agricultural production guidelines
dairying in kwazulu-natal
Dairying in KwaZulu-Natal
Dairying 5.5 1995
RATIONS FOR DAIRY CATTLE
P G Stewart
Cedara Agricultural Development Institute
The nutrient requirement tables
Practical problem solving
Computers for ration balancing
Rules of the thumb for ration
Farmers and their advisors
(consultants, extension officers, feed salesmen and veterinarians) are usually faced with
two kinds of problems concerning animal nutrition:
Is the nutrition of a
particular cow, heifer or group adequate? If not, the farmer must identify the deficiency
in the ration.
What must be fed to meet
the requirements of the animal or group of animals under consideration?
The purpose of this leaflet
is to describe a practical approach to solving these problems. The approach adopted is
based on the set of "look up" tables contained in KwaZulu-Natal Dairy Leaflet
5.12 using a calculator, and pencil and paper. Alternative approaches using computers are
THE NUTRIENT REQUIREMENT TABLES
The nutrient requirement
tables of KwaZulu-Natal Dairy Leaflet 5.12 are derived from various sources, in
particular, the British ARC system (ARC, 1980) and the American NRC system (NRC, 1989).
These sources are detailed in KwaZulu-Natal Dairy Leaflet 5.12. The tables are not as
"accurate" as the original equations but as was explained in KwaZulu-Natal Dairy
Leaflet 5.4, the sources of error in any feeding system are manifold. Therefore the
advantages of simplification outweigh any "inaccuracies" introduced by using the
At first acquaintance, the
volumes of tables may appear rather daunting but they are set out in a logical sequence
and once learnt, their use will be found to be straightforward. As explained in the
earlier leaflets on nutrition, especially KwaZulu-Natal Dairy Leaflet 5.4, the
requirements of an animal are affected by livemass, body reserves, age, production (rate
of gain and/or milk production), stage of lactation, stage of pregnancy, distance walked
and climbed and quality of the diet. In drawing up the tables the aim has been to take as
many of these factors into account as possible. The most important variable in feeding and
usually the most difficult to estimate accurately is dry matter intake. However, no ration
can be worked out unless some assumption is made about dry matter intake. The logic
followed here has been to estimate likely dry matter intake, ignoring ration quality. This
is a reasonable assumption provided that the animal is fed to meet her nutritional
requirements or, at the least is not being fed grossly in excess of requirements. The
estimates obtained may have to be modified in the light of experience, for example,
reduced if the actual feed is known to be unpalatable or in short supply.
It is far more difficult to
predict how well a heifer will grow on a particular diet than to predict milk yield and
weight change in cows. Therefore, any set of nutrient requirement tables for heifers has
to be used as a starting point and no more. The real test is frequent weighings to check
growth rates. Diets should then be adjusted in the light of this information.
The steps set out below
must be followed whether a pencil-and-paper or a computerised method of problem solving is
PRACTICAL PROBLEM SOLVING
OUTLINE OF METHOD
A: Is the nutrition
||Identify or imagine a typical
cow, or heifer, in the problem group or herd and list her attributes in terms of
live-mass, condition score, milk production (including butterfat %), stage of lactation
(days in milk) and stage of pregnancy (days in calf). If still a heifer or a first calver
the age and desired growth rate (ADG) must also be known. In addition, the distance
walked, and the height climbed each day need to be taken into account.
It may be necessary to do the calculations for
several "typical animals" such as a high producer in early lactation, an average
producer in mid-lactation, a low producer and a dry cow.
||Look up, or work out, her
potential dry matter intake and other nutrient requirements. For the sake of this example
we will use dry matter (DM), metabolisable energy (ME), crude protein (CP), crude fibre
(CF), calcium (Ca) and Phosphorus (P). The method and arithmetic is exactly the same for
any combination of nutrients.
feed or feeds eaten and their nutrient content (analyses).
the intake of the feed or feeds.
the differences between requirements and intakes.
||Assess whether any of the
differences are significant and decide on action to correct them (i.e. What should be
should be fed?
||As before identify or imagine a
||As in the previous example look
up expected dry matter intake and nutrient requirements or, if working out a ration mix,
such as a dairy concentrate, write down the specifications for the mix.
||List the feeds available or
likely to be needed to make up a suitable ration.
||Work out a suitable diet or
ration by the best means available. A pencil-and-paper method is described in the examples
that follow. However, this is the kind of problem for which programmable calculators and
microcomputers are well suited.
||Check that the proposed ration
is both practicable and economical. This is particularly important if a computerised
least-cost ration balancing program has been used for step 4.
