In late March 1995, the USA's Yellowstone National Park received a special delivery. Fourteen grey wolves, flown in two months earlier from the Canadian Rockies, were released into the park. Neighbouring Idaho received fifteen wolves. Wolves howling in the snow, not heard in these forests since their extermination 60 years before, trumpeted their return.

Scientists had intended to reintroduce and conserve grey wolves in their original habitats. They did not foresee that the wolves, with blood on their teeth and claws, would restore leaves to the trees.

In fact, it seems the return of the wolves has had remarkable consequences for the entire Yellowstone ecosystem. The story illustrates how the presence or absence of a top predator can utterly reshape an ecosystem.

In Yellowstone, the wolves quickly reclaimed their spot as top predator. Ecologist William Ripple of Oregon State University has been studying the wolves since their return. He found that, within a decade of their release, the wolves had cut the number of elk – their main prey – by half. The surviving elks avoided the wolves' core range and stayed on the periphery. Woody trees like aspen and willow, which had been chewed and trimmed by zealous elks, now grew tall and lush.

Wolves suppress coyote and release foxes, because foxes are small and do not compete with wolves

The wolves' rivals, such as coyotes, also suffered their wrath. "Coyotes are very scared of wolves," says Ripple, who has seen the wolves' aggression towards coyotes in Yellowstone. "Wolves will chase coyotes, kill them, and even consume them sometimes. Wolves do not like coyotes at all."

In the Lamar River Valley in Yellowstone, coyote densities dropped almost 40% after the wolf reintroduction. In neighbouring Grand Teton National Park, coyote densities fell by 30% in the presence of wolves. Pronghorn fawns, which coyotes prey on, survived better where there were more wolves and fewer coyotes.

Ripple and Thomas Newsome, an ecologist at Deakin University, found that across North America, coyotes retreated where wolves roamed. In turn red foxes, the prey and competitor of coyotes, increased. Ripple and Newsome derived their findings from fur harvest records from eight provinces and states.

"We found on a large geographic scale that wolves suppress coyote and release foxes, because foxes are small and do not compete with wolves," says Ripple. "So, foxes benefited from wolves suppressing coyotes."

In most ecosystems, every species eats and is eaten by various others. You can picture this as a ladder. The top predator claims the highest rung; mesopredators, which are smaller predators that are eaten by the top predator, sit one rung lower; and so on to plants on the bottom rung.

In Australia, we have lost thirty mammal species over the last 200 years

This means that top predators put a cap on the numbers of mesopredators. If top predators dwindle or disappear, the cap is lifted, and mesopredator numbers should surge. This idea is called the "mesopredator release hypothesis".

Mesopredators, released in the absence of top predators, can potentially claim dominance. If that happens, species down the ladder have to contend with an unbridled mesopredator.

In the decades after grey wolves were hunted out of the USA, coyote numbers rose while wild rabbit and hare numbers plummeted. From coast to coast, snowshoe hares, white-tailed jackrabbits, black-tailed jackrabbits and pygmy rabbits were added to species-of-concern lists. Some were wiped out locally. Evidence suggests that coyotes caused the rabbit and hare's demise.

When the released mesopredator is an invasive species, the consequences can be drastic.

"In Australia, we have lost thirty mammal species over the last 200 years," says Newsome. "That's half of the world mammal extinctions in Australia alone." Such extinction rates greatly exceed past records.

When dingoes started attacking sheep, people fought back

Europeans reached Australia in the 1600s, bringing the tides of change that eventually swept aside the continent's biodiversity. Humans and their livestock now dominate the wetter, richer environments of Australia, while introduced herbivores like camels and goats graze the arid parts. As their foods shrunk and competition spiked, native animals faltered.

However, for many of the extinct animals the final blow might likely have come from two invasive predators: the red fox and domestic cat. A 2006 study suggested that predation by foxes and feral cats was a key force driving many native rodents, marsupials and birds into decline or extinction.

Despite their deadly impact, foxes and cats are not the top predators in Australia. They are mesopredators, weighing only 6kg and 4kg on average respectively. Larger marsupial predators once reigned over Australia, but they are just bones and dirt now. Today, a canine, the dingo sits atop the ladder in mainland Australia.

Looking much like a dog, the dingo weighs about 20kg. It too was brought into Australia by humans, 3,500-5,000 years ago, and used to live across most of mainland Australia. But when dingoes started attacking sheep, people fought back.

Dingoes were trapped, shot and poisoned in and around sheep farms. Australians were so keen to exclude dingoes that they built a wire fence 5,500km long and up to 2m high. The Dingo Barrier Fence, completed in 1946, keeps dingoes out of the sheep pastures of south-east Australia. South of the fence remains an almost dingo-free sheep haven.

