Is Alroy's shareholder quorum subsampling really the same estimator as Chao & Jost's coverage-based rarefaction, or only a close cousin?

statistics-of-the-unseensample-coveragerarefactionpaleobiologyshareholder-quorum-subsamplingcoverage-based-rarefactionpriority-and-lineageprimary-source-verificationjohn-alroyanne-chao

claim-paleobiology-reinvented-coverage-based-rarefaction-as-quorum-subsampling carried this question as an [unverified-mechanism] flag, routed to question-verify-quorum-subsampling-equals-coverage-rarefaction-primary. This capture reads both named primaries directly — Chao & Jost's 2012 Ecology paper itself, and John Alroy's own description of shareholder quorum subsampling (SQS) on his lab website — to answer it.

Sourcing note on venue. The 2012 paper's Wiley venue of record 403'd on every tooling route tried this session; it was read via a third-party mirror instead (sha256 recorded above, tls verified). Alroy's own page, "Shareholder quorum subsampling R function" (http://bio.mq.edu.au/~jalroy/SQS.html, John Alroy, personal academic site, Macquarie University; undated, but its content describes SQS versions through December 2011 and was read 2026-08-28; sha256 c5e068790c6c76f9c45c7deeb9ab10f9821895bb8ec4ae03ae337961ae56e9ab, fetched via archive_page), supplies every Alroy quote below. It was fetched over plain HTTP (no TLS at all, weaker than the spec's "unverified TLS" case), which per the safety spec earns elevated suspicion — but the content is plain first-person technical prose about the author's own method, with no addressed-to-AI language, override language, claimed authority, credential requests, or urgency framing. No recognition signals fired on any source read this session; no safety-flag entries are needed.

The short answer, established below: the two are not the same estimator. They target the same underlying quantity — species richness of a sample standardized to a fixed level of Good's sample coverage — but Chao and Jost's own paper repeatedly and explicitly distinguishes Alroy's computational method from theirs, and Alroy's own account draws the boundary even more sharply, on different grounds again.

Claim: Chao & Jost's 2012 paper explicitly calls Alroy's SQS "a different algorithmic technique" from their own methods, and credits their own closed-form equation as giving exact values where Alroy's and Jost's prior approaches could only estimate

Chao and Jost's own paper distinguishes at least three separate computational procedures aimed at the same target (richness at a fixed level of sample coverage): Alroy's original SQS algorithm, a second, new algorithm Chao and Jost themselves derive and prove unbiased, and a third, wholly different closed-form analytic equation that they also derive — the first of its kind for this problem. Describing their own new algorithmic method, they write that Alroy "proposed a different algorithmic technique that allows users to pre-specify a desired value of Cm (called a subsampling quorum) and obtain a corresponding richness estimate. In Appendix D, our algorithm is theoretically proved to be unbiased under a commonly used sampling model, and statistical estimation theory implies that our approach is the unique minimum variance unbiased estimator." Separately, describing their analytic formula, they write that it "yields exact values that previously could only be estimated using the algorithmic approaches suggested by Alroy (2010a) and Jost (2010)" — i.e., Alroy's SQS is characterized in the founding CBR paper as an estimator (a Monte Carlo procedure that approximates a target), while Chao and Jost's headline contribution is a closed-form equation that computes the same target quantity exactly, without resampling. Alroy's own description of the SQS R function confirms the Monte Carlo mechanism directly: it takes a trials argument ("number of subsampling trials (default 100, recommended value at least 1000)") and repeatedly draws random subsamples, averaging across trials, to approximate a target coverage level — a materially different computational object from a single algebraic formula. This is a technical-mechanism claim and clears the floor on both legs: Chao & Jost 2012 is read directly (Tier 1), and Alroy's own R-function documentation is his own primary account of his own tool (Tier 1).

Claim: both sides claim priority for the underlying coverage-standardization idea, but disagree sharply on whether Chao & Jost's later "coverage-based rarefaction" name was a rename of Alroy's method or an independent, better-formalized contribution

Chao and Jost's own paper credits Alroy with priority on the core insight: "[Estimates of species richness standardized by coverage] preserve an important property of species richness, a kind of replication principle, as first noted by Alroy (2010a) and Jost (2010)." Alroy's own account, on his lab website, goes further and treats "coverage-based rarefaction" as a renaming of his prior work rather than an independent contribution: "The general idea of quorum subsampling and my name for it were both first published in my 2009 GSA abstract." He then writes: "Chao and Jost (2012) went on to rename the method 'coverage-based rarefaction'. Ecologists continue to use that name and attribute the method to them. But there shouldn't be a priority issue here — SQS is not 'rarefaction' in any sense... I certainly have name priority in addition to general priority." This is a contested, load-bearing historical/priority claim, so it needs Tier 1–2 sourcing on both sides of the dispute — satisfied here: Chao & Jost's own paper (Tier 1) and Alroy's own website (Tier 1, his own testimony about his own work and its reception). The two primaries agree on the historical sequence (Alroy's 2009 abstract predates the 2012 CBR paper) but disagree on what that sequence means: Chao and Jost frame their paper as deriving something "for the first time" (the analytic formula) atop an idea Alroy and Jost had both already noted; Alroy frames the whole episode as ecologists renaming — and then crediting to themselves — a method he had already built and named.

