Managing Mixing Wagons for Performance and Health

Ellen R. Jordan, Ph.D.

Texas Agricultural Extension Service
The Texas A&M University System

Tremendous strides have been mad e in • Provide plenty of clean, fresh water
improving the quality of the rations formulated for
dairy cattle. In the recently released Nutrient Desp ite addressing these issues on many farms,
Req uirem ents of Da iry Ca ttle (NRC , 2001), effort pred icted D MI may not correspond to actual DM I.
was spent on predicting dry matter intake (D M I) of The curre nt N RC (2001) e quation to pre dic t D MI
dairy cattle so that diets can be formulated to meet the includes only animal factors: fat corrected milk, body
nutritiona l needs at vario us pro duction leve ls. Yet, weight and week of lactation. The equation does not
DM I and production response to formulated rations include temp erature or hum idity adjustments to
continue to vary from herd to herd and within herd. account for heat stress periods, relying instead upon
Thus we must address the question “What are the the use of lower milk production inputs. Although
consequences of various on-farm mixing procedures this methodo logy may work for lactating animals,
and bunk management on dry matter intake?” DM I for prepartum cows may be over estimated
during perio ds of heat stress.
Bunk ma nagement has received considerable
emphasis over the years. Some of the key elements of To day’s mixer wagons allow us to cap ture da ta
feedbunk management which have been identified so that actual D MI can b e reco rded and used in
include: subsequent ration formulation. Figure 1 illustrates
the changes in actual DM I observe d over the c ourse
• Provide ample bunk space per cow of a year for animals in four different physiological
• Keep fresh feed in front of cows at all times sta te s loca te d in a Ce ntra l Texas he rd. Tra cking DMI
• Monitor feed refusals changes from year to year provides a customized
• Clean feed bunks regularly pred iction tool so tha t rations can be formulated in
• Minimize feed bunk competition anticipation of expected changes in DM I in an

47
attempt to minimize climatic impacts on nutrient virtually every T exas T MR exceed ed P enn State’s
intake. This is particularly critical in adjusting recommended long particle size (Table 1; Rippel et
nutrient density in rations formulated to meet the al., 1998). Based on the amount of long fiber (i.e.
needs of non-lactating animals. particle size) in these rations, little acidosis would be
expected , yet acido sis continues to be a p roblem.
Fiber Needs Furthermore, these samples were taken during the
summ er, when the typical ration is higher in
Sometimes rumen function is compromised by concentrate to compensate for the effects of reduced
the way rations are delivered, preventing attainment DM I. The difference may be the forage base and how
of maximum DM I. To maintain rumen function and it is stored. Texas silage-based rations are
health, NRC (2001 ) recommends that dairy rations predominantly wheatlage although some corn and
contain a minimum of 25% neutral detergent fiber sorghum silages are fed as well. Alfalfa hay is used,
(NDF) with 19% of dietary DM originating from not alfalfa haylage. Some rations have coastal
forage NDF . Adjustments to the guidelines are bermuda grass or sorghum/sudan hay. In addition,
recommended based on the effectiveness of that fiber silage and haylage are typically stored in bunker
at stimulating cud chewing, salivation and rumen rather than upright silos. Possin et al. (1994) reported
mov ement. that mean particle length and the percent of the
samp le retaine d on the upper screens was gre ater in
The effects o f insufficient effec tive fiber in forages from bunker co mpare d to upright silos,
dairy rations include: primarily due to unloader co nstraints.

