Source
doi:10.3109/09637489809086412 — original publication (opens in a new tab; the file is not redistributed)
Savage & McNeil (1998) — Composition of NZ-grown hazelnuts
The NZ hazelnut oil content — and the basis statement that makes it un-multipliable against the yield figures
International Journal of Food Science and Nutrition 49(3):199-203 · DOI 10.3109/09637489809086412. Six cultivars from 12-year-old trees at the Lincoln University experimental orchard, harvested autumn 1995. Verbatim: “The total oil content of the hazelnuts ranged from 54.6 to 63.2%”; protein 14.3–18.2%; dietary fibre 9.8–13.2%. 🔴 These are KERNEL-basis percentages — the paper’s phrase “the remaining portion of the kernel” is what establishes the basis. They therefore cannot be multiplied by OT_182’s per-hectare yields, whose in-shell-vs-kernel basis is undeclared (RT_393). ⚠ Abstract only — the publisher’s full text is paywalled.
Summary
A short compositional study of hazelnuts grown in New Zealand: total oil, protein and dietary fibre across six cultivars from mature (12-year-old) trees at Lincoln University’s experimental orchard, harvested in autumn 1995. It is the only NZ-grown hazelnut oil-content measurement in corpus.
Its role is precise and limited. It supplies the oil-content leg of the hazelnut pathway, and — through a single phrase — it supplies the basis of that leg, which is what turns a seemingly complete hazelnut calculation into a blocked one. The yield leg (OT_182: 1 t/ha by year 6, 2.5 t/ha by year 10) does not state its basis. This paper’s does. Multiplying an undeclared yield by a kernel-basis percentage risks roughly a 2× overstatement, so the two cannot be combined until the crack-out fraction is sourced (RT_393).
Key claims
- claim: "NZ-GROWN HAZELNUT COMPOSITION, verbatim: 'The total oil content of the hazelnuts ranged from 54.6 to 63.2%.' Protein ranged 14.3-18.2%; dietary fibre 9.8-13.2%. [verified verbatim against the source abstract]"
source_location: "Savage GP & McNeil DL (1998), Int. J. Food Sci. Nutr. 49(3):199-203, DOI 10.3109/09637489809086412 — abstract, compositional results."
- claim: "TRIAL BASIS: six cultivars, 12-year-old trees, Lincoln University experimental orchard, harvested autumn 1995. [verified verbatim against the source abstract]"
source_location: "Savage GP & McNeil DL (1998), Int. J. Food Sci. Nutr. 49(3):199-203 — abstract, materials."
- claim: "🔴 THESE ARE KERNEL-BASIS PERCENTAGES. The paper's phrase 'the remaining portion of the kernel' indicates the composition is stated on a kernel basis, NOT an in-shell basis. This is what prevents the oil % being multiplied by OT_182's per-hectare yields, whose basis is undeclared. [verified verbatim against the source abstract]"
source_location: "Savage GP & McNeil DL (1998), Int. J. Food Sci. Nutr. 49(3):199-203 — abstract, phrasing 'the remaining portion of the kernel'."Neobiome Intelligence relevance
Feeds d02_smart_food_systems as the oil-content leg of the hazelnut pathway — correct, NZ-measured, and currently unusable in combination with the yield leg.
- 54.6–63.2% oil is high, well above oilseed rape’s 42% (LIT_095) and comparable to NZ-grown walnut’s 62.4–68.7% (LIT_097). On a per-kilogram-of-kernel basis, nuts are far denser oil sources than oilseeds.
- But the denominator is the problem, not the numerator. Hazelnuts sit at ”⚠ blocked on a unit” precisely because this number is right and the yield’s basis is missing. See RT_393.
- The protein (14.3–18.2%) and fibre (9.8–13.2%) figures are recorded for completeness. Hazelnut is not a plant-protein candidate for the model — the anchor there stays faba bean (OT_088) — but the numbers are the NZ-measured values should a whole-food nutritional treatment ever need them.
No engine cell is auto-changed by this ingest.
Notes
- Abstract only. The publisher’s full text is paywalled; the raw is an HTML capture of the abstract record.
data_quality: mediumreflects that the figures are the authors’ own reported ranges but the methods and per-cultivar detail were not read. - The single most load-bearing word on this page is “kernel”. It is the basis statement, and it is the reason the hazelnut oil-per-hectare figure does not exist. Preserve it in any restatement.
- 1998 publication on a 1995 harvest — old, but composition is a far more stable property than yield (compare the olive census’s >2× swing in kg/tree against its stable 13.4–14.8% oil, OT_181). Vintage is a weak objection to a compositional figure.
- Lincoln University experimental orchard — the same institution as the NZ oilseed rape trial (LIT_095) and the NZ walnut trial (LIT_097). Much of NZ’s oil-crop evidence base sits in one place.
Research targets
Documents to retrieve
- (none new) — the crack-out / basis blocker is raised as RT_393 at OT_182 and is not duplicated here.
Research gaps
- Not raised as research targets (recorded here instead):
- The full text of this paper. Paywalled; the abstract carries the figures and the basis statement, which is all the model needs. Retrieving it would add per-cultivar resolution, not a new capability.
- A more recent NZ hazelnut composition study. Composition is stable enough that this is low value.
Connections
Links to
Referenced by