R07 verification B — ChatGPT citation set (live web resolution)
Date: 2026-09-10. Verifier: Claude subagent (Fable 5.1). Input: raw-inputs/chatgpt-2026-09-09.md (534 lines; references R1–R31; candidate claims CAND-01–25). Method (PROTOCOL v1.4 two-check): Crossref REST (api.crossref.org/works/<DOI>) and OpenAlex (api.openalex.org/works/https://doi.org/<DOI>) as the two independent registries for every DOI; Europe PMC REST (PMCID / PMID lookup, fullTextXML) for PMC-hosted full text; Semantic Scholar graph API for abstracts the registries did not carry; Crossref query.bibliographic + OpenAlex for the DOI-less R29; publisher pages (Frontiers, actahort.org) and direct HTTP fetches for the four agency URLs. No WebFetch/curl. Both registries agreed on title, authors, venue, year and locator for all 24 DOI-bearing tuples. Every CLAIM verdict rests on a passage retrieved live in this session; nothing is adjudicated from memory.
Verdict keys — IDENTITY: PASS / PARTIAL / FAIL (sub-types: nonexistent DOI · wrong-object DOI · fabricated tuple · author-year misattribution · wording partial). CLAIM: YES / PARTIAL / NO / UNCHECKABLE.
1. Identity table R1–R31
“Self-flag” = ChatGPT marked the entry “verify”/“partial”/“tuple note” in its own reference list (13 entries: R4, R11, R13, R14, R15, R16, R17, R19, R23, R24, R25, R28, R29).
| # | Tuple as written | Resolved tuple (Crossref = OpenAlex unless noted) | IDENTITY | Self-flag | Notes |
|---|---|---|---|---|---|
| R1 | Yep B, Gale NV, Zheng Y. 2020. Comparing hydroponic and aquaponic rootzones… Ind Crops Prod 157:112881. DOI 10.1016/j.indcrop.2020.112881 | Yep, Gale, Zheng 2020, same title, Industrial Crops and Products 157:112881 | PASS | — | Abstract via Semantic Scholar (Crossref/OpenAlex carry none). |
| R2 | Burgel L, Hartung J, Graeff-Hönninger S. 2020. Impact of Different Growing Substrates… Horticulturae 6(4):62. DOI 10.3390/horticulturae6040062 | Burgel, Hartung, Graeff-Hönninger 2020, Horticulturae 6(4):62 | PASS | — | Crossref abstract retrieved. |
| R3 | Schober T, Präger A, Hartung J, Hensmann F, Graeff-Hönninger S. 2023. Growth dynamics and yield formation… Ind Crops Prod 203:117172. DOI 10.1016/j.indcrop.2023.117172 | Schober, Präger, Hartung, Hensmann, Graeff-Hönninger 2023, Industrial Crops and Products 203:117172 | PASS | — | Abstract not in any registry; ScienceDirect 403 (PII S0926669023009378); abstract text obtained from a mirror (colab.ws) — see CLAIM notes. |
| R4 | Ortiz-Delvasto et al. 2024. Aquaporins-mediated water availability… Plant and Soil 495:469–485. DOI 10.1007/s11104-023-06341-8 | Ortiz-Delvasto N, García-Gómez P, Carvajal M, Bárzana G; Plant and Soil 495(1–2):469–485; both registries record 2023 (online); vol 495 issue 1–2 is the 2024 print issue | PASS | yes → resolved | Author list completed: Ortiz-Delvasto, García-Gómez, Carvajal, Bárzana. Year: cite as 2023 (online) / 2024 (issue). |
| R5 | Caplan D, Dixon M, Zheng Y. 2019. Coir-based growing substrates for indoor cannabis production. Acta Horticulturae 1266. DOI 10.17660/ActaHortic.2019.1266.9 | Caplan, Dixon, Zheng 2019, Acta Hortic 1266:55–62 | PASS | — | Pages 55–62 added; abstract from actahort.org (HTTP 200). |
