Original: Fri 11 Oct 2013 | Frank Burden

24 Nov 2013: By request - added postscript and Braunvieh, Charolais, Santa Gertrudis results.

Royal Adelaide Show


Analysis of results from the recent 2013 Royal Adelaide Show led steer on hoof and on hook competition provided evidence that Limousin has maintained its dressing percentage advantage over all other breeds, but also identified inconsistencies between ultra sound scan and abattoir carcase measurements, especially for eye muscle area.

Of the 180 entries, key breeds represented were Angus/Red Angus (40), Hereford/Poll Hereford (22), Murray Grey (22), Limousin (18), Shorthorn (17), Red Poll (13), Simmental (12), South Devon (10), Santa Gertrudis (9), Charolais (5), and Braunvieh (5). Other breeds were represented by less than 5 entries each.

Dressing Percentage

Average dressing percentages (carcase weight measured at the abattoir divided by liveweight measured at the show) are listed below. Dressing percentages based on empty liveweight should be increased by about 3% (assuming 5% gut fill):

Interestingly, two Murray Grey entries were reported to have the highest and lowest dressing percentages of 66.5% and 52.5% respectively!

Carcase Measurements

scan abattoir ribfatscan abattoir p8scan abattoir ema

Ultra sound scan and direct measurements of rib and P8 fat, and of eye muscle area, are shown in the scatter plots to the right.

Each point on the graphs represents the combined scan/abattoir measurements for each animal. Ideally scan and abattoir measurements for each animal should be equal and appear on the diagonal line. Errors or inconsistencies in either or both measurements will cause a point to deviate from the line, with points above the line representing over-reading by ultra sound, under-reading at the abattoir, or a combination of both.

Rib fat measurements are distributed within about +/- 3mm of the line. Although P8 fat has a similar distribution, there is a slight bias to below the line (ie under-reading ultra sound scan, over-reading abattoir measurement, or a combination of both).

In two cases, P8 and Rib Fat scan measurements exceeded abattoir measurements by 200% (3mm scanned compared with 1mm measured in the abattoir). Where fat depths were greater, scan measurements deviated from abattoir measurements by typically less than 50%.

Eye muscle area measurements showed a significant bias to average about 8.5cm2 above the line, suggesting that either ultra sound scanning exaggerated eye muscle areas, abattoir measurements under-estimated the areas, or a combination of both. In general, scan measurements were confined to the range +18cm2 and -5cm2 of the abattoir measurements.

Generally most eye muscle area inconsistencies advantaged Angus on the hoof, although Santa Gertrudis with a smaller number of entries (9 compared with 40 Angus/Red Angus) had the largest average inconsistency. Angus/Red Angus scan areas exceeded abattoir measurements by an average of just over 11%; the greatest inconsistency was close to 36% (scanned 80cm2 but 59cm2 measured in the abattoir).

Santa Gertrudis scanned on average 12.9% greater than measured in the abattoir (greatest inconsistency was 28.8%), Charolais was on average 9.9% greater, Murray Grey 9.3% greater, Simmental 7.8% greater, Braunvieh 7.4% greater, Hereford 7% greater, Shorthorn 6.7% greater, Red Poll and South Devon 5.8% greater, and Limousin 5% greater.

Comments are welcome.



The review was initiated by a show committee member who argues that judging on the hoof should not be influenced by scan results because of the observed inconsistencies.

James (see below) suggests that inconsistencies should be expected because the measurement methods are different and also that shrinkage caused by loss of condition will occur between the show ring and abattoir. The results indicate that shrinkage in eye muscle area, if this is the principal cause of the inconsistencies, appears to be greatest in Santa Gertrudis and Angus and smallest in Limousin.

An alternative or additional explanation provided for the inconsistencies is that scan and abattoir measurements were made at slightly different positions on the carcase.

Finally, according to a NSW DPI article titled Objective measurement in beef cattle showring judging:

"Eye muscle area (EMA) can also be measured by ultrasound although not as accurately as subcutaneous fat, hence measurements on individuals need to be interpreted with caution if being used by judges. The EMA scan should be related to the fat-corrected liveweight."


Sat 2 November 2013 | Ben

I breed angus and just come across the article on Adelaide Show steer results. Confirms what I've suspected for a long time. With such large errors in actual measurements of real cattle why should we believe predictions made by breedplan based on scan measurements alone?

Sat 2 November 2013 | Frank Burden

Good question. If the discrepancies are typical and caused primarily by scanning, then this would mean that initial carcase EBVs for cattle based only on scan measurements could be significantly in error. In practice, the above discrepancies are likely to be a product of both scan and abattoir measurement errors. However, it is unclear if the Adelaide Show discrepancies are normal.

Prediction error margins for EBVs are represented by their accuracy figures, and initial EBVs based on scan data alone are expected to have significant error margins. Although not obvious from EBV accuracy figures, initially typically around 50%, prediction error margins are actually only just better (about 10% smaller) than if the EBV was guessed!

As scan measurements from an animal's relatives are received, errors tend to cancel out and EBVs are adjusted up or down as predictions become more precise. Some stud or commercial producers observing large adjustments to an EBV then quote this as evidence that BREEDPLAN cannot be trusted, when the EBV's accuracy figure in fact indicates that large adjustments should always be expected. EBV error margins remain large up to accuracy figures of about 95%, but they are still about one-third those for EBVs based on a guess. In other words, EBVs should be considered to be inaccurate for the vast majority of stud cattle.

In my view, such misunderstandings are caused by BREEDPLAN's misleading method of presenting EBVs and their accuracies, which imply high precision when the opposite is usually the case. BREEDPLAN's "accuracy" figures distort and grossly exaggerate the true accuracy of EBVs.

EBVs and corresponding direct measurements of carcase traits (as demonstrated by the Show results) should always be treated as imprecise indicators of genetic merit that are subject to significant error and change.

I hope this answers Ben's question.

Sun 3 November 2013 | Ben

Thanks Frank. Your explanation confirms what I suspected.

Thu 21 November 2013 | James

It should not be surprising there is a difference between what is scanned and what is measured on a carcass as they are two different methods. Furthermore slaughter did not occur directly after scanning and there is natural variation in loss of condition, fat cover, marbling and muscle consistency between animals.

Scanning in bulls at 12 months of age is a substitute for what we really want to know 'the performance of steers suited to a particular market'. The scanning of heifers is much closer to reality, the importance of which is commonly overlooked.

As I see it the main thing that everyone seems to have overlooked is the inclusion of actual carcass data in Breedplan! All this information is available through the NLIS, as grading for MSA or Ausmeat are recorded with the NLIS. But this requires more working with commercial producers, something Stud Breed Societies forget about because they don't currently only derive revenue from registered cattle. The biggest question for breed societies is 'how to be relevant' to the beef industry?

You no longer need to be a member of a breed society to maintain pedigrees!

Post a comment

Click here.

Comments are moderated and will appear as soon as they have been approved. When posting comments you agree to be bound by our Terms and Conditions.