Science2 distinct publishers3 min readPublished
Volunteers buried more than 2,000 pairs at about 1,000 Swiss sites and dug them up two months later, and land use predicted how much fabric vanished better than most of the soil chemistry measured. That makes cotton a usable, if coarse, field assay.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
Cotton is mostly cellulose, and that is the whole trick: bury it, and anything in the soil equipped to break down a polysaccharide begins eating, so the fabric that goes missing stands in for how busy the organisms are [5]. What comes back from the lab is mass lost over a fixed window, photographed on retrieval and posted in with a soil sample [3]. That figure is a sum, not a measurement of one process. It cannot say whether earthworms or fungi or bacteria did the work, and it does not separate biology from weather. The team is explicit about the second part: soil activity tracks temperature and moisture, and soil that is too cold or too dry is barely active at all [9].
Repetition is what rescues a crude assay. About two pairs per location [19] tells you very little about any single garden. A thousand locations, all following the same protocol, is enough to sort land uses by decomposition rate [6].
The design is observational. Nobody assigned gardens and lawns to sites at random; the people who keep gardens also compost, irrigate and plant differently. So the reported result is that a site's land use category explains more of the variation in fabric loss than most measured soil properties do, with nutrient chemistry also contributing [7][8]. Co-study leader Marcel van der Heijden, a UZH professor of agroecology, reads that as management affecting both soil life and soil quality [11], and the team's suggested levers are permanent soil cover, compost, diversified rotations and restraint with mineral fertiliser and pesticide [12]. Those levers were not tested one at a time in this study. They are the plausible mechanism behind a ranking, which is a different kind of evidence.
Then there is the direction problem, which the researchers raise themselves. Franz Bender of Agroscope, who co-led the project, says maximum decomposition is not desirable in every ecosystem: in near-natural habitats such as forests, very high nutrient levels can indicate a disturbed nutrient balance, and in gardens fast decomposition may signal oversupply [10]. Underwear that has nearly vanished is not automatic evidence of a balanced system [18]. A high score has to be read against what the site is for.
One gap is worth naming. Volunteers also buried 12,000 tea bags, roughly a dozen per location [20], and neither the announcement nor the coverage of the paper in Plants, People, Planet [4] describes what the tea bags showed [22]. A second substrate with different chemistry is the obvious internal check on the cotton signal, and it is the part of this dataset I would wait for.
If you want to run the proposed underwear index on ground you manage [15], the useful form is a contrast rather than a score: same burial date, same depth, two plots differing in one thing you control. The Swiss project's contribution is standardisation at a scale that would have been difficult and expensive to buy [17], producing what the university calls one of the country's most comprehensive soil-life datasets [16]. Participants got individual soil results, evaluation tools and management advice back [14], which is also how you keep a thousand people burying things on schedule.
Ranked by verification strength, evidence, and original report placement.
In 2021, researchers from the University of Zurich (UZH) and Agroscope, the Swiss federal centre of excellence for agricultural research, launched the citizen science project Proof by Underpants (Beweisstueck Unterhose) to draw attention to soil as an overlooked ecosystem and to collect scientific data.
Following a standardized procedure, 1,000 volunteers buried more than 2,000 pairs of cotton underwear and 12,000 tea bags at around 1,000 locations across Switzerland.
Two months later the volunteers dug up the buried items, photographed them and sent them to the laboratory together with a soil sample for further analysis.
The results were published in the journal Plants, People, Planet.
Because cotton is primarily made from cellulose, soil-dwelling organisms can gradually consume the material; the more fabric that disappeared, the more biologically active the soil was likely to be.
The underwear decomposed most rapidly in private gardens, where researchers found the highest levels of organic matter; soil organisms were least active in lawns, while agricultural fields and meadows fell somewhere in between.
Distinct publishers with included, body-backed reporting in this cluster.
discovermagazine.com
1 article · August 27, 2026
phys.org
1 article · August 26, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
science
Whale-watch footage catches 1,417 humpbacks feeding along a corridor theory calls a fast1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Peer-reviewed result, but both write-ups are one release and carry no numbers
The underlying study is peer-reviewed with a DOI, the protocol is standardized and the sample is large (about 1,000 sites, national coverage), and the two publishers agree on every substantive finding. But the agreement is provenance, not corroboration: both items derive from the same UZH/Agroscope announcement and reuse the same two quotes. Neither reports a single decomposition figure, effect size or statistical measure, neither addresses bias from self-selected volunteer sites, and the 12,000 tea bags buried as a parallel measure go entirely unreported, so the strength of the central 'land use beats soil chemistry' finding cannot be checked from the supplied material.
One large national deployment; no uptake beyond the project
Real-world execution is substantial and verified: 1,000 volunteers, more than 2,000 buried pairs, roughly 1,000 sites in all Swiss regions, plus a disclosed participation model in which about 240 volunteers became co-authors and all participants received soil results and tools. That is genuine field adoption of the protocol by non-specialists. Against that, adoption stops at the study boundary — the supplied sources show no agency, farm body, monitoring programme or other country picking up the method, and the proposed 'underwear index' exists only as a researcher proposal.
Mildly overstated framing, with the caveat carried by the researchers themselves
Headline framing runs slightly ahead of what is shown: 'land use is key to soil health' and a proposed 'underwear index' imply a validated ranking instrument, while the coverage supplies no decomposition numbers, no statistics and no result from the parallel tea-bag measure. The overstatement is modest rather than severe because the researchers supply their own brake — Bender's point that maximum decomposition is not desirable everywhere — and Discover reinforces it by stating a buried pair cannot provide a complete diagnosis. So the gap sits just above alignment, driven by missing quantification rather than by inflated claims.
Institutional awareness campaign, relayed with light independent scrutiny
The project's stated purpose is partly promotional: draw attention to an overlooked ecosystem and raise public awareness, with a memorable 'underpants' hook and a proposed public-facing index. The two performing institutions are a university and a federal agricultural research centre with an interest in demonstrating both soil-policy relevance and successful public engagement. Distribution amplifies that: phys.org reproduces the announcement nearly whole, and Discover rewrites it around the same quotes while adding no outside expert. No funding source is disclosed in either item, and there is no commercial product to sell, which caps the score below the high range.
Facts stable and internally consistent; quantitative basis unverifiable here
Confidence is solid on the descriptive layer — protocol, counts, land-use ordering, co-author count, journal and DOI are consistent across both items and mutually reinforcing, and the two publishers are recent and closely spaced. It is held back by single-root provenance, the total absence of reported quantitative results, and the unexplained silence on the tea bags, which prevents independent verification of the headline finding from the supplied sources.