| R6 | Caplan D, Dixon M, Zheng Y. 2017. Optimal Rate of Organic Fertilizer during the Flowering Stage… HortScience 52(12):1796–1803. DOI 10.21273/HORTSCI12401-17 | Same | PASS | — | = S-101 (already ✅ in library). |
| R7 | Caplan D, Dixon M, Zheng Y. 2019. Increasing Inflorescence Dry Weight and Cannabinoid Content… HortScience 54(5):964–969. DOI 10.21273/HORTSCI13510-18 | Same | PASS | — | Crossref abstract retrieved. |
| R8 | Lyu D, Ruan ED, Backer R, Gagné-Bourque F, Smith DL. 2025. Changes in morpho-physiological traits… Ind Crops Prod 224:120377. DOI 10.1016/j.indcrop.2024.120377 | Lyu, Ruan, Backer, Gagné-Bourque, Smith; Industrial Crops and Products 224:120377; Crossref issued 2025, OpenAlex 2024 (online Dec 2024, vol 224 = Feb 2025) | PASS | — | Year 2025 acceptable (issue year). |
| R9 | McGehee CS, Raudales RE. 2021. Pathogenicity and Mefenoxam Sensitivity… Frontiers in Agronomy 3:706138. DOI 10.3389/fagro.2021.706138 | Same | PASS | — | = S-116. Full text from frontiersin.org (200). |
| R10 | Punja ZK, Rodriguez G. 2018. Fusarium and Pythium species infecting roots of hydroponically grown marijuana… Can J Plant Pathol. DOI 10.1080/07060661.2018.1535466 | Punja, Rodriguez 2018, Canadian Journal of Plant Pathology 40(4):498–513 | PASS | — | Volume/pages 40(4):498–513 added. Crossref abstract retrieved. |
| R11 | Allred et al. 2025. Crop steering through osmotic stress can reduce height but reduced yield in medical Cannabis. J Cannabis Res 7:92. DOI 10.1186/s42238-025-00351-2 | Allred J, Fatzinger B, Bugbee B 2025, Journal of Cannabis Research 7(1):92; PMID 41254801; PMC12625080 | PASS | yes → resolved | Full text via Europe PMC. |
| R12 | Yep B, Gale NV, Zheng Y. 2020. Aquaponic and Hydroponic Solutions Modulate NaCl-Induced Stress… Front Plant Sci 11:1169. DOI 10.3389/fpls.2020.01169. PMID 32849724; PMC7424260 | Same; OpenAlex PMID 32849724; Europe PMC PMC7424260 | PASS | — | PMID and PMCID both match. |
| R13 | Crispim Massuela et al. 2026. Less water, same buds… Front Plant Sci 17:1930650. DOI 10.3389/fpls.2026.1930650 | Crispim Massuela D, Munz S, Hartung J, Pieters A, Asch F, Graeff-Hönninger S 2026, Frontiers in Plant Science 17:1930650 (published 03 Sep 2026) | PASS | yes → resolved | = S-182 (already registered). Abstract via Semantic Scholar + frontiersin.org page. |
| R14 | Buser/Büser et al. 2025. Subsurface drip irrigation reduces weed infestation… J Cannabis Res 7:41. DOI 10.1186/s42238-025-00302-x | Büser C, Hartung J, Graeff-Hönninger S 2025, Journal of Cannabis Research 7(1):41; PMID 40640911; PMC12247431 | PASS | yes → resolved | = S-180. Spelling is Büser. Full text via Europe PMC. |
| R15 | [no authors] Environmental and social life cycle assessment of growing media for urban rooftop farming. Int J LCA. DOI 10.1007/s11367-021-01971-5 | Toboso-Chavero S, Madrid-López C, Villalba G, Gabarrell Durany X, Hückstädt AB, Finkbeiner M, Lehmann A 2021, The International Journal of Life Cycle Assessment 26(10):2085–2102 | PASS | yes → resolved | Full tuple completed from registries. |