A cow's diet
The method of using the
tables in KwaZulu-Natal Dairy Leaflet 5.12 is illustrated by means of an example. The
purpose is to assess the nutritional status of a first calver that calved on 15 July, was
confirmed pregnant to an insemination on 10 October and the date now is 1 February. Her
present production is 14 kg per day and a recent butterfat test was 4%. She weighs 450 kg
and her condition score is 1,5 on the Mulvany scale. She is being fed 2 kg of a 14% CP
dairy meal and is grazing day and night on kikuyu pasture. She walks about 4 km to and
from the pasture and climbs 35 m.
Is the nutrition
adequate? (Table 1)
||Identify or invent a
"typical" animal in the problem group or herd.
Four items need
to be worked out to complete the list of attributes viz: days in milk (Lactation stage),
days pregnant (Stage of pregnancy), days to drying off and butterfat production. From Table
1: (Leaflet 5.12):
LS = 365+TD-LCD
= 365+32-196 = 201
Therefore, Days in milk (LS) = 201 days
PS = TD+365-CD = 32+365-283
Therefore, Days pregnant (PS) = 114 days
Calculate the butterfat
production (14 x 4% = 0,56) and enter this into the worksheet at top left (see Table 1).
||Look up or work out her
potential dry matter intake and nutrient requirements. From Table 5a: Dry matter intake:
Round 0,56 to 0,55 and read the 0,55 kg BF line and 1,5 CS column. Potential or expected
dry matter intake = 14,0 kg. Enter this into the top right block of the worksheet.
Maintenance and production
(from Table 5a): It will be necessary to do some interpolation between
the 0,55 and 0,60 kg BF lines. From the ME column the requirement lies between 129 and 135
MJ. Guess the amount or calculate as
129 + [(135-129)÷((0,6-0,55)÷(0,56-0,55))] = 129 + [6÷5] = 130,2
requirements can be read off or calculated in a similar manner and entered in the
worksheet. The cow is less than 180 days pregnant so no adjustment is needed for
First calver growth:
Present mass = 450 but the cow is thin (CS<2). Therefore, if in desired condition she
would weigh 450 plus at least 50 kg. Let us assume she must put on 50 kg's for growth in
280-114 days = 166. Therefore, 50/166 = 0,30 kg/day, multiply the nutrient requirements by
0,30. Therefore an, ME = 47 x 0,30 = 14 MJ/day, add on allowance for condition will be
added. Target mass = 500, Days pregnant = 114.
From Table 5b:
Correction of condition: Read any extra needed. If the target condition at drying off is 3
then 9 MJ ME needs to be added to her requirements.
From Table 5b:
Walking and climbing: Read any extra needed. There is already an allowance for 10 m
climbing and 3 km walking in the series of tables. This covers the walking and climbing
while actually grazing or feeding. Therefore, add only 4 MJ.
The additional CP
requirement is the additional ME x 9,75. The extra protein for condition, exercise and
growth can be calculated individually or from the additional ME x 9,75 viz: (9 + 4 + 14) x
9,75 = 263. Now 1 449 + 263 = 1 712 which is higher than ME x 9,75. viz. 1 530, therefore
use the higher figure.
There are so many
uncertainties in calculating mineral requirements that the safety margins built into the
maintenance and pregnancy allowances will cover any extra which may be needed for
condition, exercise and growth.
Add up the total
requirements and transfer the numbers to the foot of the feed intake half of the main
table, the "Nutrient Requirements" line of the worksheet. Note that an amount of
2 100 g CF has been entered. As a general rule, CF should not be less than 150 g/kg, and
14,0 x 150 = 2 100.
feed or feeds eaten and their nutrient content (analyses). From KwaZulu-Natal Dairy
Leaflet 5.13, the book "Nutritive value and chemical composition of commonly used
South African cow feeds", or other source, look up the analyses and enter in the
measure/ estimate/ guesstimate the intake of the feed or feeds. For the sake of the
example, assume that the cow will eat her expected dry matter intake. In practical
farming, this is often not the case, and a more conservative figure may be more correct.
