NEOBIOME

LIT_123: Rawidech (2024) Romney lamb carcass dissection: NZ dissectible-fat coefficient

Properties5
typesource
source_typeliterature
authorRawidech R
year2024
topicsfood_systems_agriculture, nz_specific

10 min read

LIT_123: Rawidech (2024) Romney lamb carcass dissection: NZ dissectible-fat coefficient

Source

https://mro.massey.ac.nz/items/d57c77af-5b59-46dc-baae-21ae5e83385b — original source (opens in a new tab; the file is not redistributed)

Rawidech (2024) Romney vs ¾ Wiltshire lamb carcass dissection (NZ dissectible-fat coefficient)

Resolves RT_394 at the LAMB tier: the NZ separable-fat coefficient for the animal-fat leg of the D02 edible-fat model

Massey University MSc (Animal Science) thesis, 77 pp. Pasture-finished NZ ram lambs (Romney n=11, ¾ Wiltshire n=12), whole-leg physical dissection. The load-bearing figure for Neobiome: dissectible (physically separable) fat = 13.69% ± 0.69 for Romney, 11.16% ± 0.56 for ¾ Wiltshire (Table 4.1, mean ± SEM, HCW-adjusted). Verified verbatim against the thesis PDF (pdftotext -layout). ⚠ Two caveats the model MUST carry. (1) The 13.69% is dissectible fat as a proportion of the whole LEG, used by the author as “indicative of the overall carcass composition” (leg-dissection method, Purchas et al. 1991) — it is a carcass-composition proxy, not a whole-carcass measurement. (2) This is SEPARABLE fat, NOT edible-rendered tallow. The model’s edible-fat output needs a further construction (× HCW × ~0.78 render recovery); that render fraction, the national mean HCW, and the beef/pig percentages are external construction inputs, not this source’s claims (see NI relevance).

Summary

Rawisara Rawidech’s 2024 Massey University Master of Science (Animal Science) thesis compares carcass and meat-quality characteristics of Romney and three-quarter (¾) Wiltshire ram lambs, pasture-finished under identical management and slaughtered as one group. It is a NZ primary source: the animals were raised and slaughtered in New Zealand and the dissection was carried out at Massey University’s Te Ohu Rangahau Kai facility, Palmerston North. Carcass composition (muscle %, dissectible fat %, muscle-to-bone ratio, muscularity) was obtained by physical dissection of the whole hind leg into its eight muscles, with the total dissectible fat weighed to give a dissectible-fat percentage of the leg; intramuscular fat (IMF) was measured separately in the loin. Hot carcass weight (HCW) was fitted as a covariate so breeds are compared at equal carcass weight.

For Neobiome the value is a single figure: the NZ dissectible-fat percentage of a pasture-finished lamb carcass (Romney 13.69%, the crossbred ¾ Wiltshire lower at 11.16%). This is the “separable-fat %” leg of the animal-fat pathway in the D02 food model (RT_394), the NZ-sourced anchor that lets the engine estimate how much cooking fat a community’s meat lambs can physically yield, before a rendering-recovery step converts separable fat to edible tallow. Beef and pig separable-fat percentages are NOT in this source and remain flagged construction proxies (see NI relevance and Research targets). The meat-quality and bone-morphology results (pH, colour, shear force, fatty-acid profile, femur pQCT) are the thesis’s main scientific contribution but are not model-relevant to Neobiome.