In the presence of dingoes you can have fewer foxes and maybe fewer cats

However, it turns out the Dingo Barrier Fence affects more than dingoes.

In a 2011 study, scientists compared sites on either sides of the fence and found more small mammals on the dingo side. There were also fewer foxes on the dingo side. What's more, when scientists began to relate the distributions of dingoes, foxes and small animals, they found a consistent pattern elsewhere beyond the vicinity of the fence. Where dingoes live, there were fewer foxes and more small animals like greater bilbies, dusky hopping mice, rock wallabies, painted dragons and malleefowl.

Since dingoes and foxes compete for many of the same prey, it is easy to imagine the bigger dingoes dominating foxes through brute force. Like wolves limiting coyotes in North America, dingoes in Australia seem to suppress foxes. However, the evidence is less clear as to whether dingoes suppress the feral cats that are Australia's other major mesopredator.

The success of the grey wolf reintroduction into the Yellowstone ecosystem, and the positive response of the ecosystem to the return of its top predator, has prompted some scientists to consider reintroducing dingoes to their original habitats.

Of 30 extinct mammal species in Australia, "at least 20 are attributed to the predation of red foxes and feral cats in the absence of dingoes," says Newsome. Because studies have suggested that "in the presence of dingoes you can have fewer foxes and maybe fewer cats, and you have higher survival of native mammals, I think this is something we should explore further and pursue as a conservation tool."

However, the return of a top predator could also wreak havoc.

The top predator might have a severe impact on mesopredators, or on prey that are close to extinction. The story of wild dogs, and their struggle in the wake of the reintroductions of lions to parts of Africa, highlights the dire consequences.

Wild dogs are Africa's most endangered large carnivores. A 2012 census put the population at fewer than 1,400 adults. On average, one and three out of every ten wild dog adults and pups are killed by lions, respectively. Lions are death manifest for wild dogs.

Savé Valley Conservancy may end up trading wild dogs for lions

A study published in December 2016 detailed the conflict between wild dogs and lions in the Savé Valley Conservancy, Zimbabwe.

The conservancy received three lions in 1995, and had fewer than ten by 1999. Another ten were released in 2005, and the population had exceeded 100 by the late 2000s. Meanwhile, wild dogs recolonised the conservancy in the early 1990s and their numbers peaked in 2004. To understand the impact of lions on wild dogs, scientists examined how wild dog numbers and den sites changed between the periods 1996-1999 and 2010-2013, when lion numbers were low and high, respectively.

The growing lion pride knocked the wild dogs off their stride. The lions took to hunting in impala-rich areas, so to avoid them wild dogs shifted their dens to rugged areas with fewer impalas. Even so, their pack sizes shrunk by one-third and pup numbers halved. At least 30% and 70% of wild dog adults' and pups' deaths, respectively, were caused by lions in 2010-2013; there were no such casualties in 1996-1999.

The issue may be that wild dogs huddle near their dens for three months to raise newborns, making them easy pickings for lions. Wild dogs breed only once a year, and lions kill more pups than the dogs can produce. In effect, Savé Valley Conservancy may end up trading wild dogs for lions.

In Australia, reintroducing dingoes might also plunge some endangered animals into the same plight as wild dogs.

Wild greater bilbies in central Australia are often eaten by foxes and cats, and evidence suggests that dingoes can relieve bilbies by controlling foxes, says ecologist Euan Ritchie of Deakin University. But in particularly small bilby populations, "putting dingoes [in] could just tip them over the edge."

Similar concerns have been raised over dingo predation on the critically endangered northern hairy-nosed wombat. "There are surely areas in Australia where you wouldn't rush to put dingoes back, because that could contribute to extinction of critically endangered animals," says Ritchie.

On top of that, there is the conflict between sheep and dingoes. "Dingoes and sheep don't mix," says Peter Fleming, a pest researcher at Australia's Invasive Animals Cooperative Research Centre. The issue is that dingoes kill more sheep than they eat. "If you have dingoes with your sheep, they will continue to kill your sheep until your sheep's all gone or the dingoes die."

Socio-economic issues aside, scientists are yet to agree on the impact of dingoes on mesopredators.

Some scientists, like Fleming, insist that only experiments can confirm whether a mesopredator release will happen. For that, dingoes should either be added into or removed from multiple sites, and mesopredators monitored over time and compared against sites without dingo treatments. These experiments must be done across different ecosystems to reflect Australia's diverse environments.

