The paper
Improving marbling by genetics
- Author
- D. Johnston
- Published
- 2002
- In
- Busselton Weaner Production Workshop
- Collection
- Busselton Weaner Workshop
- Listed on the old library
- 8 March 2012
We know of no online copy of this paper today. A university or state library that holds the Busselton Weaner Production Workshop is the place to ask.

In this summary
- Why marbling matters to producers and markets
- Measuring marble score for improving marbling in beef cattle
- Differences between and within breeds
- Building the IMF% EBV: scanning and abattoir data
- Genetic relationships with other carcase traits
- Putting a value on marbling and reading the EBV table
- A marbling gene test and where the science still has gaps
- Sources and further reading
- Questions
Why marbling matters to producers and markets#
Johnston opens by explaining that marbling, as a sign of intramuscular fat, has become a major talking point in the Australian beef industry even though its full effect on eating quality is still not completely worked out. The paper states that marbling adds to flavour and juiciness and has a limited effect on tenderness. It also notes that Meat Standards Australia includes marble score as one of the inputs in its model for predicting palatability, with higher marble scores producing higher predicted palatability.
Beyond the plate, the paper points to the commercial weight marbling carries in grading systems overseas. It explains that both the USA and Japanese grading schemes reward higher marbling with higher prices per kilogram, reflecting demand in those markets, and that marbling also affects how carcases are classified for the restaurant trade domestically. Johnston frames this as the reason the Australian industry needs a practical way of producing the level of marbling particular markets are prepared to pay for.
The paper is careful to note that marbling outcomes depend on more than one lever. It states that manipulating the environment, such as through feedlot finishing, and improving the genetics of the animal are strategies that work together rather than as alternatives. That framing sets up the rest of the paper, which concentrates on the genetic side of the equation while acknowledging that feeding and management also shape the final result, a topic covered in a related summary on nutrition and management effects on marbling.
Measuring marble score for improving marbling in beef cattle#
According to the paper, the relationship between the visual marble score assigned to a carcase and the actual percentage of intramuscular fat in the muscle is not perfect, with a figure of around 70 percent relationship described as typical. Both measures are heritable, in the range of 28 to 40 percent, but Johnston stresses that the particular method used to score or measure marbling changes the heritability estimate obtained, which in turn affects how useful that measure is within a genetic evaluation system.
The paper also lists non-genetic influences on marbling, including management before and after slaughter, the age and weight of the animal, its nutrition, and the length of time it spends on feed. Johnston draws a clear distinction here: money spent achieving marbling through management has to be spent again each time, whereas genetic gains, once made, persist in the herd. This is presented as the reason producers look for sires whose progeny reliably marble well.
Three broad genetic routes are described for lifting marbling: switching to a breed that marbles more heavily, crossbreeding, or selecting superior sires within an existing breed. The paper also flags that molecular genetics research has identified genes thought to have a large individual effect on marbling, introducing the gene-testing material covered later in the paper, and sits alongside related genetic work summarised in a separate review of what drives fat deposition within muscle.
Differences between and within breeds#
The paper reports considerable breed differences in mean marble score. Breeds named as tending higher include Wagyu, Jersey and Chianina, while Limousin, Angus, Murray Grey, Shorthorn, South Devon and Brahman are grouped as comparatively lower, based on the table in the paper. Johnston cautions that the actual difference achieved in practice will still depend on the genetics involved, the production environment, management, and the market weight targeted, so breed averages are a starting point rather than a guarantee.
Even where a breed sits at the lower end of that list, the paper notes that considerable variation exists between sires within every breed, and that some sires within the higher marbling breeds still produce relatively poor marbling progeny. This is the basis for the paper's central argument: that within-breed selection, not just breed choice, is where ongoing improvement has to come from if seedstock herds and their commercial customers want consistent gains.
The paper connects this within-breed variation to the introduction of an estimated breeding value for intramuscular fat, the IMF% EBV, into the national BREEDPLAN evaluation run by breed societies. Johnston describes this as an important development because it gives breeders a formal tool for identifying and selecting the better-marbling animals within their own breed, rather than relying on breed averages alone.

