Eel research · provenance

Every data source we reviewed

An exhaustive, honest catalogue of the data behind this program — what we ingested and analyzed, what is in use as an ocean or climate input, what is request-gated and still pending, and what we reviewed and deliberately set aside. Licences and status are shown throughout. Where a source was excluded, the reason is stated. Nothing here is cherry-picked: the nulls, the embargoes, and the dead-ends are listed alongside the results.

Status key — INGESTED: pulled and analyzed here · IN USE: a standing input to the models · REQUESTED: access-gated, request sent or drafted · EXCLUDED: reviewed and set aside (reason given) · ANALOG: comparative / cross-species.

Adult telemetry & survey (eel life-stage)

SourceProvider / licenceHow used / why notStatus
EELIAD silver-eel satellite/archival tagsCefas Data Portal holding 20655 (Righton et al. 2016)
Open Government Licence v3
Oceanic ground truth. Real release→terminal headings cluster on the Sargasso bearing (resultant R=0.76, p≈0, robust to centroid and reconstruction subset); over-ground speed 0.28 m/s independently matches the acoustic anchor; swim-vs-current shows eels compensate for currents; predation is 50% of tag fates; energetics feasible for most eels.INGESTED
ETN / Verhelst acoustic telemetryVLIZ / Ghent University (EurOBIS)
CC0
136 female silver eels, 81,302 detections. The departure-speed anchor (~0.28 m/s) that the EELIAD tags independently reproduce.IN USE
England WFD transitional-water eel surveyCefas Data Portal holding 20473
Open Government Licence v3
10,681 surveys, 505 sites, 1981–2018. Occupancy decline is real but gear-dependent and concentrated in the South East; a worked example of how raw single series overstate cleanliness.INGESTED
Burrishoole / Newport 50-year eel trapMarine Institute Ireland ERDDAP
CC-BY
50 years of daily silver-eel counts. No robust shift in run timing (a quick OLS "earlier migration" did not survive robust estimators) — an honest null.INGESTED
EDA / SUDOANG eel density databaseZenodo 6397009 + Briand et al. 2022 report
Open
48,699 electrofishing operations + 20,101 silver eels (FR/SP/PT). Dams cut upstream presence hard (OR 0.27/sd; 85%→<5%) but reduce silver-eel escapement only 4–12% — so continental dam habitat loss is unlikely to be a primary driver of the collapse (refutes Kettle et al. 2011).INGESTED
Hanningfield reservoir acoustic telemetryCefas Data Portal holding 21846
Open Government Licence v3
Reservoir behavioural telemetry; low priority for the oceanic question. Resolves an old licensing uncertainty.INGESTED

Larvae & spawning ground

SourceProvider / licenceHow used / why notStatus
ICES Eggs & Larvae Database (leptocephali)ICES (McCleave/Miller compilation), 1862–2007
CC-BY
The open larval record. Back-tracks spawning origin (~26°N, 66–67°W) and validates the thermal anchor; the result reproduces on this open data without the restricted file.IN USE
Als et al. 2011 Sargasso larvaeDryad 10.5061/dryad.8155
Open
271 A. anguilla + 117 A. rostrata larvae with coordinates; open spawning-area source points for inversion and panmixia.IN USE
Miller et al. 2015 leptocephalus compilationM. J. Miller (pers. comm.)
Restricted — not redistributed
Corroboration only; no core result depends on it (the open ICES record carries the finding).IN USE

Ocean & climate inputs

SourceProvider / licenceHow used / why notStatus
SST — OISST v2.1, ERSST v5, COBE-SST2, OSTIANOAA NCEI / JMA / CMEMS
Open
The 1854–2025 SST record for the isotherm-vs-front divergence and its attribution; multi-product robustness.IN USE
GLORYS12 reanalysis + HYCOM currentsCopernicus Marine / HYCOM
Copernicus / open
Eddy-resolving currents for the larval-transport IBM and the swim-vs-current decomposition; front-insensitivity holds across both products.IN USE
GlobColour chlorophyll + net primary productionCopernicus Marine OCEANCOLOUR_GLO_BGC_L4
Copernicus
Route productivity along the drift corridor. Both biomass (CHL −5.3%/dec) and production (PP −3.3%/dec) decline in the drift season and survive AMO control — a warming-forced candidate mechanism, not excluded.IN USE
Subsurface T — IAP/Cheng, EN4, Argoocean.iap.ac.cn / Met Office / argopy
Open
Confirms warming reaches the ~150–200 m spawning depth across three products.IN USE
Winds — ERA5, CCMP, NCEP/NCARECMWF (ARCO-ERA5) / RSS / NOAA PSL
Open
Ekman/convergence diagnostics; the wind leg was tested and dropped (product-robust null).IN USE
NOAA Global Drifter ProgramNOAA AOML ERDDAP
Open
Real Lagrangian ground truth: observed passive delivery (~3%) and the measured dispersion that bounds the inversion horizon.IN USE
Sargassum AFAINOAA AOML / USF
Open
Floating-algae habitat index; nursery-band habitat cross-check.IN USE
CMIP6 ensemblePangeo (Google Cloud)
Open
Forward isotherm trajectory and the prospective habitat-edge pressure (+1.7/+3.3 °C by 2100).IN USE
Climate indices — AMO, NAO, GISTEMPNOAA PSL / CPC / NASA GISS
Open
Internal-variability controls and the forced/warming attribution (frozen snapshots for reproducibility).IN USE

