# Supplementary Table: Results traceability Every quantitative claim in the paper, its operative definition, the table it comes from, and the query or computation that reproduces it against the deposited corpus. Every claim listed here is valued, and every one is reproducible from the deposited package alone - no database, credentials or network access. `deposit/verification/verify_claims_offline.py` recomputes each value under the operative definition recorded beside it and fails if any differs. | ID | Claim | Value | Methods | Source table | Operative definition | |---|---|---|---|---|---| | C01 | Trade records for the 43 listed species, 1975 to 2024 | **492095** | S3 (trade integration) | `cites_trade_records` | One CITES trade record is one line in the Trade Database. | | C02 | Species in the study set | **43** | S1 (species identification) | `species` | CITES-listed species with an Arctic distribution, curated. | | C03 | Share of records for commercial purpose | **67.01** | S3 | `cites_trade_records` | purpose = 'T'. Denominator is ALL records including 1.65% with a blank purpose. | | C04 | Share of records involving Appendix I species | **27.34** | S1, S3 | `cites_trade_records` | appendix recorded ON THE TRADE RECORD, not the species' current listing. | | C05 | Share of records involving Appendix II species | **67.34** | S1, S3 | `cites_trade_records` | As C04. Note Appendix III is 5.32% and must be reported or the shares do not sum. | | C06 | Share of records involving Appendix III species | **5.32** | S1, S3 | `cites_trade_records` | As C04. | | C07 | Share of records that are wild-sourced | **28.40** | S3 | `cites_trade_records` | source = 'W'. The majority of the corpus (58.5%) is captive-bred, mostly farmed sturgeon. | | C08 | Records with a missing quantity | **0** | S3 | `cites_trade_records` | Quantity IS NULL. Reported because a per-dimension total silently drops such rows. | | C09 | IUCN assessments in the series | **370** | S2 (status compilation) | `iucn_assessments` | All assessments, all scopes, all vintages. | | C10 | Assessments using a pre-2001 Red List category | **95** | S2 | `iucn_assessments` | Categories outside LC/NT/VU/EN/CR/EW/EX/DD/RE. Not on the manuscript's ordinal scale. | | C11 | Species whose first post-listing assessment is a legacy category | **25** | S2, S6 | `iucn_assessments + cites_listings` | Baseline is the first assessment, global or European, at or after the EARLIEST listing date. Both conventions are stated here rather than assumed; under other conventions the value differs, which is the point of question 1 and 2. | | C12 | NAMMCO catch records | **958** | S2 | `catch_records` | All records as exported 11 June 2025. | | C13 | Illegal-trade seizure records attributable to a study species | **225** | S3 | `illegal_trade_seizures` | 919 source rows decontaminated to those whose reported taxon matches the attributed species. Quantities are unitless in the source, so no total exists. | | C14 | CITES listings in force | **62** | S1 | `cites_listings` | All listing rows. is_current does not select one: 8 species carry several. | | C15 | CMS listings, excluding daughter-agreement memberships | **43** | S1 | `cms_listings` | agreement = 'CMS'. Daughter agreements (ACCOBAMS etc) are memberships, not CMS listings. | | C16 | Daughter-agreement memberships | **49** | S1 | `cms_listings` | agreement <> 'CMS'. Carry no CMS appendix. | | C17 | Timeline events with no corroborating table in the deposit | **66** | S7 | `timeline_events` | cites_suspension and eu_decision. Curated assertions, flagged in the API. | | C18 | Narrative sections withheld pending expert review | **343** | S5, S7 | `species_narratives` | GATED. Machine-drafted, unservable until a named expert signs. | | C19 | Species whose post-listing trajectory improved | **13** | S6 (scoring) | `source_data.zip:figure5_source_data.csv` | Earliest and most recent assessment, global or European regional, in 1982-2025, one series per species ordered by year and assessment id, compared on the IUCN category ladder; improved means the most recent category is less threatened than the earliest. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C20 | Species whose post-listing trajectory was stable | **18** | S6 | `source_data.zip:figure5_source_data.csv` | Same comparison as C19; unchanged means the two categories sit at the same rung, which is not the same as no assessment having been made. | | C21 | Species whose post-listing trajectory deteriorated | **10** | S6 | `source_data.zip:figure5_source_data.csv` | Same comparison as C19; deteriorated means the most recent category is more threatened than the earliest. | | C22 | Category transitions between comparable global and European assessments | **43** | S6 | `source_data.zip:figure5_source_data.csv` | One step between consecutive comparable assessments, global or European, of the same species inside 1982-2025, ordered by year and assessment id; both rows of a same-year pair stay in the sequence, as the published method always counted them. A non-comparable assessment breaks the series rather than being stepped over. A transition requires a DIFFERENT category: redlist_scale.transitions appends only when the rank changes, so a pair that holds its category is not a transition and the total is not the number of consecutive comparable pairs. That condition was missing from this definition until 6 October 2026, when the definition lens of round u15 found the same looseness in the two copies in the manuscript; this file exists to stop exactly that, and it was loose in the same way. The scale is the one in ingest/redlist_scale.py, under which LR/cd reads as Near Threatened per the crosswalk the Methods cite and Regionally Extinct is not on the ladder. Recorded as 34 until 6 October 2026, when this number was found to have been computed on a second, divergent rank table that excluded LR/cd while the species split in the same sentence used one that did not; the two transitions it could not see are the polar bear's 1996 LR/cd to 2006 VU and Dall's porpoise's 1996 LR/cd to 2008 LC. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C23 | Species whose current category, on the latest global or European assessment, is CR (Critically Endangered) | **4** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C24 | Species whose current category, on the latest global or European assessment, is EN (Endangered) | **3** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C25 | Species whose current category, on the latest global or European assessment, is VU (Vulnerable) | **11** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C26 | Species whose current category, on the latest global or European assessment, is NT (Near Threatened) | **3** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C27 | Species whose current category, on the latest global or European assessment, is LC (Least Concern) | **21** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C28 | Species whose current category, on the latest global or European assessment, is DD (Data Deficient) | **0** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C29 | Improving species whose first improvement falls after their CITES listing | **12** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv; tables/cites_listings.csv` | Years from a species' first CITES listing to the first assessment, global or European, inside 1982-2025 at a lower risk rank than its baseline assessment, over the improving species whose first improvement falls after their listing. One improving species, Phoebastria albatrus, has its whole assessment series inside the window before its 2023 listing and is excluded from the interval and reported separately. | | C30 | Median years from CITES listing to first improvement | **29** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv; tables/cites_listings.csv` | Years from a species' first CITES listing to the first assessment, global or European, inside 1982-2025 at a lower risk rank than its baseline assessment, over the improving species whose first improvement falls after their listing. One improving species, Phoebastria albatrus, has its whole assessment series inside the window before its 2023 listing and is excluded from the interval and reported separately. | | C31 | Shortest listing-to-improvement interval, years | **5** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv; tables/cites_listings.csv` | Years from a species' first CITES listing to the first assessment, global or European, inside 1982-2025 at a lower risk rank than its baseline assessment, over the improving species whose first improvement falls after their listing. One improving species, Phoebastria albatrus, has its whole assessment series inside the window before its 2023 listing and is excluded from the interval and reported separately. | | C32 | Longest listing-to-improvement interval, years | **48** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv; tables/cites_listings.csv` | Years from a species' first CITES listing to the first assessment, global or European, inside 1982-2025 at a lower risk rank than its baseline assessment, over the improving species whose first improvement falls after their listing. One improving species, Phoebastria albatrus, has its whole assessment series inside the window before its 2023 listing and is excluded from the interval and reported separately. | | C33 | Species with a global or European Red List assessment published in 2005 or later | **43** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, in the deposited assessment series, over the 43 species in the study set. 'Reassessed' means the year of that assessment; subpopulation and other regional assessments (the Mediterranean porbeagle among them) are not counted, in line with the classification rule. | | C34 | Species with a global or European Red List assessment published in 2015 or later | **42** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, in the deposited assessment series, over the 43 species in the study set. 'Reassessed' means the year of that assessment; subpopulation and other regional assessments (the Mediterranean porbeagle among them) are not counted, in line with the classification rule. | | C35 | Species with a global or European Red List assessment published in 2020 or later | **37** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, in the deposited assessment series, over the 43 species in the study set. 'Reassessed' means the year of that assessment; subpopulation and other regional assessments (the Mediterranean porbeagle among them) are not counted, in line with the classification rule. | | C36 | Median years from a species' most recent global or European assessment to the recorded snapshot | **3** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, in the deposited assessment series, over the 43 species in the study set. 