Key claims

- claim: "Dissectible (physically separable) fat percentage by breed — the model-relevant NZ coefficient. VERBATIM (Table 4.1, 'mean ± SEM', row 'Dissectible fat (%)'): Romney (n=9) = 13.69 ± 0.69; ¾ Wiltshire (n=12) = 11.16 ± 0.56; P-value Breed = 0.029; P-value HCW covariate = 0.006. The table caption: 'Table 4.1. Carcass characteristics (mean ± SEM) for Romney and ¾ Wiltshire lambs. Hot carcass weight (HCW) was fitted as a covariate in the statistical model for muscle, dissectible fat, M:B, intramuscular fat (IMF)...'. So the percentages are standard-error (not SD) values compared at equal hot carcass weight; the crossbred ¾ Wiltshire is significantly leaner. [verified verbatim, pdftotext -layout]"
  source_location: "Table 4.1, Chapter 4 Results §4.1 'Carcass Characteristics' (printed p.38)"
- claim: "Definition/basis of the dissectible-fat percentage — it is a WHOLE-LEG figure, not a whole-carcass figure. VERBATIM (§3.3 Dissection of Lamb Leg): 'Each leg was dissected into its eight muscles: Gracilis, Sartorius, Pectineus, Semimembranosus, Adductor, Bicep femoris, Semitendinosus and Quadriceps femoris. Individual muscles were weighed and recorded. Muscle percentage was expressed as a proportion of total muscle weight to the whole leg weight. The total dissectible fat was obtained from the whole leg to calculate the dissectible fat percentage.' So 13.69% is dissectible fat ÷ whole-leg weight (n=21 legs dissected). [verified verbatim, pdftotext -layout]"
  source_location: "§3.3 'Dissection of Lamb Leg and Carcass Characteristics Analysis' (printed p.32)"
- claim: "The leg dissection is a carcass-composition PROXY, not a whole-carcass measurement — the author's own stated limitation. VERBATIM (Discussion, study limitations): 'The dissection of muscles to analyse carcass characteristics was conducted on the leg, which might not accurately represent the muscle and fat compositions of the entire carcass. Despite this, the use of the leg dissection method (Purchas et al., 1991) ensures that the results are indicative of the overall carcass composition. An ideal dissection would include a whole carcass or a side (half) of the carcass for more precise results. However, due to practical constraints of time and resources, this was not feasible.' Applying the leg dissectible-fat % as a whole-carcass separable-fat fraction is therefore an indicative approximation the source explicitly supports but does not itself perform. [verified verbatim, pdftotext -layout]"
  source_location: "Chapter 5 Discussion, study-limitations paragraph (printed p.46)"
- claim: "Dissectible fat = the two physically separable depots (subcutaneous + intermuscular); intramuscular fat (IMF) is measured and reported separately. VERBATIM (§2.2.3.2 Fat Partitioning): 'As an animal grows, they deposit fat in three different carcass fat depots orderly: subcutaneous, intermuscular, and intramuscular depots (Warris, 2010). ... Subcutaneous fat locates under the skin ... The intermuscular fat is found between muscles while intramuscular fat (IMF) lies within the muscle (Johnson et al., 2016).' The dissectible fat physically removed at dissection is the subcutaneous + intermuscular depots; IMF is reported separately in Table 4.1 (Romney 1.74 ± 0.34%, ¾ Wiltshire 2.51 ± 0.34%, breed P=0.211). This matches the model's 'separable fat' definition, which excludes intramuscular fat. [verified verbatim, pdftotext -layout]"
  source_location: "§2.2.3.2 'Fat Partitioning', Literature Review (printed p.19) + Table 4.1 IMF row (printed p.38)"
- claim: "Study design and the study's own carcass weights (context for the construction). VERBATIM (Abstract): 'Data were collected from Romney (n=11) and ¾ Wiltshire (n=12) ram lambs, managed and fed on pasture under the same conditions, and sent to slaughter, as one group when all lambs reached a minimum of 42 kg liveweight.' Table 4.1 hot carcass weights: Romney (n=9, two excluded for pleurisy) = 17.0 ± 0.7 kg; ¾ Wiltshire (n=12) = 21.8 ± 0.6 kg (breed P<0.0001); dressing-out 37.6 ± 0.5% vs 42.3 ± 0.4%. These are pasture-finished NZ ram lambs; the study's Romney HCW (17.0 kg) is close to but not the national mean (the ~19.4 kg B+LNZ figure used in the construction is external to this source). [verified verbatim, pdftotext -layout]"
  source_location: "Abstract (printed p.1) + Table 4.1 HCW / Dressing-out rows (printed p.38)"

Neobiome Intelligence relevance

Resolves RT_394 at the LAMB tier and feeds the animal-fat pathway (④f) of d02_smart_food_systems. RT_394 (“NZ rendered tallow / lard yield per carcass”) was CR_055’s first declared negative, carrying the instruction “Do not ingest as NZ coefficient yet”, held at verification because no NZ source publishes rendered fat yield per carcass directly. This source closes the lamb leg by supplying the one thing that WAS missing and IS NZ-sourced: the separable-fat percentage of a pasture-finished NZ lamb carcass (Romney 13.69%, ¾ Wiltshire 11.16%). The engine can now estimate whether a community’s meat lambs can supply cooking fat, which was previously unanswerable in either direction.

The construction the model needs (state it explicitly — most inputs are NOT this source’s claims):

edible rendered tallow (kg/carcass)
    ≈ HCW (kg)  ×  separable-fat fraction  ×  rendering-recovery fraction
 
LAMB worked example (Romney):
    HCW              = 19.4 kg     ← EXTERNAL: B+LNZ 2022-23 NZ mean lamb HCW (NOT this source; this study's Romney HCW was 17.0 kg)
    separable-fat    = 0.137       ← THIS SOURCE (13.69% dissectible fat, leg-basis used as carcass proxy)
    → separable fat  ≈ 19.4 × 0.137 ≈ 2.7 kg   (separable, NOT yet edible-rendered)
    render recovery  = ~0.78       ← EXTERNAL construction input (NOT this source; no NZ figure — residual gap)
    → edible tallow  ≈ 2.7 × 0.78 ≈ ~2.1 kg/carcass

