Meat Quality Grading TodayPaul Allen

What is meat quality?

• Can mean different things at each point in the supplychain

• Consumer is ultimate arbiter – if it doesn’t satisfyexpectations then they may not purchase again

• At point of sale – appearance is important – colour,fatness, marbling, lack of drip, packaging

• After cooking – tenderness, juiciness, flavour, overallexperience

• Can be assessed by trained panels – more descriptorscan be used

• Can be assessed by untrained consumers - largenumbers

Why grade on palatability?

• Inconsistent eating quality –AUS, US, IRL

• Beef consumption declining

• Increased competition

• EUROP grading unrelatedto palatability

• Consumers willing to payfor quality

Current meat quality grading Most countries DO NOT grade on eating quality

Notable exceptions are US and Australia

US – Carcasses graded on yield and quality

Quality grade is based on visual assessment of marbling (loin) and maturity(hand held camera systems can be used)

Marbling is the amount and distribution of visible flecks of fat within the eyemuscle at 12th/13th rib

Marbling is primary factor in determining quality grade

Maturity (physiological age) is assessed visually

Degree of ossification of cartilage on vertebrae and spinal processes, colourof bones

Colour and texture (fineness of grain) of loin muscle (less emphasis thanossification)

All these are combined to give an overall quality grade – Prime, Choice, Select

Australia - Measures to improve tenderness known but not interactions - alsobased on expert panels not consumers

MSA solution – predictive model using PACCP approach

The PACCP approach

Conception

Consumption

Consumerfeedback

GeneticsNutrition

Pre-slaughter factors

Post-slaughter factors

Chilling/ageing

Processing

Cooking

Packaging

Critical Control Points

MSA grading

Assess effect of pre and postmortem factors to producepredictive model

Effects measured as responseof consumers

Large database – 65,000consumers, 420,000 samples

Very detailed protocols forsampling, cooking etc.

Cuts based model

• Original model graded carcasses

• Became clear that cuts weredifferent – can’t predictpalatability of all cuts by gradingcarcass as cuts responddifferently to various factorsparticularly ageing, carcasssuspension and cooking method

• Therefore developed cuts basedmodel – palatability of individualcuts predicted for range ofcooking methods

Factors in the model Predictors

Breed (BI)

sex

growth rate

Electrical stimulation

hanging method

Marbling

Ossification

ageing

cooking method

pH

rib fat

Basic criteria

minimum stress

• Thresholds for

– ossification score

– pHu

– colour

– rib fat

Components of palatability

Combination of all factors that make beefenjoyable to eat, assessed by sensoryanalysis and weighted to give quality score

Main factors are

tenderness x 0.4

juiciness x 0.1

flavour x 0.2

overall liking x 0.3

= Meat Quality Score

Consumer grades

Ungraded 3 star 4 star 5 star

Grindingbeef

Everydayquality

Better thanEveryday

Premiumquality

MeatQualityScore

< 48 48 - 63 64 - 79 80+

Format Name Input ?Aged cut muscle GRL RST SFR TSL

% or X if doubt EPBI 0 spinalis SPN081 77 67 77 73

M/F Sex M tenderloin TDR034 82 76

Y or ? / N HGP N tenderloin TDR062 78 77 80 74

Y/N MFV N tenderloin TDR063 73

Y/N

SlYrd

N cube roll CUB045 69 69 69 71

striploin STA045 65 65 67 67

Y/N RnFl N striploin STP045 63 64 66 67

Weight in Kg

HSCW

268 oyster blade OYS036 63 60 66 69

AT/TS/TL/TC/TX

Hang

TX?

