Published online 23 December 2008 | Nature | doi:10.1038/news.2008.1331


'Proto-spiders' made silk, but not webs

An arachnid with no talent for weaving may have excreted the first known spider silk 386 million years ago.

spiderThe fossilised proto-spider, Permarachne novokshonovi.PNAS

The earliest known spider silk was apparently made by ancient arachnids that were not really spiders and lacked the equipment to weave webs, scientists have found. These proto-spiders may have used sheets of silk to line burrows, wrap eggs or even to have sex.

The finding recasts a species that palaeontologists had formerly seen as the oldest spider, Attercopus fimbriunguis. The species was described on the basis of scraps of fossilized cuticle found inside 386-million-year old shale rocks from upstate New York. Among the scraps were what seemed to be a single spinneret, the dextrous appendage that spiders use to shape silk into webs and other structures1.

More Attercopus fragments were found in New York during the 1990s, but researchers studied the samples for years before new features began to emerge, says palaeontologist Paul Selden of the University of Kansas in Lawrence. "It's kind of like trying to do a jigsaw puzzle with only half the pieces and without the picture on the box," he says.

Selden eventually realized that the creature did not have a spinneret. The tiny hollow hairs that excrete spun silk, called spigots, are arranged in a double row on plates lining Attercopus 's belly, and what had been identified as a spinneret was actually a plate folded over.

Without spinnerets, the creatures could not have precisely controlled the emerging silk. "It would have been much less manoeuvrable," says Selden. Possibly, he says, the silk dragged out from under them in sheets as they crawled along. Silk sheets could have been used to reinforce the walls of sandy burrows, or to make a breadcrumb trail to help the spiders find their way home after hunting.

Today, female spiders wrap eggs in silk, and aroused males deposit sperm onto a silk structure called a sperm web. Attercopus might have done the same, says Selden. He and his colleagues report their discovery in the Proceedings of the National Academy of Sciences of the USA2.

Spinning a tail

Another surprise is that this species, unlike any spider worthy of the name, has a tail. Attercopus measures about 5 millimetres long, and 1-millimetre tails, or flagella, had been found among the first set of fragments. These were assumed to belong to another species that died nearby. Now the researchers have found a flagellum attached to the animal's rear end, as seen in a group of modern arachnids called whip scorpions.

silkThis image, about 0.25 mm across, shows a strand of what appears to be silk emerging from an Attercopus spigot. It could be the oldest known silken strand in the fossil record, at about 380 million years old.Selden et al / PNAS

"What gelled," says Selden, "was that I looked at this again, and Permarachne had come along." Permarachne novokshonovi, reported in 2005, was a largely intact spider-like fossil with a striking tail.

Putting together Permarachne and Attercopus, "the idea of a tailed spider made more sense", says Selden. The researchers believe these organisms are from a previously unknown order of proto-spiders, named Uraraneida, which are part of a separate evolutionary branch from true spiders.

The creation of a new taxonomic order is rare, says arachnid expert Jeffrey Schultz at the University of Maryland in College Park. And because spiders are key predators in many ecosystems the report of these early relatives is very important, he adds. It could help scientists to learn more about spiders' ancestors: "The origin of spiders and the evolution of spider silk are probably major events in the history of life," he says.


Selden says that the findings change the definition of spiders, formerly identified as arachnids that produce silk from their abdomens. Since proto-spiders also fit this description, modern spiders should more correctly be described as those that use spinnerets and don't have a tail, he says.

The accumulated evidence supports that rethink, Selden says. "You keep wanting more material. You keep hanging on for that crucial piece. But we now think that we've got enough crucial pieces." 

  • References

    1. Shear, W. A., Palmer, J. M., Coddington, J. A. & Bonamo, P. M. Science 246, 479–481 (1989).
    2. Selden, P. A., Shear, W. A. & Sutton, M. D. Proc. Natl Acad. Sci USA 105, 20781–20785 (2008).
Commenting is now closed.