Sie sind auf Seite 1von 42

Several decisions are made daily on a dairy farm.

Some involve cow and labor management, others address crop and
business concerns and there are a host of others in between. Unfortunately, dairy producers do not always have
enough information to make the best decisions for current circumstances. One could argue time constraints
signifcantly hamper the ability to make proftable decisions all the time, noted Victor E. Cabrera, University of
Wisconsin-Madison/Extension dairy systems management specialist.
Oftentimes there are too many variables that simply cannot be controlled due to the nature of the business. A
combination of highly volatile market prices and increasingly complex technology and management strategies do not
leave much margin for error, either, he added.
In response, a renewed shift toward advanced information management is currently taking place within our dairy
industry. Previously foreign words such as modeling and simulation are becoming commonplace, thanks largely in
part to advancements in computer technology.
Cabrera said, With the help of computers and some advanced mathematics, we are not only able to manage the
herd events happening today, but efectively forecast changes in herd structure as well.
The Dairy Expansion Decision Support System a decision support system designed to explore dairy farm
production and expansion scenarios and simulate specifc metrics of their performance provides critical information
dairy producers need to actively manage risk on their dairies.
The versatility of the programs structure ofers potential use in several other areas, including providing a tool for risk
management in times of great uncertainty, particularly during periods of dairy expansion; accounting for future herd
growth when considering livestock housing needs; and matching the proper facility design with specifc user-defned
goals in mind. The Dairy Expansion Decision Support System program and supporting documentation can be
accessed via the UW-Extension Dairy Management website at:
UW-Madison dairy scientists began working on this project by drafting a few simple, yet very important goals: First,
efectively simulate the natural biological progression of a real dairy herd and accurately forecast herd structure at a
future point in time. Then, create a robust, yet user-friendly economic decision support computer program that could
be widely adaptable to complex decision-making scenarios involving dairy production and expansion. With these
ideas in mind, the latest application in what promises to be a powerful lineup of risk management tools was created.
What is herd structure? Take a closer look at the people working on a 1,000-cow dairy, for example. There are several
diferent workers who each have a specifc skill set tailored to their role in keeping the dairy operational from day-to-
day and beyond. Milking personnel are in charge of properly milking the cows, feeding staf keep the cows fed, crop
employees ensure high-quality feedstufs are grown for the dairy, and managers ensure the whole farm remains in
business well into the future. There are several other team members who support the mission of a dairy; this
shortened list will sufce for our illustration.
Clearly, the people in each respective category depend on each other to keep the whole farm operating. Without
crops, the feeding staf would not be able to keep the cows fed. This directly afects the milkers, who know cows will
not produce much milk if their energy requirements are not met, and so on. A similar arrangement is present within a
herd of dairy cattle; however, we categorize them diferently to enhance our management capabilities.
Herd structure defnes the number and age of both individuals and specifc groups of cows within any collective cohort
at a defnite point in time. If one were to look at the animal inventory on a dairy, it is easy to see how this breakdown
occurs. There are heifers up to twenty-four months of age, cows in their frst lactation, second, third, and beyond. At
the beginning of each lactating cycle, a calf is born. If it is female, she most likely will be raised on-farm and eventually
enter the herd two years later. At any time during the lactation, a cow can be removed from the herd and her spot is
(hopefully) taken by a more proftable replacement animal. When you have any number of cows with all of these
events occurring at once, management of the whole herd becomes quite a challenge.
A decision support system can be described as an organization of information that allows a person to make otherwise
complex decisions, and in this case, forecasts, with relative ease. Of course, the accuracy of these forecasts is highly
dependent upon the quality of information entered by the user. In its current state, the program incorporates over 120
input variables from various aspects of the dairy. Each input variable is defned by the user, meaning the program can
be adapted to nearly any farm. Some inputs relate to specifc cow information, such as milk production; others
describe related economic and fnancial data, including milk price.
Once these essential pieces of information are entered into the Dairy Expansion Decision Support System program,
the model predicts heifer growth over time, cow movement, and monthly cash fows. Specifc information related to
milk production, feed intake, and labor requirements are also utilized to improve accuracy in generating monthly cash
fow fgures. In the end, a present value analysis adds meaning to the projected values and ofers solid supporting
evidence to make informed decisions regarding many what-if situations. PD
Excerpts from University of Wisconsin Cooperative Extension news release
Example of how the Dairy Expansion Support System tool works
Lets follow an example with a 200-cow dairy. Suppose John wants to double his herds size within three years to
accommodate his sister, Jane, and her family, as they want to move back to the farm and carry on the family tradition.
Currently, John and his family are milking 170 cows in a parlor/freestall setup that was updated 10 years ago.
Although there are other concerns that need to be addressed (manure storage, land, feed, etc.), John and Jane want
to fnalize plans for their new freestall barn, which is scheduled to be fnished six months from now. It will house the
entire milking herd and include extra space for calving pens. The existing freestall barn will house heifers and dry
cows. Their current facilities are already overcrowded, with only 150 stalls available for lactating cows. They have
been saving up to purchase bred heifers, but are unsure of how many to buy once the new 400-cow freestall barn is
John and Jane are also concerned about the number of calvings per month, fguring out correct pen sizes, and
ensuring they have enough labor available when the milking herd is moved to the new barn.
This is the exact point where the Dairy Expansion Decision Support System tool goes to work. After flling in the input
values to describe the herd, a simulation was run on the computer within a matter of minutes. The results suggest a
total of 114 heifers should be purchased in months 7, 8, and 9 (38 heifers per month) to achieve just over a 100
percent stocking density at the end of three years. At that time, there would be 402 total cows and an estimated 22
percent increase in income over variable costs per cow.
However, other potential scenarios could be explored. What would happen if John and Jane were to wait until months
34, 35 and 36 to purchase heifers instead? When comparing scenarios, it is useful to look at the total net present
value and make a decision based upon which option generates a higher value. A comparison of the results indicates
John and Jane could attain a 16.5 percent higher net present value over the 3-year period by purchasing heifers in
months 7, 8, and 9 rather than in months 34, 35 and 36. PD
P G Stewart
Cedara Agricultural Development Institute
Lactation can be initiated with drugs but, in general, if a cow does not produce a calf, she
will not produce milk. Calves are also necessary to provide replacements for the dairy herd,
and surplus heifers can form a very important part of the dairy farmer's income. Therefore,
reproduction management is vital to successful dairying.
The cow's early life, from birth until she is mated for the first time, is dealt with in detail in
Natal Dairy Leaflets .! and .". #t is important to understand that a heifer must not be
mated at too light a mass. $rom %& to '( ) of her e*pected mature mass is normally
considered to be a desirable mating mass. +fter mating, the heifer must continue to grow so
that she is as close to her mature mass as possible at first calving ,-ee Table . in Dairy
Leaflet ." for growth standards for the main dairy breeds/. The age at which the heifer
becomes a cow, that is her age at first calving, depends on the rearing system ,she must
meet the target weights mentioned above/ and the desired generation interval. The sooner
a heifer calves down, the sooner she contributes to the economy of the farm in the form of
milk and calves. Therefore, the farmer must weigh up the possible e*tra rearing costs of,
say, calving heifers at close to two years of age, which is the earliest practicable age,
against the loss of income from milk and calves by delaying calving to a later age. #f a
farmer is wanting to e*pand his herd, then the lower the age at first calving, the faster is
the rate of increase in the herd. #nterestingly, in a stable herd, the age at first calving has
no effect on the number of surplus animals available for sale each year. + greater age at
first calving merely means that there are more heifers on the farm. + final consideration is
that the older the heifer, the lower her fertility. #n other words, the later a heifer is mated
the more difficult it is to get her into calf. $or this reason, heifers should not calve down
later than three years of age.
0nce a heifer has calved for the first time, she becomes a cow, and takes her place in the
milking herd. #t can be shown that the farmer should aim for a calf per cow per year.
Therefore, the cow must be got back into calf so that she re1calves as close to one year
after her previous calf as possible. #n practice, cows not pregnant after a reasonable period
are culled for infertility. +lthough a cow may not be truly infertile at culling, the farmer may
decide that the e*tra cost of keeping her and continuing to try to get her into calf is not
2ustified. +s a generalisation, it can be said that the average -outh +frican dairy cow
produces three calves in her life time. This average is calculated from a large population of
cows, some of which have fewer than three lactations and others which may go on for as
many as ten lactations. 3*pected age structures in e*panding and stable herds are shown in
Table !. The number of heifers is dependant on the age at first calving ,+$C/ and the age
when surplus heifers are sold.