Building the IMF% EBV: scanning and abattoir data#
The paper explains that the IMF% EBV draws on live animal ultrasound scanning of intramuscular fat together with abattoir carcase data, which itself includes chemically extracted intramuscular fat records from the Beef CRC and MSA marble scores collected from industry progeny tests, plus marbling EPDs from overseas evaluations. Australian scanning research covered several thousand animals from seedstock herds of three breeds, Angus, Hereford and Santa Gertrudis, with a heritability estimate for intramuscular fat percentage of 25 percent in heifers but only 15 percent in bulls. The equivalent US estimate for yearling bulls was 37 percent, a difference the paper attributes to US bulls generally carrying more fat at scanning time than Australian bulls.
A finding Johnston describes as now confirmed in both countries is that the fat picked up by scanning young breeding stock is genetically linked in a positive direction to the fat those same genetics later produce in steers taken through to finishing. This matters commercially because it means scanning data gathered on young seedstock bulls and heifers can be used to predict how their progeny will marble once finished, well before any carcase data exists.
The paper sets out the practical case for ultrasound scanning: it is cheaper than collecting abattoir data, reduces the time lag before data become available, and allows more animals to be measured, which increases selection pressure. It notes tens of thousands of animals already carry scanned intramuscular fat records on breed society databases. The trade-off is that scanning has a lower heritability than the chemically measured trait, so more data are needed for the same accuracy, and the paper states scanning is good enough for ranking sires genetically but not precise enough for judging an individual animal.
Genetic relationships with other carcase traits#
The paper reports several specific genetic parameters from this work. Heritability of IMF% is given as 0.38 in both temperate and tropically adapted breeds, while the paper states that temperate breeds carry about twice the genetic variance for this trait compared with the tropically adapted group. IMF% is negatively genetically correlated with retail beef yield percentage, at -0.40, and positively correlated with carcase fat depth, at 0.2 to 0.3. It is also described as very highly positively correlated with marble score itself and positively correlated with eating quality traits.
Drawing on the Beef CRC's collection of chemically extracted intramuscular fat records from over 7,500 carcases, the paper reports that, measured on a standard 300 kg carcase, the genetic relationship between IMF% and retail beef yield comes out slightly negative, meaning selection purely for more intramuscular fat tends to nudge retail yield down. Johnston's response to this is not to avoid selecting for marbling, but to point out that EBVs exist for both traits, so breeders can identify sires that lift IMF% without excessive cost to yield.
Non-genetic results from the same CRC dataset are also reported: heavier, Japanese-finished carcases had more IMF% than lighter, domestic-weight carcases; grain-finished cattle had higher IMF% than pasture-finished cattle; and heifers carried somewhat more intramuscular fat than steers finished to the same live weight. The paper also reports very high genetic correlations for IMF% expression across markets (0.92-1.0), across grain and pasture finishing systems (0.96-1.0), and across temperate versus subtropical finishing regions for tropically adapted breeds (0.94), which the paper interprets as evidence that the genes controlling IMF% behave similarly regardless of the production system.
Putting a value on marbling and reading the EBV table#
Because IMF% is genetically linked to other carcase traits, the paper argues that selection decisions cannot sensibly be based on marbling alone; the weight given to marbling in a breeding objective should reflect its economic value alongside other profit traits. Johnston points to work by Barwick and Henzell, cited for a detailed treatment of valuing marbling in Australian breeding objectives, which found that the economic value of improved marbling changes depending on a herd's starting average marble score and on which production and market system is being supplied. That cited work found marbling carried its greatest relative economic value of all profit traits in a particular self-replacing breeding operation aimed at a long-fed export market in Japan.
The paper also explains a change in how carcase EBVs are presented: they are now adjusted to a standard 300 kg carcase weight endpoint rather than the earlier fixed age of 450 days, a change intended to better reflect sire differences at a commercial endpoint and to allow abattoir data from both steers and heifers to be used.