| R16 | Paoli et al. 2022. Peat production for horticultural use in Latvian context… J Cleaner Prod. DOI 10.1016/j.jclepro.2022.134559 | Paoli R, Feofilovs M, Kamenders A, Romagnoli F 2022, “Peat production for horticultural use in the Latvian context: sustainability assessment through LCA modeling”, Journal of Cleaner Production 378:134559 | PASS | yes → resolved | Title completed; no abstract in any registry (not needed for a claim). |
| R17 | Migoni Alejandre et al. 2026. Life cycle assessment of growing media: a baseline dataset for European horticulture. J Cleaner Prod 575:149269. DOI 10.1016/j.jclepro.2026.149269 | Alejandre EM, Sentinella AT, Ameloot N, Schmilewski G, Weststrate J 2026, Journal of Cleaner Production 575:149269 | PASS (wording note) | yes → resolved | Registries index the first author as “Alejandre, Elizabeth M.” — cite as Alejandre et al., not “Migoni Alejandre”. |
| R18 | Vergara D, Ruterbories KA. 2026. High-THC Cannabis sativa in a New York greenhouse… J Cannabis Res 8:85. DOI 10.1186/s42238-026-00429-5. PMID 42087198; PMC13374136 | Same; OpenAlex PMID 42087198; Europe PMC PMC13374136 | PASS | — | = S-011. Full text via Europe PMC. |
| R19 | Carlile WR et al. 2015. Organic Growing Media: Constituents and Properties. Vadose Zone J. DOI 10.2136/vzj2014.09.0125 | Carlile WR, Cattivello C, Zaccheo P 2015, Vadose Zone Journal 14(6):1–13 | PASS | yes → resolved | = S-107. Crossref abstract retrieved. |
| R20 | Abad M et al. [no year] Physico-chemical and chemical properties of some coconut coir dusts… Bioresource Technology. DOI 10.1016/S0960-8524(01)00189-4. PMID 11991072 | Abad M, Noguera P, Puchades R, Maquieira A, Noguera V 2002, Bioresource Technology 82(3):241–245; OpenAlex PMID 11991072 | PASS | — | Year 2002 and locator added; PMID matches; abstract via Europe PMC. |
| R21 | Verhagen JBG. 1999. CEC and the saturation of the adsorption complex of coir dust. Acta Hortic 481. DOI 10.17660/ActaHortic.1999.481.14 | Verhagen JBGM 1999, Acta Hortic 481:151–158 | PASS | — | Pages added; abstract from actahort.org (200). |
| R22 | Velechovský J, Malík M, Šenkyřík JB, Tlustoš P. 2024. Effect of augmented nutrient composition and fertigation system… Front Plant Sci 15:1322824. DOI 10.3389/fpls.2024.1322824. PMID 38328699; PMC10847352 | Same; OpenAlex PMID 38328699 | PASS | — | = S-205. Full text via Europe PMC. |
| R23 | [no authors] Evaluating Propagation Techniques for Cannabis sativa L. Cultivation: A Comparative Analysis of Soilless Methods and Aeroponic Parameters. 2024. PMC11085668 | Weingarten M, Mattson N, Grab H 2024, Plants (Basel) 13(9):1256; DOI 10.3390/plants13091256; PMID 38732471; PMC11085668 (Europe PMC) | PASS | yes → resolved | Full tuple completed; full text via Europe PMC. |
| R24 | [no authors] Closing the Yield Gap for Cannabis: A Meta-Analysis… Front Plant Sci 2019. PMID 31068957; PMC6491815 | Backer R, Schwinghamer T, Rosenbaum P, McCarty V, Eichhorn Bilodeau S, Lyu D, Ahmed MB, Robinson G, Lefsrud M, Wilkins O, Smith DL 2019, Front Plant Sci 10:495; DOI 10.3389/fpls.2019.00495 (Europe PMC) | PASS | yes → resolved | = S-012. |
