Skip to content

Publication attribution for the used Nordic Neotoma datasets

The source-bound ledger joins the Neotoma dataset entity decisions to 85 exact provider publication-link responses retained privately under neotoma/publication-links/2026-09-29. These are the datasets used in the selected four-country direct observed-pollen development study. The ledger binds the entity report, relational dataset part, capture manifest and every consumed response by SHA-256. It keeps source dataset DOI strings, investigators, all linked publications and provider primary flags. It also preserves the source dataset notes, including an explicit unpublished statement for Kuivajärvi. Provider publication flags remain separate from the independent source review below.

Neotoma's 2025 Data Use and Embargo Policy states CC BY 4.0 use with attribution to Neotoma, the constituent database, the original investigators and primary publications, and asks for a site/dataset table with dataset DOIs. All 85 used rows identify the European Pollen Database, have at least one dataset DOI and name an investigator. The provider links give 79 datasets exactly one flagged primary publication with citation, author and year metadata. This is provider metadata readiness, not a qualified citation decision, source-specific rights review, or LRA admission.

Country Used datasets One flagged primary Primary-link review Distinct source DOIs Either review
Sweden 3 3 0 2 2
Denmark 2 2 0 0 0
Norway 52 49 3 9 11
Finland 28 25 3 2 5
Total 85 79 6 13 18

The six primary-link cases have exact source questions:

Dataset Country Provider state Citation decision needed
4168 FI No publication link Match the original report or article to the deposited profile before assigning a primary publication.
4472 FI No publication link Cite the independently matched 1995 original study; retain the missing provider link.
4510 FI Two links, neither primary Cite the original Vasikkasuo I report lead; do not treat the linked Vasikkasuo II follow-up as this core's primary work.
45347 NO Two links flagged primary Confirm whether both works apply and cite each applicable primary work.
52255 NO Two links flagged primary Confirm both primary works; the two DOI strings differ only in ASCII case.
53098 NO Three links flagged primary Confirm applicable primary works and resolve its two distinct dataset DOI names.

Sixteen datasets carry more than one literal DOI string. The DOI Handbook §3.3.4 treats ASCII case variants as equivalent identifiers: three of those 16 have only case variants. The other 13 have distinct DOI names, including two Swedish, nine Norwegian and two Finnish datasets. Their exact strings remain in the ledger; no DOI was silently selected or discarded. The official registry comparison below resolves three identity relations and leaves ten source-owner citation decisions. The union of primary-link and distinct-DOI review cases is 18 datasets.

Finnish original-work review

The two Finnish no-link datasets have been checked against the Neotoma source rows and a bounded set of external records. Exact copies of the 1994 EPD review, the 1995 original article and four dataset metadata responses are retained in the private source store at neotoma/publication-evidence/finland. Its six-file manifest has SHA-256 dd52deaced69bacd6bd41394cf9e01048c6aad8c4897376808fcc8901ca90c78. The source dataset and collection-unit identities below are from the governed Neotoma relational surface, not an inference from a search result.

Dataset Source and original-work finding Citation status
4168, Kuivajärvi (KUIVAJAR) Neotoma's dataset note explicitly says the data are unpublished. Hicks and Grönlund's 1994 EPD inventory, Table 2, also lists this profile as unpublished as of March 1994. The Finnish National Library catalogues a 1992 Geological Survey report on Kuivajärvi, and a Forssa municipal source identifies a Vuorela and Kankainen article in a 1993 local yearbook. Their full contents have not been matched to Neotoma's 112 samples at 2.5–400 cm. Original-work leads; no verified primary-publication assignment. Cite the Neotoma dataset DOI and Vuorela for the used dataset. The 2010 PANGAEA profile is a separate archival dataset citation, not proof that the original analysis was published.
4472, Syrjälänsuo (SYRJALAN) The same 1994 EPD inventory lists the profile as unpublished at that time. Vuorela's 1995 original article, Pollen evidence of Stone Age and Early Metal Age settlement in Taipalsaari, Southern Karelia, Eastern Finland, Fennoscandia archaeologica XII:207–214, presents the Syrjälä mire pollen diagrams. It describes June 1992 sampling in an esker kettle-hole bog and a 0–285 cm profile. Neotoma's bog core is dated 1992-06-15 and has 81 samples spanning 0–285 cm. The 2010 PANGAEA profile agrees on handle, site coordinates and collection date. The 1995 article is supported as the original study citation alongside the Neotoma dataset DOI and Vuorela. The article calls the location Syrjälä mire while Neotoma calls it Syrjälänsuo; approximate coordinates and elevations differ slightly, and printed diagrams were not matched row by row. Neotoma still has no linked primary publication for this dataset.

The 1994 unpublished designations are historical observations, not claims that the datasets lack later archival publication. The independent 1995 article match resolves one original-study citation for reporting while leaving the provider's six missing or ambiguous primary-link cases unchanged. The other five cases still require review.

Vasikkasuo core identity

The provider links two unflagged works to Finnish dataset 4510. The linked Geological Survey report P34.4.100 has a scanned title page dated 1991, although its Neotoma publication string says 1990. Its introduction explicitly identifies its newly studied core as Vasikkasuo II and the earlier, cultivated-field core as Vasikkasuo I, sampled about 120 m away. It attributes the first core's original investigation to Vuorela's 1990 report, Siitepölyanalyyttinen tutkimus Puolangan kunnan kasvillisuuden kehityksestä, also catalogued by the Finnish National Library. The linked 1991 Suo article reviews regional peat evidence and does not by itself establish the original record publication for the first core.

