Content
Citation(s) & Footnote(s)
Book(s)Whiskey Knowledge Databases › Literature Notes
Main article reviewed; full-source completion reopened September27 because five supplementary tables and two supplementary figures are missing. Expanded independent assessment and exact recovery gap follow below.
Scope and core summary
Complete synthesis of a 23-page controlled study testing two barley varieties grown in two Irish environments over two seasons and distilled at laboratory scale.
Author argument
The authors report that variety, environment, season, and their interactions affected sensory attributes and aroma-active volatiles, with environment and variety-by-environment interaction often stronger than variety alone.
Researcher synthesis
This is evidence that place-linked agricultural variation can survive into new make when process conditions are controlled. It does not prove that consumers can identify place in matured whiskey or that ‘terroir’ overrides fermentation, distillation, and cask effects.
Evidence assessment
Primary controlled study combining GC-MS-olfactometry, descriptive analysis, and chemometrics.
Limitations and open questions
Limitations: Only two varieties, two sites, two seasons, and laboratory spirit; no maturation stage; researchers included distillery and malt-supplier affiliates. Generalization and commercial claims require independent replication.
Open questions: Do the reported site signals persist, change, or disappear after standardized American-style barrel maturation and dilution?
Connected records
- Contributors: 6
- Verified Excerpts: 3
- Citations: 1
- Zettels: 4
Completion record
September27 independent review covers all23 supplied main-article pages,64reference entries,Figure1 andTables1–4. Five supplementary tables and two supplementary figures remain unread and inaccessible; the former unconditional completion claim is withdrawn.
Independent review and reopened supplement gap — September 27, 2026
All 23 supplied PDF sheets, including all 64 reference entries, the complete 42-row Table1, Tables2–4 and Figure1, have now been independently read. Table and figure renders on sheets7–9,14,16,18 were inspected. The original is preserved. The earlier unconditional completion claim is reopened: five supplementary tables and two supplementary figures are explicitly cited on sheet20 but absent from this PDF and its existing Source attachments. Recovery found the public PMC supplementary PDF link; it returned403 through the retrieval tool and an HTML response rather than PDF through direct retrieval. Publisher access also failed. This gap does not prevent reading other accessible library sources.
Contribution and experimental boundary
Kyraleou et al. (Foods2021,10,443; DOI10.3390/foods10020443) compare Olympus and Laureate barley at Athy and Bunclody in 2017 and2018. This is useful primary evidence of differences associated with growing conditions in laboratory unaged spirit. Both barley varieties share Quench ancestry, so the smaller variety effect is specific to this comparison. It does not establish a general hierarchy in which geography outweighs genetics.
The field trials are described as randomized complete blocks, but the number of blocks, independent field lots carried through malting/distillation, pooling and exact statistical replication are insufficiently reported in the main text. The coordinates change between years, and sowing dates shift from March2017 to lateApril2018. Season therefore does not mean weather alone; field and calendar changes need to remain visible. Soil association, nutrients, climate and management vary together. No single soil mineral was experimentally manipulated, and no causal soil-to-aroma mechanism was isolated.
Micro-malting used16kg samples and mashing approximately10kg, with two water additions, 72-hour fermentation at32.5°C and a laboratory spirit still. The main text gives an upper spirit strength of76%ABV but omits detailed cuts, recovery volumes, batch run order and full distillation parameters. Standardized stated conditions are a strength, not proof that all processing variation was eliminated. Waterford and Boortmalt employees participated and the commercial collaboration is disclosed.
What the sensory measurements mean
Six GC-O panelists were selected from nine volunteers after ten training sessions, including adapted odor vocabulary and a greater-than12/16 identification screen. Training and standards improve consistency but also define this panel's perceptual language. Different descriptors for the same molecule are not necessarily chemical contradictions.
GC-O used5mL spirit diluted to20%ABV, headspace SPME at40°C for40minutes and three-way MS/FID/olfactory detection. Modified frequency is the square root of detection-frequency percentage times intensity percentage; it is neither percent composition nor percent contribution to whole-drink flavor. Higher MF does not itself establish hedonic quality. FD values come from injector split-ratio dilution, not a calibrated concentration or odor-activity ratio. Changing the split after extraction does not eliminate all headspace-matrix effects.