'Reassessed' means the year of that assessment; subpopulation and other regional assessments (the Mediterranean porbeagle among them) are not counted, in line with the classification rule. RE-ANCHORED 6 OCTOBER 2026: this was measured to 2025 while the deposited source-release table records the IUCN Red List snapshot as 2026-07-22, so the median lag was understated by a year. The verifier now reads the snapshot year out of that table (C108) instead of carrying a literal. | | C37 | Share of trade records that are Siberian sturgeon | **39.78** | S3 (trade integration) | `cites_trade_records` | Count of records for Acipenser baerii over all 492,095 records. The unit is the trade record; no quantity is summed. | | C38 | Share of trade records that are gyrfalcon | **17.64** | S3 (trade integration) | `cites_trade_records` | Count of records for Falco rusticolus over all 492,095 records. The unit is the trade record; no quantity is summed. | | C39 | Share of Siberian sturgeon records whose source code is C (captive-bred) | **99.40** | S3 (trade integration) | `cites_trade_records` | source = 'C' over all Acipenser baerii records. C is the Convention's captive-bred code; F (born in captivity) and D (Appendix-I captive-bred for commercial purposes) are counted separately and not merged into it. | | C40 | Unchanged species in a threatened category (VU, EN or CR) at the end of the window | **8** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Species classed unchanged whose most recent assessment, global or European, inside 1982-2025 is VU, EN or CR (C40) or CR alone (C41). 'Unchanged' means the rank of the latest comparable assessment equals the rank of the earliest, which says nothing about the level it is held at. | | C41 | Unchanged species at Critically Endangered at the end of the window | **2** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Species classed unchanged whose most recent assessment, global or European, inside 1982-2025 is VU, EN or CR (C40) or CR alone (C41). 'Unchanged' means the rank of the latest comparable assessment equals the rank of the earliest, which says nothing about the level it is held at. | | C42 | Distinct purpose codes present in the corpus | **12** | S3 (trade integration) | `cites_trade_records` | Distinct purpose values present in the corpus, blank included. The Convention's purpose code list has no subsistence or Indigenous-harvest category, so this count is the fact itself: there are twelve values and none of them denotes subsistence. | | C43 | Narwhal wild-origin records whose purpose code is P (personal) | **15129** | S3 (trade integration) | `cites_trade_records` | Wild-origin records (source = 'W') for the named species, and of those the records whose purpose code is P, personal. Purpose and source are read exactly as the Trade Database records them and are never merged. The pair is reported because for these species the dominant wild-origin purpose is P rather than T, which is what makes the absence of a subsistence code consequential. | | C44 | Narwhal wild-origin records | **23866** | S3 (trade integration) | `cites_trade_records` | Wild-origin records (source = 'W') for the named species, and of those the records whose purpose code is P, personal. Purpose and source are read exactly as the Trade Database records them and are never merged. The pair is reported because for these species the dominant wild-origin purpose is P rather than T, which is what makes the absence of a subsistence code consequential. | | C45 | Atlantic walrus wild-origin records whose purpose code is P (personal) | **15468** | S3 (trade integration) | `cites_trade_records` | Wild-origin records (source = 'W') for the named species, and of those the records whose purpose code is P, personal. Purpose and source are read exactly as the Trade Database records them and are never merged. The pair is reported because for these species the dominant wild-origin purpose is P rather than T, which is what makes the absence of a subsistence code consequential. | | C46 | Atlantic walrus wild-origin records | **20998** | S3 (trade integration) | `cites_trade_records` | Wild-origin records (source = 'W') for the named species, and of those the records whose purpose code is P, personal. Purpose and source are read exactly as the Trade Database records them and are never merged. The pair is reported because for these species the dominant wild-origin purpose is P rather than T, which is what makes the absence of a subsistence code consequential. | | C47 | Beluga wild-origin records whose purpose code is P (personal) | **2236** | S3 (trade integration) | `cites_trade_records` | Wild-origin records (source = 'W') for the named species, and of those the records whose purpose code is P, personal. Purpose and source are read exactly as the Trade Database records them and are never merged. The pair is reported because for these species the dominant wild-origin purpose is P rather than T, which is what makes the absence of a subsistence code consequential. | | C48 | Beluga wild-origin records | **2895** | S3 (trade integration) | `cites_trade_records` | Wild-origin records (source = 'W') for the named species, and of those the records whose purpose code is P, personal. Purpose and source are read exactly as the Trade Database records them and are never merged. The pair is reported because for these species the dominant wild-origin purpose is P rather than T, which is what makes the absence of a subsistence code consequential. | | C49 | Trade records involving species currently Critically Endangered | **196140** | S3 (trade integration) | `cites_trade_records + source_data.zip:figure5_source_data.csv` | Trade records of the species whose most recent assessment, global or European, inside 1982-2025 is Critically Endangered - the same current-category rule as C23-C28 - as a count (C49) and as a share of all 492,095 records (C50). The four species are Acipenser baerii, Eubalaena glacialis, Leucogeranus leucogeranus and Numenius borealis, and 99.8% of the block is the first of them, captive-bred. The share is 39.86%, which the paper reported as 'over 39.9%': it is under it. | | C50 | Share of all records involving species currently Critically Endangered | **39.86** | S3 (trade integration) | `cites_trade_records + source_data.zip:figure5_source_data.csv` | Trade records of the species whose most recent assessment, global or European, inside 1982-2025 is Critically Endangered - the same current-category rule as C23-C28 - as a count (C49) and as a share of all 492,095 records (C50). The four species are Acipenser baerii, Eubalaena glacialis, Leucogeranus leucogeranus and Numenius borealis, and 99.8% of the block is the first of them, captive-bred. The share is 39.86%, which the paper reported as 'over 39.9%': it is under it. | | C51 | Items traded, wild-origin, 15 species in Figure 7 | **25710439** | S3 (trade integration) | `source_data.zip:figure8_source_data.csv` | The 15 wild-origin species Figure 7 draws, summed over the shipped Source Data rows for that figure: items traded (C51, the total_quantity column) and trade records (C52, the record_count column). The two are different units and are never added together; the figure's panels (a) to (c) are in items and panels (d) to (f) in records. The set is the 15 largest by items traded among records carrying CITES source code W, 1985-2024. | | C52 | Trade records, wild-origin, 15 species in Figure 7 | **133497** | S3 (trade integration) | `source_data.zip:figure8_source_data.csv` | The 15 wild-origin species Figure 7 draws, summed over the shipped Source Data rows for that figure: items traded (C51, the total_quantity column) and trade records (C52, the record_count column). The two are different units and are never added together; the figure's panels (a) to (c) are in items and panels (d) to (f) in records. The set is the 15 largest by items traded among records carrying CITES source code W, 1985-2024. | | C53 | Records with a non-Arctic importer | **345603** | S3 (trade integration) | `cites_trade_records` | Trade records whose importer is not one of the nine Arctic reporter codes, out of the 492,095 in the corpus; no condition on reporting side applies and none is needed, since no record carries a blank importer, exporter or year and none post-dates 2024 (asserted, not assumed) (CA, US, RU, NO, IS, DK, SE, FI, GL): the population Figure 9 draws. C53 is that denominator, C54 and C55 the United Arab Emirates as a count and a share of it, and C56/C57 the lowest and highest share among the five additional importers the Results name (UK, Hong Kong, Qatar, South Korea, Singapore). The paper gave that range as 1.5-2.7%; it is 2.1-3.9%. Denmark and Greenland are separate codes in the Trade Database and both count as Arctic. | | C54 | United Arab Emirates import records | **83131** | S3 (trade integration) | `cites_trade_records` | Trade records whose importer is not one of the nine Arctic reporter codes, out of the 492,095 in the corpus; no condition on reporting side applies and none is needed, since no record carries a blank importer, exporter or year and none post-dates 2024 (asserted, not assumed) (CA, US, RU, NO, IS, DK, SE, FI, GL): the population Figure 9 draws. C53 is that denominator, C54 and C55 the United Arab Emirates as a count and a share of it, and C56/C57 the lowest and highest share among the five additional importers the Results name (UK, Hong Kong, Qatar, South Korea, Singapore). The paper gave that range as 1.5-2.7%; it is 2.1-3.9%. Denmark and Greenland are separate codes in the Trade Database and both count as Arctic. | | C55 | United Arab Emirates share of non-Arctic imports | **24.05** | S3 (trade integration) | `cites_trade_records` | Trade records whose importer is not one of the nine Arctic reporter codes, out of the 492,095 in the corpus; no condition on reporting side applies and none is needed, since no record carries a blank importer, exporter or year and none post-dates 2024 (asserted, not assumed) (CA, US, RU, NO, IS, DK, SE, FI, GL): the population Figure 9 draws. C53 is that denominator, C54 and C55 the United Arab Emirates as a count and a share of it, and C56/C57 the lowest and highest share among the five additional importers the Results name (UK, Hong Kong, Qatar, South Korea, Singapore). The paper gave that range as 1.5-2.7%; it is 2.1-3.9%. Denmark and Greenland are separate codes in the Trade Database and both count as Arctic. | | C56 | Lowest share among the five additional importers named | **2.09** | S3 (trade integration) | `cites_trade_records` | Trade records whose importer is not one of the nine Arctic reporter codes, out of the 492,095 in the corpus; no condition on reporting side applies and none is needed, since no record carries a blank importer, exporter or year and none post-dates 2024 (asserted, not assumed) (CA, US, RU, NO, IS, DK, SE, FI, GL): the population Figure 9 draws. C53 is that denominator, C54 and C55 the United Arab Emirates as a count and a share of it, and C56/C57 the lowest and highest share among the five additional importers the Results name (UK, Hong Kong, Qatar, South Korea, Singapore). The paper gave that range as 1.5-2.7%; it is 2.1-3.9%. Denmark and Greenland are separate codes in the Trade Database and both count as Arctic. ON 6 OCTOBER 2026 THE BAND LEFT THE PROSE: the Results now give each of the five jurisdictions its own count and share (C86-C95), so C56 and C57 no longer check a reported sentence. They are kept as the range guard behind those ten - if a release pushes one of the five outside 2.09% to 3.92%, the five named in the sentence are no longer the five it is about, and C86's ranking guard and these two fail together. | | C57 | Highest share among the five additional importers named | **3.92** | S3 (trade integration) | `cites_trade_records` | Trade records whose