Sensitivity Analysis of Milk2000 for Measuring Corn Hybrid Forage Performance
July 22, 2004 11(19):124-126
Joe Lauer, Corn Agronomist
Milk2000 is based upon National Research Council (National Research Council, 2001)
guidelines for dairy nutrition. Milk2000 is used as a tool for ranking corn hybrids
based upon standard forage analysis measurements. Farmers and agronomists have inquired
about the sensitivity of Milk2000 and the changes that can occur to milk per ton
and milk per acre when yield, moisture, NDFD and starch content are changed.
To get an idea for the amount of change that can occur among hybrids within a trial,
data from the University of Wisconsin corn performance trials conducted between
1995 and 2003 were summarized. Table 1 reports average forage quality measurements
needed to calculate milk per ton and milk per acre using Milk2000. An average hybrid
from these trials yields 7.91 tons dry matter per acre and has forage moisture of
62.3%, NDFD of 61.6% and starch content of 31.4%. To calculate the typical range
of these measurements in the real world, I multiplied the standard deviation of
the mean by 1x and 2x to represent 67 and 95% of the hybrids tested. Values were
adjusted one at a time between these ranges to provide a sensitivity analysis on
milk per ton and milk per acre (Table 2). The Milk2000 spreadsheet (Schwab et al.,
2003) was used for these calculations and can be downloaded at
http://www.uwex.edu/ces/forage/pubs/milk2000.xls.
In Milk2000, forage moisture is an indirect measurement of the digestibility (hardness)
of kernels and reduces the amount of energy derived from grain as forage moisture
decreases (i.e. becomes more mature). Based on previous experience crude protein
has not affected milk calculations significantly (<1% change). In this analysis
NDF was not adjusted and an average value of 46.4% was used in the spreadsheet.
Other measurements were adjusted to NDFCP= 1.3%, Ash= 4.3% and Ether extract= 3.2%.
Table 1. Average values of corn hybrid means (reps per location=3) in University
of Wisconsin performance trials conducted between 1995 and 2003 (n=3255).
|
Measurement
|
Mean
|
Standard deviation
|
Minimum
|
Maximum
|
Forage yield (T/A)
|
7.91
|
1.69
|
2.87
|
13.5
|
Forage moisture (%)
|
62.3
|
6.10
|
37.4
|
84.8
|
NDFD (%)
|
61.6
|
3.61
|
46.4
|
80.6
|
Starch (%)
|
31.4
|
5.33
|
5.06
|
47.2
|
Crude protein (%)
|
7.25
|
0.78
|
5.00
|
10.9
|
NDF (%)
|
46.4
|
4.39
|
34.6
|
61.2
|
Using the mean + 2x standard deviation
of the mean, typical of 95% of the hybrids tested, the minimum and maximum values
for forage yield, moisture, NDFD and starch content were 4.53 and 11.3 T dry matter/A,
50.1 and 74.5%, 54.4 and 68.8%, and 20.7 and 42.1%, respectively. Changing forage
yield from 4.53 to 11.3 T dry matter/A did not change forage quality (milk per ton)
and increased milk yield (milk per acre) by 24 500 lbs milk/A to 40 900 lbs milk/A
(change = 149%). Changing forage moisture from 50.1 to 74.5% increased milk per
ton from 3410 to 3640 lbs milk/T and increased milk per acre from 27 000 to 28 800
lbs milk/A (change = 6.7%). Changing NDFD (digestibility of the stover) from 54.4
to 68.8% increased milk per ton from 3330 to 3920 lbs milk/T and increased milk
per acre from 26 300 to 31 000 lbs milk/A (change = 17.7%). Finally, changing starch
content from 20.7 to 42.1% increased milk per ton from 3300 to 3630 lbs milk/T and
increased milk per acre from 26 100 to 28 800 lbs milk/A (change = 10.0%).
Clearly forage yield has the largest impact on milk per acre and does not affect
milk per ton. Of the forage quality measurements affecting milk per ton, NDFD has
the largest impact among hybrids (17.7%), followed by starch content (10.0%) and
moisture (6.7%).
Table 2. Sensitivity analysis of Milk2000 when the forage measurements of yield,
moisture, NDFD or starch content were changed. One measurement was changed 1x and
2x of the standard deviation while the others remain at average levels.
|
Situation
|
Forage yield
|
Forage moisture
|
NDFD
|
Starch
|
Milk per ton
|
Milk per acre
|
|
(T/A)
|
(%)
|
(%)
|
(%)
|
(lbs milk/T)
|
(lbs milk/A)
|
|
|
|
|
|
|
|
Average
|
7.91
|
62.3
|
61.6
|
31.4
|
3630
|
28 700
|
Standard deviation
|
1.69
|
6.10
|
3.61
|
5.33
|
----
|
----
|
|
|
|
|
|
|
|
Changing yield
|
4.53
|
62.3
|
61.6
|
31.4
|
3630
|
16 400
|
|
6.22
|
62.3
|
61.6
|
31.4
|
3630
|
22 600
|
|
9.60
|
62.3
|
61.6
|
31.4
|
3630
|
34 800
|
|
11.3
|
62.3
|
61.6
|
31.4
|
3630
|
40 900
|
Changing moisture
|
7.91
|
50.1
|
61.6
|
31.4
|
3410
|
27 000
|
|
7.91
|
56.2
|
61.6
|
31.4
|
3490
|
27 600
|
|
7.91
|
68.4
|
61.6
|
31.4
|
3640
|
28 800
|
|
7.91
|
74.5
|
61.6
|
31.4
|
3640
|
28 800
|
Changing NDFD
|
7.91
|
62.3
|
54.4
|
31.4
|
3330
|
26 300
|
|
7.91
|
62.3
|
58.0
|
31.4
|
3480
|
27 500
|
|
7.91
|
62.3
|
65.2
|
31.4
|
3770
|
29 800
|
|
7.91
|
62.3
|
68.8
|
31.4
|
3920
|
31 000
|
Changing starch
|
7.91
|
62.3
|
61.6
|
20.7
|
3300
|
26 100
|
|
7.91
|
62.3
|
61.6
|
26.1
|
3620
|
28 700
|
|
7.91
|
62.3
|
61.6
|
36.7
|
3630
|
28 700
|
|
7.91
|
62.3
|
61.6
|
42.1
|
3640
|
28 800
|
Literature Cited
National Research Council. 2001. Nutrient requirements of dairy cattle. Seventh
edition (revised): National Academy of Sciences, Washington, DC.
Schwab, E. C., R. D. Shaver, J. G. Lauer, and J. G. Coors. 2003. Estimating
silage energy value and milk yield to rank corn hybrids. Animal Feed and Science
Technology 109:1-18.