The numbers in this figure
44 rows from figureA_source_data.csv, read out of source_data.zip in the deposit. Not a copy: the page reads the archive a reader downloads, so the two cannot disagree. The filename carries the pipeline's figure number, which is not this figure's number; the archive's index states the binding.
The archive's digest: 3558736147e731a25c113309a5af49e534421701d97c1324ddf703aed3d45266
All 44 rows, 5 columns
The table scrolls inside its frame; the header stays in place.
Scientific namescientific_name | Common namecommon_name | Appendixappendix | Split listedsplit_listed | Appendix as recordedappendix_as_recorded |
|---|---|---|---|---|
| Acipenser baerii | Siberian sturgeon | II | no | II |
| Alopias vulpinus | Common Thresher Shark | II | no | II |
| Asio flammeus | Short-eared Owl | II | no | II |
| Balaena mysticetus | Bowhead Whale | I | no | I |
| Balaenoptera acutorostrata | Common Minke Whale | I | no | I |
| Balaenoptera borealis | Sei Whale | I | no | I |
| Balaenoptera musculus | Blue Whale | I | no | I |
| Balaenoptera physalus | Fin Whale | I | no | I |
| Berardius bairdii | Baird's Beaked Whale | I | no | I |
| Branta hutchinsii | Aleutian Cackling Goose | II | no | II |
| Branta ruficollis | Red-breasted Goose | II | no | II |
| Bubo scandiacus | Snowy Owl | II | no | II |
| Buteo lagopus | Rough-legged Buzzard | II | no | II |
| Cetorhinus maximus | Basking Shark | II | no | II |
| Delphinapterus leucas | Beluga Whale | II | no | II |
| Enhydra lutris | Sea Otter | I | yes | I/II |
| Enhydra lutris | Sea Otter | II | yes | I/II |
| Eschrichtius robustus | Gray Whale | I | no | I |
| Eubalaena glacialis | North Atlantic Right Whale | I | no | I |
| Eubalaena japonica | North Pacific Right Whale | I | no | I |
| Falco peregrinus | Peregrine Falcon | I | no | I |
| Falco rusticolus | Gyrfalcon | I | no | I |
| Globicephala melas | Long-finned Pilot Whale | II | no | II |
| Grus canadensis | Sandhill Crane | II | no | II |
| Hyperoodon ampullatus | Northern Bottlenose Whale | I | no | I |
| Lagenorhynchus acutus | Atlantic White-sided Dolphin | II | no | II |
| Lagenorhynchus albirostris | White-beaked Dolphin | II | no | II |
| Lamna nasus | Porbeagle Shark | II | no | II |
| Leucogeranus leucogeranus | Siberian Crane | I | no | I |
| Lontra canadensis | North American River Otter | II | no | II |
| Lynx canadensis | Canada Lynx | II | no | II |
| Megaptera novaeangliae | Humpback Whale | I | no | I |
| Mesoplodon stejnegeri | Stejneger's Beaked Whale | II | no | II |
| Monodon monoceros | Narwhal | II | no | II |
| Numenius borealis | Eskimo Curlew | I | no | I |
| Odobenus rosmarus | Atlantic Walrus | III | no | III |
| Orcinus orca | Killer Whale | II | no | II |
| Phocoena phocoena | Harbour Porpoise | II | no | II |
| Phocoenoides dalli | Dall's Porpoise | II | no | II |
| Phoebastria albatrus | Short-tailed Albatross | II | no | II |
| Physeter macrocephalus | Sperm Whale | I | no | I |
| Rhodiola rosea | Arctic Rose | II | no | II |
| Ursus maritimus | Polar Bear | II | no | II |
| Ziphius cavirostris | Cuvier's Beaked Whale | II | no | II |
The manuscript's tally, re-derived from the deposited dataset: 16 species on Appendix I, 25 on Appendix II, 1 on Appendix III, and 1 split-listed, summing to 43.
Counting each species once by its primary listing, with the one split-listed taxon (the common minke whale) broken out as its own category, gives 16, 25, 1 and 1. Counting that taxon in both Appendix I and Appendix II instead gives 17 and 26; the split-out convention is the one used here.