importer is not one of the nine Arctic reporter codes, out of the 492,095 in the corpus; no condition on reporting side applies and none is needed, since no record carries a blank importer, exporter or year and none post-dates 2024 (asserted, not assumed) (CA, US, RU, NO, IS, DK, SE, FI, GL): the population Figure 9 draws. C53 is that denominator, C54 and C55 the United Arab Emirates as a count and a share of it, and C56/C57 the lowest and highest share among the five additional importers the Results name (UK, Hong Kong, Qatar, South Korea, Singapore). The paper gave that range as 1.5-2.7%; it is 2.1-3.9%. Denmark and Greenland are separate codes in the Trade Database and both count as Arctic. ON 6 OCTOBER 2026 THE BAND LEFT THE PROSE: the Results now give each of the five jurisdictions its own count and share (C86-C95), so C56 and C57 no longer check a reported sentence. They are kept as the range guard behind those ten - if a release pushes one of the five outside 2.09% to 3.92%, the five named in the sentence are no longer the five it is about, and C86's ranking guard and these two fail together. | | C58 | Seizure records after taxonomic reconciliation | **225** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The seizure records compiled by Stringham et al. and reconciled to CITES-listed Arctic species: 225 records over 26 species, retained from 919 source rows. C59 to C61 are the three species the Results name, counted as records of that species in that table. This is NOT the corpus Figure 8 draws, which is CITES records with source code I; the two have different denominators and must not be mixed. | | C59 | Atlantic walrus seizure records | **33** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The seizure records compiled by Stringham et al. and reconciled to CITES-listed Arctic species: 225 records over 26 species, retained from 919 source rows. C59 to C61 are the three species the Results name, counted as records of that species in that table. This is NOT the corpus Figure 8 draws, which is CITES records with source code I; the two have different denominators and must not be mixed. | | C60 | Sperm whale seizure records | **31** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The seizure records compiled by Stringham et al. and reconciled to CITES-listed Arctic species: 225 records over 26 species, retained from 919 source rows. C59 to C61 are the three species the Results name, counted as records of that species in that table. This is NOT the corpus Figure 8 draws, which is CITES records with source code I; the two have different denominators and must not be mixed. | | C61 | Canada lynx seizure records | **20** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The seizure records compiled by Stringham et al. and reconciled to CITES-listed Arctic species: 225 records over 26 species, retained from 919 source rows. C59 to C61 are the three species the Results name, counted as records of that species in that table. This is NOT the corpus Figure 8 draws, which is CITES records with source code I; the two have different denominators and must not be mixed. | | C66 | Records reported from both ends on the exact eight-field match | **226046** | S3 (trade integration) | `cites_trade_records` | Records whose eight-field key -- species, year, importer, exporter, term, unit, purpose and source -- occurs under both reporter sides, counted as records rather than as keys. C66 is the exact join Methods describes. C67 is the same join with the year dropped, and it is larger, which is why the paper reports C66 as a LOWER BOUND on the both-ends overlap: the two reporter sides disagree on exactly these fields -- a consignment leaving in December and clearing in January is two different years -- so an exact match misses the hard cases and leaves them looking independent. Relaxing term, unit or purpose instead also enlarges the set (281,875 / 269,158 / 270,342); the year is named in the paper because it is the largest. | | C67 | Records reported from both ends when the year is not required to match | **340991** | S3 (trade integration) | `cites_trade_records` | Records whose eight-field key -- species, year, importer, exporter, term, unit, purpose and source -- occurs under both reporter sides, counted as records rather than as keys. C66 is the exact join Methods describes. C67 is the same join with the year dropped, and it is larger, which is why the paper reports C66 as a LOWER BOUND on the both-ends overlap: the two reporter sides disagree on exactly these fields -- a consignment leaving in December and clearing in January is two different years -- so an exact match misses the hard cases and leaves them looking independent. Relaxing term, unit or purpose instead also enlarges the set (281,875 / 269,158 / 270,342); the year is named in the paper because it is the largest. | | C68 | Trade records before 1985 | **7362** | S3 (trade integration) | `cites_trade_records` | Trade records with a year before 1985, out of the 492,095 in the corpus. C69 counts those carrying CITES source code W and C70 those carrying no source at all. C69 is zero and C70 is 94.6% of C68: the absence of wild-origin trade before 1985 is a property of the reporting format, not of the trade. The verifier fails if C69 ever becomes non-zero, because the manuscript's stated reason for the 1985 window would then be wrong. | | C69 | Pre-1985 records carrying source code W | **0** | S3 (trade integration) | `cites_trade_records` | Trade records with a year before 1985, out of the 492,095 in the corpus. C69 counts those carrying CITES source code W and C70 those carrying no source at all. C69 is zero and C70 is 94.6% of C68: the absence of wild-origin trade before 1985 is a property of the reporting format, not of the trade. The verifier fails if C69 ever becomes non-zero, because the manuscript's stated reason for the 1985 window would then be wrong. | | C70 | Pre-1985 records carrying no source code | **6963** | S3 (trade integration) | `cites_trade_records` | Trade records with a year before 1985, out of the 492,095 in the corpus. C69 counts those carrying CITES source code W and C70 those carrying no source at all. C69 is zero and C70 is 94.6% of C68: the absence of wild-origin trade before 1985 is a property of the reporting format, not of the trade. The verifier fails if C69 ever becomes non-zero, because the manuscript's stated reason for the 1985 window would then be wrong. | | C71 | Latest year Russia appears as exporting Party in an exporter-reported record | **2023** | S3 (trade integration) | `cites_trade_records` | Russia and Iceland as the exporting Party. C71 and C72 are the latest year each appears in an exporter-reported record; C73 and C74 count records with that exporter reported from the IMPORTING side. The paper gives 2017 and 2010 as the Parties' last reports and said CITES-LEX offsets the gap; CITES-LEX holds legislation and policy and no trade records, while the trade corpus shows both Parties reporting later than those dates and the mirror importer side covering them in addition. The mitigation is the reporter side, not the legislation database. | | C72 | Latest year Iceland appears as exporting Party in an exporter-reported record | **2021** | S3 (trade integration) | `cites_trade_records` | Russia and Iceland as the exporting Party. C71 and C72 are the latest year each appears in an exporter-reported record; C73 and C74 count records with that exporter reported from the IMPORTING side. The paper gives 2017 and 2010 as the Parties' last reports and said CITES-LEX offsets the gap; CITES-LEX holds legislation and policy and no trade records, while the trade corpus shows both Parties reporting later than those dates and the mirror importer side covering them in addition. The mitigation is the reporter side, not the legislation database. | | C73 | Importer-reported records whose exporting Party is Russia | **1152** | S3 (trade integration) | `cites_trade_records` | Russia and Iceland as the exporting Party. C71 and C72 are the latest year each appears in an exporter-reported record; C73 and C74 count records with that exporter reported from the IMPORTING side. The paper gives 2017 and 2010 as the Parties' last reports and said CITES-LEX offsets the gap; CITES-LEX holds legislation and policy and no trade records, while the trade corpus shows both Parties reporting later than those dates and the mirror importer side covering them in addition. The mitigation is the reporter side, not the legislation database. | | C74 | Importer-reported records whose exporting Party is Iceland | **246** | S3 (trade integration) | `cites_trade_records` | Russia and Iceland as the exporting Party. C71 and C72 are the latest year each appears in an exporter-reported record; C73 and C74 count records with that exporter reported from the IMPORTING side. The paper gives 2017 and 2010 as the Parties' last reports and said CITES-LEX offsets the gap; CITES-LEX holds legislation and policy and no trade records, while the trade corpus shows both Parties reporting later than those dates and the mirror importer side covering them in addition. The mitigation is the reporter side, not the legislation database. | | C75 | Narwhal records in the seizure corpus | **10** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | Records of Monodon monoceros in the 225-record seizure compilation reconciled from Stringham et al. Ten records over the whole compilation. The manuscript's "typically hundreds per year" for narwhal is the order of magnitude of the LEGAL, quota-limited harvest and of the 4,923 tusks legally exported 1987-2009 that the paper reports from TRAFFIC - about 230 a year - not of trafficking. This claim exists so the sentence can never again be written as though the seizure record supported it. | | C76 | Beluga trade records in 2007 | **15** | S3 (trade integration) | `cites_trade_records` | Trade records of Delphinapterus leucas by year, out of the 492,095 in the corpus. C76 is 2007, the baseline the Results compare to; C77 is 2019; C78 is 2022, the peak of the series; C79 is 2023. The manuscript said "128 beluga trade records were reported in 2023, an increase of 11.6% over 2007 levels": 128 is the 2019 count, 2023 is 25, and 11.6% matches no pair of years. The verifier fails if 2022 stops being the peak, because the corrected sentence names it as one. | | C77 | Beluga trade records in 2019 | **128** | S3 (trade integration) | `cites_trade_records` | Trade records of Delphinapterus leucas by year, out of the 492,095 in the corpus. C76 is 2007, the baseline the Results compare to; C77 is 2019; C78 is 2022, the peak of the series; C79 is 2023. The manuscript said "128 beluga trade records were reported in 2023, an increase of 11.6% over 2007 levels": 128 is the 2019 count, 2023 is 25, and 11.6% matches no pair of years. The verifier fails if 2022 stops being the peak, because the corrected sentence names it as one. | | C78 | Beluga trade records in 2022 (series peak) | **129** | S3 (trade integration) | `cites_trade_records` | Trade records of Delphinapterus leucas by year, out of the 492,095 in the corpus. C76 is 2007, the baseline the Results compare to; C77 is 2019; C78 is 2022, the peak of the series; C79 is 2023. The manuscript said "128 beluga trade records were reported in 2023, an increase of 11.6% over 2007 levels": 128 is the 2019 count, 