Remember that the tables are based on the assumption that the animals are not underfed.
the differences between requirements and intakes.
||Assess whether any differences
exist and take action to correct the differences. In this particular example the slight
excess of phosphorus can be ignored; the missing 15 g of calcium could easily be provided
in a lick; the major problems are shortage of energy and excess of protein. One kg of
live-mass loss is equivalent to about 30 MJ ME. However, this cow is already thin, and is
a first calver which still needs to grow, therefore, if practicable and economical, the
deficiency should be made up (see below) - Excess protein could be reduced by feeding a
low protein meal.
What should be fed?
The same worksheet and same
cow will be used to illustrate the method. In practice, it is usually necessary to do two
or three recalculations to achieve a satisfactory result. Therefore, work lightly in
pencil and only use ink as a final step.
Table 1. Example
worksheet for dairy ration calculations
||Identify a "typical"
animal. This has already been done.
As in the
exercises above look up her potential dry matter intake and nutrient requirements.
||List the feeds available, or
likely to be needed to make up a suitable ration. The protein should be reduced if
possible and the energy increased. Assume that a mixture of maize meal and minerals (a
maximum energy, minimum protein meal) could be mixed. Enter its analysis in the worksheet.
||Work out a suitable diet.
Without a computer, the most practicable method is to solve simultaneous equations for two
feeds and one nutrient. In practice the easiest way of setting up and solving simultaneous
equations is to use the "Pearson's square" method. It is not possible to work
with the energy, protein and minerals simultaneously. Begin with the energy or protein;
assess the result; recalculate if necessary and, finally, balance the minerals. In order
to reduce the excess protein, a low protein, high energy mix of maize meal and minerals
will be tried. The details are in Table 2.
The desired amount of energy in the final ration is 157 MJ ME in 14,0 kg
ration. Therefore, we need an energy concentration or density of 157 ÷ 14,0 = 11,2 MJ/kg
ME. The maize analysis must be expressed on a 100 % dry-matter basis for the calculation.
Because of the need to convert to 100 % dry matter, many nutritionists prefer to do all
their calculations on a 100 % dry-matter basis and only convert to the "as fed"
basis as a final step. The MM+mins has 11,7 ÷ 88 x 100 = 13,3 MJ/kg ME on a dry
matter basis. Note that, for some purposes, it may not be necessary to convert to a
dry-matter basis, for example, when calculating a ration which uses only dry (about 90 %
dry-matter) feeds. If we only needed 4,6 kg of the ration then 2,1 kg kikuyu plus 2,5 kg
MM+mins would give as a concentration of 11,2 MJ ME in the total ration.
Figure 1. An example of
Pearson's square for calculating simultaneous equations
Table 2. Example worksheet for
dairy ration calculations
we need 14,0 kg. Therefore, these amounts must be converted using simple proportion viz.
2,1 x 14,0 ÷ 4,6 = 6,4 kg kikuyu and
2,5 x 14,0 ÷ 4,6 = 7,6 kg MM+mins on 100 % dry basis
Therefore, convert to "as fed"
7,6 ÷ 88 x 100 = 8,6 kg. Enter these amounts into the worksheet and calculate the
||Check that the proposed ration
is both practicable and economical. Each situation will be different and it is very
difficult to lay down hard and fast rules for ration practicability. In this case, the
result is very close to what was wanted. The protein and the fibre are marginal with both
the calcium and the phoshorus well within acceptable limits. There is also no substitute
for experience in this field. A few useful rules for ration balancing are set out at the
end of this leaflet.
A concentrate mixture (Example 2)
It is usual to have more than two
ingredients in a ration. Pearson's square can still be used. However, all except two
ingredients must be fixed. Some may argue that this is not a "least cost"
ration or an "optimum" solution. This is true, but in practice most farmers have
only a limited number of ingredients available for mixing, and the method to be outlined
will give perfectly satisfactory results. Assume that a farmer wishes to mix a dairy meal
containing not less than 140 g/kg CP, 11 MJ/kg ME, 7 g/kg Ca and 5 g/kg P on an as fed
basis. He has his own maize meal but wishes to ration it out over a longer period by using
hominy chop. Other ingredients available include groundnut oilcake, fish meal, lucerne
meal, dicalcium phosphate, feedlime and salt. The mixer unit holds about 1 000 kg. It is
therefore convenient to work out a mix to make up about a ton.