A third reading gives 15 and 2, and it is not wrong either. split_listings holds two current rows for the common minke whale, one per population, while cites_listings marks only the Appendix I row current for the species. Counting from the first table makes the minke whale split AND counts the sea otter's compound I/II value as split, which gives 15 Appendix I and 2 split-listed. An inspector reached exactly that from this deposit and read it as the figure being wrong. It is a different convention, not a different corpus, and the three integers it produces still sum to 43. The one below is the one the manuscript uses, and it is stated here so the two can be told apart rather than argued over.
The convention, because the corpus carries split listings in two shapes. A species with listing rows under more than one appendix is counted here as split-listed: that is the common minke whale, and it is the one the tally breaks out. A species with a single row whose appendix value is itself compound is counted under the first appendix that value names: that is the sea otter, Enhydra lutris, whose row reads I/II and which is counted in the 16. Its own species page describes that value as a split-population listing, which it is, so the two statements are counting different things and this says which. Under any other treatment of these two the four figures move, and the total stays 43.
Binding derived
appendix-membership · pipeline artifact figureA · binding proposed, unconfirmed
Binding evidence, as the record states it: figureA_source_data.csv holds one row per (species, appendix) membership: 44 rows over 43 species, 17 Appendix I, 26 Appendix II, 1 Appendix III. That is the figure's own tally, which the caption states against the 16/25/1-split/1 of Section 4.2; the appendix_as_recorded column carries the split listing so both readings come out of the same table. The tally reproduces deposit/sources/curated/cites_listings.csv. The caption of 18 October states the convention itself: the boxes count the split-listed minke whale in both Appendix I and Appendix II, so they read 17, 26 and 1 and sum to 44 rather than 43.
No named person has checked that pipeline artifact figureA is the code that produced this manuscript figure. The evidence for the proposal is stated below. The four fields that follow describe figureA. If the binding is wrong, they describe the wrong figure, and that possibility is disclosed here rather than absorbed.
Evidence: figureA_source_data.csv holds one row per (species, appendix) membership: 44 rows over 43 species, 17 Appendix I, 26 Appendix II, 1 Appendix III. That is the figure's own tally, which the caption states against the 16/25/1-split/1 of Section 4.2; the appendix_as_recorded column carries the split listing so both readings come out of the same table. The tally reproduces deposit/sources/curated/cites_listings.csv. The caption of 18 October states the convention itself: the boxes count the split-listed minke whale in both Appendix I and Appendix II, so they read 17, 26 and 1 and sum to 44 rather than 43.
Confirming this requires the person who produced the figures.
1. Data inputs derived
Computed from the figure pipeline's declared tables and query.
- Database schema
arctic_v2- Tables
- Database release
- CITES Trade Database v2025.1
- Extraction date
- 13 June 2025
- Redistributable
- Yes
2. Processing steps derived
Each step is a path in the figure pipeline's own repository, which this site does not host. Regenerate and verify commands are literal. 1 of the 2 artifacts below is served here, and carries a download link; the rest are named so the chain can be read, not followed.
- Read the frozen intermediate.
data/raw/1_DATA_cites_effectiveness.json - Compute and write the frozen numeric source data.
visualizations/source_data/figureA_source_data.csv, download
3. LLM involvement: none
No generative model participated in producing this figure. The generator imports only csv, json, sys, collections, pathlib and the project's redlist_scale module, which is the one Red List scale and holds no model call either. No model API is called at any step, and there is no prompt because there is no model.
This is asserted mechanically rather than by declaration. visualizations/tests/test_no_model_calls.py reads this manifest, opens every script and generator it names, searches for model SDKs and API clients, and fails if it finds one. It covers every script and generator the manifest names.
The generative model in this project drafted 343 narrative prose sections about the 43 species. None of that prose reaches any figure, and none of it is displayed on this site: it is withheld pending expert review.
No review has begun, so there is no agreement statistic to report and no page reporting one. What can be checked is that the withholding is complete: the drafts are deposited unrendered as supplementary/species_narratives_no_citations.csv, every species page states how many of its sections are withheld, and the dataset page marks the table as unreviewed content.
4. Validation
Automated gate: passed. Byte-exact regeneration from the live database against the frozen source data.
Human review: generated 2026-10-06; no named person has signed the binding and it is not yet independently confirmed. withheld The automated gate establishes that the numbers are reproducible from the database. It establishes nothing about whether they are the right numbers to plot. That judgement requires a person, and no person has recorded it.
Known defects: none recorded. The automated regeneration gate passed and no defect has been recorded against this figure. That is a statement about what is known, not a guarantee that none exists.