It is almost 30 years to the day since Chernobyl became synonymous with nuclear disaster.
In the early hours of 26 April 1986, an experiment designed to investigate the safety of the nuclear reactor went badly wrong. Radiation spilled into the environment.
Within weeks, hundreds of thousands of people in a 30km exclusion zone around the plant had been evacuated. To this day the zone remains largely uninhabited.
It is far from lifeless, though. Visit the exclusion zone today - which some bold tourists do - and you will find that the local wildlife is thriving.
The big question: does this mean that the environment can cope with a nuclear disaster even on the scale of Chernobyl?
No one doubts that the immediate ecological effects of the Chernobyl disaster were devastating.
One to two years after the accident populations began to recover
In one area of forest covering between 4 and 5 sq km (1.5-1.9 sq miles), many coniferous trees died. The dying needles turned rusty red, earning the region a new name: the Red Forest.
"In that first year, in the most contaminated areas many soil invertebrates were killed, and the small mammal population plummeted," says Nick Beresford at the Centre for Ecology & Hydrology in Lancaster, UK.
However, in large areas of the exclusion zone radiation levels dropped dramatically within months, says Jim Smith at the University of Portsmouth, UK. Wildlife began to bounce back, taking advantage of the absence of people.
Exactly why, how and even when that recovery began is unclear.
Within 10 years of the disaster, the small mammal populations were apparently showing no ill effects
The Iron Curtain still separated east from west at the time, and the Chernobyl region was accessible only to Soviet scientists. "Understandably their early research focus was on exposure of humans and how this could be reduced," says Beresford.
Even so, in the late 1980s local scientists did start a programme of aerial survey counts of three wild species: elk, roe deer and wild boar. The surveys showed slow but steady rises in the abundance of all three species.
"That helicopter data isn't exactly brilliant," says Smith. "The counts were done only once a year, and they didn't cover the whole area. But it does suggest that one to two years after the accident populations began to recover."
In the mid-1990s came better evidence of what was going on.
A team of US and Ukrainian ecologists set up traps to explore how small mammals were responding. They caught a range of voles, mice and shrews – and found that the abundance of animals and the diversity of species was more or less identical both inside and outside the exclusion zone.
If anything, wolves are faring better at Chernobyl than at the other reserves
In other words, within 10 years of the disaster, the small mammal populations were apparently showing no ill effects from the radiation.
We should not be surprised by these findings, says Smith. "There have been a lot of radiobiological studies over the decades to find out what it takes to really damage animal populations, to do some serious reproductive damage. And across most of the exclusion zone, the doses aren't really high enough to have that effect."
Last year Smith and his colleagues in Belarus, the UK, Russia and Germany published details of the most in-depth survey to date into the mammalian populations in the exclusion zone. Again, they suggest that the radiation now has only a limited impact on wildlife.
Between 2008 and 2010 they surveyed hundreds of kilometres of animal trackways, to assess population densities of elk, wolf, wild boar, roe deer and foxes. They found that the track densities were similar to those recorded at four radiation-free nature reserves in Belarus.
There is a clear signal of the negative effects of radiation on wild populations
If anything, wolves are faring better at Chernobyl than at the other reserves. The data suggests they might be seven times as abundant.
Beresford, with his colleague Mike Wood of the University of Salford, UK, has also reported encouraging results.
They set up motion-activated camera traps in the exclusion zone, as part of an ongoing project to better understand the risk to humans and wildlife associated with exposure to radioactivity. They found evidence of an extraordinary abundance of species. There were beavers, badgers, lynx and bison: even a brown bear made an appearance.
But it would be wrong to say there is universal scientific agreement that the wildlife of Chernobyl is hale and hearty.
Anders Møller at University of Paris-Sud and Timothy Mousseau at the University of South Carolina in Columbia have spent 15 years exploring the impact of wildlife in the area. They have reached a very different set of conclusions.