We weren't dealing with negative press about buffoons trying to reintroduce dingoes to kill all the sheep

"I'm quite happy to accept whatever evidence it is, as long as it's causative rather than just correlative," says Fleming. "You should not make decisions based just on correlative evidence."

Other scientists, like Ritchie and Newsome, see compelling evidence in various studies showing that fewer dingoes correlate with more red foxes. Although correlational studies cannot pinpoint underlying processes like experiments can, they capture patterns across large swaths of land and time, says Ritchie.

"If we continue to take dingoes out of the system, or compare between areas with and without dingoes, we don't really know to what extent dingoes suppress foxes and cats," says Newsome. Reintroducing dingoes as an experiment would assess the full ecological effects of dingoes. "It would help us make an informed decision about the next step going forward."

To that end, in 2015 Newsome, Ritchie, Fleming, Ripple and others proposed reintroducing dingoes as an experiment.

They suggested Sturt National Park, an arid area of 3,000 sq km tucked in the north-west corner of New South Wales. Red foxes, feral cats and several large herbivore species live in the Park, and the Dingo Barrier Fence runs along the north and west borders of the Park.

At the moment, it doesn't look like there's any evidence of suppression

If the fence was moved inwards to run along the south and east borders instead, dingoes should naturally recolonise the Park. Scientists could then compare sites inside and outside the dingo-recolonised Park to see how an increase in dingoes affect their communities. Newsome is currently talking to stakeholders and sponsors about the project.

When Newsome first proposed his dingo reintroduction experiment, he was surprised by the press's encouraging response. "We weren't dealing with negative press about buffoons trying to reintroduce dingoes to kill all the sheep," he says. "That was the more common rhetoric about 15 years ago."

Meanwhile, Fleming and his team are about to end a four-year dingo-removal experiment in east Australia. "At the moment, it doesn't look like there's any evidence of suppression," says Fleming. After "eye-balling the data", Fleming says dingoes, foxes and cats all increased when they stopped removing dingoes. If dingoes suppress foxes, then "dogs would go up but the foxes wouldn't. But that doesn't occur [in their experiment]." Fleming aims to publish the results by the end of 2017.

If we wish to enlist dingoes to control red foxes and feral cats, we would need to do more than test the mesopredator release hypothesis. We must also protect the welfare of sheep farmers and their flock from dingoes. For that, we may have to turn to another introduced canine: livestock guardian dogs.

Linda van Bommel, an ecologist at Australia National University, has been studying the use of guardian dogs to protect Australian livestock. Maremma sheepdogs, a white woolly dog of 40kg, are one of the more common guardian dogs used in Australia. Maremma bond well with sheep. "Often, the sheep will run towards the dogs, and the dogs will stand between the sheep and the threat," says van Bommel.

Most farmers benefited from having the dogs. In van Bommel's survey, published in 2012, 66% of livestock guardian dog owners said sheep predation stopped after they started using the dogs. Another 30% reported lower sheep predation.

One farmer used to lose all his lambs and 100 sheep a year, either to wild dogs or dingoes. But with guardian dogs among his sheep, he kept most of his lambs and did not lose a single adult sheep. "For that farmer," says van Bommel, "it's the difference between going broke and continuing to profit from running sheep."

In another study, published in 2016, van Bommel found that red foxes and grey kangaroos avoid Maremma sheepdogs. Hence Maremma sheepdogs not only protect livestock from dingoes, they may even function like top predators and deter mesopredators and large herbivores.

If dingoes were ever to be reintroduced into areas around farms, farmers would need ways to protect their flock without killing the dingoes. "Livestock guardian dogs would be prominent candidates for that," says van Bommel.

However, she says there is still a mountain to climb. "Most farmers would rather not see dingoes in the environment altogether."

Today, more than ever before, scientists are studying predators and unraveling their influence. "It is a new science. I expect a lot more to be known," says Ripple, whose grey wolf studies in Yellowstone popularized the positive role of predators.

A century ago, people were enthusiastically killing lions, wolves, and dingoes. When we eliminated the top predators, we unknowingly released mesopredators that have led to unexpected, and still underappreciated, consequences. Now, though we have yet to fully appreciate the dynamics of predators, we are at least debating alternatives to killing them. Nature is red in tooth and claw, and it seems that can be a good thing.

Join over six million BBC Earth fans by liking us on Facebook, or follow us on Twitter and Instagram.

If you liked this story, sign up for the weekly features newsletter called "If You Only Read 6 Things This Week". A handpicked selection of stories from BBC Capital, Culture, Earth Future and Travel, delivered to your inbox every Friday.