A worked example in the paper compares two named Angus sires from the January 2002 Angus BREEDPLAN run, one carrying an IMF% EBV of plus 1.8 and the other at 0.4. The paper calculates that the higher sire's progeny would be expected to carry about 0.7% more intramuscular fat than the lower sire's progeny, worked out on the same 300 kg carcase basis under matching conditions, which comes close to one third of a single AUS-MEAT marble score step, while the lower sire's progeny were predicted to have a larger eye muscle, leaner carcases, and higher retail beef yield. Johnston is explicit that a negative EBV does not mean an animal marbles poorly in absolute terms, only that it sits below the current breed mean, which itself can shift as a breed makes genetic progress.

A marbling gene test and where the science still has gaps#
The paper describes the discovery, by Bill Barendse and colleagues at CSIRO in Brisbane with funding from Meat and Livestock Australia, of a gene association with marble score in long-fed Angus, Shorthorn and Wagyu cattle fed for 200 days. The gene in question affects thyroglobulin, which the paper describes as the body's storage form for two thyroid hormones that influence how fat cells grow and develop. The paper notes that levels of those hormones have been tied to intramuscular fat deposition in Wagyu cattle through work done by Canadian researchers, but states plainly that the biological pathway connecting the thyroglobulin gene, the hormones and marbling score is not understood.
GeneSTAR Marbling, the commercial DNA test for this gene sold by a Brisbane-based livestock genetics company, had been on the market for more than 12 months at the time of writing and had already found users among cattle breeders across several overseas beef-producing countries as well as in Australia. The paper reports that long-fed Angus, Shorthorn and Wagyu cattle carrying two copies of the favourable form of the gene reached AUS-MEAT marble scores of 4 and 5 twice as often as animals carrying zero or one copy.
Johnston is clear about the test's current limits: it is not suitable for drafting cattle at a feedlot, and its value to a breeder depends on how common the favourable gene form already is in their particular breed. The paper describes work underway to combine such gene test results with EBVs and indexes. We would treat the headline doubling of top marble scores as an early, breed- and feeding-regime-specific result rather than a figure that automatically transfers to every herd or finishing system, a caution worth bearing in mind alongside the wider discussion collected in the Busselton Weaner Production Workshop proceedings.
Sources and further reading#
- Trove library search: find a library that holds the paper
- Meat & Livestock Australia: red meat industry research and marketing body
- BREEDPLAN genetic evaluation: genetic evaluation system for Australian beef cattle
- Animal Genetics and Breeding Unit
Questions#
Is marble score the same thing as intramuscular fat percentage?
Not exactly. The paper describes marble score as a visual assessment that relates to actual intramuscular fat percentage at around a 70 percent relationship, so the two measures track together closely but are not identical. Both are heritable, in the paper's figures between 28 and 40 percent, but the method used to measure or score marbling changes the heritability estimate obtained.
Can choosing a higher-marbling breed guarantee better marbling?
The paper lists breeds such as Wagyu, Jersey and Chianina as tending to score higher and others such as Limousin, Angus, Murray Grey, Shorthorn, South Devon and Brahman as tending lower, but it stresses that considerable variation exists between sires within every breed, including the higher-marbling ones. Actual results also depend on environment, management and market weight, so breed choice alone is not a guarantee.
Does selecting for marbling cost yield or tenderness?
The paper reports a negative genetic correlation of -0.40 between IMF% and retail beef yield percentage, and a positive correlation of 0.2 to 0.3 with carcase fat depth, meaning selecting hard for marbling alone tends to reduce yield and add fat cover. It also reports a positive genetic correlation between IMF% and eating quality traits, and recommends using EBVs for multiple traits together rather than selecting on marbling in isolation.
What does the GeneSTAR marbling test actually tell a breeder?
According to the paper, animals with two copies of the favourable thyroglobulin gene form reached AUS-MEAT marble scores of 4 and 5 twice as often as animals with zero or one copy, in long-fed Angus, Shorthorn and Wagyu cattle. The paper notes the test is not yet suited to drafting cattle at a feedlot and that its usefulness depends on how frequent that gene form already is in a given breed.
Written by the Livestock Library team from the published paper by D. Johnston (2002), and released on 2 October 2026. It is our account of the research in our own words, not the paper itself. For anything you plan to act on, read the original.
Are you an author or publisher of this paper? If something here is wrong, or you would like it changed or removed, tell us and we will fix it.