More than 100 million years ago, dinosaurs roamed Maryland. So did our ancestors - small mammals the size of squirrels or badgers - and the flying reptiles known as pterosaurs.

Amazingly, the footprints of all these creatures of the Cretaceous era were preserved on a single 2.5-metre-long slab of sandstone unearthed on the grounds of NASA's Goddard Space Flight Centre in Greenbelt, Maryland, not far north of Washington, DC.

"It's unusual to have such a large concentration of different kinds of tracks and small tracks in such a small space," said Martin Lockley, an emeritus geology professor at the University of Colorado at Denver, who studied the tracks.

Lockley and his colleagues described the findings in an article published in the journal Scientific Reports on Wednesday.

The slab offers unique insights into the behaviour of dinosaurs and early mammals; possibly some of the dinosaurs were looking to make a meal of the mammals.

All this might never have been discovered if an amateur dinosaur fossil hunter hadn't gone to lunch with his wife not long before the construction of a new building obliterated the site.

Even back when dinosaurs ruled the world, the Washington, DC, area was a swamp. Somehow, the right sequence of events allowed the traipsing of the animals across a muddy surface to be preserved in stone. Millions of years later, the rock happened to be poking out to reveal its paleontological bounty.

At one end of the slab, there is a single footprint of a juvenile sauropod, a long-necked plant-eating dinosaur. At the other end is a print from a nodosaur - an armoured plant-eater as heavy as a small elephant. Alongside are smaller footprints, a baby nodosaur following its parent.

There are also tracks of four theropods - relatives of the tyrannosaurus rex that lived tens of millions of years later, but these were smaller, roughly the size of a large raven. Elsewhere on the slab, pterosaurs walked around and even left indentations in the ground where they were pecking for something to eat.

The scientists even spotted at least one clump of what appears to be a coprolite, or fossilised faeces. (They don't know who pooped, but it may have been the sauropod.)

Most intriguing were the mammal tracks.

"The mammal track shape is very distinctive," Lockley said. "Actually they look a little bit like very, very small pads."

For most of the mammals that lived during this era, scientists have rarely found complete skeleton fossils. Instead, their knowledge is, more often than not, based on a scattered bone or tooth. Footprints have been found before, but usually a single impression on a stray piece of rock.

Here, there are pairs of prints that show the left and right feet of the mammal in a sitting position. The scientists gave to these prints the name of Sederipes goddardensis, which "means sitting footprint", Lockley said.

Lockley said this was one of only two known sites where dozens of dinosaur-era mammal footprints had been found.

John Foster, executive director of the Museum of Moab in Utah, who was not involved in the research, said the findings were "very interesting to me, especially the mammal tracks and what they show about structure and probable behaviour".

The discovery was made by serendipity, and almost lost forever.

Ray Stanford, an amateur palaeontologist who has become an expert on dinosaur tracks, had just dropped off his wife, Sheila, who worked at Goddard, after the two went for lunch in 2012.

A few years earlier on the grounds of Goddard, Stanford had come across a loose rock with the footprint of a small three-toed theropod, and the brownish stone was the type of iron-rich sedimentary material that often preserves such prints.

As he was leaving the parking lot, he noticed rock of a similar colour sticking out of the grass on a hill about 30 metres away. He stopped the car and went to take a look, and spotted a prominent dinosaur footprint.

"Lo and behold," he said. "It's a perfect large nodosaur. This one was beautiful. I was in ecstasy as a tracker."

But there was a problem. Goddard was about to rip up the parking lot and the hill and put a $US31 million office building there.

Officials called Compton Tucker, a Goddard scientist who has participated in geophysical surveys to find buried ruins at archaeological sites.

Tucker recalled that, as he listened, he thought, "This sounds kind of strange but it sounds interesting."

Additional examination revealed the baby nodosaur's footsteps and the sauropod print at the other end of the slab.

Before construction of the building started, Tucker used ground-penetrating radar to search for other promising pieces of sandstone. Then an army of volunteers dug up those areas. But none of the other pieces turned out to be as interesting as the one Stanford had first spotted.

Stephen Godfrey, curator of paleontology at the Calvert Marine Museum in Solomons, Maryland, cleaned the rock and made a fibreglass cast of it. A couple of years later, the cast ended up in Stanford's basement, where he took a more careful and closer look.

He spotted the tracks of the four small theropods, all headed in the same direction, walking slowly.

"Why were the theropods walking slowly?" Stanford said. "These things were moving at less than half a mile per hour."

He then saw the steps of the mammals. "This was the real thing that hit me for a loop," Stanford said. "To see them with their potential predators."

Lockley came to Maryland about a year ago and stayed with Stanford to pore over the slab. After 80 footprints had been mapped, Stanford thought he was done. Then his wife walked up behind him, and pointed out a pterosaur footprint he had missed. She is a co-author on the paper.

None of the footprints on the slab overlap, indicating that the animals all passed in a short period of time, perhaps over a few hours or days, not likely more than a week or two. The patterns are suggestive of what each animal was doing, but it is impossible to know for certain that the dinosaurs were hunting the mammals.

"That's definitely a possibility," Lockley said.

The hill is now gone. So far, no other such slabs have been found elsewhere at Goddard.

"Studying dinosaurs is not NASA's primary, secondary or tertiary forte," Tucker said.

On Wednesday, a replica of the slab will be unveiled in the atrium of the centre's earth sciences building.

NASA usually does not search for signs of life in its backyard, but instead elsewhere in the solar system and universe.

"The fact this is found right under their nose," Stanford said.

"Maybe it's an omen they're going to start finding fossil and extant life out there."

The New York Times