There are measurable genetic consequences of exposure to low-dose-rate radiation
"In almost all cases, there is a clear signal of the negative effects of radiation on wild populations," says Mousseau. "Even the cuckoo's call is affected."
For instance, in 2009 Møller and Mousseau performed their own mammal track count in the exclusion zone – albeit on a much smaller scale than the study Smith and his colleagues undertook. The results, published in 2013, suggested that track abundance is low where radiation levels are high.
Another study the pair published in 2009 suggested that insects and spiders are less abundant in areas of the exclusion zone where radiation levels are high.
The study suggested there was an impact on insect abundance even in areas of the exclusion zone where radiation levels are now extremely low.
"Based on Chernobyl studies, not just our own, most of the rigorous scientific reports indicate that there are measurable genetic consequences of exposure to low-dose-rate radiation," says Mousseau. Those consequences come either in the form of damage to chromosomes or elevated mutation rates.
We didn't find any of these effects even in the most contaminated lakes
"They're publishing evidence of effects at radiation levels within the range of UK background radiation levels," says Wood.
How is it possible for radiation levels far below those considered harmful to have a significant impact on animal health?
Møller and Mousseau argue that the studies used to establish safe levels of radiation are largely performed under laboratory conditions. Out in the real world, animals face a multitude of ecological pressures that are not replicated in those lab studies. In natural settings, animals might be weaker and consequently more vulnerable to the effects of low-level radiation.
However, Møller and Mousseau are virtually alone in their views. Most of the other researchers who have spent time working at Chernobyl take issue with their findings.
"I very much question some of the dose rates where effects are being claimed on the insects," says Smith. "We did some studies on aquatic invertebrates. We didn't find any of these effects even in the most contaminated lakes."
Møller and Mousseau are not quite alone in reporting damaging effects from very low-level radiation exposure
It is not just Møller and Mousseau's findings that are questioned. Their research methods have also come in for criticism. It does not help that, in 2002, the Danish Committees on Scientific Dishonesty ruled that Møller had fabricated the data for an unrelated study, a claim that he has always denied.
One of their latest studies claims that bank voles in the exclusion zone have unusually high numbers of cataracts, and that this is linked to radiation levels. It was published in January 2016.
Smith, Beresford, Wood and several of their colleagues posted a lengthy comment beneath the online version of the paper. They highlight what they claim are a multitude of problems and flaws with the science. In particular, they argue that freezing animals for later study can inadvertently cloud their eyes in a way that could give the false impression that they had cataracts.
However, Møller strongly rejects these criticisms of his research. "There is no effect of freezing, neither has such an effect ever been shown," he says.
There does not seem to be a way to explain all the results
It is also fair to say that Møller and Mousseau are not quite alone in reporting damaging effects from very low-level radiation exposure. A study by a Japanese research team published in 2012 looked at the biological impact of the 2011 Fukushima nuclear accident. It concluded there were effects on the pale grass blue butterfly even at low-dose exposures.
In response Beresford, with David Copplestone at the University of Stirling, UK, published an article to highlight what they say are basic errors and misunderstandings about how such studies should be performed.
This ongoing controversy means that, on the 30th anniversary of Chernobyl, there are two views on its legacy, and they are poles apart.
Either the evidence from the exclusion zone shows that ecosystems are far more robust to the fallout from nuclear disasters than we thought, or they are more vulnerable than anyone imagined possible. There does not seem to be a way to explain all the results from both sides of the debate, so one side or the other must simply be wrong.
To this day the zone remains largely uninhabited
This matters, because in order to figure out how to manage the exclusion zone, we need to know how its wildlife is responding. "It needs to get resolved from the perspective of having good and credible robust regulation," says Beresford.
With that goal in mind, even more research is now underway around Chernobyl. The emphasis now is on making sure that the studies involve scientists with expertise in radiation studies as well as in ecology, so that there can be no doubting either the quality of the research or the conclusions that are reached.
The hope must be that by the time the next significant Chernobyl anniversary rolls by, the legacy of the disaster is clear and unambiguous.