| R25 | Punja et al. 2021. Several Pythium species cause crown and root rot on cannabis plants grown under commercial greenhouse conditions. Can J Plant Pathol. DOI 10.1080/07060661.2021.1954695 | Punja ZK, Scott C, Lung S; Canadian Journal of Plant Pathology 44(1):66–81; both registries year 2021 (online; issue 44(1) = 2022) | PASS | yes → resolved | Authors and locator completed; Crossref abstract retrieved. |
| R26 | U.S. NIOSH. NIOSH Pocket Guide — Mineral wool fiber. REL TWA 3 fibers/cm³ and 5 mg/m³ total. cdc.gov/niosh/npg/npgd0432.html | HTTP 200; title “CDC - NIOSH Pocket Guide to Chemical Hazards - Mineral wool fiber” | PASS | — | Page text: “NIOSH REL TWA 3 fibers/cm3 (fibers with diameter ≤3.5 µm & length ≥10 µm) TWA 5 mg/m3 (total); OSHA PEL TWA 15 mg/m3 (total) TWA 5 mg/m3 (resp)”. Figures as stated. |
| R27 | U.S. OSHA. Synthetic Mineral Fibers — Exposure Limits / Health Effects. osha.gov/synthetic-mineral-fibers/exposure-limits and /health-effects | Both HTTP 200 with a browser User-Agent (403 to a default fetch client); titles “Synthetic Mineral Fibers - Exposure Limits…” and “… - Health Effects” | PASS | — | Exposure-limits page lists OSHA PEL-TWA as nuisance dust (5 mg/m³ respirable, 15 mg/m³ total) and ACGIH TLV 1 f/cc for rock wool; health-effects page: rock (stone) wool “not classifiable as to their carcinogenicity to humans”. |
| R28 | [no authors] Elevated root-zone P and nutrient concentration do not increase yield or cannabinoids in medical cannabis. Front Plant Sci 16:1433985 (2025). DOI 10.3389/fpls.2025.1433985 | Hershkowitz J, Westmoreland FM, Bugbee B 2025, Front Plant Sci 16:1433985; PMID 40051879; PMC11882401 | PASS | yes → resolved | = S-104. |
| R29 | [no authors, no DOI] Rethinking Efficiency: How Increased Electricity Use Can Reduce Environmental Impacts in Controlled Hemp Cultivation—A LCA Study. Agronomy 15(10):2400 (2025) | Kalkušová A, Neumann J, Veselovská N, Kůrková E, Konvalina P, Neugschwandtner RW, Bernas J 2025, Agronomy 15(10):2400; DOI 10.3390/agronomy15102400 (Crossref bibliographic search rank 1; OpenAlex by DOI agrees) | PASS | yes → resolved | = S-185. |
| R30 | CDFA. OCal Cannabis Certification Program. cdfa.ca.gov/is/ocal.html (accessed 2026-09-09) | HTTP 200; title “CDFA - IS - OCal Cannabis Certification Program” | PASS | — | = S-215. |
| R31 | CDPH. OCal Registration Program / OCal FAQs. cdph.ca.gov/…/OCal.aspx | HTTP 200 once TLS verification relaxed (site’s chain fails default Node verification — same as R06 #21); title “OCal” | PASS | — | Bundled in S-215. |
Identity totals: 31 PASS, 0 PARTIAL, 0 FAIL (0 nonexistent DOI, 0 wrong-object DOI, 0 fabricated tuple, 0 author-year misattribution, 0 wording partial). All 13 self-flagged entries resolved to real, correctly described objects; the flags were honest and every gap (missing authors, truncated title, missing DOI) was filled from the registries. Wording notes only: R17 first-author form; R4/R8/R25 online-vs-issue year.
