewardship.

“Rather than trying to return land to an unattainable pristine state, conservation efforts would achieve more by empowering traditional and Indigenous societies and supporting local, community-based sustainable ecosystem management,” Boivin argues.

Rebecca Shaw, chief scientist at the World Wildlife Fund in San Francisco and a co-author of the paper, agrees that “the perspectives of Indigenous and local peoples should be at the forefront of global negotiations to reduce biodiversity loss.” But some of her fellow authors go further. They want not just Indigenous “perspectives,” but Indigenous control. “These findings have particular salience for contemporary Indigenous rights and self-determination,” says co-author Darren Ranco, an anthropologist at the University of Maine and a citizen of the Penobscot Indian nation in that state.

Certainly, the paper’s analysis will add fuel to the criticisms of some human-rights activists about current calls by conservationists to increase the areas of land protected by governments from the current 16 percent to 30 percent. Survival International this month called this goal “the biggest land grab in history,” arguing that it is both unjust and counter-productive, because it will strip land rights from the planet’s best ecological protectors.

Those activists have ammunition. They point out that one of the intact protected ecosystems singled out by Plumptre, the Nouabale-Ndoki National Park in the Republic of Congo, was created in the 1990s by expelling several thousand Bayaka “pygmy” people, who had lived there sustainably for generations. Another “intact” protected area, the Serengeti, has seen widespread expulsions of the Maasai, who for centuries nurtured it.

The three papers clearly have very different ideas about ecological intactness and the best way of conserving wild lands. They start from different approaches: a love of species for their own sake; a desire to maintain the systems that ultimately sustain life; and a more human-centered view of what is, and what is not, possible.

But there are some areas of agreement. Notably, Plumptre concedes in his paper that many of the areas he has identified as ecologically intact “coincide with territories managed by Indigenous communities, who have played a vital role in maintaining the ecological integrity of these areas.” His definition of intactness may rely a lot on his choice of a baseline in AD 1500, by which time the cultural landscapes that Ellis identifies were well established.

Back then, the Amazon was at its peak of human occupation, planted with economically valuable trees and underlain by human-made soils. Back then too, forest elephants were living in profusion with local tribes in Central Africa, only disappearing when European colonialists moved in to plunder their ivory. And the Serengeti had long been shared between Maasai cattle herders and Africa’s most charismatic megafauna.

“Natural history is human history,” says Ellis. “Conserving biodiversity is not about finding places without people, but about conserving the biodiverse cultural landscapes that people have shaped and sustained.”