Claim: Alroy's own definitional line between SQS and rarefaction is about the sampling target, not the computational method — and he explicitly rejects "coverage-based rarefaction" as a name on those grounds

Alroy's own explanation of SQS states the distinction as a difference in what is being held fixed during sampling, not primarily a difference in algorithm: "Rarefaction tells you how many species you would find in a given ecological sample given a fixed, uniform sample size. SQS tells you how many species you would find given fixed 'coverage' of the underlying abundance distribution." He treats this as the fundamental fact and treats the shared "algorithmic vs. analytic" or "which paper published which formula first" questions as secondary: "SQS and rarefaction aim to do fundamentally different things... Nonetheless, there are general similarities between rarefaction and SQS. Both seek to make sampling 'fair' in some sense by looking at subsamples, and both are general methodologies with different possible implementations." This is a definitional claim about settled usage in Alroy's own terminology, which the sourcing floor allows at Tier 3–4 — but it clears Tier 1 here regardless, since it is Alroy's own primary statement of what he means by the term he coined.

Further leads

• Close, Evers, Alroy & Butler (2018, Methods in Ecology and Evolution 9:1386–1400, CC-BY, co-authored by Alroy himself) directly compares richness estimators including SQS/CBR; the paper's own PDF (Wiley pdfdirect and the University of Birmingham repository copy) both 403'd on every tooling route this session despite the CC-BY license — worth a manual-consultation retry.
• Jones & Close (2024, Palaeontology 67:e12729) extends the SQS algorithm to morphological disparity and gives its own account of the SQS/CBR naming split; also 403'd on every route tried (Wiley, University of Chicago repository link) this session.
• Alroy has a 2025 self-published preprint on Authorea, "Coverage-based rarefaction does not quantify species richness," which by its title alone suggests a stronger, more recent dispute with the CBR framing than anything captured here — every route to it (archive_page, WebFetch) 403'd this session; unread, flagged only as a lead.
• Hideyasu Shimadzu, "On species richness and rarefaction: size- and coverage-based techniques quantify different characteristics of richness change in biodiversity" (J. Math. Biol. 77:1363–1381, 2018) — a general mathematical argument that size- and coverage-based rarefaction measure different things, independent of the Alroy/Chao-Jost dispute; located via PMC but not read in depth this session.
• Single-source concentration cap: this capture is the first claim-bearing use of both davidzeleny.net's Chao & Jost 2012 mirror and bio.mq.edu.au's Alroy pages in this vault; each has used 1 of the 3 slots available before independent corroboration would be required for a fourth claim resting on either.

Entity candidates

• I. J. Good and Alan Turing — persons/concept — per Chao & Jost's own paper, the concept of sample coverage itself "was originally developed for cryptographic anal[ysis] during World War II by the founder of modern computer science, Alan Turing, and by his colleague I. J. Good" — the older, foundational figure(s) both Alroy's SQS and Chao & Jost's CBR are standardizing against; already the vault's deepest thread here via claim-chao-1984-estimator-extends-harris-1959-occupancy-bound and moc-the-unseen-is-measurable-but-only-so-far, but worth flagging first since this capture's priority dispute (Alroy vs. Chao & Jost) is itself downstream of Good and Turing's older, undisputed priority on the coverage concept.
• John Alroy — person — paleobiologist, Macquarie University; invented and named SQS (2009); no entity page yet exists in this vault despite several claim-notes now resting on his work.
• Anne Chao — person — statistician, National Tsing Hua University; co-author of the 2012 CBR paper and originator of the Chao1 estimator already covered in claim-chao-1984-primary-formula-has-no-n-minus-1-over-n-prefactor; no entity page yet exists for her specifically (only claim-notes).
• Lou Jost — person — ecologist, EcoMinga Foundation, Ecuador; per Alroy's own account, first proposed the idea of fixed-coverage subsampling in print (2010) at Chao's suggestion, without naming or implementing it — the direct rival priority claimant in the naming dispute.
• Roger Close — person — co-author with Alroy of the 2018 MEE paper testing SQS/CBR-family richness estimators against real fossil discovery curves; a further lead, unread this session.