2. Claim-support table CAND-01–25
| # | Type | Claim (condensed) | Cited | CLAIM | Basis / quoted passage |
|---|---|---|---|---|---|
| CAND-01 | measured | Three root-zone/nutrient bundles differed significantly; synthetic-fertigated 11-L peat gave 42–116% more inflorescence biomass than aquaculture/aquaponic | R1 | YES | Abstract: “1) an 11-litre pot containing a peat-based growing substrate, top fertigated with a synthetic fertilizer solution (‘hydroponic’); 2) an 11-litre pot … aquaculture effluent solution (‘aquaculture’); 3) a 3-litre pot … lower portion of the roots submerged in a conventional aquaponics deep-water culture system (‘aquaponics’)”; “‘Nordle’ and ‘Sensi-Star’”; “significant rootzone-effects on most growth and physiological traits”; “producing 42–116% greater inflorescence biomass than the other rootzone systems”. Magnitude, cultivars, stage and system descriptors all match. Abstract-level. |
| CAND-02 | methodological | Yep cannot support generic “hydro > soil” because the “hydroponic” treatment was peat in a pot and treatments differed in nutrient source and geometry | R1 | YES | Same abstract passage: the treatment labelled “hydroponic” is an 11-L peat pot; aquaponics is 3-L + DWC; nutrient sources differ. Descriptor reading is correct. |
| CAND-03 | measured | Genotype-specific substrate response: KANADA large floral-yield loss in coir vs peat; 0.2x stable | R2 | YES | Abstract: “KAN indicated a reduced floral yield of plants grown in G30 (4.94 g plant−1) and CC (3.84 g plant−1) compared to PM (8.56 g plant−1). 0.2x indicated stable high floral yields of 9.19 g plant−1 (G30) to 7.90 g plant−1 (CC)”. Indoor pot culture, 36 plants, Latin square. |
| CAND-04 | measured | Rockwool, peat-perlite, coir differed in vegetative growth/partitioning under medium-specific fertigation; CBD concentration insensitive to medium | R3 | YES (abstract-level, mirror copy) | Abstract: “rockwool (RW), peat-perlite (PP), and coconut-coir (CC) … chemotype III medical cannabis strain in pot cultivation, under GM-specific fertigation schedules”; “the effect of GM was negligible for CBD concentration”; “LAI and biomass accumulation of leaves and stems were impaired in PP when compared with RW and CC”; “effect of GM on assimilate partitioning between vegetative plant organs was marginal but significant”; “differences could be mainly explained by the differences in water-holding capacities and irrigation management”. Caveat: abstract retrieved from a third-party mirror, not the publisher (ScienceDirect 403); wording matches the DDG snippet of the publisher page. |
| CAND-05 | measured | Substrate available water was a major correlate of growth/production across five coir/peat/perlite substrates | R4 | YES | Crossref abstract: “cultivated on 5 substrates with different physical–chemical characteristics”; “The substrates available water (AW) was the main factor affecting growth and production”; “S2 and S5, composed by coconut fiber and perlite, showed the lowest growth”; var. Tiborszallasi. |
| CAND-06 | measured | Lower-container-capacity coir formulation gave ~29% greater floral dry weight in the flowering trial | R5 | YES | actahort.org abstract: “U2 having higher CC than U2-HP”; “in the flowering stage, the drier coir-based substrates (U2-HP) may be preferred as it delivered 29% higher yield”; “dry floral weight (yield)”. Walk-in growth chambers; liquid organic fertilizer. Abstract-level; irrigation rule detail not checked. |
| CAND-07 | measured | Fertilizer-rate optima differed between two coir substrates (substrate × fertility interaction) | R6 | YES | S-101 library record (R08 full-text verification): flowering optimum 212–261 mg N/L “substrate-dependent; U2 peaked at highest rate”; Crossref abstract (R06 verify-B #2) confirms two coir-based substrates × five rates, cultivar ‘Wappa’ not named in abstract but recorded at R06/R08. |
| CAND-08 | measured | Drought/crop-steering responses are not directionally uniform across Caplan 2019, Allred 2025, Crispim Massuela 2026 | R7/R11/R13 | YES | R7 abstract: drought “increased concentrations of … THCA and CBDA by 12% and 13%”; THCA yield 43% higher. R11 full text: “Flower yield decreased by 20% in cv. ‘Cherry’”; “There was no difference in cannabinoid concentrations among treatments in either cultivar”. R13 abstract: “Moderate DS reduced water consumption without negatively affecting biomass or CBD yield”; “Severe DS consistently impaired plant performance and provided no agronomic advantage”. Heterogeneity is real; the claim does not pool. |