Table 1. Typical age structure ! e"pa#$i#g a#$ stable $airy %er$s
(er ce#t ! c)s i#
i# eac% lactati#



",4 ,.

+s mentioned above, some cows are culled for fertility, and some may be culled for other
reasons,e.g. mastitis. +lso, some cows die through accident or disease. The annual culling
rate ,including mortality/ is typically from "( to (). This means that to maintain a herd of
!(( cows, "( to ( replacement heifers must calve into the herd each year. #n the
remainder of this leaflet the interactions between these factors in the life of the dairy cow
and dairy herd will be e*amined in some detail.

Culli#g Rate +S,
The percentage of cows culled per year. This is usually "& to () of the herd. #n other
words, in a herd of !(( cows only '( to '& of the cows will remain in the herd after one
year. This implies that at least "& to () heifers have to be brought into a herd every year
to maintain herd si6e.

Replace'e#t Rate +R,
The percentage of cows brought into the herd per year e*pressed as a percentage of the
number of cows in the herd at the start of the year.

Rate ! I#crease +E,
7eplacement 7ate less Culling 7ate e8uals 7ate of #ncrease. $or e*ample, if there were !((
cows at the start of year, "& cows were culled and heifers brought in, then91
Culling 7ate : "&)
7eplacement 7ate : )
7ate of #ncrease :
1"& :

Age at *irst Cal-i#g +A*C,
The time from birth to calving, i.e. when a heifer becomes a cow, is called the age at first
calving. The mean age at first calving is the generation interval in the herd. The younger the
+$C, the greater the potential rate of increase and the fewer the heifers which have to be
fed and cared for. #t should not normally be less than two years of age or greater than three
and heifers should have achieved about 5() of their average mature mass when weighed
2ust before calving.
Note the following three points very well. $irst, if a farmer wishes to e*pand his herd then
the younger the +$C, the greater the potential rate of e*pansion. -econd, the younger the
+$C, the sooner the heifer becomes a cow and therefore, the fewer heifers of all ages in the
herd. Third, in a stable herd, +$C has no effect on the number of surplus heifers for sale;
+$C only affects the total number of heifers of all ages.

Nu'ber surplus %ei!ers +.,
3ither the sale of surplus heifers, or herd e*pansion by retaining all the heifers are very
important parts of dairy farm income. The number of surplus heifers is affected by the
average intercalving period ,see -ection on <The interaction between +$C, #C= and other
reproduction parameters/. The shorter the intercalving period the more calves that will be
born and the greater the probability that more heifers will be born.

Ttal #u'ber ! %ei!ers +H,
The ma*imum number of heifers is deterined by both the +$C and the intercalving period.
The minimum number which must be kept on the farm for herd replacements depends on
the age at +$C and the culling rate plus an allowance for infertility and mortality.

I#tercal-i#g (eri$ +IC(,
The period in a cow's life between the birth of one calf and her ne*t calf. The #C= is so
important that an entire section of this leaflet, <The intercalving period<, is devoted to it.

Ttal ser-ices t c#cepti# a#$ ser-ices per c#cepti# +TS(C / S(C/
The total number of services per conception ,T-=C/ is the number of inseminations in the
entire herd divided by the number of confirmed pregnancies. #n practice this is not easy to
calculate, and the usual figure 8uoted is services per conception ,-=C/ for pregnant cows
only. This is an important distinction because it could happen that cows which fall pregnant
do so readily, and the use of #C= could be masking a real problem amongst those which do
not. Note, too, that the -=C and T-=C figures are influenced by the rate of culling. 7uthless
culling of apparently infertile cows may result in a good -=C figure but hide a real problem.
The figure for percentage take to first +.#. helps in interpretation ,see below/.
Targets9 -=C !. to !.%;
T-=C > "

Ta0e t !irst i#se'i#ati#
?sually calculated as number of cows pregnant to one insemination divided by the number
of first inseminations. $or e*ample, '( first +.#.'s and .5 pregnancies would be91
.5@'( * !(( : '() take to first +.#.
Target9 A%( )

Heats sptte$ r %eats 'isse$
The number of heats spotted divided by the total possible heats. The total possible heats is
usually calculated by assuming that all open cows ,non1pregnant/ cycle every "! days. Due
to cystic ovaries, and many other causes, not all cows cycle and the length of the oestrus
cycle is very variable. Therefore it is impossible to achieve !(( ) success.
Target9 A'&) of theoretical possible number of heats.

This is the period in a cow's life between the birth of one calf and her ne*t calf. The
recommended targets for #C= are9
+ mean of %& days.
No cows should have an #C= less than ( days,
5() from ( to .(( days,
and less than !() an #C= of more than .(( days.

*igure 1. Diagra''atic represe#tati# ! t%e peri$s a#$ e-e#ts i# t%e
i#tercal-i#g peri$! t%e $airy c)

The #C= is best understood if broken up into its component periods and events as illustrated
in $igure !. The reader is advised to commit this diagram to memory because successful
dairy farming is ine*tricably linked to a thorough understanding of the implications of the
factors affecting the intercalving period and their management.
The intercalving period can be broken down into the following9

&actati# (eri$
The time from calving until the cow is dried off, i.e. the time during which the cow is
producing milk.
Target9 (( to (& days ,. weeks/;
7ange9 "%& to .( days ,4 to .5 weeks/;
The length of the lactation period depends on the open period and level of production.

Ope# (eri$ +O(,
The time from calving to re1conception. The length depends on the voluntary waiting period
and the breeding period.
Target9 Bean of 4& days ,!" weeks/;
7ange9 .& to !"( days ,' to !' weeks/.