2023 is 25, and 11.6% matches no pair of years. The verifier fails if 2022 stops being the peak, because the corrected sentence names it as one. | | C79 | Beluga trade records in 2023 | **25** | S3 (trade integration) | `cites_trade_records` | Trade records of Delphinapterus leucas by year, out of the 492,095 in the corpus. C76 is 2007, the baseline the Results compare to; C77 is 2019; C78 is 2022, the peak of the series; C79 is 2023. The manuscript said "128 beluga trade records were reported in 2023, an increase of 11.6% over 2007 levels": 128 is the 2019 count, 2023 is 25, and 11.6% matches no pair of years. The verifier fails if 2022 stops being the peak, because the corrected sentence names it as one. | | C80 | Species in EN or CR in 1996 | **7** | S2 (conservation-status compilation) | `figureB_source_data.csv (Source Data, Figure 11 panel C)` | Species in EN or CR in each year of the assessment window, counted on the global-and-European series (from 21 October 2026, the lead author's decision) by the rule Figure 11 panel C is drawn on: a species counts once it has a comparable assessment, at its latest comparable category, the higher assessment id standing in a tied year. The years are those of the panel's Source Data (figureB_source_data.csv, panel C, metric high_risk_en_cr); until that file is regenerated on this series the verifier reports on stderr that the shipped panel still carries the global-only counts. C80 is the first year of the window, C81 the peak count, C82 the first year at that peak, C83 the last year of the window. The manuscript said only "more species assessed as EN or CR through the 1990s-2010s", a comparative with no second term; it now names the rise and the fall. The verifier fails if the series stops rising to a peak and falling away from it. | | C81 | Peak number of species in EN or CR in the window | **13** | S2 (conservation-status compilation) | `figureB_source_data.csv (Source Data, Figure 11 panel C)` | Species in EN or CR in each year of the assessment window, counted on the global-and-European series (from 21 October 2026, the lead author's decision) by the rule Figure 11 panel C is drawn on: a species counts once it has a comparable assessment, at its latest comparable category, the higher assessment id standing in a tied year. The years are those of the panel's Source Data (figureB_source_data.csv, panel C, metric high_risk_en_cr); until that file is regenerated on this series the verifier reports on stderr that the shipped panel still carries the global-only counts. C80 is the first year of the window, C81 the peak count, C82 the first year at that peak, C83 the last year of the window. The manuscript said only "more species assessed as EN or CR through the 1990s-2010s", a comparative with no second term; it now names the rise and the fall. The verifier fails if the series stops rising to a peak and falling away from it. | | C82 | First year at that peak | **2015** | S2 (conservation-status compilation) | `figureB_source_data.csv (Source Data, Figure 11 panel C)` | Species in EN or CR in each year of the assessment window, counted on the global-and-European series (from 21 October 2026, the lead author's decision) by the rule Figure 11 panel C is drawn on: a species counts once it has a comparable assessment, at its latest comparable category, the higher assessment id standing in a tied year. The years are those of the panel's Source Data (figureB_source_data.csv, panel C, metric high_risk_en_cr); until that file is regenerated on this series the verifier reports on stderr that the shipped panel still carries the global-only counts. C80 is the first year of the window, C81 the peak count, C82 the first year at that peak, C83 the last year of the window. The manuscript said only "more species assessed as EN or CR through the 1990s-2010s", a comparative with no second term; it now names the rise and the fall. The verifier fails if the series stops rising to a peak and falling away from it. | | C83 | Species in EN or CR in the last year of the window | **7** | S2 (conservation-status compilation) | `figureB_source_data.csv (Source Data, Figure 11 panel C)` | Species in EN or CR in each year of the assessment window, counted on the global-and-European series (from 21 October 2026, the lead author's decision) by the rule Figure 11 panel C is drawn on: a species counts once it has a comparable assessment, at its latest comparable category, the higher assessment id standing in a tied year. The years are those of the panel's Source Data (figureB_source_data.csv, panel C, metric high_risk_en_cr); until that file is regenerated on this series the verifier reports on stderr that the shipped panel still carries the global-only counts. C80 is the first year of the window, C81 the peak count, C82 the first year at that peak, C83 the last year of the window. The manuscript said only "more species assessed as EN or CR through the 1990s-2010s", a comparative with no second term; it now names the rise and the fall. The verifier fails if the series stops rising to a peak and falling away from it. | | C84 | Unchanged species whose last category in the window is threatened | **8** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | Of the 18 species whose Red List category, on global and European assessments, is unchanged across 1982-2025, those whose last comparable category in the window is VU, EN or CR (C84), and those whose last is CR (C85). The two CR species are Eubalaena glacialis and Numenius borealis. The Abstract carried both numbers and the Results carried neither, so a reader who asked which species could not be answered; Section 4.7 now states them. The check refuses to run unless this module's ladder produces the same 18 unchanged species that verify_trajectories_offline and the manuscript carry - one scale, one source. | | C85 | Of those, species whose last category is Critically Endangered | **2** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | Of the 18 species whose Red List category, on global and European assessments, is unchanged across 1982-2025, those whose last comparable category in the window is VU, EN or CR (C84), and those whose last is CR (C85). The two CR species are Eubalaena glacialis and Numenius borealis. The Abstract carried both numbers and the Results carried neither, so a reader who asked which species could not be answered; Section 4.7 now states them. The check refuses to run unless this module's ladder produces the same 18 unchanged species that verify_trajectories_offline and the manuscript carry - one scale, one source. | | C86 | Figure 9 records whose importer is the United Kingdom | **13540** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C87 | Figure 9 records whose importer is Hong Kong | **12101** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C88 | Figure 9 records whose importer is Qatar | **9973** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C89 | Figure 9 records whose importer is South Korea | **7246** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C90 | Figure 9 records whose importer is Singapore | **7217** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C91 | Share of the Figure 9 population imported by the United Kingdom | **3.92** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C92 | Share of the Figure 9 population imported by Hong Kong | **3.50** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C93 | Share of the Figure 9 population imported by Qatar | **2.89** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C94 | Share of the Figure 9 population imported by South Korea | **2.10** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C95 | Share of the Figure 9 population imported by Singapore | **2.09** | S3 (trade integration) | `cites_trade_records` | Records in the Figure 9 population (importer not one of the nine Arctic reporter codes; C53 is the denominator, 345,603) whose importer is the jurisdiction named. C86-C90 are the counts, C91-C95 the shares of C53. The Results name these five because they are the sixth to tenth largest importers in that population, and the verifier refuses to run unless they still are - a reordered release would otherwise publish a count against the wrong country. The paper reports each share to one decimal, as it already does for the UAE (C55 is 24.05 and the prose says 24.1%). Before 6 October 2026 the Results gave these five no figure of their own, only the band 'between 2.1% and 3.9%' spanning nearly a factor of two, so a named Party could not read its own attributed number out of the paper. The band itself was correct and C56/C57 still check it. | | C96 | CITES trade records for the polar bear, 1975-2024 | **24730** | S3 (trade integration) | `cites_trade_records` | CITES trade records for Ursus maritimus over 1975-2024: all records (C96), those whose reporter side is the exporting Party (C97), and those carrying source code W (C98). The Results cite 24,590 polar bear 'exports' with no unit, against the Methods' undertaking that the unit is stated explicitly wherever results refer to individuals, products or mass. No reading of this corpus yields 24,590, so the paper now reports these three counts with their units beside the cited figure and asks the authors where it comes from (Q66). The verifier fails if one of the three ever equals 24,590, because the sentence saying the corpus does not reproduce it would then be false. | | C97 | Of those, records reported by the exporting Party | **16464** | S3 (trade integration) | `cites_trade_records` | CITES trade records for Ursus maritimus over 1975-2024: all records (C96), those whose reporter side is the exporting Party (C97), and those carrying source code W (C98). The Results cite 24,590 polar bear 'exports' with no unit, against the Methods' undertaking that the unit is stated explicitly wherever results refer to individuals, products or mass. No reading of this corpus yields 24,590, so the paper now reports these three counts with their units beside the cited figure and asks the authors where it comes from (Q66). The verifier fails if one of the three ever equals 24,590, because the sentence saying the corpus does not reproduce it would then be false. | | C98 | Of those, records carrying source code W | **21117** | S3 (trade integration) | `cites_trade_records` | CITES trade records for Ursus maritimus over 1975-2024: all records (C96), those whose reporter side is the exporting Party (C97), and those carrying source code W (C98). The Results cite 24,590 polar bear 'exports' with no unit, against the Methods' undertaking that the unit is stated explicitly wherever results refer to individuals, products or mass. No reading of this corpus yields 24,590, so the paper now reports these three counts with their units beside the cited figure and asks the authors where it comes from (Q66). The verifier fails if one of the three ever equals 24,590, because the sentence saying the corpus does not reproduce it would then be false. | | C99 | Seizure records in the reconciled sub-dataset carrying any date | **0** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The reconciled Stringham et al. seizure sub-dataset, 225 records across 26 species. C99 is the number carrying any seizure date or year, and it is zero: the deposited table's seizure_date and seizure_year are empty for every row, so NO time window