Not only did Texas rations exceed the

• Acidosis (subacute or acute)
recommended long forage particle length, they
• Erratic dry matter intakes
differed substantially from the TM R averages
• Decreased milk yields
reported by Penn State researchers (Table 1). The
• Lowered milk fat production
major differences were more long particles on screen
• Health problems (laminitis, ketosis, displaced
1 and less of the TMR on the bottom pan. W e have
abomasum)
not tested as m any rations as the Penn State
researchers; however, based on these results and
Allen (1996) identified eight factors which
additional rations evaluated since, producers feeding
could be used to adjust the optimal level of NDF
TM Rs using different forages from those in the
needed in early lactation rations to maximize energy
Northeast region must be cautious in using the Penn
intake. These factors included: forage particle size,
State Separator guidelines.
quantity of NDF from by-products, frequency of
concentrate feeding, starch digestibility, use of
buffers, rate of fiber digestibility and passage, Particle Size as a Tool
addition of fat and variability of forage dry matter and
quality. Besides assessing the adequacy of fiber in high
producing dairy rations, particle size provides another
Most of these factors can be readily evaluated. method to evaluate key components of bunk
However, the determination of forage particle size management which might impact rumen function and
has been a tedious task. As a result of the increased prevent attainm ent of m aximum D MI. Namely:
interest in evaluating ration particle size to assess the
adequacy of fiber in high producing dairy rations, the • Is a uniform ration delivered to each group?
Nasco Fo rage Particle Separator was designed at The • Is over mixing occurring?
Pennsylvania State University to physically separate a • W hat is the optimal loading sequence?
feed sample into three portions: particles greater than • Does the available ration change over time?
0.75 inches, between 0 .31 and 0.75 inche s, and those • How do feed refusals differ from the feed
smaller than 0.31 inches. Based on their analysis of delivered?
total mixed rations they recommended 6-10% of
particles should remain on the upper sieve, 30-50%
should be on the middle sieve and 40-60% on the
bottom pan.

W e began analyzing TM Rs and found that

48
Table 1. Compa rison o f particle size of T exas farms and those in the northeast.

Ration n >0.75 inches 0.31-0.75 inches <0.31 inches

Silage TM R 7 18.4 42.7 38.7

Hay T M R 12 22.2 31.8 45.9

T M Ra 367 6.1 35.5 58.4

a
Lammers et al., 1996.

Table 2. Ration U niformity by C.V. ranges and interpretations.

C.V. Screen 1 Screen 2 Screen 3 Interpretationa

< 10 % 4.7 38.0 47.9 Desirable

10-20 % 38.0 57.0 38.0 Needs improve ment

> 20% 57.0 4.7 14.0 Cause for concern

a
Recommendation from Behnke, 1996.

Ration Uniformity evaluated with Behnke’s (1996) recommendations are

shown. If we evaluate o nly screen three, the majority
There is very little published data on the of the herds were within an acceptable variance
effects of nutrient uniformity on animal performance. range. This, coupled with chloride ion analysis we
W hat work has b een d one focused on non-ruminants. conducted, suggested that the concentrate portion of
Thus there are few benc hmarks for using this the rations were mixed adequately. But the high
technique with dairy rations. If mixing is poor percent of herds with greater than 20% variation for
enough to alter nutrient intake of certain individuals, screen one, indicated the forage portion of the ration
performance may be altered. The animals receiving was not being mixed adequately with the concentrate.
excess nutrients will be inefficient in feed conversion.
In extreme cases, suc h as with urea inclusion, this Uniformity C heck Proced ures: To
situation m ay even bec ome toxic. Conve rsely, determine whether the forage in the ration delivered
anima ls receiving rations deficient in nutrients will is uniform, take a minimum of five samples (ten
have performance comp romised. Feed intake and would be better), evenly spaced along the entire
body size influences susceptibility to ration length o f the bunk where the m ixer wagon d ischarges.
imbalances in that smaller animals consuming smaller Next use the Penn State Particle Size Separator to
meals are more likely to be influenced by improper sort each sample. Determine the average percent on
ration mixing than are larger animals consuming more each screen. Loo k at the va riation fro m the average.
dry matter. The influence of DM I on nutrient Statistically compute the coefficient of variation by
consum ption is most noted in transition and fresh using the formula:
cows with lower levels of intake, and p erhap s it is
with this group that ration uniformity is of most CV = (s/0) 100 = 100/s²
impo rtance. 0
where S = /S 2
Behnke (199 6) has recommende d we use the
statistical calculation coefficient of variation (C .V.) n

as a measure of ration uniformity. In Table 2 the and S2 = [ 3 (xi - 0)2 ] / n-1

1
percent of herds in our study (Rippel et al., 1998)
within the C.V. ranges and their interpretations as

