Spring is the season of blossoms, birds and bees, especially in the temperate parts of the northern hemisphere.
It was also the beginning of the end for all non-avian dinosaurs and many other prehistoric reptiles, a new study suggests.
Spring was when the dino-killing asteroid slammed into Earth 66 million years ago, according to a study of fish fossils published today in Nature.
Melanie During and colleagues studied fossils excavated from the Tanis site in North Dakota.
It was "unimaginable" to be able to pin down the timing of the event, said Ms During, who led the project while a masters student at the University of Amsterdam.
"If we are extremely lucky as a palaeontologist, maybe we get into the millennial timescale, but nowhere near this high resolution.
"This study has showed that yes, you can read bones, and infer seasonality from them."
Kate Trinajstic, a fish palaeontologist at Curtin University who was not involved in the research, said evidence presented in the study supported the idea that fish were wiped out in spring.
"To say the asteroid hit in spring, is really quite amazing to get that accuracy," Professor Trinajstic said.
But others were cautious of findings about fossils excavated from the controversial site.
A car crash frozen in place
The Tanis fossil site is on private property in the state of North Dakota in the US.
It contains stunningly preserved fish fossils with shards of glass in their gills and the sediment surrounding them.
The discovery divided palaeontologists when it was first announced in 2019, and questions were raised about access to the site for independent assessment.
"I know there's a lot of scepticism about this site," said Ms During, who was not part of the original team.
According to the team that originally excavated the site, it captures the mass death of a range of animals living in a freshwater river in the hours after the asteroid slammed into the Yucatan Peninsula some 3,500 kilometres away.
As the shockwave reverberated through the continent, fish in the water were pelted by searing hot glass shards and buried alive by layers of mud slooshing back and forth over them.
Ms During was given access to the site in 2017 by Robert DePalma, who led the original team.
"I compare the site to a car crash frozen in place," Ms During said.
Above this scene of carnage it is still possible to see a pink layer of iridium, a signature found when meteorites hit Earth.
Ms During dug up the jawbones of two ancient paddlefish and shoulder fin spines from four sturgeon and took them back to the Netherlands for analysis.
Bones read like tree rings
The bones were scanned using the European synchrotron to reveal growth patterns in the bone like tree rings, right down to individual cells.
"I could literally see the bone cells increase and change shape from spring to summer to autumn to winter over multiple years in multiple fishes," said Ms During, who is now doing her PhD at Uppsala University.
Bone growth in fish is related to temperature, said Professor Trinajstic, who studies fish fossils from much earlier in Earth's history.
"As the seasons warm up, the fish grow more and you get wider rings," she said.
The fish in this study died during a growth period.
"If it was further along it could be summer, but the fact is it's just been caught at that moment where that cycle of rapid growth is just starting again," Professor Trinajstic said.
"They were wiped out basically at the time they were going to go through maximum growth and maximum reproduction."
Food was flush when fish died
An analysis of the carbon isotopes in the fishes' bones also indicated they died when food was abundant.
The idea is that fish eat other animals such as crustaceans, which eat phytoplankton that produce energy through photosynthesis.
When sources of food that rely on photosynthesis are booming, carbon-13 builds up relative to carbon-12 and can be traced through the food chain.
The results showed that patterns in bone growth matched the times of year when C13/C12 ratio increased or decreased.
"You could see every time the fish stop growing, they stop eating as much food, it [the C13/C12] ratio dips down," Ms During said.
Kliti Grice, a geochemist at Curtin University not involved in the research, agreed that the isotope data indicated the fish died when there was lots of food.
"It's a very nice study, it's got lots of merit and is quite a different approach," said Professor Grice, who has studied the site of the Chicxulub asteroid impact, as well as fossils in other highly preserved sites.
But she said more work was needed to pin down when the asteroid wiped out the majority of dinosaurs.
"I think you'd definitely need to see this data on a global [scale] at more than one site," she said.
Why does seasonality matter?
Pinning down just when the asteroid hit could help scientists understand the enduring mystery of why some animals died and others didn't, in what is known as the Cretaceous-Palaeogene (C-Pg) mass extinction.
Following the initial asteroid impact, the world spiralled into a hellish scenario of fires, acid rain and a deep nuclear winter that lasted for thousands of years.
The team speculate that if the asteroid hit Earth during the northern hemisphere spring, it may help explain why this region was slower to recover than the southern hemisphere.
The idea is that while plants and animals in the north would have been going through their peak breeding season when the asteroid hit, those in the southern hemisphere autumn were already winding down and may have escaped the immediate hit.
"The recovery phase was much faster [in the southern hemisphere]," Ms During said.
But, she added, there could also be a bias in the data.
It is extremely hard to find sites in the southern hemisphere that document the C-Pg (also known as the K-T) boundary, said Steve Salisbury, a palaeontologist at the University of Queensland not involved in the study.
But, he said, there are some parts of Antarctica such as Seymour Island that have late-stage Cretaceous deposits.
"Seymour Island is probably the best place on the planet to see continuous deposition before, during and after the KT boundary," said Dr Salisbury, who has done extensive work in the area.
Unlike the area around Tanis River, where four seasons are etched in tree rings, the southern hemisphere was very different 66 million years ago.
Back then Australia was still connected to Antarctica.
"There was still no Southern Ocean, so seasonality was not as extreme as it is now because a lot of the key drivers for global climate related to circulation of the Southern Ocean around Antarctica weren't present."
And even if there was some seasonality, connecting it to the random nature of the extinction event was "a hard thing to do," he said.
"Why did big marine lizards like mosasaurs, plesiosaurs and ichthyosaurs that had global distribution disappear, yet marine turtles and crocodiles didn't?
"It doesn't seem to be anything to do with the physiology of those groups that would explain why they did or didn't make it."
Question over site isn't settled
The original team led by Mr DePalma also conducted a separate analysis of fish fossils from the site, publishing their results late last year.
Ms During said that while the two papers reached similar conclusions, her study, which was accepted before the other study, was independent and went into more depth using 3D technology to view the bones.
"Consistent insights from independent studies will always benefit the advance of research as a whole, as it allows researchers to corroborate new findings with independently obtained data," she said.
All the data from her research, including the full set of isotopic measurements and high-resolutions scans, would be made available for other researchers to interrogate, she added.
"We found the season, that's the data, everything else is interpretation trying to see where this can take us in the future," Ms During said.
Dr Salisbury said that while the ideas in the research could be used to test other fossil sites, questions still remained about the Tanis River site itself.
"If this is what they say it is, then it is pretty exciting, but I think there really needs to be independent verification of the sedimentology of the site to demonstrate its the sort of event they are proposing," Dr Salisbury said.
A case in point, he said, was evidence of a massive fish kill on Seymour Island around the time of the K-T boundary.
"Initially, a lot of people back in the '90s thought this was an impact layer, but trying to demonstrate a fish layer coincides with the impact has been impossible to prove."
Dr Salisbury said the community would continue to be sceptical until the Tanis River site was verified, or evidence was found at another site, because "these sorts of things have been proposed before and haven't stood up".