can be stated for this corpus. C100 to C102 are what can be stated instead - the source database each record comes from: CITES, LEMIS and WTP. The verifier fails if any record gains a date, because a window could then be given, and if the three databases stop covering all 225. | | C100 | Of the 225 seizure records, those from the CITES source database | **125** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The reconciled Stringham et al. seizure sub-dataset, 225 records across 26 species. C99 is the number carrying any seizure date or year, and it is zero: the deposited table's seizure_date and seizure_year are empty for every row, so NO time window can be stated for this corpus. C100 to C102 are what can be stated instead - the source database each record comes from: CITES, LEMIS and WTP. The verifier fails if any record gains a date, because a window could then be given, and if the three databases stop covering all 225. | | C101 | Of the 225 seizure records, those from LEMIS | **79** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The reconciled Stringham et al. seizure sub-dataset, 225 records across 26 species. C99 is the number carrying any seizure date or year, and it is zero: the deposited table's seizure_date and seizure_year are empty for every row, so NO time window can be stated for this corpus. C100 to C102 are what can be stated instead - the source database each record comes from: CITES, LEMIS and WTP. The verifier fails if any record gains a date, because a window could then be given, and if the three databases stop covering all 225. | | C102 | Of the 225 seizure records, those from WTP | **21** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures` | The reconciled Stringham et al. seizure sub-dataset, 225 records across 26 species. C99 is the number carrying any seizure date or year, and it is zero: the deposited table's seizure_date and seizure_year are empty for every row, so NO time window can be stated for this corpus. C100 to C102 are what can be stated instead - the source database each record comes from: CITES, LEMIS and WTP. The verifier fails if any record gains a date, because a window could then be given, and if the three databases stop covering all 225. | | C103 | Improving species with no post-listing improvement, curated ladder | **3** | S2 (conservation-status compilation) | `cites_listings.csv, figure5_source_data.csv` | Of the 13 species whose Red List category, on global and European assessments, improves across 1982-2025, the number with NO improvement measurable at or after their CITES listing. The trajectory classification itself never reads a listing date - it compares the earliest and latest comparable assessment inside the window - so 'after listing' is a label the method does not earn, and whether it is false depends on which listing ladder is read. C103 uses the curated listings table, which Sections 4.2 and 4.7 follow, and gives 3: Phoebastria albatrus (listed 2023-02-23, every comparable assessment 1994-2018), Phocoenoides dalli (listed 2003-02-13, improvement straddles the listing with nothing moving after it) and Orcinus orca (listed 2003-02-13, only one comparable assessment, a European one, falls at or after the listing, so no improvement is measurable from it). C104 uses first_listed_on in the events table, the anchor Table 3's sensitivity note uses, and gives 0. The paper uses both ladders, which is what Q6 and Q24 ask about; Section 4.7 already concedes the albatross case while the Results headline still lists it among those that improved 'after listing'. The verifier fails if the two ladders ever agree, because the disagreement is the whole point of the pair. | | C104 | The same count under first_listed_on in the events table | **0** | S2 (conservation-status compilation) | `cites_listing_events.csv, figure5_source_data.csv` | Of the 13 species whose Red List category, on global and European assessments, improves across 1982-2025, the number with NO improvement measurable at or after their CITES listing. The trajectory classification itself never reads a listing date - it compares the earliest and latest comparable assessment inside the window - so 'after listing' is a label the method does not earn, and whether it is false depends on which listing ladder is read. C103 uses the curated listings table, which Sections 4.2 and 4.7 follow, and gives 3: Phoebastria albatrus (listed 2023-02-23, every comparable assessment 1994-2018), Phocoenoides dalli (listed 2003-02-13, improvement straddles the listing with nothing moving after it) and Orcinus orca (listed 2003-02-13, only one comparable assessment, a European one, falls at or after the listing, so no improvement is measurable from it). C104 uses first_listed_on in the events table, the anchor Table 3's sensitivity note uses, and gives 0. The paper uses both ladders, which is what Q6 and Q24 ask about; Section 4.7 already concedes the albatross case while the Results headline still lists it among those that improved 'after listing'. The verifier fails if the two ladders ever agree, because the disagreement is the whole point of the pair. | | C105 | CITES trade records dated 2024 | **138** | S3 (trade integration) | `cites_trade_records` | The CITES trade corpus's final year against its recent level. C105 is the number of records dated 2024, C106 the number dated 2023, C107 the number of the 2024 records reported by the exporting Party. Release v2025.1 was downloaded on 28 July 2025, before the 2024 annual reports were due, so 2024 holds 138 records against a 2019-2023 average of 29,712 - 0.5% of it - and 131 of the 138 come from the importing side, which is what an unsubmitted exporter reporting year looks like. The corpus is 1975-2024 because that is what was downloaded; 2024 is not a year of data, and any trend, decline or peak read to that edge is reading a reporting gap. The verifier fails if a later release fills 2024 to more than a quarter of the recent average, because the sentences calling it incomplete would then be wrong. | | C106 | CITES trade records dated 2023 | **27432** | S3 (trade integration) | `cites_trade_records` | The CITES trade corpus's final year against its recent level. C105 is the number of records dated 2024, C106 the number dated 2023, C107 the number of the 2024 records reported by the exporting Party. Release v2025.1 was downloaded on 28 July 2025, before the 2024 annual reports were due, so 2024 holds 138 records against a 2019-2023 average of 29,712 - 0.5% of it - and 131 of the 138 come from the importing side, which is what an unsubmitted exporter reporting year looks like. The corpus is 1975-2024 because that is what was downloaded; 2024 is not a year of data, and any trend, decline or peak read to that edge is reading a reporting gap. The verifier fails if a later release fills 2024 to more than a quarter of the recent average, because the sentences calling it incomplete would then be wrong. | | C107 | Of the 2024 records, those reported by the exporting Party | **7** | S3 (trade integration) | `cites_trade_records` | The CITES trade corpus's final year against its recent level. C105 is the number of records dated 2024, C106 the number dated 2023, C107 the number of the 2024 records reported by the exporting Party. Release v2025.1 was downloaded on 28 July 2025, before the 2024 annual reports were due, so 2024 holds 138 records against a 2019-2023 average of 29,712 - 0.5% of it - and 131 of the 138 come from the importing side, which is what an unsubmitted exporter reporting year looks like. The corpus is 1975-2024 because that is what was downloaded; 2024 is not a year of data, and any trend, decline or peak read to that edge is reading a reporting gap. The verifier fails if a later release fills 2024 to more than a quarter of the recent average, because the sentences calling it incomplete would then be wrong. | | C108 | The snapshot year the deposit records for the IUCN Red List | **2026** | S2 (conservation-status compilation) | `source_releases.csv, figure5_source_data.csv` | How recent the assessment record is, anchored on the snapshot year the deposited source-release table records for the IUCN Red List (C108, read from that table rather than written into the verifier). C109 is the species whose most recent assessment, global or European, falls inside the ten years before the snapshot, C110 those ten years or more before it - on the global-and-European series only Rhodiola rosea, at 2014. Until 6 October 2026 the recency paragraph and C36 were anchored on 2025 while the recorded snapshot was 2026-07-22, so the paper said 'the 2025 snapshot', 'five years before the snapshot', '42 within the past decade' and 'only one is older than a decade'. Re-anchored on 6 October 2026 those were the 2026 snapshot, six years, 41 and two; on the series of 21 October 2026, which counts European assessments, they are the 2026 snapshot, three years, 42 and one. The verifier reads the year from the deposit so the next move of the snapshot carries these claims with it. | | C109 | Species last assessed inside the ten years before the snapshot | **42** | S2 (conservation-status compilation) | `source_releases.csv, figure5_source_data.csv` | How recent the assessment record is, anchored on the snapshot year the deposited source-release table records for the IUCN Red List (C108, read from that table rather than written into the verifier). C109 is the species whose most recent assessment, global or European, falls inside the ten years before the snapshot, C110 those ten years or more before it - on the global-and-European series only Rhodiola rosea, at 2014. Until 6 October 2026 the recency paragraph and C36 were anchored on 2025 while the recorded snapshot was 2026-07-22, so the paper said 'the 2025 snapshot', 'five years before the snapshot', '42 within the past decade' and 'only one is older than a decade'. Re-anchored on 6 October 2026 those were the 2026 snapshot, six years, 41 and two; on the series of 21 October 2026, which counts European assessments, they are the 2026 snapshot, three years, 42 and one. The verifier reads the year from the deposit so the next move of the snapshot carries these claims with it. | | C110 | Species last assessed ten years or more before the snapshot | **1** | S2 (conservation-status compilation) | `source_releases.csv, figure5_source_data.csv` | How recent the assessment record is, anchored on the snapshot year the deposited source-release table records for the IUCN Red List (C108, read from that table rather than written into the verifier). C109 is the species whose most recent assessment, global or European, falls inside the ten years before the snapshot, C110 those ten years or more before it - on the global-and-European series only Rhodiola rosea, at 2014. Until 6 October 2026 the recency paragraph and C36 were anchored on 2025 while the recorded snapshot was 2026-07-22, so the paper said 'the 2025 snapshot', 'five years before the snapshot', '42 within the past decade' and 'only one is older than a decade'. Re-anchored on 6 October 2026 those were the 2026 snapshot, six years, 41 and two; on the series of 21 October 2026, which counts European assessments, they are the 2026 snapshot, three years, 42 and one. The verifier reads the year from the deposit so the next move of the snapshot carries these claims with it. | | C111 | Earliest year