49
 Figure 2: Sam ple
ration with a mean on
the top sieve = 15.7% and CV = 55.22%

This statistical method results in one number which After analyzing the samples we could not make
can be used to decide if too much variation in the any generalizations as to whether over-mixing by 15
total mixed ration exists. Use B ehnke’s minutes affected particle size or if one type of mixer
recommend ations to characterize the variation in the reduced particle size more than another. On
herd’s ration. individual dairies it did appear that over mixing
reduced particle size. In figure 3, the results from one
One example herd is shown in figure 2. This dairy, feeding an alfalfa hay-based ration, are shown.
herd had a CV of 55.2% o n screen 1. Although the
avera ge pa rticle size fo r the 10 samp les was adeq uate In a W isc onsin survey of 49 he rds, TM R
based on the Penn State recommendations, some rations were grouped based on whether the ration was
individual animals did not receive sufficient long mixed over or under 15 m inutes for their standard
fiber. In this herd, tho se cows had no oppo rtunity to mixing procedures (Possin et al., 1994). Using the
eat the longer fiber at another meal because 5 W isconsin Forage Particle Size Separator 7.6% of the
different pens of cattle were fed from the same load ration was on the top two screens (>1.5 inches long)
of feed in the same order. in rations mixed less than 1 5 minutes, while only
4.8% was on the top two screens in rations mixed
Evaluating Particle Size Reduction more than 15 minutes. The herd s were also
categorized based on incidence of laminitis. Herds
One of the reasons we evaluated Texas TMRs with a high-incidence had 3.5% o f the particles on the
was to determine if over mixing resu lted in particle top screens, while herds with a low incidence had
size reduction and whether either a vertical or 7.9%. In addition, herds with a higher incidence of
horizontal m ixer wagon caused more deg radation. laminitis had 42.9% of the ration dry matter from
W e identified ten dairies using vertical-type mixer forage compared to 49.5% in the herds with reduced
wagons and ten dairies using horizontal mixers. Each laminitis.
dairy mixed their feed using their standard mixing
procedures. We placed ten sample containers at
equal distances down the feed bunk and then
unloaded the mixer wagon. Next, an identical ration
was over mixed by mixing the load for an additional
15 minutes.

50
Figure 3: Effect of over mixing by 15 minutes in an example alfalfa hay based TMR.

In 2000, we collected samples from 10 herds and density appear to have the greatest impact on mix
and conducted bo th particle size and wet chemistry uniformity. The addition of forage and the level of
analysis o n those samp les to evaluate w hether their forage inclusion in dairy rations presents a unique
was a relationship between variability in the challenge to determining adequate mixing times. The
compo sition of the diets delivered between various differences in forage and concentrate particle size
nutrients and particle size. Table 3 illustrates the alone present a pro blem. Differences in particle
mean, range for various nutrients evaluated and density between ingredients add another
correlation of the nutrient CVs with the particle size consideration. On a dry matter basis, corn silage and
3
CV in the ana lysis. In this sample, the variation in haylage are fairly eq ual in bulk density (kg/m );
particle size does explain some of the variation found however, on an as-fed basis, corn silage tends to have
in various nutrients, while very little in others. a 33% greater bulk density than haylage. In addition,
mineral density can be two to three times that of grain
Although no t the original purpose of this trial, and protein, making it difficult to maintain a random
one item of particular interest is the amount of distribution.
phosphorus in these rations. As nutrient management
becomes mo re critical on these farms, feeding As a general rule, lighter and larger particles
nutrients above re quirements will become less tend to mo ve upward while the smaller, mo re dense
acceptable. Since these samples were taken and particles gravitate downward. Traditionally it has
results shared with individual producers, feeding been recommended to load larger particle size
phosphorus above requirements has been curtailed. ingredients first (forage) and heavier, smaller
particles last. However, with the use of individual
Loading Sequence com mod ities and rations c ontaining many ingred ients
with a large variation in size, shape , and d ensity,
Several ingredient properties can influence determination of loading sequence has become a
mixing: particle size, particle shap e, density, metho d of trial and error on man y farms.
hygroscop icity, static charge, and adhesivene ss
(Be hnke, 199 6). Fro m this list, particle size, shape,