.lu#tary 2aiti#g (eri$ +.2(,
The minimum time the farmer decides to allow between calving and the first mating. ?sually
.& to %( days.
Creeding before %( days ,4 weeks/ after calving is not normally recommended. #f the DE=
is less than %( days, then it must not be less than .& days and the cows must be clean and
showing at least the second observed heat.

3ree$i#g (eri$
C= : 0= 1 DE=, i.e., the time from the end of the DE= until conception. Length depends on
the number of services per conception and the per cent heats observed.

1estati# (eri$
The time from conception to calving. ?sually "4( to "4& days. This varies slightly between
breeds, and depending on the se* of the calf, but for practical purposes can be taken to be
"4( days ,.( weeks/.

Dry (eri$
The time from the end of lactation until the cow calves again. ?sually &% days ,4 weeks/
calculated from the conception date, i.e. conception date plus "". days gives the date on
which to dry off the cow.
Target9 A.& but >&' days.
Low producers, cows which have been sick, or cows with e*tended service ,open/ periods
may reach uneconomically low levels of production before &% days from when the ne*t calf
is e*pected, and thus have a dry period longer than &% days. The length of the gestation
period is also variable and may not be e*actly "4( days, causing the length of the dry
period to vary by appro*imately a week. The dry period is necessary to allow the cow time
to replenish body reserves, especially minerals which were depleted during lactation and to
allow the udder tissue to involute, and to be renewed before the ne*t lactation. + dry period
of less than .& days will have a markedly deleterious effect on the ne*t lactation.

(re4cal-i#g (eri$
-tarting about % weeks before the anticipated calving date, the cow may be fed e*tra feed
for replenishment of body reserves to ensure that she calves in good condition ,the cow
should not be over1conditioned/ and to ensure that her rumen micro1organisms are adapted
to concentrate feeding.

Calculati# ! a-erages
-ee section on <$ertility analysis< in FwaGulu1Natal Dairy Leaflet ..! for full details on the
calculation of averages and other summaries of herd data.
Mean Intercalving Period and Breeding Period
Bean intercalving period is an indication of calving percentage. The lower the #C=, the
greater the number of calves born per year. #ntercalving periods are not normally ,evenly/
distributed. The distribution about the mean is skew. Thus a geometric rather than an
arithmetic average must be calculated. Bean #C= is calculated as follows9
Bean #C= : Bean C= H DE= H "4(
Bean C= : e*pIJln,C=/! H....H ln,C=/nK@nL
where e*p : e*ponent ,natural antilogarithm/;
ln : natural logarithm; and
n : number of cows.

Services per conception (SPC)
The fewer the services per conception the shorter the breeding period ,see section
<Creeding period</ provided that heat spotting is good. -ee section on <$ertility analysis< in
FwaGulu1Natal Dairy Leaflet ..! for details on how to calculate -=C.

Percentage heats spotted
?nless cows are seen to be in season they cannot be inseminated. Therefore, the better the
heat spotting the shorter the breeding period. -ee <$ertility analysis< in FwaGulu1Natal Dairy
Leaflet ..! for methods of estimating heat spotting success.

T%e i#teracti# bet)ee# A*C5 IC( a#$ t%er repr$ucti# para'eters.
Dariations in the age at first calving ,+$C/, the inter1calving period ,#C=/ and the culling rate
,-/ in a dairy herd has profound effects on the possible replacement rate ,7/, the net rate of
increase in the herd ,3 : 7 1 -/ and the number of surplus heifers ,D/ which could be sold
annually. These parameters also affect the make up of the herd in that an e*tended +$C
means more heifers ,M/ in relation to cows, an e*tended #C= means more dry cows, and so
The interactions between these factors can be illustrated in a simple way by considering the
life histories of two hypothetical cows ,$igure "./. #n this e*ample it was assumed that the
two cows had identical milk productions. Thus the difference in lactation production and
calves born is entirely due to the different +$C's and #C='s.
Table " e*tends the argument by showing e*amples of the interactions between the age at
first calving ,+$C/, intercalving period ,#C=/, culling rate ,-/, rate of replacement ,7/, rate
of increase ,3/, the number of surplus heifers ,D/ and the number of heifers of all ages ,N/
in stable and e*panding herds. The data in the table make allowance for heifer infertility and
Note the dynamic interaction of these factors. #n particular, note how important +$C is to
the e*panding herd. Therefore, in the well managed herd a short #C= means more calves
and, as will be shown, more milk. #n the e*panding herd the stockman must aim for the
shortest possible +$C commensurate with good husbandry ,well grown1out heifers whose
masses meet or e*ceed breed targets/, and the possible need to produce milk for seasonal
demands ,e.g. the need to build up the 8uota/.

Table 6. E"a'ples ! t%e i#teracti# bet)ee# %er$ repr$ucti# para'eters i#
stable +S, a#$ e"pa#$i#g +C, %er$s
N A*C IC( R S E . H Mil0

Ntes9 Merds +1- to D1- are kept constant in si6e through sale of all surplus stock.
Merds 31C to M1C are increasing at !() per annum ,3/.
N : Number of cows in the herd.
+$C : +ge at first calving ,in months/
#C= : #ntercalving period ,in days/
7 :
7eplacement rate i.e. number of heifers
introduced per annum as a ) of N.
- : Culling rate i.e. ) of N culled each year.
3 : 7ate of increase in the herd i.e. 7 1 - : 3
D :
Number of heifers available for sale per
M :
Ba*imum number of heifers in the herd from
age ( to age +$C months.
Bilk inde* : 3*pected production relative to first herd

*igure 6. T%e li!e %istry ! t) 7a-erage7 c)s

Note that, with reference to Table ", changing the +$C has no effect on the number of
surplus heifers for sale ,D/, if herd si6e is constant but a changing +$C has a dramatic effect
on the number of heifers ,M/ in the herd ,herd +1- versus C1- in Table "/. The effect of
increasing #C= in stable herds is to reduce the number of saleable surplus heifers ,D/ ,Merd
C1- versus D1-/. The difference between %& and ."( days represents % heifers per year
per !(( cows 9 a very large loss in gross income. #f this loss is considered along with the
declining milk production with increasing #C=, it is clear that farmers cannot afford e*tended
Ehere the farmer is attempting to increase herd si6e as in herds !C to .C, the +$C has a
very significant effect; compare D in herds 31C and $1C. #f the effect of #C= is added to a
poor +$C it is clear that a rate of increase of even !() per annum is not possible ,Merd M1
C/. +verage +$C's in e*cess of ( months and #C='s greater than .(( days are common in
-outh +frican herds and e*plains why many farmers struggle to increase their herds. +ctual
+$C's and #C='s in -outh +frican herds are illustrated in Table ! in FwaGulu1Natal Dairy
Leaflet ".!.
?nderstanding the very important concepts illustrated by Table " might be helped if it is
remembered that the value of N in the stable herds is always !((, whereas in the changing
herds it was not !(( but 5! one year earlier and 4 the year before that. This e*plains why
the values for 3 plus D do not stay the same i.e. 3 H D for herd +1- was !4 but only !' for
herd 31C and so on. -imilarly, the values for M are lower in the changing herds. These
principles illustrate the importance of seeking the underlying causes of, say, a high
percentage of dry cows or a high proportion of heifers to cows. To simply say that the ratio
of heifers to cows is wrong or that there are too many dry cows is unhelpful. The reasons
for the undesirable figures must be identified and will be found in the interactions between
the factors discussed in this section.