The governed Neotoma collection-unit crosswalk resolves the potential mix-up: 4510 is VASIKKAS, core 1, collected 1990-05-01, with 59 samples from 0–290 cm and included in the selected direct-pollen study. The nearby PUOLANKA, core 2, belongs to dataset 4363, excluded from that study. The 1991 report is direct evidence for core II and a first-party pointer to the 1990 original report for core I. The 1990 report's full contents have not been compared with the 59 deposited samples, so the original-work match for 4510 is still a documented lead rather than a verified numerical identity. The scanned report is retained privately at neotoma/publication-evidence/vasikkasuo; its one-file manifest has SHA-256 e9168e64d725867ba1e5e0be4365bb27ba8c667e5f8af233835d217b77d772ce. Neotoma's two false primary flags remain unchanged.

Norwegian multiple-primary review

Four exact external metadata responses for the three Norwegian cases are retained in the private source store at neotoma/publication-evidence/norway; the manifest SHA-256 is e9d72ba7b0f7d22f8a8e1f77d1e608c3f69477adbb11018ff5227d6d7f651803. This review keeps provider flags and DOI strings intact.

Dataset Source match and citation decision Remaining distinction
45347, Råtåsjøen (OTTRAT) The deposited lake core was collected in 1999 and has 68 Neotoma samples at 801–998 cm. Velle and colleagues' 2005 lake study is the original site-specific publication identified by the provider and publisher DOI metadata. Cite it with the dataset DOI. The other flagged work is a 2012 excursion guide. Its direct use of these 68 pollen samples was not verified; retain it as a linked contextual work pending a source-specific check.
52255, Lake Skogstjern (SKOGSTJE) The Neotoma core is 418 cm and has 236 samples at 2–418 cm. PANGAEA's independently published pollen record identifies the same 418 cm lake core, Usinger corer and location, and explicitly relates both the 2017 landscape study and 2018 Mesolithic study. Cite both original applications with the Neotoma dataset DOI. PANGAEA describes 245 pollen spectra, while the governed Neotoma surface contains 236 samples; no complete row-level equivalence is claimed. Neotoma's two literal DOI strings differ only in ASCII case and refer to one DOI name under the DOI Handbook rule above.
53098, Lake Store Finnsjøen (FINNSTOT) The 2015 original lake study describes the pollen sequence, and the 2019 cold-event study analyzes pollen at this lake and Flåfattjønna. They are source-supported site publications. The 2017 forest-line study compares megafossils and pollen and cites the 2015 study; its specific contribution of new pollen data to the deposited FINNSTOT sequence remains unverified. Neotoma exposes 70 samples at 1483–2197 cm. The 2015 paper reports 45 sampled pollen levels, so exact sample-to-paper lineage still needs reconciliation. DataCite has two separate registrations, 10.21233/5XR8-GV26 created in 2022 and 10.21233/sxnn-hq25 created in 2025, both pointing to dataset 53098 but without a declared equivalence relation. Preserve both and request a source-owner citation choice.

The Norwegian source review supports named citations for the actual lake studies but does not turn the three provider multiple-primary states into three unique primary flags. The 2012 guide, 2017 Finnsjøen comparison, sample-count differences and distinct Finnsjøen DOI registrations remain explicit.

Distinct DOI registry review

The source-bound DataCite comparison consumes exact registry responses for both DOI names of all 13 used datasets in the distinct-DOI class. The 26 responses and their byte hashes are retained privately under neotoma/doi-metadata/used-nordic; the capture manifest SHA-256 is 207b791bed9d3d004fc55f407fb7a840efd81bd9180fed17f24b187ac0e69edb. The comparison checks the literal DOI pair, registration metadata, landing dataset identity and any relation to the counterpart DOI. DataCite's guidance defines IsIdenticalTo for two DOIs of the same research output.

Registry finding Used datasets Citation treatment
One DOI record explicitly declares IsIdenticalTo its counterpart FI 20; NO 317, 977 The three pairs have registry-supported identity. Preserve both literal names; a preferred citation format still needs a source-owner choice.
Both DOI records land on the same numbered Neotoma dataset, with no DOI-to-DOI identity or version relation SE 53995, 54001; NO 53002, 53081, 53098, 53119, 53121, 53127, 53131; FI 58530 Ten pairs remain distinct registrations. A common target URL, similar title or shared version number does not establish registered equivalence. Preserve both until the owner identifies the intended citation.

For example, the two Store Finnsjøen records were created in 2022 and 2025, both lead to dataset 53098, and neither declares an IsIdenticalTo relation to the other. This review does not change the separate three ASCII case-only pairs, the provider publication flags, or the direct-pollen response set.

None of these citation records supplies LRA basin geometry, PPE, or independent vegetation truth.

The existing direct-pollen predictions and scores remain development results; this attribution review does not change their response denominator.

Replay with the retained private capture:

PYTHONDONTWRITEBYTECODE=1 .venv/bin/python -m \
  bijux_pollenomics_dev.source_completion.neotoma_publication_readiness \
  --repository . \
  --capture-root /path/to/private/source/store/neotoma/publication-links/2026-09-29 \
  --output artifacts/jobs/neotoma-publication-readiness/replay/report.json
cmp docs/report/regions/nordic/source-capability/neotoma-attribution/report.json \
  artifacts/jobs/neotoma-publication-readiness/replay/report.json