|Step 1 &
||Have been done viz: the
desired specifications for the ration have been decided. These must be entered in the
"Desired Specifications" line of the worksheet (See Table 3).
||Enter the analysis of the
||Work out a suitable ration. As
a first approximation, fix all except two ingredients. In practice, protein is the most
expensive component of a ration, hence it is usual to balance for protein and hope that
there is sufficient energy. In other words, if a series of feeds of similar energy values
but increasing protein contents is worked out, the cost will increase sharply as the
protein content goes up. Also, excess protein can restrict intakes and depress production.
Therefore, the usual practice is to balance for protein; then to check the energy level
and, if that is too low, to try again with different constraints on the ingredients.
In this example it was decided to include 25 %
hominy chop (250 kg), 3 % fish meal (this is expensive and is only included to provide
some undegradable protein (rumen bypass protein). Whether this is necessary or not will
depend on the level of production of the cows and the quality of the other protein
sources). Adding 1 % feedlime, 0,5 % dicalcium phosphate and 0,5 % salt should balance out
the mineral needs. Commercial HPC's contain minerals, but the Feeds Act (Act 36/1947) does
not require the manufacturer to state the exact quantities on the label. Therefore, if a
commercial HPC is used it will be necessary to find out the actual specifications from the
seller. The mineral inclusion rates in the example were decided on because it is usual,
when using maize and oilcakes (which have inverse Ca:P ratios) to need 0,5 to 1 %
feedlime. Salt is an essential nutrient and should always be included at the rate of at
least 0.5 %. These and other rules of thumb for ration balancing are set out at the end of
this leaflet. Note that no lucerne meal has been included. It is too expensive in relation
to it's nutrient content for inclusion in concentrate mixes except in very special
circumstances. If in doubt, work out the costs per unit of CP and ME and compare those
with the costs of these derived from the other ingredients.
Having fixed everything except the maize
and groundnut oilcake, it is now possible to set up a "Pearson's square" and
determine the amounts needed of these two ingredients.
- Determine the total amount of the remainder
1 000 - (250 + 30 + 10 + 5 + 5) = 700 kg
- Add up the amount of nutrient (in this
example, the protein) contributed by the fixed ingredients.
26,5 + 18,3 = 44,8 kg. Therefore the amount to be supplied by the
floating ingredients is 140 - 44,8 = 95,2 kg
- Determine the concentration required
95,2 ÷ 700 x 1 000 = 136 g/kg
- The Pearson's square
359 x 700 ÷ 406,3 = 619 kg maize
495 47,3 x 700 ÷ 406,3
= 81 kg GNOC
- Enter these numbers in the worksheet and
complete the calculations
- Check whether the specifications have been
met. In this example all differences are probably insignificant and the ration can
now be mixed.
Table 3. An example of formulation of a
CONCENTRATE / TOTAL MIXED
Ration: 14% CP Dairy Meal
FOR RATION BALANCING
To use a computer effectively, it is
necessary to have some understanding of its operating principles. But use of the computer
should be preceded by a thorough understanding of the pencil-and-paper method of solving
Two methods are commonly used viz:
spreadsheets and linear programs (commonly called LP's). The first is so similar to the
pencil-and-paper methods outlined in the previous sections that a user who has mastered
the pencil-and-paper technique will feel at home with spreadsheets, and merely has to
learn how to use the programs. It is then possible to build one's own spreadsheet or to
get a matrix from someone else. The Cedara Agricultural Development Institute, Private Bag
X9059, Pietermaritzburg, 3200 has developed suitable matrices for use on Apple and MS-DOS
or PC-DOS compatible micro-computers. The advantage of these spreadsheets is the speed of
calculation. By using a database linked to the spreadsheet a library of feed analyses can
be built up, thus eliminating the need to look up and enter analyses from sets of tables.
Table 4 illustrates part of the output from
a spreadsheet developed for the MS-DOS micro-computer (Lotus 1.2.3 or Supercalc). Note
that in the "kg" column, maize meal and groundnut oilcake (GNOC) have -1 and -2
values. These are flags that tell the program to use these two ingredients to achieve the
protein target (157 g/kg) set out at the foot of the "CP" column. The final mix
is reported in the "MIX THIS" column. The logic is similar to that in the
Pearson's square pencil-and-paper method. The big advantage is that a variety of mixes and
ideas can be tested very rapidly.