Her$ Structure a#$ Culli#g Rates
#t has already been shown that the higher the +$C the more heifers in the herd ,Table "/.
Logically, the higher the culling rate, the lower the average age of the herd or, the more
culls, the fewer the average number of lactations per cow. Cows seldom have twins and the
generation time ,+$C/ usually e*ceeds two years. +lso, cows' productive lives are seldom
longer than about ten or twelve lactations. These biological facts have important
implications on the highest and lowest possible culling rates.

*igure 8. T%e i#teracti# bet)ee# culli#ga#$ t%e a-erage #u'ber ! lactati#s

$igure shows that, even under good management, the ma*imum culling rate is about (
to &), giving an average of , to ",4 lactations per cow, and the minimum possible
culling rate is about "() which means an average of & lactations, some cows having to live
for !. to !4 lactations. These comments are based on a constant rate of culling between
lactations. #n practice, culling rates will be higher as cows get older and the real average
number of lactations and ma*imum ages necessary will be lower than the figures 8uoted
above. #ntercalving period also affects the culling rate. The greater the #C=, the higher the
average number of lactations will be for a particular culling rate. -table and e*panding herds
will have slightly different age structures as illustrated by Table !. The differences in age
structures must be considred when making comparisons between herds. $or this reason, it
is usual in some countries, notably the ?-+, to convert all milk yields to mature e8uivalents
before calculating herd averages. The difference between a first lactation and a mature
lactation is about "&) in most herds, so the conversion will make a significant difference.
-hould the good herd manager aim for a high or low culling rateO The ob2ectives of the herd
manager will determine the answer. $acts to be considered are9
The most rapid genetic progress is made at ma*imum culling rate.
The most milk is produced at minimum culling rates ,+bout to &)
more milk at "() than at () culling/.
There are more surplus heifers for sale at low culling rates. This is
perhaps the most important consideration. #f there is a very big
difference between the value of pregnant heifers and cull cows then it
may pay to cull leniently. Cut, if the slaughter value of cull cows and
the sale value of pregnant heifers is similar, then it probably pays to
ma*imise culling.
#n practice, it seems that most farmers cull "& to () annually. #n well1managed herds
there is more voluntary culling for production, temperament and conformation ,type,
milkability etc./ whereas in badly managed herds most culls are involuntary ,i.e. cows cull
themselves because of infertility, mastitis, footrot etc./. The good manager seeks to
minimise involuntary culls in order to place emphasis on production and type. + common
management mistake when trying to build up dairy herds is to cull too leniently. 7emember
that many undesirable traits, such as very bad udders, are highly heritable and hanging on
to poor cows to build up numbers is a recipe for building up a mediocre herd.
#f three lactations is the average life span of a dairy cow then beef bulls should never be
used on dairy heifers. #f inferior bulls, or beef bulls, are used on dairy replacement heifers,
the first calf, if a heifer, will be of no use as a dairy cow. 0f the second and third calves, one
will probably be a bull. Therefore, in a lifetime production of three calves, only one will be
useful as a herd replacement, which means that no e*pansion of the herd is possible.

H) t ac%ie-e g$ !ertility
+s implied in definition <#ntercalving =eriod< the spread of #C='s in a herd is important. The
early part of lactation is the most productive. Therefore, the shorter the #C=, the more milk
that a cow will produce in her lifetime. Conversely, e*tended #C='s mean either long dry
periods or at best an e*tended lactation with low milk production. $igure " illustrates these
There are four fundamental points which must considered to improve intercalving periods91
Her$ %ealt% a#$ )ellbei#g i#clu$i#g culli#g
$igures derived from study groups have shown that there is a positive correlation between
veterinary costs and profit. Boney spent on regular herd checks, which result in early
identification of treatable cows and the cows which should be culled, is money well spent.
-imilarly, good nutrition and stockmanship, which lead to contented cows, are essential to
good fertility. Bineral and vitamin nutrition is as important as ade8uate supplies of protein
and energy.
A.I. tec%#i9ue
The best time to inseminate in relation to standing heat ,see FwaGulu1Natal Dairy Leaflet
"../ and other details of +.#. in practice are outside the scope of this guide. Mowever,
correct techni8ue is very important and declining success with +.#. often results from lack of
attention to detail, e.g. careless thawing of straws, or failure to deposit semen in, or 2ust
through, the cervi*. #f there appears to a fertility problem in a herd this is the easiest area
to check on and should be eliminated before going into aspects such as nutrition.
Heat sptti#g
This 2ob is critically important and must not be left to the lowest paid worker on the farm.
Too often, complete responsibility for this task is left to the young <herdboy<. #f it is
assumed that all open cows cycle every "! days ,they don't/ then more than '&) of the
possible number of heats should be observed. -ee section on <=ercentage heats spotted< for
further details.
Ti'e t !irst A.I. a!ter cal-i#g
#t is clear from numerous studies that, provided that cows are ade8uately fed and are given
at least &( days dry, the shorter the #C= the better, from both a milk and a calf production
point of view. Cut there are sound biological reasons why attempting to shorten #C=s simply
by shortening the DE= leads to an increase in the number of inseminations per conception
and the number of e*tended lactations. Cows may e*hibit signs of oestrus as early as
twelve days post partum. Mowever, the uterus re8uires at least .&, and preferably && to %(,
days before it has involuted sufficiently for the successful implantation of a new foetus. $or
this reason, many early inseminations are either entirely unsuccessful or worse, appear to
be successful but the cow is pregnant only for a short while, then the foetus is resorbed
and, after a few weeks, the cow re1cycles. Cecause no abortion was observed, the stockman
tends to assume that the cow was not pregnant, simply because he failed to observe her in
season. The net result of too short a DE= is that there is an increase in the number of
e*tended lactations. The average #C= may be improved but the spread of intercalving
periods has increased. This situation is commonly seen in herds where bulls are permitted to
run with the cows. Cows which fall pregnant very early do not have sufficient time to regain
body condition before the ne*t lactation, and cows which take a long time to settle have
long periods of low or no milk production.

To sum up9
pay attention to herd health and welfare ,especially regular veterinary
ensure that +.#. techni8ue is correct
be really concientious about heat spotting, even on those cold, wet spring
ensure that every cow to be inseminated has been given a clean bill of health
at the veterinary check
do not inseminate too soon, allow at least .& days ,but preferably %( days/ to
lapse after calving, while ensuring that the cow is inseminated only on her
second observed heat.
P G Stewart
Cedara Agricultural Development Institute
$arm records are kept for all or some of the following reasons91
T satis!y t%e Recei-er ! Re-e#ue
This is an essential re8uirement of record keeping but should not be the sole reason, and a
record system can be designed which satisfies the 7eceiver and is also useful for other
T assist i# !i#a#cial pla##i#g $ecisi#s
$inancial records, in more detail than those re8uired for the 7eceiver, can be
used for cash flow planning, enterprise analysis and other purposes.
T c#trl labur
This is usually a wages book recording days worked, wages paid, money
owed, leave etc.
T assist i# la#$ 'a#age'e#t $ecisi#s
These include farm maps and gra6ing, irrigation, fertili6er use, crop yield,
areas and management operations records.
T assist i# li-estc0 'a#age'e#t $ecisi#s
These are the records of individual animals and groups of animals, their
production, health, feed use etc.
N lgical reas#
+ lot of useless information is often kept which is never, or can never, be
converted into useful information.