a comparable trajectory opens | **1988** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | The years a trajectory can actually open. The assessment record is labelled 1982-2025 and called 'roughly four decades', but the pre-1994 codes V and E are off the ordinal scale, so no trajectory opens before a comparable category exists. C111 is the earliest opening year across the 41 species that carry a trajectory and C112 the number that open in 1996, the first global assessment round on the modern criteria. So trajectories span at most 37 years rather than 43, and 25 of the 41 span 29. The verifier fails if a trajectory ever opens before 1988, which would mean the scale had admitted a pre-1994 code, and if the species count stops being 41. | | C112 | Trajectories that open in 1996 | **25** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | The years a trajectory can actually open. The assessment record is labelled 1982-2025 and called 'roughly four decades', but the pre-1994 codes V and E are off the ordinal scale, so no trajectory opens before a comparable category exists. C111 is the earliest opening year across the 41 species that carry a trajectory and C112 the number that open in 1996, the first global assessment round on the modern criteria. So trajectories span at most 37 years rather than 43, and 25 of the 41 span 29. The verifier fails if a trajectory ever opens before 1988, which would mean the scale had admitted a pre-1994 code, and if the species count stops being 41. | | C113 | First year any record carries source code I | **1980** | S3 (trade integration) | `cites_trade_records` | Figure 8's window. The caption gives 1980-2023 with no reason while Figure 7's caption explains its own, and it needs none: C113 and C114 are the first and last years in which any record in the corpus carries CITES source code I. The boundary is a property of the data, not a choice, and the caption now says so. | | C114 | Last year any record carries source code I | **2023** | S3 (trade integration) | `cites_trade_records` | Figure 8's window. The caption gives 1980-2023 with no reason while Figure 7's caption explains its own, and it needs none: C113 and C114 are the first and last years in which any record in the corpus carries CITES source code I. The boundary is a property of the data, not a choice, and the caption now says so. | | C115 | CITES trade records whose exporting Party is Russia | **1919** | S3 (trade integration) | `cites_trade_records` | Records whose exporting Party is Russia (C115) or Iceland (C116), the two Arctic Parties whose national reporting lapsed. Together 2,443 records, 0.50% of the 492,095, which is the bound on the objection that corpus-level decline partly measures their reporting cessation rather than trade. The verifier also recomputes the mirror-reported counts the Results state, 1,152 and 246, and fails if they move - they are the importer-reported records for those exporters across the whole window, which is what a reader could not previously tell. | | C116 | CITES trade records whose exporting Party is Iceland | **524** | S3 (trade integration) | `cites_trade_records` | Records whose exporting Party is Russia (C115) or Iceland (C116), the two Arctic Parties whose national reporting lapsed. Together 2,443 records, 0.50% of the 492,095, which is the bound on the objection that corpus-level decline partly measures their reporting cessation rather than trade. The verifier also recomputes the mirror-reported counts the Results state, 1,152 and 246, and fails if they move - they are the importer-reported records for those exporters across the whole window, which is what a reader could not previously tell. | | C117 | Share of assessed species in EN or CR in 1996 | **18.40** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | The share of species with a current assessment, global or European, that are in EN or CR, at 1996 (C117), at the 2015-2017 peak (C118) and from 2019 (C119). The Results reported this series as raw counts (seven, thirteen, nine) while the figure cited plots a proportion and the assessed set was still growing: 19 species carry a current assessment in 1988, 38 in 1996 and 43 from 2015. As a share the series is 18.4%, 30.2%, 20.9% - the denominator grew 13% between 1996 and 2015 while the count grew 86%, and the fall after 2017 happens at a fixed denominator of 43 - so the rise and the fall hold on either measure. The verifier fails if the share series stops rising to a peak and falling away. | | C118 | The same share at the 2015-2017 peak | **30.20** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | The share of species with a current assessment, global or European, that are in EN or CR, at 1996 (C117), at the 2015-2017 peak (C118) and from 2019 (C119). The Results reported this series as raw counts (seven, thirteen, nine) while the figure cited plots a proportion and the assessed set was still growing: 19 species carry a current assessment in 1988, 38 in 1996 and 43 from 2015. As a share the series is 18.4%, 30.2%, 20.9% - the denominator grew 13% between 1996 and 2015 while the count grew 86%, and the fall after 2017 happens at a fixed denominator of 43 - so the rise and the fall hold on either measure. The verifier fails if the share series stops rising to a peak and falling away. | | C119 | The same share from 2019 to the end of the window | **20.90** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | The share of species with a current assessment, global or European, that are in EN or CR, at 1996 (C117), at the 2015-2017 peak (C118) and from 2019 (C119). The Results reported this series as raw counts (seven, thirteen, nine) while the figure cited plots a proportion and the assessed set was still growing: 19 species carry a current assessment in 1988, 38 in 1996 and 43 from 2015. As a share the series is 18.4%, 30.2%, 20.9% - the denominator grew 13% between 1996 and 2015 while the count grew 86%, and the fall after 2017 happens at a fixed denominator of 43 - so the rise and the fall hold on either measure. The verifier fails if the share series stops rising to a peak and falling away. | | C120 | Seizure records naming a bone product | **36** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures, illegal_trade_products` | What the 225-record reconciled seizure corpus holds, grouped by a substring on the deposited product name so a reader can redo it: names containing 'bone' (C120), 'skin' (C121), 'ivory' or 'tusk' (C122) and 'fur' (C123), with C124 the number of fur entries in the product vocabulary. Until 6 October 2026 the Results said 'Seized products were dominated by ivory, furs, and skins'. All three are wrong on this corpus: bone-named products lead at 36 records, skin-named follow at 22, ivory and tusk together are 20, and NO record names a fur product at all - while the vocabulary carries two fur entries, so the zero is an absence in the record and not a missing code. These are record counts, not quantities: the corpus has no usable quantity and no dates (C99). The verifier fails if bone stops leading or if a fur record appears. | | C121 | Seizure records naming a skin product | **22** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures, illegal_trade_products` | What the 225-record reconciled seizure corpus holds, grouped by a substring on the deposited product name so a reader can redo it: names containing 'bone' (C120), 'skin' (C121), 'ivory' or 'tusk' (C122) and 'fur' (C123), with C124 the number of fur entries in the product vocabulary. Until 6 October 2026 the Results said 'Seized products were dominated by ivory, furs, and skins'. All three are wrong on this corpus: bone-named products lead at 36 records, skin-named follow at 22, ivory and tusk together are 20, and NO record names a fur product at all - while the vocabulary carries two fur entries, so the zero is an absence in the record and not a missing code. These are record counts, not quantities: the corpus has no usable quantity and no dates (C99). The verifier fails if bone stops leading or if a fur record appears. | | C122 | Seizure records naming ivory or a tusk | **20** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures, illegal_trade_products` | What the 225-record reconciled seizure corpus holds, grouped by a substring on the deposited product name so a reader can redo it: names containing 'bone' (C120), 'skin' (C121), 'ivory' or 'tusk' (C122) and 'fur' (C123), with C124 the number of fur entries in the product vocabulary. Until 6 October 2026 the Results said 'Seized products were dominated by ivory, furs, and skins'. All three are wrong on this corpus: bone-named products lead at 36 records, skin-named follow at 22, ivory and tusk together are 20, and NO record names a fur product at all - while the vocabulary carries two fur entries, so the zero is an absence in the record and not a missing code. These are record counts, not quantities: the corpus has no usable quantity and no dates (C99). The verifier fails if bone stops leading or if a fur record appears. | | C123 | Seizure records naming a fur product | **0** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures, illegal_trade_products` | What the 225-record reconciled seizure corpus holds, grouped by a substring on the deposited product name so a reader can redo it: names containing 'bone' (C120), 'skin' (C121), 'ivory' or 'tusk' (C122) and 'fur' (C123), with C124 the number of fur entries in the product vocabulary. Until 6 October 2026 the Results said 'Seized products were dominated by ivory, furs, and skins'. All three are wrong on this corpus: bone-named products lead at 36 records, skin-named follow at 22, ivory and tusk together are 20, and NO record names a fur product at all - while the vocabulary carries two fur entries, so the zero is an absence in the record and not a missing code. These are record counts, not quantities: the corpus has no usable quantity and no dates (C99). The verifier fails if bone stops leading or if a fur record appears. | | C124 | Fur entries in the deposited product vocabulary | **2** | S3 (data compilation: seizure sub-dataset) | `illegal_trade_seizures, illegal_trade_products` | What the 225-record reconciled seizure corpus holds, grouped by a substring on the deposited product name so a reader can redo it: names containing 'bone' (C120), 'skin' (C121), 'ivory' or 'tusk' (C122) and 'fur' (C123), with C124 the number of fur entries in the product vocabulary. Until 6 October 2026 the Results said 'Seized products were dominated by ivory, furs, and skins'. All three are wrong on this corpus: bone-named products lead at 36 records, skin-named follow at 22, ivory and tusk together are 20, and NO record names a fur product at all - while the vocabulary carries two fur entries, so the zero is an absence in the record and not a missing code. These are record counts, not quantities: the corpus has no usable quantity and no dates (C99). The verifier fails if bone stops leading or if a fur record appears. | | C125 | Species currently in a threatened category | **18** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | Species among the 43 whose most recent Red List category, on the latest global or European assessment, is VU, EN or CR (C125) and that number as a share of the 43 (C126). The Introduction cites 'around 40% of CITES-listed species remain threatened' as the one