Every non-lamb-fraction input above is a flagged construction assumption, NOT a claim of this source. The ~0.78 rendering-recovery multiplier, the ~19.4 kg national mean HCW (B+LNZ 2022-23), and especially the beef ~20% (NZ grass-fed, adjusted DOWN from a North-American 27-29% dissection proxy) and pig ~22% separable-fat percentages are construction inputs carried from the batch RETRIEVAL_REPORT, corroborated only by non-NZ support files (USDA ERS, AHDB pork, MPI dressing-out, NRA rendered-fats). They must be presented in the engine as clearly-flagged proxies, never as verified NZ figures. Only the lamb 13.69% is a verbatim NZ value from a retrieved primary.

Two caveats stack on the lamb figure itself (be conservative in the engine):

  1. Leg → carcass is a proxy. The 13.69% is dissectible fat ÷ whole-leg weight, which the author states is “indicative of the overall carcass composition” but “might not accurately represent … the entire carcass”. Treating it as a whole-carcass separable-fat fraction is an indicative approximation, so carry the lamb tallow output as a ballpark, not a precise coefficient.
  2. Breed spread is real and large. The crossbred ¾ Wiltshire is a full ~2.5 percentage points leaner (11.16% vs 13.69%, breed P=0.029). A shedding-breed flock (increasingly common in NZ low-input systems) yields materially less separable fat, so the coefficient should be a range (~11-14%), not a point, and ideally breed-selectable.

Residual gap (documented, not a new RT): no NZ source publishes edible-rendered fat per carcass directly, so the ~0.78 render step and the beef/pig percentages stay as construction proxies. Per the retrieval report this is treated as a known-unclosable gap rather than an actionable retrieval target (see Research targets).

Research targets

Resolved

  • RT_394 (gap — NZ rendered tallow/lard yield per carcass) — RESOLVED at the LAMB tier by this source. The NZ separable-fat coefficient for lamb (Romney 13.69% ± 0.69, ¾ Wiltshire 11.16% ± 0.56) is now sourced verbatim from an NZ primary; combined with the model’s rendering-recovery construction it answers whether a community’s meat lambs can supply cooking fat. Residual (documented, no new RT opened): beef (~20%, NZ grass-fed adjusted from a N-American proxy) and pig (~22%) separable-fat percentages remain flagged construction proxies, and no NZ source publishes edible-rendered fat per carcass directly (the ~0.78 render fraction stays a construction input). Per the batch retrieval report and minimise-new-RTs, these residuals are recorded here as a known gap rather than reopened as a fresh target.

Documents to retrieve

  • None.

Research gaps

  • No new gap opened. (The beef/pig-proxy and edible-rendered residual is captured in the RT_394 resolution note above rather than as a speculative new RT.)

Notes

  • NZ primary, data_quality: high. The raw is the genuine Massey University MSc (Animal Science) thesis (77 pp), downloaded as an authoritative PDF; the headline figure was re-verified verbatim against the thesis PDF (pdftotext -layout) at draft. high rather than verified because it is a single NZ dissection study (n=9 Romney) with no second NZ source corroborating the 13.69%; high rather than medium because it is the retrieved primary itself, not an AI synthesis, and its figure is read verbatim.
  • Capture format: directly-retrieved authoritative PDF (first-party Massey University research-repository thesis), so no ## Retrieval provenance block and no upstream_source/retrieved_via frontmatter — the file is the authoritative document, not an AI extraction. Retrieved in the rt_cost_recovery_2026-07-25 batch (RETRIEVAL_REPORT.md, RT_394); the report rates RT_394 overall “MODERATE” because it is a construction, but the lamb value it anchors is clean.
  • Page citations are the thesis’s own printed page numbers (footers). Named tables/sections (Table 4.1, §3.3, §2.2.3.2) make each claim unambiguous regardless of the PDF-viewer page offset.
  • Filed to source-files/01_literature/ (single-file, renamed to the page stem per L26). sha256 recorded in frontmatter; recompute after move.
  • Not model-relevant (excluded): meat-quality results (pH, colour, water-holding capacity, shear force, fatty-acid profile) and femur pQCT bone morphology — the thesis’s main scientific contribution, but outside the NI food model. IMF (1.74-2.51%) is recorded only to fix the boundary of “dissectible” fat, not as a model input.
  • cr-needs-a-primary: this thesis is the retrievable NZ primary anchoring the RT_394 lamb leg of the CR_055 dairy/fats compiled research; that linkage is noted on CR_055, not restated here.

Connections

Referenced by

EDT domains (1): D02: Smart Food Systems & Agriculture

Graph View

Backlinks