The brief summary of record types given above illustrates that several sets of records must
be kept, inevitably involving much of the farmer's time. Mence, if records are not to be more
trouble than they are worth, they should satisfy the following criteria9
T%ey 'ust be use!ul
?nless data which is being recorded will at some future time be used ,turned
into information/ in making management decisions it should not be recorded
at all.
Recr$s 'ust be 0ept i# suc% a !r' t%at t%ey ca# be easily
c#-erte$ i#t i#!r'ati#
Cefore keeping a record, the eventual end use must be decided upon so that
the form in which the data are recorded will facilitate later analysis and
interpretation. Too often the end use is not considered, and the usefulness of
the data is severely impaired.
Recr$ 0eepi#g syste's 'ust be si'ple
Dairy farmers have enough to do without burdening themselves with comple*
record keeping systems, that are difficult to understand and time consuming
to complete, and therefore nearly impossible to delegate to employees.
Duplicati# 'ust be a-i$e$ as 'uc% as pssible
-ome data may have to be recorded more than once in different forms, but
this must be reduced to a minimum. #n other words, if a record is to be made
in the field, the recording system should be such that data can be
conveniently entered in the field and does not have to be re1entered back at
the office.
Recr$s 'ust lea$ t acti#s bei#g ta0e#
?nless a record is specifically intended to be used for some future action or in
management planning it should not be kept.

A# -er-ie)
#t is not the purpose of this guide to describe a complete farm record keeping system. +
simple, effective dairy herd record system, which satisfies the criteria listed above, will be
This convenient system can be used to supply information for91
3ffective monitoring of animal performance right from birth.
3valuation of management and feeding systems.
#ndividual animal comparisons to assist in breeding, culling and other
Creed society, milk recording and computer program usage.
3*traction of useful herd indices for evaluation and comparison.
=roduction of +ction Lists for management
The physical re8uirements for the system are cow byre sheets ,or milk recording book/,
individual animal cards, ring1binder file@s with plastic filing sleeves, cow calendar board,
heifer1rearing board, an insemination book, daily diary and last, but not least, a comfortable
office. 0ther records which may be added include a separate mastitis record and
chalkboards for drying1off and feeding lists.

C) byre s%eets/Mil0 recr$i#g b0
These are used for9
7ecording of live mass
7ecording of condition scores
7ecording of butterfat percentages
7ecording of daily milk yields
7ecording concentrate allocations
7ecording concentrate use
Noting drying1off dates
Noting oestrus
7oughage record ,camps gra6ed, 8uantity silage fed etc./
Bastitis presence and treatments
7unning lactation totals.
#t is not possible to reproduce an e*ample sheet here as it would be too detailed, but the
essential layout is illustrated in $igure !. This layout can be achieved using some of the
sheets which are given away by feed companies, or certain bookkeeping analysis books sold
by commercial stationers can be used.
The essential points are as follows91
Cow names are recorded vertically in alpha1numeric order. Bilk
recording sample bottle numbers, butterfat percentages, live masses
and condition scores are recorded monthly.
Condition scores may be recorded fortnightly
Days of the month are recorded hori6ontally, ideally with four columns
per day. This is usually 8uite simple as many of the sheets which are
available have space for three milkings per day plus a total and, as
most herds are milked twice daily, the spare column can be used for
two purposes. Note that in $igure ! totals have not been added since
most users of this system find that there is little point in recording
totals since these daily totals can be calculated with a minimum of
mental effort.
The far right hand side is used for running lactation totals. Bost people
find it far easier to add vertical columns of numbers than hori6ontal
columns. The most fre8uently done additions are thus arranged
vertically. The only hori6ontal addition that has to be done is the
running lactation total. This is done because official Bilk 7ecording
results fre8uently come back too late to be used for culling decisions
and for other analyses.
There are very few farms where the milk is recorded at every single milking. #n any event
adding every single day's yield for the purpose of estimating a cow's yield is unnecessarily
laborious ,0fficial Bilk 7ecording only uses one day per month/. Therefore, it is suggested
that three days of the month are selected; for e*ample, the rd, !.th, and ".th. +dd the
three days together, and multiply by !(. #f it's a thirty day month the total for the month
has been arrived at, which can be added to the total for the previous month. #f it is a thirty1
one day month, add the three days together, multiply by !( and add the middle day to get
an estimate for the month. $or $ebruary add the three days together, multiply by !( and
subtract two days yield. $or cows which calved or were dried off during the month, add the
available days, divide by the number of records used and multiply by the number of days
the cow was in milk.
An example:-
"Bett!s" ields were "##.$%l & ##.%l & '$.'%l( x )* + ,-%l .or a #* da mont/ or0 ,-%l &
##.%l + ,$1l i. it were a #) da mont/.
Anot/er example:-
"A%," calved on t/e )*t/0 /er ields .or t/e availa2le das were )$ and '%. It was a #) da
mont/0 s/e was t/ere.ore in mil3 .or ') das and /er estimate would 2e )$l & '%l + -' x
') 4 ' + --)l.
#t is useful, whenever possible, to add the vertical columns of yields as a check against milk
sales or the inaccuracy of the recording 2ars or milk meters. #t is normal for the sum of
individual yields to be about ) higher than bulk tank readings. This is owing to the
cumulative effect of the tendency to read upwards when reading yields in the parlour. Larger
differences should be investigated. The other items need only be added weekly to get an
estimate of concentrate use, number of cows in milk, cows in herd, dry cows etc., as
illustrated in $igure !.
-ome farmers enter the cows' names on the sheets in the order in which they calved. This
has a marginal advantage once a month when samples are taken for 0fficial Bilk 7ecording,
but for the rest of the month it is unnecessarily time consuming to search for cow names,
especially in large herds. +lso, the dry cows may not, and seldom do, calve e*actly on their
e*pected date and can be entered in the wrong order. The preferred system of entering is
alpha1numeric ,or logical/ and would read thus9
Cetty, Cugs, Point, Bystery, 7"!, E'5, !", !"4, ..

*igure 1. E"a'ple c) byre s%eet

I#$i-i$ual a#i'al car$s
+ commercially available card which fits into a standard filing tray, or filing folder, which can
be carried conveniently and used in both the office and milking parlour is illustrated in
$igures " to &. This card was specifically designed to satisfy the criteria listed on page
!. -ide ! ,$igure "/, the calf record card, is used first vi59 from birth to calving. The
emphasis on this side is on getting the heifer to grow out well. The outlines are for $riesland
and +yrshire breeders to record markings ,very useful if ear tags are lost/. Meats and
inseminations are recorded on side " ,$igure /, and the card would normally be turned to
this side once the first insemination has been done.

