2010 10 14_repro

School of Veterinary Medicine and Science

Can we use milk recording data to predict reproduction?

An improvement on the fat to protein ratio

A. Madouasse, J.N. Huxley, W.J. Browne, A.J. Bradley, M.J. Green


Background

• Start of lactation:– Massive increase in energy demand– Limited appetiteNegative Energy Balance

• 41.5 days (de Vries and Veerkamp, 2000)• Up to 8 weeks for most cows (Heuer et al, 2000)


Background

• Poor reproductive performance in dairy cows:– Later resumption of luteal activity in cows in

NEB (Heuer et al, 1999)

– Milder oestrus expression with increasing milk yield (Cutullic et al, 2009)

– Negative correlation between milk yield and calving interval (Haile-Mariam et al, 2003)

– Genetic selection on yield and composition has been accompanied by poorer reproduction (Lucy, 2001)


Background

• Link between energy balance and milk composition at the start of lactation

• NEB associated with:– Increase in milk butterfat percentage (de Vries and Veerkamp, 2000)– Decrease in milk protein percentage (Duffield, 1997)

– Increase in the fat to protein ratio (Grieve, 1986)


Question

Can we predict the calving to conception interval from milk

recording data?


Materials

• Selection from NMR data collected in 2004-2006:

– 8,211,988 individual cow recordings– 992,625 lactations– 483,474 cows– 2,128 herds


What do we want to look at?

Milk yieldButterfat (% or kg)Protein (% or kg)Fat to Protein ratioSomatic cell count


When?

• First 2 test-days of lactation


How ?

• Do you really want to know?


Outcome: calving to conception interval

• Calculated from recalving dates• Gestation length of 280 days assumed

• Intervals categorised as:– [20-60] ; [61-81] ; [82-102] ; [103-123] ; [124-

144]

Calving 1 Calving 2Conception280 days

Calving to Conception

Interval


Cumulative Probability of Conception per Interval


How does it work?

Herd Lactation Interval Conception Milk 1 Milk 2 …

1 1 1 0 33.2 47.2 …

1 1 2 0 33.2 47.2 …

1 1 3 0 33.2 47.2 …

1 1 4 0 33.2 47.2 …

1 1 5 0 33.2 47.2 …

1 2 1 0 32 39 …

1 2 2 1 32 39 …

1 3 1 1 33.6 36.6 …

1 4 1 0 38.8 41.2 …

… … … … … … …


Can we include raw data in our model?

• Variation in the percentage of butterfat



• Variation in the percentage of protein



• Variation in the fat to protein ratio


Model Results

Variable Test-day Corrected β Standard Error Odds-Ratios

Interval 1 -2.527 0.039 0.080

Interval 2 -1.876 0.027 0.153

Interval 3 -1.556 0.024 0.211

Interval 4 -1.574 0.022 0.207

Interval 5 -1.636 0.024 0.195

Milk (kg) 2 dim -0.075 0.010 0.928

Milk (kg)2 2 dim -0.023 0.006 0.977

Protein 2 dim 0.152 0.018 1.164

Protein2 2 dim -0.010 0.005 0.990

Protein * ln(Interval) 2 dim -0.061 0.005 0.941

Lactose 1 dim + seas 0.092 0.018 1.096

Lactose2 1 dim + seas 0.004 0.002 1.004

Lactose * ln(Interval) 1 dim + seas -0.048 0.016 0.953

Protein 1 dim 0.041 0.010 1.042

Protein2 1 dim -0.014 0.005 0.986

Cell count 2 dim -0.040 0.010 0.961

Cell count 1 dim -0.025 0.010 0.975

Butterfat 1 dim + seas -0.023 0.009 0.977


Does the model work?


Model Predicted Effects


Interlude

The parable of the jet assisted take-off

Darwin award, 1995


Interlude

– A US sergeant had fitted a jet-assisted take-off unit onto his car

– Went to a long straight road in the Arizona desert

– Put the gas on


Interlude

– The vehicle quickly reached a speed of between 250 and 300 mph

– Continued at that speed, under full power, for an additional 20-25 seconds


Interlude

– The car remained on the straight highway for approximately 2.6 miles

– The driver applied the brakes• Completely melting them• Blowing the tires• Leaving thick rubber marks on the road surface


Interlude

– The vehicle then became airborne for an additional 1.3 miles

– Impacted a cliff face at a height of 125 feet

– Left a blackened crater 3 feet deep in the rock


Interlude

– The Arizona Highway Patrol were mystified when they arrived at the site

– The folks in the lab tried to figure it out

Blackened crater 3 feet deep in the

rock

Impact on a cliff face at a height of 125

feet

Rubber marks on the road

Straight road in the

Arizona desert


End of the interlude


Have we fitted jet engines on our cows?

Poor reproduction

Changes in milk constituents

Ketosis

Negative energy balance

Uncoupling of the somatotropic axis

Displaced abomasum

Impaired immunity

Delayed uterine involution

Insulin resistance

Lameness

Low T3/T4

Steatosis

Early lactation


Have we fitted jet engines on our cows?

• Cows selected on milk production for decades

Selected for yield and constituents Much of the cow resources directed toward milk

production as opposed to reproduction and other functions

Range of metabolic perturbations


What happens?

• Massive quantity of energy mobilised at the start of lactationMore milk More energy

• Fat mobilised to get energy– Fat present in milk

• But – Fat percentage influenced by time of the year

So is the fat to protein ratio

Fat and fat to protein ratio unreliable as predictors of conception


What happens?

• Percentages of protein and lactose have much stronger associations with the probability of conception

– Probably because of endocrine perturbations (Insulin ???)

– They are less dependent on season


Conclusion


Take home!

• Cows producing more milk at the peak synthesise less lactose and protein and mobilise more fat at the start of lactation

• These cows: – Could be more likely to get mastitis and other

health disorders– Conceive later … or not!


Acknowledgments

This work was funded by the University of Nottingham

Thanks to the National Milk Records for providing the data


Thanks to…SVMS Population Health

Cattle• James Breen• Marnie Brennan• Andrew Bradley• Tracey Coffey• Richard Emes• Martin Green• Chris Hudson• James Husband

(Special Lecturer)• Jon Huxley• Jasmeet Kaler• Nigel Kendall• Jamie Leigh• Wendela Wapenaar

Special Profs / Lecturers:

Sheep• Jasmeet Kaler

Dairy Herd Health Group:

Horses• John Burford• Sarah Freeman• Emma Shipman

Dairy Herd Health Group:

Post graduate researchers

• Simon Archer• David Black• Peter Down• Helen Higgins• Chris Hudson• Aurélien Madouasse• Sarah Potterton• Adam Spencer• Jenny Wills• Paula Wilson

[email protected]://aurelienmadouasse.wordpress.com/

mailto:[email protected]

2010 10 14_repro

Documents