| CAND-09 | measured | 2025 stone-wool crop-steering study: cultivar-dependent height reduction and ~20% flower-yield penalties under high osmotic stress, no compensating cannabinoid increase | R11 | PARTIAL | Full text: “Plant height was reduced 15% in both the high and hybrid treatments” — height reduction was NOT cultivar-dependent. What was cultivar-dependent is yield: “The higher osmotic stress treatments (8/4 and 8 mS cm−1) resulted in a 20% reduction in flower yield for ‘Cherry’”; “in ‘Trump,’ only the 8/4 mS cm−1 treatment significantly reduced yield, while the 8 mS cm−1 treatment was not different”. Stone wool confirmed (“15 × 15 cm stone wool blocks (Grodan)”); “no difference in cannabinoid concentrations among treatments in either cultivar” confirmed. Correction: “uniform ~15% height reduction; cultivar-dependent ~20% yield penalty”. |
| CAND-10 | measured | Moderate drought reduced water use without reducing biomass or CBD yield in two chemotype III genotypes; severe drought no advantage | R13 | YES | Abstract: “two chemotype III (CBD-dominant) medicinal cannabis genotypes”; “Moderate DS reduced water consumption without negatively affecting biomass or CBD yield”; “Severe DS consistently impaired plant performance and provided no agronomic advantage”. Genotypes Kanada and Terra Italia (page tables). Indoor confirmed by title. |
| CAND-11 | measured | SDI cut irrigation water use 18.6% vs surface drip in an outdoor-tunnel experiment and modestly increased inflorescence yield | R14 | YES | Full text: “SDI resulted in a reduction of irrigation water usage by 18.6% compared to DI” (231.23 vs 188.19 L m−2); “inflorescence yield increased by 5% and CBD concentration by 9%”; “three CBD-rich cannabis chemotype III genotypes (Kanada, Terra Italia, FED) in an outdoor foil tunnel”. “Modestly” = 5%. |
| CAND-12 | measured | Pathogenic oomycete/fungal isolates recovered from commercial cannabis in coir and rockwool; isolates caused disease under inoculation | R9 | YES | Full text: “recovered from coconut coir and rockwool substrates where marijuana (cv. Silver and Citron) plants with root rot and wilt symptoms were grown in a commercial growing facility in Connecticut”; “All the isolates tested were pathogenic to hemp plants in vitro and in vivo, with disease incidence between 6.7 and 100%”. |
| CAND-13 | measured | Hydroponic cannabis develops Fusarium/Pythium root disease; recirculating solution is a plausible spread route | R10 | YES | Crossref abstract: “hydroponically grown C. sativa plants were sampled over a 3-year period (2014–2017)”; “Root rot was found to be caused by … Pythium dissotocum and P. myriotylum”; “The potential for spread of F. oxysporum through the hydroponic system was confirmed by its detection in the recirculating nutrient solution”. |
| CAND-14 | absence | No controlled study comparing root-zone systems on cycle count / days-to-harvest with genotype and environment fixed | — | YES (bounded) | Worded “Within the documented R07 search … through 2026-09-09”; search-stopping section explicitly disclaims ontology. Nothing in the 31 resolved sources contradicts it (R18 varies flowering strategy, not root zone). |
| CAND-15 | absence | No controlled root-zone study with trained sensory panel or blinded consumer test | — | YES (bounded) | Bounded as above; none of R1–R31 reports a sensory panel. |
| CAND-16 | methodological | Chemistry differences do not by themselves establish flavor/preference differences | R1/R2/R3/R8 | YES | Synthesis rule; the cited papers report cannabinoid/terpene concentrations only (R1, R2, R8 abstracts), no sensory endpoint. |
| CAND-17 | general | Coir is chemically/physically variable across sources; broad ranges of salinity, CEC, exchangeable ions | R19–R21 | YES | R20 abstract: 13 coir dusts from Asia, America, Africa; “All properties studied differed significantly between and within sources”; “salinity varied dramatically between 39 and 597 mS m−1”; “cation exchange capacity … ranged from 31.7 to 95.4 cmol(c) kg−1”. R21 abstract: 17 sources; “high contents of exchangeable K, Na, Ca and Mg”. R19 abstract: general review of coir vs peat. |
| CAND-18 | absence | No representative cannabis spent-media mass-flow inventory | — | YES (bounded) | Bounded; none of the resolved LCAs (R15–R17, R29) reports cannabis spent-media mass per unit flower. |