*igure 6. T%e %ei!er recr$ car$ +Si$e 1,

*igure 8. T%e bree$i#g recr$ car$ +Si$e 6,

The complete fertility history is recorded here, and bull codes for use at first and subse8uent
inseminations are instantly accessible when needed. -ide " flaps open to reveal sides and
,$igure ./, and on these two sides the complete production history of the cow can be
recorded. -ide opens up to reveal sides & and % ,$igure &/. -ide & is used to record the
cow's physical and psychological attributes and is used in con2unction with the bull brochure
in selecting suitable mates for the cow. -ide % is used to record any health problems.
=articular emphasis is on mastitis control.

*igure ;a. (r$ucti# recr$s +Si$e 8,

*igure ;b. (r$ucti# recr$s +Si$e ;,
These cards also are filed in alpha1numeric order so that they can be used conveniently.
$urther, their portability lends the system to field use. Coloured tags are available which
can be clipped to the bottom of the card to signify, for e*ample, cows to be pregnancy
checked. The file or tray is carried in the shed and notes can be made directly onto the
card. Different colours can, if desired, be used to indicate <dry<, <pregnant< etc.
Lactation totals ,up to (( days and complete lactations/ from the cow byre sheets are
entered on these cards. #f desired, these can be entered lightly in pencil and the <official<
records superimposed when these are received.
These cards are not designed to be kept in shoe bo*es, and proper file drawers or folders
are well worth the cost involved.

*igure <a. Type classi!icati# +Si$e <,

*igure <b. .eteri#ary recr$s +Si$e =,

Ring-binder files and plastic filing sleeves
#n almost any herd the farmer collects pieces of paper relating to individual animals. The
more common ones are breed society certificates and pedigrees, and 0fficial Bilk recording
forms such as from LTD !55& ,the record of production and breeding performance received
after each completed lactation/. These forms, together with any other bits and pieces
,photos etc.,/, are most conveniently kept, again in alpha1numeric order, in transparent
plastic filing sleeves in ring1binder files. ?nder this system the latest performance record
,the old one having been discarded/, or any other information can be referred to with the
minimum of effort. + separate file@s should be used for culled or sold animals as these
records are very seldom needed.

Cow and heifer calendars
Darious <whole herd at a glance< methods are available and are essential ancillaries to the
records listed above. #n other words, these are not substitutes for the individual animal
cards, but are supplementary to them, and are used to identify problem breeders, poor
producers, cows due to show oestrus, and for other purposes. + commercially available cow
calendar is illustrated in $igure %. 3ach cow is represented by a two1 or si*1colour magnet
which is placed on the board in a position and of a colour relevant to the latest event in her
life. $ull instructions on using these calendars are supplied by the manufacturers. The two
most useful makes presently available differ slightly, therefore their use will not be
described in detail here. Cut the necessity for the calendar is its usefulness in highlighting
actions and for planning decisions. -ome e*amples of these are cows for pregnancy
diagnosis ,=D/ and numbers of cows e*pected to calve each week for the ne*t % months.
+ heifer1rearing board can be constructed using <peg1board<. This is marked similarly to
the calf record card for mass1for1age graph ,$igure "/, e*cept that the calf outlines are
omitted. 3ach heifer is represented by a wooden golf tee. 3ach month, when the heifers are
weighed, the tees are moved to the correct mass and age position. The background should
be painted with green schoolboard paint and dosings and inoculations can be marked on the
board at appropriate ages. Coloured washers ,castrator rings are good/ can be put onto the
tees to record inseminations and pregnancy or even important dosings and inoculations,
especially if a whole age group is not done simultaneously. There is also a commercially
made heifer1rearing calendar using steel magnets. ?nfortunately, it does not permit masses
to be recorded.

Insemination / PD notebook
#ndividual cow cards, unfortunately, do not lend themselves to certain important analyses
and some duplication is unavoidable as will be clear from the following section on <+nalysis
of Merd 7ecords<. The insemination @ =D notebook is simply a listing, in chronological order,
of inseminations and their success or failure as determined at the regular veterinary
inspection. + suggested layout for an insemination book is given in $igure '. + pocket1
si6ed notebook which can be used for this purpose is available from an +.#. co1op.

Figure 6 !he la"out of a commercial cow calendar

!he dail" diar"
+ large format <page a day< diary should always be on the office table. This has many
uses. #n particular, it is used for recording events not recorded elsewhere. $or e*ample,
7eserve the first few lines for appointments such as <Det due at 5h((< and the rest for
events. <Dosed all calves in small shed with Linte*<; or <-tarted silage making today. .
loads<; or <Top %( cows to camp +! ,( * '( metres/< and so on. The diary is particularly
important on large farms where more than one person may be involved in the
management. The cow byre sheet can serve some of the daily diary functions, but not all.

#ther records
$iles need to be maintained in the farm office for various bits of information which apply to
the herd as a whole, such as leucocyte counts, milk recording summaries and T.C. tests.
#t should be apparent that, if these records are going to be effectively kept and the
ma*imum value e*tracted from them, a comfortable farm office is absolutely essential.

No record keeping system is complete unless it leads to actions. Therefore the data
collected and recorded in the sheets, cards, books and calendars must be processed
,converted from data to information/ before it is of real value.

Figure $ %n e&ample %I / Insemination 'ook recorded in chronological order to facilitate a (-sum anal"sis

'reeding and selection
T%e i#$i-i$ual c) car$s lea$ $irectly t i$e#ti!icati# ! pr pr$ucers5 $i!!icult
bree$ers5 'astitis prble's etc. I! t%e li#ear classi!icati# is use$ i# c#>u#cti#
)it% t%e A.I. cp4p?s bull brc%ure5 suitable 'ates a#$ alter#ati-e 'ates ca# be
c%se# as $escribe$ i# :)a@ulu4Natal Diary &ea!let 6.6 a#$ recr$e$ # t%e c)?s
car$. I# t%is )ay5 )%e# a c) is rea$y !r i#se'i#ati#5 last 'i#ute c#!usi#
-er )%ic% bull t use # %er )ill be a-i$e$5 as t%is )ill %a-e bee# $eci$e$ at
leisure5 per%aps i# c#>u#cti# )it% a# A.I. c4p r bree$ sciety !iel$ !!icer.

T%is ter's re!ers t lists ! #a'es ! a#i'als re9uiri#g particular atte#ti#. T%e
're i'prta#t lists are $iscusse$ bel).