51
Table 3: Correlation o f the particle size CV to the CV for various nutrients.
Correlation
Particle Size 0 Range CV top : CVnutrient

To p, % 20.6 3.0 - 35.6

Middle, % 37.5 17.7 - 51.0

Bottom % 41.9 30.7 - 58.4

Nu trient An alysis

Moisture, % 44.6 37.8 - 48.7 .42

CP , % 17.8 16.4 - 19.5 .45

AD F, % 27.4 23.3 - 29.9 -.44

ND F, % 40.3 35.8 - 45.4 -.08

Ca, % .71 .37 - .99 .26

P, % .62 .50 - .78 .51

Mg, % .33 .25 - .42 .06

K, % 1.64 1.21 - 1.95 .41

Na, % .44 .24 - .66 .49

Zn, ppm 97.7 73.2 - 124 .4 .17

Cu 33.8 16 - 8 0.8 -.22

Mn, ppm 79.7 57.3 - 108 .8 .56

% Cl ion .51 .39 - .66 .11

Adjusting Particle Size for Feed Refusals

In an attempt to reduce the variation in ration
uniform ity in one herd we chan ged the mixing ord er. On occasion, initial particle size evaluation
The prod ucer had already changed to chopping the indicates that ad equate long fiber is b eing provid ed in
hay to increase uniformity. The ration consisted of the T MR, yet acidosis continues to be a problem in
wheatlage, alfalfa, and various concentrate the herd. This is the time to evaluate what the cows
components. In this particular herd with a horizontal are refusing and reevaluate the amount of long fiber
mixer, loading the wheatlage first followed by the provided based upon a value ad justed for refusa ls.
concen trate and cho pped hay resulted in the most Example 1 illustrates how to make adjustments for
uniform ration m ix base d on forage particle size. feed refusals.
W hile reversing the orde r resulted in the highest
variation (Figure 4). Furthermore, the average As this example demonstrates, a ration which
particle length w as longer whe n the chopp ed alfalfa app eared to have sufficient lo ng fiber when initially
was ad ded last. fed, actually had inadequate amounts of fiber
consumed as a result of so rting by the cattle.

52
 Example 1: Adjusting Forage Particle Size Measurements for Feed Refusals.

Step 1: Determine the percent of long fiber in the initial TMR.

A 2,000 kg batch had 12% on screen 1, 40% on screen 2 and 48% on the
bottom pan

Step 2: Determine the percent of long fiber in the feed refusals.

In the 200 kg of refusals 50% was on screen 1, 28% was on screen 2 and 7%
was on screen 3.

Step 3: Adjust the percent on each screen/pan by the refusals.

kg originally kg refused kg consumed % as consumed
Screen 1 240 100 140 7.8
Screen 2 800 56 744 41.3
Bottom Pan 960 44 916 50.9
Total 2000 200 1800

Figure 4: Effect of changing mixing order and hay processing on variation in particle size.