A separate experienced six-person panel rated15 descriptive attributes on0–5 scales. The main text does not clearly specify its serving ABV, whether each attribute was assessed by sniffing or oral tasting, complete presentation randomization or independent repeat counts. Do not transfer the GC-O20% dilution automatically to this panel. The term oily finish alone does not resolve the missing tasting procedure. Neither panel was a consumer preference or geographic-identification trial.
Main findings with precise limits
Table3 reports stronger Athy malty/biscuity, feinty/earthy and oily-finish means and stronger Bunclody floral/fresh-fruit/dried-fruit means across the two years. Many other attributes do not differ significantly; this is not a complete sensory fingerprint or a blind origin classifier. Olympus had stronger pungent/fresh-fruit main effects within this narrow design. Interaction claims must be assessed from individual cells rather than assuming a main effect holds within every site.
Table1 separates identified, tentatively identified and ten unidentified odor regions. Two key high-MF regions remain unknown. The paper's strength is combining chemical separation with human detection: abundant chromatographic signals need not dominate aroma. Even so, separated-compound detection cannot reconstruct mixture masking/synergy without recombination/omission experiments. Associations in Table4 do not demonstrate that a molecule causes a whole-spirit descriptor, and pooled season/site differences can drive correlations.
The sensory ANOVA reports Bonferroni adjustment and the MF comparisons Tukey HSD. The main text does not adequately define the corrected families or show a corresponding multiplicity correction for the many Pearson correlations. Missing uncertainty intervals, denominator/sample-size detail and full raw observations limit reanalysis. Repeated analytical runs and panelists are not independent agricultural replicates.
Errors and discrepancies to preserve in reuse
- Page6 says ethyl decanoate was excluded from Table1 for negligible odor impact, yet Table1 includes it atMF39.6. Page11 acknowledges it again; do not repeat the categorical exclusion.
- Page13 calls the medium-MF group fifteen compounds but enumerates eighteen. Table1 supports eight above80 and eighteen between50 and80. The conclusion says eight most influential but lists only six. Keep counts tied to an explicit threshold and table.
- Page11 attributes season/environment effects to ethyl butyrate although it is absent from the affected-compounds Table2. Page12 says heptanal changes by season, while both seasonal means in Table2 have the same letter. These remain unresolved.
- Page17's claim that Olympus is more pungent at both sites in both years is contradicted by the 2018 Bunclody means: Olympus2.07 and Laureate2.21, both letterc. A significant interaction is not universal within-site superiority.
- Table3's 2017 Laureate pungent main mean1.85 differs from the simple average of its two listed cells1.85 and1.56. In2018 oily finish, Athy main mean2.15 conflicts with two cells both1.97; the listed variety means also fail to reconcile. These may involve errors or unexplained model weighting; do not silently recalculate published effects.
- Table4 prints2-butylfuran/stale-mouldy r=0.077 with a significance star despite the small experimental design; treat it as suspect, not proof. Unknown8/sour r=0.784 has two stars while several larger correlations have one; the observation basis and flags require reconciliation. Page19 reverses a p-sign in one significant-correlation description.
- Figure1 uses broken vertical scales and averages of discrete FD steps; compare labeled values rather than raw bar heights. It contains FD values below1, although the methods assign splitless detection1 and never explicitly define nondetection handling. The large nonenal150 and damascenone75 values are detection-dilution results, not concentrations.
- The data-availability statement retains a placeholder mdpi.com/xxx link. The soil-pH discussion cites reference60, a whisky-flavor review, rather than an explicit soil-method source. Mechanistic soil claims need independent corroboration.
Proposed Academy use and linked synthesis
Use a paired reading with 
Proposed exercises: compare a chromatogram to an aromagram; distinguish chemical abundance, MF, FD, descriptive intensity and liking; and check whether a main effect survives examination of an interaction table. Link malting review
Exact outstanding material
Recover and read S1 soil nutrients, S2 weather, S3 adapted Sniffin’Sticks descriptions, S4 GC-O odor vocabulary, S5 standard-mixture recipes, S1 chromatogram and all eight S2 treatment/year aromagrams. Until then, supplement coverage is incomplete and no full-source completion claim is warranted. Author data and model details remain a separate reproducibility issue, not proof that unavailable raw data were reviewed. Proton byte identity and browser rendering are unverified.