| CAND-19 | modeled | Growing-media rankings sensitive to geography, functional unit and impact category; no universal coir/peat/rockwool hierarchy | R15–R17 | YES | R15 abstract: per-m³ growing media for Spain-imported perlite/peat/coir; “coir-based growing medium exerted the lowest environmental burden in 5 out of 8 impact categories”; peat best socially. R17 abstract: “Single score results show a reversal in ranking for white peat (7.7 mPt/m3) and coir constituents (9.0 to 10.5 mPt/m3), emphasising the need for multi-indicator assessments”. R16 title confirms Latvian-context peat LCA. Note: rockwool is not a constituent in R15/R17 abstracts — the claim’s “rockwool” rests on R29/general framing. |
| CAND-20 | absence | No matched cannabis CAPEX/OPEX/labor/failure-risk dataset across systems with annualized output | — | YES (bounded) | Bounded; R18 is the only economics paper and it does not vary root zone. |
| CAND-21 | measured | 2026 NY greenhouse economics: cycle count, labor and yield assumptions matter; compares autoflower vs photoperiod, not root zone | R18 | YES | Full text: “autoflower planting scenarios with five two-month cycles annually … photoperiod … four three-month cycles”; “Labor represented the largest share of variable costs … 52% … autoflower and 34% … photoperiod”; “Photoperiod cultivation generated higher annual returns per square foot”. Consistent with S-011. |
| CAND-22 | regulatory | California uses OCal as comparable-to-organic; cannabis cannot use USDA organic designation | R30/R31 | YES | CDFA: “OCal is a statewide certification program that establishes and enforces comparable-to-organic cannabis standards”; “Since cannabis is illegal under federal law, cannabis products are not eligible to carry the ‘organic’ designation”. CDPH: “cannabis operations cannot be certified organic under the National Organic Program, therefore the word organic cannot be used to refer to cannabis grown or manufactured in California”. |
| CAND-23 | measured | Organic-derived / biologically mediated nutrition is technically feasible in coir and aquaponic cannabis | R6, R1/R12 | YES | R6 = liquid organic fertilizer in two coir substrates (S-101). R12 full text: aquaponic solution with “nitrifying bacteria … supply of NO3− from organic sources of NH4”; R1 aquaponic treatment grew cannabis to harvest. Feasibility only, as bounded. |
| CAND-24 | measured | Recirculation vs drain-to-waste altered total THC yield / harvest timing without a universal recirculation yield advantage | R22 | YES | Full text: “neither the altered nutritional regime of the ET variant nor the type of fertigation system had a statistically conclusive effect on the final inflorescence harvest mass”; “the recirculation system consistently outperformed the drain-to-waste system, especially after the ninth week, resulting in significantly higher total THC yields”; “we propose the 11th week of cultivation as the suitable harvest time for the recirculation system”. Matches S-205 / R06 #11 correction exactly. |
| CAND-25 | methodological | Index results by substrate class × delivery architecture because “hydroponic” is used for incompatible root environments | R1/R3/R22 | YES | R1 labels an 11-L peat pot “hydroponic”; R3 compares rockwool/peat-perlite/coir as “growing media”; R22 compares recirculation vs drain-to-waste. The taxonomy problem is documented in the sources themselves. |
Claim totals: 24 YES (4 of them bounded absences), 1 PARTIAL (CAND-09), 0 NO, 0 UNCHECKABLE.
3. S-number cross-check and reuse map (rule 11)
ChatGPT’s export uses no S-numbers anywhere (checked: zero S-nnn strings in the 534 lines), so there are no attribution mismatches to report against S-003, S-007, S-096, S-099, S-100, S-101, S-104, S-116, S-134, S-175, S-205, S-208 or S-209. Of those thirteen, only S-101, S-104, S-116 and S-205 are cited (by DOI) in R07; S-003, S-007, S-096, S-099, S-100, S-134, S-175, S-208, S-209 are not used by ChatGPT this cycle. Note: S-099 (Massuela et al. 2023, fpls.2023.1233232) is a different paper from R13 (Crispim Massuela et al. 2026, S-182) — do not conflate. ChatGPT did not repeat the R06 “Cockson” misattribution (S-208 not cited).
Already registered — reuse, do not re-register (12 tuples → 11 S-numbers):
| R# | Reuse |
|---|---|
| R6 | S-101 |
| R9 | S-116 |
| R13 | S-182 |
| R14 | S-180 |
| R18 | S-011 |
| R19 | S-107 |
| R22 | S-205 |
| R24 | S-012 |
| R28 | S-104 |
| R29 | S-185 |
| R30, R31 | S-215 (CDFA + CDPH bundled) |
New registrations needed (19), with the resolved tuples above: R1, R2, R3, R4, R5, R7, R8, R10, R11, R12, R15, R16, R17, R20, R21, R23, R25, R26, R27. Burgel 2020 (R2) appears in a refuted-claims row (Gemini R08) but has no S-number yet.