prior quantitative finding this paper's headline is comparable to, and the paper never returned to it. It can: the current distribution the Results already report, 4 CR, 3 EN and 11 VU, is 18 of 43, which is 41.9%. The comparison is left to the reader because the cited 40% has no stated denominator in the source as cited, and a share of an unknown base cannot be compared arithmetically to a share of 43. The verifier refuses unless the distribution is still the 4, 3, 11 the Results report. | | C126 | That number as a share of the 43 | **41.90** | S2 (conservation-status compilation) | `figure5_source_data.csv (Source Data, assessment series)` | Species among the 43 whose most recent Red List category, on the latest global or European assessment, is VU, EN or CR (C125) and that number as a share of the 43 (C126). The Introduction cites 'around 40% of CITES-listed species remain threatened' as the one prior quantitative finding this paper's headline is comparable to, and the paper never returned to it. It can: the current distribution the Results already report, 4 CR, 3 EN and 11 VU, is 18 of 43, which is 41.9%. The comparison is left to the reader because the cited 40% has no stated denominator in the source as cited, and a share of an unknown base cannot be compared arithmetically to a share of 43. The verifier refuses unless the distribution is still the 4, 3, 11 the Results report. | | C127 | Species first listed in the 1970s, events ladder | **38** | S1 (corpus assembly and listing ladders) | `cites_listing_events.csv, cites_listings.csv, supplementary_table_1_post_listing_window.csv` | Section 4.2's listing enumeration, measured on both deposited ladders because the sentence reported one ladder's counts and named three of its five species from the other. C127/C128 are species whose earliest is_listing event in cites_listing_events.csv falls in 1970-1979 and in or after 1998; C129/C130 are the same two counts over the earliest listing_date in the curated cites_listings.csv, which carries the current instrument rather than the first. C131 is the year of the short-tailed albatross's earliest listing event, which is 1975 in Appendix I; its 2023 events are a transfer out of Appendix I into Appendix II, and Section 4.2 had called that a listing. The verifier refuses unless the two ladders still disagree, unless the albatross's first event is still Appendix I and its 2023 events are still a transfer, and unless Supplementary Table 1's first_listed_on agrees with the events table. Q6 still asks the authors which ladder the paper should use throughout. | | C128 | Species first listed in or after 1998, events ladder | **5** | S1 (corpus assembly and listing ladders) | `cites_listing_events.csv, cites_listings.csv, supplementary_table_1_post_listing_window.csv` | Section 4.2's listing enumeration, measured on both deposited ladders because the sentence reported one ladder's counts and named three of its five species from the other. C127/C128 are species whose earliest is_listing event in cites_listing_events.csv falls in 1970-1979 and in or after 1998; C129/C130 are the same two counts over the earliest listing_date in the curated cites_listings.csv, which carries the current instrument rather than the first. C131 is the year of the short-tailed albatross's earliest listing event, which is 1975 in Appendix I; its 2023 events are a transfer out of Appendix I into Appendix II, and Section 4.2 had called that a listing. The verifier refuses unless the two ladders still disagree, unless the albatross's first event is still Appendix I and its 2023 events are still a transfer, and unless Supplementary Table 1's first_listed_on agrees with the events table. Q6 still asks the authors which ladder the paper should use throughout. | | C129 | Species first listed in the 1970s, curated listings table | **20** | S1 (corpus assembly and listing ladders) | `cites_listing_events.csv, cites_listings.csv, supplementary_table_1_post_listing_window.csv` | Section 4.2's listing enumeration, measured on both deposited ladders because the sentence reported one ladder's counts and named three of its five species from the other. C127/C128 are species whose earliest is_listing event in cites_listing_events.csv falls in 1970-1979 and in or after 1998; C129/C130 are the same two counts over the earliest listing_date in the curated cites_listings.csv, which carries the current instrument rather than the first. C131 is the year of the short-tailed albatross's earliest listing event, which is 1975 in Appendix I; its 2023 events are a transfer out of Appendix I into Appendix II, and Section 4.2 had called that a listing. The verifier refuses unless the two ladders still disagree, unless the albatross's first event is still Appendix I and its 2023 events are still a transfer, and unless Supplementary Table 1's first_listed_on agrees with the events table. Q6 still asks the authors which ladder the paper should use throughout. | | C130 | Species first listed in or after 1998, curated listings table | **15** | S1 (corpus assembly and listing ladders) | `cites_listing_events.csv, cites_listings.csv, supplementary_table_1_post_listing_window.csv` | Section 4.2's listing enumeration, measured on both deposited ladders because the sentence reported one ladder's counts and named three of its five species from the other. C127/C128 are species whose earliest is_listing event in cites_listing_events.csv falls in 1970-1979 and in or after 1998; C129/C130 are the same two counts over the earliest listing_date in the curated cites_listings.csv, which carries the current instrument rather than the first. C131 is the year of the short-tailed albatross's earliest listing event, which is 1975 in Appendix I; its 2023 events are a transfer out of Appendix I into Appendix II, and Section 4.2 had called that a listing. The verifier refuses unless the two ladders still disagree, unless the albatross's first event is still Appendix I and its 2023 events are still a transfer, and unless Supplementary Table 1's first_listed_on agrees with the events table. Q6 still asks the authors which ladder the paper should use throughout. | | C131 | Year of the short-tailed albatross's earliest listing event | **1975** | S1 (corpus assembly and listing ladders) | `cites_listing_events.csv, cites_listings.csv, supplementary_table_1_post_listing_window.csv` | Section 4.2's listing enumeration, measured on both deposited ladders because the sentence reported one ladder's counts and named three of its five species from the other. C127/C128 are species whose earliest is_listing event in cites_listing_events.csv falls in 1970-1979 and in or after 1998; C129/C130 are the same two counts over the earliest listing_date in the curated cites_listings.csv, which carries the current instrument rather than the first. C131 is the year of the short-tailed albatross's earliest listing event, which is 1975 in Appendix I; its 2023 events are a transfer out of Appendix I into Appendix II, and Section 4.2 had called that a listing. The verifier refuses unless the two ladders still disagree, unless the albatross's first event is still Appendix I and its 2023 events are still a transfer, and unless Supplementary Table 1's first_listed_on agrees with the events table. Q6 still asks the authors which ladder the paper should use throughout. | | C132 | Genuine category transitions on global and European assessments, 1982-2025 | **43** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C133 | Transitions with both endpoints in 2001 or later | **31** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C134 | Improvements among transitions that cannot cross the 2001 revision | **15** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C135 | Deteriorations among transitions that cannot cross the 2001 revision | **16** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C136 | Transitions straddling the 2001 revision | **7** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C137 | Transitions falling wholly before the 2001 revision | **5** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C138 | Species improved on assessments from 2001 onward only | **5** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C139 | Species deteriorated on assessments from 2001 onward only | **7** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C140 | Species without a comparable trajectory from 2001 onward | **4** | S4 (trajectory classification) | `figure5_source_data.csv (reconstituted via REBUILD-IUCN.md)` | Sensitivity of the trajectory balance to the 2001 revision of the IUCN criteria, which is the strongest objection put to the paper's near-even split of 22 improvements and 21 deteriorations. Instrument 1 holds the criteria fixed WITHIN each comparison by bucketing every genuine transition on the global-and-European series in 1982-2025 on where the 2001 revision falls relative to its two endpoints: C133 to C135 are transitions whose endpoints are both 2001 or later, which cannot cross the revision, with their two directions; C136 straddles it and C137 falls wholly before it. C132 is the total and must equal the three buckets. Transitions come from redlist_scale.transitions on the global-and-European series, so a non-comparable assessment breaks the chain rather than being skipped over and same-year ties go through the one tie-break. Instrument 2 (C138 to C140) restricts each species' global-and-European series to assessments published in 2001 or later and reclassifies it; it is harsher and it does NOT isolate the revision, because restricting the start year also shortens the window from 43 years to 24, which reduces movement independently of any criteria change. Neither instrument bounds category change arising from improved knowledge or from taxonomy within one criteria era: that needs IUCN's reason-for-change field, which the deposited series does not carry. The verifier refuses if fewer than half the transitions are post-revision, since the instrument would then bound nothing. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | | C141 | Wild-origin share of United Arab Emirates import records | **0.96** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C142 | Gyrfalcon and peregrine share of United Arab Emirates import records | **91.84** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C143 | Lowest wild-origin share among the ten named non-Arctic importers | **0.96** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C144 | Highest wild-origin share among the ten named non-Arctic importers | **52.27** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C145 | Named non-Arctic importers whose largest single taxon exceeds 70% of their records | **6** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C146 | Share of records for commercial purpose, excluding Siberian sturgeon and gyrfalcon | **43.28** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C147 | Records remaining after excluding Siberian sturgeon and gyrfalcon | **209518** | S3 (trade integration) | `cites_trade_records` | Composition of the shares Figure 9 reports, over the ten non-Arctic importers the Results name (AE, DE, CN, CH, JP, GB, HK, QA, KR, SG). Source code W is the Convention's wild code, read exactly as recorded and never merged with C, D or F. Dominated means the single largest taxon exceeds 70% of that importer's records. C146 and C147 move the denominator with the numerator by design: C03's 67.01% commercial share is over all 492,095 records, and the comparison is what shows how much of it is a property of the composition. | | C148 | Species whose current category, on the latest global or European assessment, is RE (Regionally Extinct) | **1** | S2 (status compilation) | `source_data.zip:figure5_source_data.csv` | Most recent assessment per species, global or European, within 1982-2025, counted by category. Each species contributes exactly once, so the seven counts (C23 to C28 and C148, Regionally Extinct) sum to the 43 species in the study set. Categories are not mapped: the most recent assessment of every species uses a 2001-system category, so no legacy crosswalk applies here. Where a species carries several assessments in its final year the higher assessment id stands (redlist_scale.pick_in_year) and the is_latest flag must agree with it; two species are decided this way (Bubo scandiacus 2021, LC against VU; Odobenus rosmarus 2025, NT against VU). Until 6 October 2026 the tie was broken by CSV row order, which reported the superseded assessment as current for those two species and gave VU 7, NT 4, LC 22. From 21 October 2026 European regional assessments are included (the lead author's decision): for each species the Global and Europe assessments form one series ordered by year and assessment id, the higher assessment id standing in a year with several, and the Mediterranean porbeagle assessment (16527805) is excluded. | ## Queries **C01** ```sql select count(*) from arctic_v2.cites_trade_records ``` **C02** ```sql select count(*) from arctic_v2.species ``` **C03** ```sql select round(100.0*count(*) filter (where purpose='T')/count(*),2) from arctic_v2.cites_trade_records ``` **C04** ```sql select round(100.0*count(*) filter (where appendix='I')/count(*),2) from arctic_v2.cites_trade_records ``` **C05** ```sql select round(100.0*count(*) filter (where appendix='II')/count(*),2) from arctic_v2.cites_trade_records ``` **C06** ```sql select round(100.0*count(*) filter (where appendix='III')/count(*),2) from arctic_v2.cites_trade_records ``` **C07** ```sql select round(100.0*count(*) filter (where source='W')/count(*),2) from arctic_v2.cites_trade_records ``` **C08** ```sql select count(*) from arctic_v2.cites_trade_records where quantity is null ``` **C09** ```sql select count(*) from arctic_v2.iucn_assessments ``` **C10** ```sql select count(*) from arctic_v2.iucn_assessments where status not in ('LC','NT','VU','EN','CR','EW','EX','DD','RE') ``` **C11** ```sql with fl as (select species_id, min(listing_date) f from arctic_v2.cites_listings group by 1) select count(*) from (select s.id, (select i.status from arctic_v2.iucn_assessments i where i.species_id=s.id and coalesce(i.scope_code, '1') in ('1', '2', 'EU') and coalesce(i.assessment_id, 0) not in (16527805) and i.year_published >= extract(year from fl.f) order by i.year_published, i.assessment_id limit 1) bs from arctic_v2.species s join fl on fl.species_id=s.id) x where bs is not null and bs not in ('LC','NT','VU','EN','CR','EW','EX','DD','RE') ``` **C12** ```sql select count(*) from arctic_v2.catch_records ``` **C13** ```sql select count(*) from arctic_v2.illegal_trade_seizures ``` **C14** ```sql select count(*) from arctic_v2.cites_listings ``` **C15** ```sql select count(*) from arctic_v2.cms_listings where agreement='CMS' ``` **C16** ```sql select count(*) from arctic_v2.cms_listings where agreement<>'CMS' ``` **C17** ```sql select count(*) from arctic_v2.timeline_events where event_type in ('cites_suspension','eu_decision') ``` **C18** ```sql select count(*) from arctic_v2.species_narratives ``` **C19** `see deposit/verification/verify_trajectories_offline.py` **C20** `see deposit/verification/verify_trajectories_offline.py` **C21** `see deposit/verification/verify_trajectories_offline.py` **C22** `see deposit/verification/verify_trajectories_offline.py` **C23** `see deposit/verification/verify_trajectories_offline.py` **C24** `see deposit/verification/verify_trajectories_offline.py` **C25** `see deposit/verification/verify_trajectories_offline.py` **C26** `see deposit/verification/verify_trajectories_offline.py` **C27** `see deposit/verification/verify_trajectories_offline.py` **C28** `see deposit/verification/verify_trajectories_offline.py` **C29** `see deposit/verification/verify_trajectories_offline.py` **C30** `see deposit/verification/verify_trajectories_offline.py` **C31** `see deposit/verification/verify_trajectories_offline.py` **C32** `see deposit/verification/verify_trajectories_offline.py` **C33** `see deposit/verification/verify_trajectories_offline.py` **C34** `see deposit/verification/verify_trajectories_offline.py` **C35** `see deposit/verification/verify_trajectories_offline.py` **C36** `see deposit/verification/verify_trajectories_offline.py` **C37** `see deposit/verification/verify_claims_offline.py` **C38** `see deposit/verification/verify_claims_offline.py` **C39** `see deposit/verification/verify_claims_offline.py` **C40** `see deposit/verification/verify_claims_offline.py` **C41** `see deposit/verification/verify_claims_offline.py` **C42** `see deposit/verification/verify_claims_offline.py` **C43** `see deposit/verification/verify_claims_offline.py` **C44** `see deposit/verification/verify_claims_offline.py` **C45** `see deposit/verification/verify_claims_offline.py` **C46** `see deposit/verification/verify_claims_offline.py` **C47** `see deposit/verification/verify_claims_offline.py` **C48** `see deposit/verification/verify_claims_offline.py` **C49** `see deposit/verification/verify_claims_offline.py` **C50** `see deposit/verification/verify_claims_offline.py` **C51** `see deposit/verification/verify_claims_offline.py` **C52** `see deposit/verification/verify_claims_offline.py` **C53** `see deposit/verification/verify_claims_offline.py` **C54** `see deposit/verification/verify_claims_offline.py` **C55** `see deposit/verification/verify_claims_offline.py` **C56** `see deposit/verification/verify_claims_offline.py` **C57** `see deposit/verification/verify_claims_offline.py` **C58** `see deposit/verification/verify_claims_offline.py` **C59** `see deposit/verification/verify_claims_offline.py` **C60** `see deposit/verification/verify_claims_offline.py` **C61** `see deposit/verification/verify_claims_offline.py` **C66** `see deposit/verification/verify_claims_offline.py` **C67** `see deposit/verification/verify_claims_offline.py` **C68** `see deposit/verification/verify_claims_offline.py` **C69** `see deposit/verification/verify_claims_offline.py` **C70** `see deposit/verification/verify_claims_offline.py` **C71** `see deposit/verification/verify_claims_offline.py` **C72** `see deposit/verification/verify_claims_offline.py` **C73** `see deposit/verification/verify_claims_offline.py` **C74** `see deposit/verification/verify_claims_offline.py` **C75** `see deposit/verification/verify_claims_offline.py` **C76** `see deposit/verification/verify_claims_offline.py` **C77** `see deposit/verification/verify_claims_offline.py` **C78** `see deposit/verification/verify_claims_offline.py` **C79** `see deposit/verification/verify_claims_offline.py` **C80** `see deposit/verification/verify_claims_offline.py` **C81** `see deposit/verification/verify_claims_offline.py` **C82** `see deposit/verification/verify_claims_offline.py` **C83** `see deposit/verification/verify_claims_offline.py` **C84** `see deposit/verification/verify_claims_offline.py` **C85** `see deposit/verification/verify_claims_offline.py` **C86** `see deposit/verification/verify_claims_offline.py` **C87** `see deposit/verification/verify_claims_offline.py` **C88** `see deposit/verification/verify_claims_offline.py` **C89** `see deposit/verification/verify_claims_offline.py` **C90** `see deposit/verification/verify_claims_offline.py` **C91** `see deposit/verification/verify_claims_offline.py` **C92** `see deposit/verification/verify_claims_offline.py` **C93** `see deposit/verification/verify_claims_offline.py` **C94** `see deposit/verification/verify_claims_offline.py` **C95** `see deposit/verification/verify_claims_offline.py` **C96** `see deposit/verification/verify_claims_offline.py` **C97** `see deposit/verification/verify_claims_offline.py` **C98** `see deposit/verification/verify_claims_offline.py` **C99** `see deposit/verification/verify_claims_offline.py` **C100** `see deposit/verification/verify_claims_offline.py` **C101** `see deposit/verification/verify_claims_offline.py` **C102** `see deposit/verification/verify_claims_offline.py` **C103** `see deposit/verification/verify_claims_offline.py` **C104** `see deposit/verification/verify_claims_offline.py` **C105** `see deposit/verification/verify_claims_offline.py` **C106** `see deposit/verification/verify_claims_offline.py` **C107** `see deposit/verification/verify_claims_offline.py` **C108** `see deposit/verification/verify_claims_offline.py` **C109** `see deposit/verification/verify_claims_offline.py` **C110** `see deposit/verification/verify_claims_offline.py` **C111** `see deposit/verification/verify_claims_offline.py` **C112** `see deposit/verification/verify_claims_offline.py` **C113** `see deposit/verification/verify_claims_offline.py` **C114** `see deposit/verification/verify_claims_offline.py` **C115** `see deposit/verification/verify_claims_offline.py` **C116** `see deposit/verification/verify_claims_offline.py` **C117** `see deposit/verification/verify_claims_offline.py` **C118** `see deposit/verification/verify_claims_offline.py` **C119** `see deposit/verification/verify_claims_offline.py` **C120** `see deposit/verification/verify_claims_offline.py` **C121** `see deposit/verification/verify_claims_offline.py` **C122** `see deposit/verification/verify_claims_offline.py` **C123** `see deposit/verification/verify_claims_offline.py` **C124** `see deposit/verification/verify_claims_offline.py` **C125** `see deposit/verification/verify_claims_offline.py` **C126** `see deposit/verification/verify_claims_offline.py` **C127** `see deposit/verification/verify_claims_offline.py` **C128** `see deposit/verification/verify_claims_offline.py` **C129** `see deposit/verification/verify_claims_offline.py` **C130** `see deposit/verification/verify_claims_offline.py` **C131** `see deposit/verification/verify_claims_offline.py` **C132** `see deposit/verification/verify_claims_offline.py` **C133** `see deposit/verification/verify_claims_offline.py` **C134** `see deposit/verification/verify_claims_offline.py` **C135** `see deposit/verification/verify_claims_offline.py` **C136** `see deposit/verification/verify_claims_offline.py` **C137** `see deposit/verification/verify_claims_offline.py` **C138** `see deposit/verification/verify_claims_offline.py` **C139** `see deposit/verification/verify_claims_offline.py` **C140** `see deposit/verification/verify_claims_offline.py` **C141** `see deposit/verification/verify_claims_offline.py` **C142** `see deposit/verification/verify_claims_offline.py` **C143** `see deposit/verification/verify_claims_offline.py` **C144** `see deposit/verification/verify_claims_offline.py` **C145** `see deposit/verification/verify_claims_offline.py` **C146** `see deposit/verification/verify_claims_offline.py` **C147** `see deposit/verification/verify_claims_offline.py` **C148** `see deposit/verification/verify_trajectories_offline.py`