Recruitment & stock

SourceProvider / licenceHow used / why notStatus
ICES WGEEL glass-eel recruitment indexICES Working Group on Eels
Report annexes
The recruitment series whose link to spawning climate does not survive de-trending.IN USE
SUDOANG silver-eel escapement (EDA 2.3)Briand et al. 2022 (GT4 deliverable)
Open report
The dam counterfactual: escapement with vs without dams (4–12% reduction); total SW-Europe stock 374.3 M eels.IN USE

Requested / access-gated

SourceProvider / licenceHow used / why notStatus
Reconstructed EELIAD daily tracksCefas / D. Righton
Embargoed (EIR 12(4)(d))
The dense reconstructed routes are withheld pending an unfinished synthesis paper (expected Dec 2026). We use the open release/terminal/geolocation subset instead.REQUESTED
Righton et al. 2022 "direct evidence" Sargasso tracksCorresponding authors (Cefas)
On reasonable request
Newer PSAT tracks where some eels reached the spawning area; request drafted to the Cefas data channel.REQUESTED
Béguer-Pon et al. 2015 American-eel PSATAuthors (DFO / U. Laval)
On request
Would enable a same-method orientation comparison across species (European vs American). Request sent.REQUESTED
Australian long-finned eel PSAT (2025)Authors
On request
A third-species comparative (Anguilla reinhardtii → Coral Sea). Request drafted.REQUESTED
Continuous Plankton RecorderMarine Biological Association
Data-use agreement
Plankton 1958–present to extend the corridor-productivity series back before the satellite era (does the decline predate the 1980s collapse?). Request sent.REQUESTED
Japanese-eel larval / catch dataAoyama / Tsukamoto
On request
Comparative spawning-area larvae; noted as not yet compiled by the authors.REQUESTED

Reviewed & deliberately excluded

SourceProvider / licenceHow used / why notStatus
European eel reference genome + population genomicsENA / GenBank (GCA_013347855.1; Jacobsen/Enbody comparative)
Open
Open but large and specialized (VCFs + selection scans); a separate pop-gen project. The panmixia / two-species result is already established, so out of scope for this program.EXCLUDED
OTN American-eel acoustic telemetryOcean Tracking Network
Login-gated
Estuarine/riverine, not the oceanic release→terminal geometry needed for an orientation comparison.EXCLUDED
DBEEL rivers spatial database (PostGIS dumps)Zenodo 6384022
Open
The GIS substrate (river-network geometry, basin polygons) behind the EDA tables; ~4 GB of database dumps with no new eel observations. Escapement numbers recovered from the published report instead.EXCLUDED
Otolith microchemistry (Sr:Ca life history)Various (HAL, theses, reports)
Mixed / paywalled
Reviewed; mostly extract-from-PDF rather than clean deposits. Held for a future life-history slice.EXCLUDED
Movebank oceanic-turtle / eel tracksMovebank
Login + data-use agreement
No open Anguilla oceanic tracks; used open drifters and Dryad turtle sets instead.EXCLUDED

Comparative / analog

SourceProvider / licenceHow used / why notStatus
Anguilla rostrata occurrenceOBIS / GBIF
Open
American-eel distribution for the comparative "same gyre, two continents" framing; modeled origins converge near the Sargasso (~1.9° from the European-eel origin).ANALOG
Cross-species reproductive-climate databaseSix public databases, GBIF-resolved
Open (compiled)
A 14k-species climate-reproduction asset built to test a generalization of the eel pattern — which then failed independent testing (reported openly, nulls included).ANALOG

This catalogue is maintained alongside the research notebook. Data redistributed here retain their original licences; restricted and embargoed sources are named but not redistributed. Last reviewed July 2026.

← Back to the research notebook