Time Capsules: Seeds sprout 120 years after going underground

An experiment designed by a botany professor to last longer than his own life has demonstrated that seeds of two common flowers still sprout and blossom despite more than a century in a bottle.

This work ranks as the longest-running test of seed dormancy in soil, says the current generation of researchers, Frank Telewski and Jan Zeevaart of Michigan State University in East Lansing.

William James Beal started the experiment in 1879 by burying several feet deep in a sandy knoll on campus 20 bottles containing sand and seeds from 21 common plants. Scientists have been retrieving bottles at various intervals. Telewski and Zeevaart exhumed the 15th bottle in 2000 and tried to coax the aged seeds to life. Two species sprouted–moth mullein and a mallow called cheeses (Malva rotundifolia)–the researchers report in the August American Journal of Botany.

“This kind of experiment is very useful,” comments Jane Shen-Miller of the University of California, Los Angeles. In 1995, she sprouted a lotus seed that she estimated by radiocarbon dating was almost 1,300 years old. Because tales of seeds sprouting after millennia in the pyramids have now been discredited, Shen-Miller’s lotus holds the record for oldest viable seed. Very little is known about the length of dormancy in most species under natural conditions, she says, so she welcomes a near-natural experiment with precise records.

Beal packed every bottle with 50 seeds of each of the plants, including primrose, plantain, wild clover, and cedar. He buried all the bottles uncorked and upside down in a row. His original plan called for digging up one every 5 years, and he lived to test five bottles’ contents. Later experimenters extended the time intervals.

By 1920, seeds of eight species of seeds could still sprout. But in 1940, only the primrose, curly dock, and moth mullein came up. The second most recent bottle, examined in 1980, produced one cheeses plant and 21 moth mulleins.

Telewski’s crew retrieved the next bottle on an April night in 2000, not so much for secrecy as for keeping a blast of daylight from triggering sprouts in the five bottles left for the future. The same species sprouted and bloomed that had come up in 1980, and the researchers also demonstrated that the 2000 crop set viable seeds.

Telewski says he hopes to collect the next bottle’s data in 2020 himself, but finishing the experiment will probably take at least one more generation of seed scientists.