Dail" action lists

Cows likely to show heat:
T%ese ca# be i$e#ti!ie$ $irectly !r' t%e up4t4$ate c) cale#$ar. C)s )it%i# 1A
t 6; $ays ! t%eir last %eat r i#se'i#ati# s%ul$ be acti-ely sug%t ut. It is
9uite li0ely t%at a c) i# seas# )ill be 'isse$ )%e# ge#erally l0i#g !r %eat
sig#s i# t%e %er$ as a )%le. S'all sig#s ! %eat suc% as a bull stri#g r a
rug%e#e$ tail %ea$ ca# be pic0e$ up by %a-i#g a g$ l0 at eac% c) # t%e 1A
t 6; $ay list5 especially i# t%e 'r#i#g5 because #ig%t4ti'e acti-ity cul$ be5 a#$
usually is 'isse$. Suc% clse e"a'i#ati# ! e-ery a#i'al i# t%e %er$ is #eit%er
practicable #r #ecessary.
Cows due for drying-off:
T l#g a $ry peri$ is '#ey )aste$. T%ere!re %ig%er pr$ucers
'ust #t be $rie$ !! t s# r t late. O#ce a c) %as bee#
c#!ir'e$ i# cal!5 %er $ry4!! $ate ca# be e#tere$ # %er car$ a#$ at
t%e start ! eac% '#t% $ry !! $ates ca# be 'ar0e$ # t%e c) byre
s%eet. T%e c) cale#$ar is #t accurate e#ug% t be use$ as t%e
pri'ary surce ! $ryi#g !! i#!r'ati#. Use t%e cal-i#g cale#$ar !
Table 1 t )r0 ut t%e $ryi#g4!! $ates5 a#$ e#ter t%ese # t%e c)
byre s%eet s t%at c)s are $rie$ !! # t%e crrect $ay.

Cows and heifers due to calve:
A#i'als )%ic% are e"pecte$ t cal-e 'ust be care!ully )atc%e$ a#$ a#y prble's
'ust recei-e atte#ti#. Due $ates are easily rea$ !r' Appe#$i" Table 1 a#$
e#tere$ # t%e i#$i-i$ual c) car$s a#$ # t%e c) byre s%eet. T%e c) cale#$ar
ca# als be use$ t ge#erate t%is list. O#e s%ul$ e"pect c)s t cal-e up t a
)ee0 early r late. T%ere!re5 c)s s%ul$ appear # t%is list !r' se-e# $ays
prir t t%e e"pecte$ cal-i#g $ate5 a#$ re'ai# # it u#til t%ey cal-e.

2ee0ly acti# lists
Eeekly activities will include dosings of younger animals, changes in feeding groups of
growing animals and of cows, heifers now ready for insemination etc. Buch of this can be
read off the cow and heifer calendars. The alternative is regular perusal of the cards and the
compiling of lists for the dairy.

M#t%ly acti# lists
+n essential monthly activity ,possibly fortnightly in large herds/ is the routine veterinary
e*amination. The following animals must be identified for this monthly visit91
Cows and heifers for pregnancy (P) testing0 i.e. cows which have gone
." days from last insemination without recycling ,list from the cow calendar/.
=ut heifers onto the calendar when first inseminated. ?se, say, the outer
section of the circle so that they can be identified easily and not confused with
the cows. $or large herds of more than !&( cows, buy two calendars and put
heifers and first1calvers on one and the rest of the herd on the other.
Cows for speculu! e"a!inationB4 +ll cows calved or aborted within the
last month ,list from the cow calendar/.
Cows for speculu! re-e"a!inationB4 +ll cows treated ,by uterine
installation, i.e. douched/ by vet. or farmer within the last month. ,List from
cow cards flagged at the time of treatment or diary/.
Cows which have gone #$ days fro! calving without having %een
reported on heat ,list from cow calendar/. These cows will retain their
calving colour on their cow calendar magnets, amongst the rest whose
magnets will have been turned to a different colour to reflect that they have
shown heat.
Cows with open periods e"ceeding &&' days and still not in calf. ,0pen
period : period from calving to successful +.#./ ,list from the cow calendar/.
They will stand out because their cow calendar magnets will be a different
colour to the other cows whose magnet colour will reflect their status as in1
calf. 7eserve part of the calendar, say the centre area, for barren cows ,cows
on the cull list/ so that they do not keep appearing in the action lists, or
remove them from the active area of the board and put them in the top left
hand corner.
Cows with irregular cycles ,listed from the cow cards and +.#. book/.
Cows confir!ed in calf and which have cycled again ,listed from the
calendar and +.#. book/.
(ny other pro%le!s ,listed from the daily diary/.

*ertility a#alysis
The pen and paper system described ,or any other pen and paper system/ re8uires a lot of
hard work to produce herd summaries. These are critically important, as discussed in
FwaGulu1Natal Diary Leaflet ".. Therefore, despite the hard work, a certain minimum
number of analyses are too important to avoid. This is the one area where computers are
of real value. +ll other aspects which have been discussed so far in this chapter do not
re8uire a computer for efficiency. #n fact, for most of the information, a computer could be
less efficient than pen and paper methods. $or e*ample, by the time the computer owner
has loaded the right program, found the right cow, and found out which bull to use to
inseminate her, the cow card user will be home drinking coffee, having completed his
insemination. This e*ample is only slightly e*aggeratedQ

Intercalving period (ICP)
Bean #C= : e*p I,#n C=! H .... H #n C=n/ @ nL H "4( H DE=
where9 C= : breeding period : #C= 1 DE= 1 "4(
n : number of cows
DE= : voluntary waiting period
The best way to check the length of the DE= is to use the open period of the cow with the
second shortest #C=. #n other words, assume that the cow with the shortest #C= was bred
too soon by mistake ,maybe the neighbour's bull got to her/ and use the days open of the
ne*t shortest as the indicator of the DE=. #f it is shorter than e*pected then someone has
been inseminating too soon after calving. ?se +ppendi* Table " to calculate days between
dates. +n e*ample work sheet is illustrated in Table !.

Table 1. S'e e"a'ple calculati#s usi#g t%e rules !r $ays bet)ee# $ates at t%e
!t ! Appe#$i" Table 6
D 8=<
T$ E
Nu'ber ! $ays

*r'ula Arit%'etic
TD 1 LCD H %&
'( H TD 1 LCD
!4" 1 ! : !4!
""" 1 '4 : !..
%& H "&& 1 &5 : "%!
".. 1 !(& H %& : &(.
'( H !4" 1 "& : &4'
NteB LCD : Last calving date but it could be the date of any event
Pulian dates are calculated from +ppendi* Table "

*igure F. Deri-i#g perce#t %eats bser-e$ !r' i#tercal-i#g peri$5 -lu#tary
)aiti#g peri$ a#$ #u'ber ! ser-ices per c#cepti#

Services (SPC) and )otal ()SPC) services per conception
-=C : No. of inseminations to pregnant cows
No. of pregnant cows
T-=C : No. of inseminations performed
No. of calves born
T-=C data are not easy to collect. -ome computer programs use a mathematical techni8ue
to estimate the T-=C. The method is not practicable without a computer. Therefore, the
usual figure 8uoted is services per conception ,-=C/ based on pregnant cows only, a
difference which could be significant. $or e*ample, the cows which fall pregnant could do so
very easily but this could be only a small percentage of the herd, thus -=C on its own could
mask a real problem in the rest of the herd.

