Three was the magic number for the very first flowering plant. The largest study into their early evolution has concluded that its flowers probably had petal-like tepals and pollen-bearing stamens arranged in layered whorls of three. It bore similarities with magnolias, buttercups and laurels – but was unlike any living flower.

The origin of flowering plants and their rapid conquest of the world’s habitats has been a puzzle for nearly a century and a half. In 1879, Charles Darwin described it as an “abominable mystery” that flowers had evolved so late in the history of life yet were still able to take over from the more ancient seed-bearing pines and cycads.

Today, flowering plants account for nine out of every 10 plants – meaning they far outnumber the once-dominant seed plants like conifers that emerged between 350 and 310 million years ago.

Studying their evolutionary roots is tricky, though: the delicacy of flowers means they rarely become fossilised. The oldest so far discovered is the 130- million-year-old aquatic plantĚýunearthed in Spain in 2015. However it is thought that flowering plants first appeared much earlier than this, sometime between 250 and 140 million years ago.

Picking flower traits

To unravel what the very first flower was like, a 36-strong team led by of the University of Paris-South, France, spent six years analysing the anatomical evidence of nearly every type of flowering plant to identify their most ancestral traits.

They calibrated their results with dates derived from molecular analyses and constructed evolutionary trees that modelled the earliest stages in flower evolution.

“We looked at the big bang of flowering plant evolution when they first evolved,” says Sauquet.

They discovered that the first flower probably had 11 or more tepals and stamens, generally grouped in threes and carried both male and female reproductive structures. It was arranged in a unique way unlike any living flower. It’s unclear how large the first flower was, but it may have been just 1 centimetre or less in diameter.

One surprise is how many petal-like tepals the first flower had compared with most living flowers. Reducing their number allowed later flowers to develop a dazzling array of specialised shapes and sizes and consequently diversify along with their animal pollinators into the enormous range of ecosystems Ěýthey occupy today. There are some 300,000 living flowering plants.

The findings mean that the living flowers identified as being most ancient, such asĚýAmborellaĚýfrom New Caledonia, and water lilies, are actually quite evolved compared with their ancient ancestors.

Read more: How flower power paved the way for our evolution

The oldest compelling fossil evidence for cellular life has been discovered on a 3.43-billion-year-old beach in western Australia. Grains of sand there provided a home for cells that dined on sulphur in a largely oxygen-free world.

The rounded, elongated and hollow tubular cells – probably bacteria – were found to have clumped together, formed chains and coated sand grains. Similar sulphur-processing bacteria are alive today, forming stagnant black layers beneath the surface of sandy beaches.

The remarkably well-preserved three-dimensional microbes will help resolve a fierce and long-running debate about what is the oldest known fossil – or at least add to it. In 1993, of the University of California, Los Angeles, claimed he had discovered fossils that are 35 million years older than the present find in nearby deposits known as the Apex chert. He thought they were probably photosynthetic bacteria.

His arguments were widely accepted until 2002 when a reanalysis by a team led by Martin Brasier of the University of Oxford found that the bacteria-like shapes could have formed in a mineral process that had nothing to do with life. ĐÓ°ÉÔ­´´s have been split between the camps ever since.

Brasier has now come back to the table with the new piece of the puzzle: slightly younger rocks that he claims do, this time, hold convincing evidence of cells.

Reasons to believe

He and his team offer multiple strands of evidence, including the physical structure of the putative microfossils and the geology and chemistry of the rocks, to prove that the forms they have discovered are biological structures. They appear to have cell walls that are consistent with bacterial life, and to have clustered and even split like modern bacteria.

The fossils were excavated from an ancient beach – now a sandstone formation near the Strelley pool in the Pilbara region – by Brasier’s colleague from the University of Western Australia in Crawley.

“Nobody had looked at fossil beach deposits because it was thought oxygen had caused the decay of all traces of life there,” says Brasier. “In fact, there was minimal oxygen in the atmosphere at this time, meaning that the fossils could preserve well.”

Indeed, grains of unoxidised iron pyrite were found among the microbes, showing that there was no oxygen present at the time. Over time, the microbes became coated in silica, forming a glassy crust that was preserved in the sandstone.

Real old

The team’s findings have been well received so far. “The multiple lines of evidence have convinced me,” says of the University of New South Wales in Sydney, Australia, who thinks that the Apex chert fossils are real.

“These structures appear to be true microfossils,” says Emmanuelle Javaux from the University of Liège in Belgium, who remains to be convinced that the Apex chert fossils are biological. Neither fossil, however, is the oldest trace of life: indirect chemical evidence suggests life may go back to 3.8Ěýbillion years.

So is it just a matter of time before another, older group of cells is found? Not necessarily. Fossils older than 3.5Ěýbillion years are unlikely, as sedimentary rocks from that time are exceptionally rare and likely to have metamorphosed beyond recognition.

But the new find does open a new line of inquiry for astrobiologists looking for life on Mars. Sand and clay deposits on the Red Planet have been altered far less than equivalent Earth rocks. “This will be an interesting target in looking for ancient life,” says Javaux.

Journal reference: , DOI: 10.1038/ngeo1238

Editorial: “We shouldn’t mourn the demotion of Archaeopteryx“

FOR 150 years Archaeopteryx has been iconic as the earliest bird. The fossil sports feathered wings but a dinosaur’s teeth and tail. Now the discovery of a feathered dinosaur in China has prompted a reassessment that has left Archaeopteryx squarely in dinosaur territory.

The diminutive new fossil, Xiaotingia zhengi, recently acquired by the Shandong Tianyu Museum of Nature from a fossil dealer, was excavated from 160-million-year-old rocks in Liaoning province (see the fossil here). It shares several key anatomical features with Archaeopteryx, including a “killing claw” on its second toe, and long and robust arms that probably allowed it to glide.

However, a team led by Xing Xu at the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing deemed it sufficiently distinct from birds to be classified as a theropod dinosaur known as a deinonychosaur – and because of the similarities with Archaeopteryx, Xu’s team concluded that the “first bird” is a deinonychosaur too (Nature, ).

“We used to think Archaeopteryx was so different from other dinosaurs that it was ancestral to birds, but recent discoveries show that this is no longer the case,” says Xu. “Our main conclusion is that Archaeopteryx is no longer a bird.”

Other palaeontologists give Xu’s findings a cautious welcome. “I am not surprised,” says Gareth Dyke of University College Dublin in Ireland. “Flight may have evolved many times among small bodied theropod dinosaurs.”

If Xu’s analysis holds up it will create quite a headache for taxonomists as Archaeopteryx is used to define the base of the birds. One solution would be to include deinonychosaurs in the birds, says Luis Chiappe of the Natural History Museum of Los Angeles County, California.