VISION2000:

Upgrade of the Multiple Trait Prediction Method (MTP)

 

Context

Since the implementation of VISION2000, many clients have noticed that the projections for some cows were overestimated.

Two situations explain this overestimation:

1: Some cows had similar milk weights for two or more tests (ex: 50 kg on test 1, and 50 kg on test 2). Also, some cows had their highest daily production on the 3^rd or 4^th test of the lactation. MTP estimates the shape of the lactation curve based on the individual animal’s data and the reference curve. In the above mentioned cases, the data from the animal was more important than the reference curve and MTP projected the lactation curve as a flat line or with an upward slope. The overall effect was to get high and often overestimated projections.

2: Other cows had very high production on their first test. Production would then rapidly decline. Also, some cows had no observations in the ascending phase of the curve. In these situations, MTP determined that the peak of the lactation occurred at day 5, with a very high peak milk production (up to 100 kg and more, in some cases). In these cases, we would see an inverted lactation curve lacking an ascending phase.

Solution

A solution was proposed by Dr. Schaeffer and Dr. Moore. The modification puts more emphasis on the reference curve. Concretely, MTP now estimates the animal’s production at 5 days in milk using her first test production and the appropriate reference curve. This value is used throughout the lactation. The overall effect is to generate an ascending phase in the lactation curve. This modification greatly diminishes the tendency towards an inverted curve with very high peak milk production occurring at day 5 of lactation.

Also, based on each new test for the animal (the latest available test), MTP uses the production and the reference curve to establish the production level that the cow will have at day 305. This modification generates a descending phase, thus avoiding flat line lactation curves with high projections.

In this document, the modification is called MTP ++. Its overall effect is illustrated on the following page.

Before

Now

Results

The following two examples present situations where the cow maintains a fairly steady or slightly increasing production from one test to the next. The corresponding projections from MTP and MTP ++ are shown.

Cow n^o 1 (37 months cow)

Days in Milk	Kg of milk	MTP Projection	MTP ++ Projection
27	45.5	11,150	10,908
56	43.2	11,696	10,719
91	47.2	14,111	11,627
118	42.8	12,985	11,574

Cow n^o 2 (26 months cow)

Days in Milk	Kg of milk	MTP Projection	MTP ++ Projection
15	31.5	8,144	8,957
44	29.4	7,643	7,829
79	33.8	10,261	8,632
106	34.2	10,829	8,985





In these two cases, the MTP ++ method gives results that are closer to what is expected, and appear to be more stable.

The two following examples illustrate the case of a cow with a high volume of milk at the first test and a rapid decline afterwards.




Cow n^o 3

Days in Milk	Kg of milk	MTP Projection	MTP ++ Projection
60	55.4	16,739	14,927
88	45.4	13,244	13,393
115	42.4	13,045	12,985
Day of lactation peak		5	40
Amount of milk at peak ( kg)		97,5	49,8




Cow n^o 4

Days in Milk	Kg of milk	MTP Projection	MTP ++ Projection
49	49.5	13,196	11,857
77	40.1	10,589	10,791
114	31.7	8,970	9,918
143	35.5	11,139	10,370
177	31.3	11,408	10,285
Day of lactation peak		5	5
Amount of milk at peak ( kg)		100,9	47,3





In these two situations, projections are less variable and the estimation of the lactation peak production is at an expected level.

The above mentioned cases are situations in which production occurs in a manner different from what would normally be expected. In cases where a cow’s production behaves in the usual way, results obtained by MTP ++ are not much different from results by MTP.

Cow Profile

When reviewing the cow’s profile, the differences between the shape of the curves become evident (next graphs represent the same cow). The cumulative production varies due to the fact that the MTP estimation at the onset of lactation was clearly overestimated. With MTP, the peak occurred at day 5 with 61 kg of milk. With MTP ++, the peak of 38 kg of milk is on day 30. In both cases, the lactation persistency is similar.






Implementation

Following an analysis of the results for projections as well as for cumulative production, milk recording agencies throughout Canada decided to implement MTP ++ on April 3, 2000.

Transition from MTP to MTP ++

The month of April is a transition period between the two methods. It is possible that some cows will have large differences in their projections. In large part, this will affect those cows whose production followed a pattern similar to that previously described.

Also, cumulative production for the cows will be computed according to the new method. Remember that the MTP approach uses every test after calving, and this each time that we compute the cumulative production on a test day. So, in the extreme cases of cows that declined rapidly in production after the first test, it may be possible to have cows with cumulative production lower than it was the month before. However, this should be limited to a small number of cows.

Advantages of MTP ++

MTP ++ brings :

• More accurate and stable projections
• Better estimation of the cow’s lactation peak
• Easier to understand cow profiles
• Better consistency between cumulative production and projections
• More accurate and more stable current averages by parity, and for the herd

* * *

* * *