Soil DNA survey

Sources

What this study stands on

external works: the software that turned DNA into names, the reference databases that supplied the names, the primers, every statistical test, the soil biology behind what a DNA census can and cannot see, the nitrogen biology the headline depends on, the public map products describing this location, and the public sequence data used for comparison.

Each one says what it is used for here, because a bare list of citations does not let you check anything. The point of the third line in every entry is that you can follow it back to the claim it supports.

How this page stays true

A bibliography is the easiest page on a site to let rot. Nothing breaks when a citation goes stale — a tool gets upgraded, a database release changes, an analysis is dropped — and the list keeps confidently naming whatever was true the year it was written. Nobody notices, because a bibliography is never recomputed.

So this one is. Entries declare the exact names the site prints for the thing they cite, and the build compares those declarations against what the pages actually say in both directions: a tool the site names with no source here is a failure, and a source here for something the site no longer names is also a failure. The first protects the reader. The second keeps this file from silently accumulating citations to software the project stopped using.

What that check cannot cover. It compares names against names. The entries under the nitrogen biology and soil life support claims in prose — that archaeal ammonia oxidisers have unusually high ammonia affinity, that one archaeal lineage is an obligate acidophile, that some Nitrospira oxidise ammonia completely, that much of the DNA in soil belongs to organisms already dead. No number in the payload can confirm any of that, so nothing automatic verifies those. They are the reader's to check, which is why the link is there.

And what is not on this page at all. Every entry here is something this study used. A citation is not a defence: none of these works was consulted to see whether it disagreed with the result, and a reference list assembled from what an argument leans on will always look supportive. The limits section on the research page is where the case against is made, and it is made from this study's own data rather than from anybody else's.

The links were checked by request rather than by assumption: every URL below was fetched and had to resolve. Two entries carry no link at all — a 1986 book whose registered identifier now points at a dead publisher page, and this project's own analysis package. A link that hands you a 404 is worse than a plain name, so those are printed without one.

The list

Loading the sources…