53
Ration Changes in Bunk size evaluation of de livered feed can be a tool to
improve feed b unk management.
Evaluating the ration when initially distributed
may be insufficient. Martin et al. (2000) reported on From our work and that of others, it is evident
the changes found in the feed remaining in the bunk that many factors can affect the particle size in the
with time after delivery. A well-managed herd mixed ration, such as:
feeding a one-group TM R was used to evaluate the
effect of selective consumption by dairy cows on • condition o f mixer,
particle length available over time. Feed was pushed • amo unt of feed mixed at o ne time,
up at 6 hour intervals and the remaining ration was • loading ord er,
sampled for particle size analysis. The percent of • type of feed, and
particles on the top screen increased from 9.3% at the • particle length prior to mixing.
time of initial distribution of the ration to 58.7% after
23.5 hours in the bunk. Thus during the first 12 hours Forage p article size evaluation can indicate
after distribution, cows consumed very little of the when a ration contains insufficient forage particle
fiber portion of the ration, increasing their risk for length to maintain proper rumen function. However,
acidosis. the interpretation of particle length depends upon the
type of forage included in the ration. In addition, the
Martin et al. (2000) suggested four factors results from pa rticle size evaluatio n can help
were important to minimize sorting: determine optimal loading sequences and mixing time
for individual dairies so that both adequate forage
• Keep ration dry matter between 46 and 52%. particle length and ration uniformity can be
• Restrict dry hay usage to <4 lbs/head. maintained.
• Maintain consistent particle size in the ration.
• Process corn silage to minimize the effects of When using particle size evaluation on a herd,
husks and cobs in the TMR. take a minimum of five samples, evenly spaced along
the entire discharge of the mixing wagon. Maintain a
Although these recommendations may be log of results from the farm. Then use the
practical for silage and haylage based rations, many information to evaluate how uniform the ration is
of the rations fed in the South and Southwest are hay delivered over time, determine if pre-processing hay
based. Thus it is not feasible to restrict the dry hay is needed and monitor the condition of the mixer. By
used. With hay based rations, adding additional continually monitoring the herd, a baseline for
water to the ration may minimize sorting as might particle size can be established and used for
pre-chopping. Another key component is delivering a comparison when metabolic or production problems
uniform mix as d iscussed previously. arise.

Many larger herds are fed multiple times per Acknowledgments

day, which may reduce the impact of sorting. Some
herd owners elect to start discharging the mixer The author expresses appreciation to the
wagon at a different location during each feeding in a Ho uston L ivestoc k Sho w and Rodeo and P rotiva, a
day which may com pensate for the lack of uniformity unit of M onsanto, for their financ ial supp ort of this
in the mix. Since the number of times per day cows study. Also, the author is grateful for the cooperation
are fed varies depending upon group size, number of of the C ounty E xtension Age nts and producers in
different rations being fed, size of the mixer wagon, Archer, Comanche, Co oke, Erath, Ellis, Hamilton,
seasonal impacts, owner preferences, etc. it is critical Hopkins and Johnson counties.
to conduc t evaluations in light o f these variables.
References
Conclusions
Allen, M. 1996. Fiber requirements for dairy cattle: How low can
Managing mixer wagons is an integral part of you go? In Proceedings of the 1996 Mid-South Ruminant
Nutrition Conference. p.12.
feed m anagement on today’s large da iry. Data
collected electronically can be use d to calculate Behnke, K.C. 1996. Mixing and nutrient uniformity issues in
actual DM I to imp rove ration fo rmulation. Particle ruminant diets. In: Mid-South Ruminant Nutrition Conference
Proceedings, p. 6.

54
Heinrichs, J. 1996. Evaluating forages and TMRs using the Penn National Research Council. 1989. Nutrient Requirements of
State Particle Size Separator. Penn State Cooperative Extension Dairy Cattle. 6th rev. ed. Natl. Acad. Press Washington, D.C.
Service. DAS 96-20.
National Research Council. 2001. Nutrient Requirements of
Lammers, B., J. Heinrichs and D. Buckmaster. 1996. Method Dairy Cattle. 7th rev. ed. Natl. Acad. Press Washington, D.C.
helps in determination of forage, TMR particle size requirements
for cattle. Feedstuffs 68(41):14. Possin, I.R., C. DeCorte, R.D. Shaver, and R.T. Schuler. 1994.
Survey of forage particle length and metabolic disorders on
Lammers, B., J. Heinrichs, R.T. Ward and D.R. Buckmaster. commercial dairies. University of Wisconsin-Madison.
1996. Forage and TMR particle size distribution used by farmers
in the northeast. J. Dairy Sci. 79(Suppl.1):153. Rippel, C.M., E.R. Jordan and S.R. Stokes. 1998. Evaluation of
particle size distribution and ration uniformity in total mixed
Martin, R., L. Armentano, and C. Leonardi. 2000. Feed sorting rations fed in Northcentral Texas. The Professional Animal
and acidosis: Is there a link? Hoard’s Dairyman 145:625. Scientist 14:44.

55
56