4. Editor-caught notes
- CAND-09 descriptor swap — height reduction was uniform (15%, both cultivars); the cultivar-dependent effect was yield (Cherry −20% in both high-EC treatments; Trump −20% only in the hybrid 8/4 treatment). Correct before ledger entry.
- R17 author form — registries give “Alejandre EM”; “Migoni Alejandre” is not the indexed family name. Cite as Alejandre et al. 2026.
- R4 year — Crossref/OpenAlex issue date 2023 (online); print issue 495(1–2) is 2024. Either is defensible; record both.
- R8 year — Crossref 2025 / OpenAlex 2024 (online Dec 2024, vol 224 Feb 2025). ChatGPT’s 2025 is the issue year.
- R25 year — online 2021, issue 44(1) 2022. ChatGPT’s 2021 is acceptable; add “44(1):66–81”.
- R3 abstract provenance — obtained from a mirror site (ScienceDirect blocks non-browser clients); if the ledger quotes R3 numerically, re-confirm against the publisher PDF.
- CAND-19 “rockwool” — R15 and R17 abstracts cover perlite/peat/coir/wood fibre/compost, not rockwool; the rockwool leg of the “no universal hierarchy” statement is not supported by those two abstracts (R29 hemp LCA may carry it; not checked at full-text level).
- CAND-06 / CAND-07 / CAND-01 / CAND-04 / CAND-05 / CAND-13 are abstract-level (HortScience, Acta Hortic, Elsevier, Taylor & Francis paywalls). The claimed figures are stated verbatim in the abstracts, so YES stands; finer numbers would need full text.
- R27 and R31 fetch conditions — OSHA returns 403 to non-browser user agents; CDPH’s TLS chain fails default Node verification (same as R06 #21). Both are live and match the stated content.
- Honesty record — all 13 self-flags were warranted-but-resolvable (missing author lists, a truncated title, a missing DOI); none concealed a wrong object. ChatGPT also correctly left R14’s spelling open rather than guessing (it is Büser).
5. Summary
- Items: 31. IDENTITY: 31 PASS, 0 PARTIAL, 0 FAIL (0 nonexistent DOI · 0 wrong-object DOI · 0 fabricated tuple · 0 author-year misattribution · 0 wording partial). Both registries agreed on every DOI; all four PMIDs (32849724, 42087198, 11991072, 31068957) and all four PMCIDs (PMC7424260, PMC13374136, PMC11085668, PMC6491815) map to the stated papers; the DOI-less R29 resolved to a unique Crossref/OpenAlex object (10.3390/agronomy15102400).
- Self-flagged partials: 13 of 13 resolved.
- CLAIM: 24 YES (20 measured/methodological/regulatory + 4 bounded absences), 1 PARTIAL (CAND-09), 0 NO, 0 UNCHECKABLE.
6. Recommended disposition
- Proceed to ledger as written: CAND-01, 02, 03, 04, 05, 06, 07, 08, 10, 11, 12, 13, 14, 15, 16, 17, 18, 20, 21, 22, 23, 24, 25 (23 claims). CAND-19 proceeds with the rockwool qualifier in note 7.
- Needs correction before ledger: CAND-09 — restate as “uniform ~15% height reduction in both cultivars; ~20% flower-yield penalty that was cultivar-dependent (consistent in ‘Cherry’, only in the hybrid treatment in ‘Trump’); no cannabinoid-concentration change”.
- Tuple corrections for registration: R17 first-author form; add resolved authors/locators for R4, R5, R10, R11, R13, R14, R15, R16, R20, R21, R23, R25, R28, R29 as tabulated.
- ChatGPT identity record this cycle: 0 failures (31/31 PASS). This is the first R-cycle in the library with zero identity failures from a ChatGPT export; the R06 wrong-object (#16 “Cockson”) pattern did not recur.