Percentage heats spotted
<?nless we see our cows in season the best +.#. techni8ue with the most fertile cows and
bulls won't get our cows pregnant<.
) Meats spotted : ,-=C * "!((/ @ ,#C=1DE=1"'(/
where9 -=C : services per conception ,pregnant cows only/
or read directly from $igure 5

G4su' a#alysis
The R1sum is a simple method of monitoring successes and failures and making
comparisons between seasons, inseminators or even different bulls as well as identifying
trends. + running R1sum should be kept on the wall of the farm office and brought up to
date at each regular veterinary visit. No other analysis gives an earlier warning of fertility
The method is very simple. +ssume that on a particular farm we were able to collect a list
of insemination dates; who did the +.#., and the results of subse8uent pregnancy diagnosis.
+n e*ample of such data is set out in Table " and more fully in $igure '.

Table 6. I#se'i#ati# $ate #ee$e$ !r a G4su' a#alysis
Date !
Result ! (D I#se'i#atr 3ull

These data could probably be read directly from the farmer's usual records, but they must
be in chronological order so that trends can be detected. Mence the +.#. book ,$igure '/.
The data are then plotted as set out in $igure !(, i.e. if an event was successful it is plotted
above the 6ero line andvice versa6 each successive event is added to, or subtracted from,
the one before.

*igure 1H. G4su' +all e-e#ts pltte$,

#n $igure !( the general trend was positive until the !"th even ,insemination/ and then it
began to decline. The trick is to find the reasons for changes in trend, and in the rate of
incline or decline, was it a change of season, a new batch of semen etc.O #n large herds
with several inseminators, keep R1sums for each inseminator as well as for the herd as a
The big advantage of the R1sum is that warnings are sounded within about two months of a
change occurring. This will be a far better indicator that something is going wrong than
with the usual calculations of average intercalving periods, days open, or inseminations per
conception. These other parameters should still be calculated because they are important
#f your reaction is9 <Ehat more recordsQ<; then stop recording something less important.
$or e*ample, how much use do you really make of all that pedigree information so patiently
recorded back to the great grandsire on every cow's cardO

A# O-er-ie)
#t has become very fashionable to computerise dairy her record keeping. + common result
is a mountain of printout which is only looked at when an adviser calls. Bany enthusiastic
computer users have never tried a good pen and paper system. The writer is an
enthusiastic computer user and has one personal computer in his office and another at
home. Nevertheless, it is his firm conviction that, for many farmers, the pen and paper
system described is the better system and that, unless a good pen and paper system is
thoroughly mastered, no farmer or adviser is in a position to evaluate a computerised record
keeping system. Too many computer systems can pull out e*tended pedigrees which could
be looked up more 8uickly in a good filing system, but do not even calculate intercalving
periods for the herd or if they do, do it incorrectly.
The only area where a computer system is superior to a well designed pen and paper and
cow calendar system is in the production of action lists in ,see section <+ction lists</ and the
analyses described in FwaGulu1Natal Dairy leaflet "., the farmer intending to buy a
computerised record keeping program must ensure that features such as a mass1for1age
analysis in the heifers, predicted future daily milk production and future calvings ,very
useful for cash flow budgeting/ are included on the program. +n on1farm computer system
should be able 9
link up with a feeding program
link to out of parlour feeders
produce e*pected milk yields for the ne*t milking
and other practical outputs.
#t is a simple truth that computers do not make life easier. #n fact, they put more demands
on management and, especially, do not take the place of competent advisers. $or these
reasons, many farmers are better served by not owning their own computers, but by
making use of a pen and paper system for everyday management and by subscribing to a
bureau service for analyses and feeding recommendations. +n e*ample of a herd analysis,
part of the C3D+7+ dairy feeding program, is illustrated in $igure !!.

*igure 11. (art ! t%e %er$ a#alysis !r' t%e CEDARA prgra'
+ssessing the value of a cow and making replacement decisions in dairy farming have
important economic implications. The value of a cow is the difference between the future
economic value of a cow and her potential replacement. + positive value means that the
farm will have a greater profit by keeping the cow rather than replacing the cow. + negative
value means that replacing the cow is economically more advantageous than keeping the
cow. $uture economic net benefits that are considered in this difference are9 milk sales, feed
costs, calf value, non1reproductive cull, mortality cost, reproductive cull, reproduction costs
and the cost of the replacement transaction.
Normally, the value of a cow has a greater value early in the lactation and it decreases as
time goes by. This value would continue a decreasing trend and likely become negative if
the cow does not get pregnant within a reasonable time. +n important factor in determining
this value is the assessment of milk productivity in the current and in successive lactations9
the greater the e*pectancy of productivity the more the value of the cow. +lso important is
the e*pectancy of productivity of the replacement animal9 the greater the e*pectancy of
productivity of the replacement ,better genetic 8uality/ the lower the value of the cow. Note
that !(() of milk productivity means productivity at the average cow level and () genetic
gain means that the replacement will produce as an average herd cow.
The present decision support system calculates the value of a cow by responding to factors
such as lactation, months after calving, and months in pregnancy according to e*pected
milk production of the evaluated cow and potential genetic gain of the replacement within
defined herd bio1economic characteristics that can easily be edited in the tab with name
<-ingle Cow +nalysis<. $urthermore, the tab with name <Merd +nalysis< calculates the value
of every single cow in a herd. #n this last tab, the user can download a template
spreadsheet to later upload the information of every single cow in a herd including9 Cow#D,
lactation, days in milk, days in pregnancy, and e*pected milk production for this lactation
and for ne*t lactations. These variables are used in con2unction with the e*pected genetic
improvement of the replacement animal and the herd bio1economic variables. The result of
the herd analysis is a sorted list ,form lowest to largest/ of the economic values of cows in
the current herd.
Heifer Replacement Tool
Herd Size
Min(0) Max(000)
!al" Hei"er !ulling #ate
Min(0) Max($0)
%&erage %ge to 'resh
Min(() Max()0)
%dult !ow !ulling #ate
Min(*) Max(*0)
#estore +e"ault ,alues
#e-uired #eplacement %nimals

1rap%ical Represe#tati#
Tabular Represe#tati#
This application calculates the number of heifers needed as
replacement to maintain constant the herd
si6e in the long run.
#nput variables ,white cells/ re8uired are9
!. The herd si6e in number of adult cows.
". The calf1heifer culling rate including deaths..
. The average month time to first freshen.
.. The adult cow's culling rate including deaths. .
The application calculates the re8uired replacement animals ,blue cell/.
The tool includes sensitivity tables and figures varying three main
!. Cow culling rate
". Calf1heifer culling rate
. +ge to first freshen
Nuidelines for using the application tool9
The values displayed in white bo*es are inputs and can be edited.
The values in grey bo*es are generated based on input values
and cannot be changed directly.
#n order to ease the process of entering the inputs, sliders have
been provided for each input bo*. The sliders can be moved to change
the input values ,white bo*es only/. The slider handles can either be
clicked and dragged or the direction keys can be used to move it left or
3very time an input value changes new values are calculated and
updated in the grey bo*es and graphs.
7eset buttons have been provided in each page to restore default
values for the sliders.
Cow culling rate is varied & and !() above and below input data, calf1
heifer is varied " and .) above and below input data, and age to first
freshen is varied " and . months above and below input data. The
middle point in tables represent the input data ,white cells/, which will
increase with higher cow culling rates, higher calf1heifer culling rates,
and higher months to first freshen. Likewise, it will decrease with lower
culling rates and lower months to first freshen.