For conservationists, it is an alarming and rapidly worsening problem: Natural areas continue to be degraded and species lost at an unprecedented rate, even in the most remote forests—and even as the extent of parks, refuges and other protected areas has dramatically increased. One explanation, a study published this week in Nature Ecology & Evolution argues, is that policy-makers working to designate protected areas need to have far more precise targets.

Since signing the Convention on Biological Diversity in 1993, the 168 member nations have almost doubled the size of their protected areas in their efforts to meet that treaty’s goal: that each nation protect at least 17 percent of its land area and 10 percent of its oceans by 2020. But these simple numerical targets have resulted in too much emphasis on the quantity of the acreage being protected—and too little on the environmental quality of what is being protected there—according to James E. M. Watson, a scientist at the Wildlife Conservation Society and senior author of the new study. Moreover, the 17 and 10 percent goals have led to the “implicit assumption,” Watson and his co-authors contend, that in the rest of the unprotected world—83 percent of the land and 90 percent of the oceans—“natural values will be lost or severely degraded,” though that would be “an unmitigated disaster for both biodiversity and humanity.”

The new study proposes instead that along with designating new protected areas, policy-makers should also stipulate a set of science-based, site-specific “nature retention targets” aimed at keeping natural systems intact and functioning. “What we need to do in the next rounds of targets,” says lead author Martine Maron, a conservation policy scientist at the University of Queensland in Australia, “is to put numbers not just on biodiversity conservation targets, but also on all the other things we care about from nature.” She says examples should include targets for wilderness retention, for human-nature interactions, for carbon capture and storage to slow the rate of climate change, and for watershed protection to battle flooding and drought. The lack of such targets now, Maron adds, risks creating “a situation where we continue blindly losing nature without any idea of how to stop it. This is not just what we need to preserve for other species. This is what we need for humanity.”

The study argues that such targets need to specify an intended end state—for instance, exactly how much intact forest needs to be retained for each hoped-for function. By contrast, the current Convention on Biological Diversity merely aims to halve the rate of loss of natural areas. While that sounds good, the study authors write, it “does not describe the desired state that the reduction is to achieve” and leaves “wiggle room” for nothing much to change.

The subject of how to stem the loss of natural areas is timely because scientists and policy-makers are already discussing the next phase of the Convention on Biological Diversity, to be negotiated at a 2020 meeting in Beijing. A conference is also taking place this week at the University of Oxford, co-sponsored by the Wildlife Conservation Society, to draw attention ahead of the Beijing meeting to protecting “intact forest landscapes.” These are defined as “large, unbroken swaths of forests whose ecological functions remain unharmed by human activity.” About 9.3 percent of such landscapes have been degraded or lost since 2000, according to a recent analysis led by the University of Maryland—and the rate of loss appears to be accelerating as logging roads bring new human pressures into formerly isolated forests.

Watson argues that protecting intact forests is the most efficient and least expensive way to capture key natural values. Having “nature retention targets” in place for such areas might, he suggests, have prevented the forests of Indonesia and Malaysia from being “smashed” over the past 20 years by palm oil plantation development. Designating upland forests as “no go” zones for logging might have avoided subsequent flooding in downhill areas from Haiti to Kashmir. Looking to the future, he says Colombia would benefit from specific targets for its intact forests, to help it address economic growth following a 2016 peace agreement to end rebel activity that had slowed development there.

“I can see a certain strong argument for targets. They can be useful in giving guidance, even if you are overshooting them,” says Yadvinder Malhi, a tropical forest ecologist at Oxford who was not involved in the current study. “The question is, what is the scientific basis? It’s very challenging if you are looking, for instance, at watershed services, to say this is the minimum forest you need to protect.”

Frances Seymour of the World Resources Institute, who also was not involved in the study, likewise expresses caution about its proposed “headline target”—essentially a global figure for how much nature needs to be retained. A natural service like carbon capture and retention, which produces a global benefit no matter where it happens, might easily feed into a headline target, Seymour says. But “others are specific to a location, like protecting an endangered species or keeping a watershed functioning, and doing more in one place doesn’t make up for doing less in another,” she says.

“Often it takes catastrophic events to get the attention of politicians,” Seymour adds, “because it gets the attention of the public.” She believes the challenge for conservationists is to get government leaders at all levels to think—years in advance of any potential catastrophic event—about “what it means to lose your forest. We need to align the incentives to make retaining forests politically attractive. So far conservationists haven’t done a very good job of making that connection.”

Watson and his co-authors argue that their nature retention targets could change that. The alternative, Watson warns, is a natural world left seemingly without end “in freefall.”