Linear programs are a lot more difficult to
use than spreadsheets, partly because the linear programming technique is difficult to
understand. A linear program is a mathematical technique for setting up and solving a very
large number of simultaneous equations containing many variables. The linear program does
not answer questions of the form; "What will be the result if 50 kg of maize meal and
13 kg of cottonseed are mixed together?" A linear program answers questions of the
form; "Given this set of ingredients, how much of each must be mixed to get the
cheapest ration which will still meet a particular set of specifications such as a minimum
of 11 MJ/kg and 120 g/kg CP?" A linear program is therefore the appropriate technique
where there are a number of possible ingredients and a complex set of nutrient
specifications have to be considered and hence, the "best" or
"optimum" solution is not at all obvious. For example, a linear program is the
most practical way of computing rations for monogastric animals such as pigs, where the
specifications have to include amino acid and vitamin levels as well as energy, minerals
and others. In ruminant nutrition a linear program is the method of choice in the milling
industry, where a wide variety of raw materials of varying prices is available. At the
farm level, linear programs are useful in some circumstances, particularly if the farmer
is faced with a wide choice of ingredients and it is not obvious which to use to make the
least expensive ration. Linear programs can be dangerous in the hands of inexperienced
users because rations which meet the specifications may in fact be completely impractical,
or even dangerous. For example, a linear program could suggest a mix containing 20%
feedlime simply because feedlime was the cheapest ingredient available to top up a mix and
the user had forgotten, or did not know, to limit feedlime to a maximum of 2% of the
ration. Therefore, in using a linear program the constraints (maximum and minimum amounts
of each ingredient) and specifications (maximum and minimum amounts of each nutrient) must
be very carefully drawn up.
Linear programs are not only used for least
cost rations. They can be used, for example, to calculate a maximum energy ration, or for
working out fertilizer blends and in many other ways. When purchasing a linear program,
make sure it is easy to use, because some are so powerful and complicated that only
dedicated specialists can use them. Ensure also that it has a post-optimal analysis. For
example, such a program should include a report on how sensitive the ingredient
proportions are to changes in prices and the prices at which rejected ingredients would be
considered for inclusion. Don't buy a linear program which has built-in feeding standards.
You may want to set your own specifications. Perfectly adequate linear programs for farmer
use are available in the public domain. Cape Software Library, P.O. Box 2462, Cape Town,
can supply one called KLP.
RULES OF THUMB FOR RATION BALANCING
Set out in the following sections
are a few guidelines on ration balancing. They are by no means complete, nor are they a
substitute for personal experience. If in any doubt, consult standard reference works or
an experienced nutritionist.
Table 4. An example of an
output from a ration balancing program on a spreadsheet.
- Protein is expensive. Don't overfeed
- Low energy rations merely replace roughages,
attempt to maximize energy.
- Maize meal should not be too finely ground
but must be cracked or rolled.
- Grains are very low in calcium and must
always be balanced for calcium and phosphorus.
- Ensure fat/oil level in the concentrate does
not exceed 10%, unless rumen protected fats are used.
- Limit barley and wheat grain to a maximum of
50% of the concentrate.
- Molasses, in any form, is an excellent
binding agent in rations, reducing dust, as well as improving palatability. Do not exceed
10 % molasses in the concentrate.
- Include a minimum of 40 % roughage in the
- Provide at least 14 to 16% or 19 to 21 %
crude (CF) or acid detergent fibre (ADF), respectively, in the ration.
- Provide 28 to 30 % neutral detergent fibre
(NDF) in the ration, of which a minimum of 70 to 75 % should be derived from roughages.
- Optimal total mixed ration dry matters are
between 50 and 75 %.
- Forage should supply the equivalent of 0,9 %
live-mass as NDF.
- Provide a 35 to 40 % non-structural
carbohydrates (NSC) in the ration. An excess of starch, or sugars, in the diet can cause
acidosis and low butterfat.
PREVIOUS CHAPTER | CONTENTS
| NEXT CHAPTER >> ]