blade BLD095 43

blade BLD096 56 59 60 61

mm

Hump

45 chucktender CTR085 49 52 54

USDA measure uoss 140 rump RMP131 60 68 65 71

USDA measure umb 220 rump RMP231 63 71 70 69

mm RbFt 5 rump RMP005 64 72 72

Metered pH UpH 5.58 rump RMP032 71 75

Metered Temp C

Utmp

3 rump RMP087 59 64 62

knuckle KNU066 53 66 61 65

Days Aged Age 14 knuckle KNU098 61 66

Testing the MSA model

• Funding from DAFM – FIRMprogramme

• AUS – Ireland comparison

• Irish commercial sample

• Experiments to test factors

• Ageing

• Stimulation –LVES andHVES

• Breed

• Hanging method

• Boning time –24 v 48h

AUS – IRL comparison

“Matched” set of samplesfrom Ireland and AUS – 18carcasses from each country,6 muscles from each carcass,2 cooking methods

AUS samples tasted by AUSand Irish consumers

Irish samples tasted by Irishconsumers

Compare responses of AUSand Irish consumers

Test fit of model to Irish beefand Irish consumers

Consumer testing Individual muscles removed and

trimmed

Samples prepared and frozen

Cooked in standard way (grilled,roasted, yakiniku) to medium done

Groups of 20 consumers (60 forroasts) – social clubs, sports clubs,charities etc.

Rate for tenderness, juiciness,flavour, overall like

Assign quality category = stars

0

10

20

30

40

50

60

70

80

90

100

Pala

tability

score

s

Tenderness

Juiciness

Flavour

Overall

Relationship between palatabilityscores and quality category

unsatisfactory good everyday better than everyday premium

Scores for all palatability attributes increased with quality category

% of cuts falling in quality categories

0%

20%

40%

60%

80%

100%

unsatisfactory good

everyday

better than

everyday

premium

Beef

Cut

f illet

striploin

rump

blade

outside round

round

Considerable variability in quality for striploin, rump and round

IU YIU GIT YIT GIS YIS GIR YIR GIO YIO GIBYIBG

40

30

20

10

0

-10

-20

-30

-40

0

Iris h R e s idua ls (Ir-M )

Irish consumers v model

Deviations from model significant only for grilled striploin andYakiniku topside

Irish v AUS consumers

U YU GT YT GS YS GR YR GO YO GB YB G

40

30

20

10

0

-10

-20

-30

-40

0

D iffe re nc e s (Ir-A u)

Deviations significant only for Yakiniku, rump and tenderloin

Ageing and stimulation –Effect on MQS

0

10

20

30

40

50

60

70

80

90

LVES NON LVES NON

14 days 28 days

Striploin

Topside

Outside

At 28 days LVES tended to improve MQS of striploin but reducedMQS for outside. Significant negative effect of ageing on OR.

Overall conclusions

• Irish beef fits model at least as well as AUS beef

• Model fits Irish consumers at least as well as AUS

Irish consumers score beef in similar way to AUSconsumers, but not identical (Irish more weight onflavour) and model may need optimising

Model tested over wide range of factors with moderatelylarge database - over 1100 samples

Accounts for different factors reasonably well in mostcircumstances

Some exceptions may be due to electrical inputs on linenot accounted for

Success of MSA in AUS

Number of carcasses graded annually

0

100

200

300

400

500

600

700

99/00 00/01 01/02 02/03 03/04 04/05 05/06

Year

No.ofcarc

asses

MSA grading pays

Average prices ($/kg, Real Dec'05)

15

16

17

18

19

20

21

Nov-04 Dec-04 Jan-05 Feb-05 M ar -05 M ar -05 Apr -05 M ay-05 Jun-05 Jul -05 Aug-05 Aug-05 Sep-05

MSA NonMSALinear (MSA) Linear (NonMSA)

What’s the future for meat qualitygrading?

USDA model not appropriate since grading occurs atquartering

Rapid methods (such as NIR) have promise but arealso most likely to be applied at quartering

MSA predictive model could be adopted

May not be optimised for Irish beef

MSA model also tested in NI, France, Poland –international effort to derive a European model

Could include age, breed etc. from ID, genetics, NIR,images of loin etc.

Thank youfor

listening!