Defining health standards through economic optimisation: The example of colostrum management in beef and dairy production
Defining health standards through economic optimisation: The example of colostrum management in beef and dairy production
Didier Raboisson 0 1
Pauline Trillat 0 1
Marie DervilleÂ 0 1
CeÂ lia Cahuzac 0 1
Elise MaigneÂ 0 1
0 1 IHAP, UniversiteÂ de Toulouse, INRA, ENVT , Toulouse, France, 2 LEREPS, UniversiteÂ de Toulouse, ENSFEA, IEP Toulouse, Toulouse, France, 3 ODR, INRA, Toulouse , France
1 Editor: Juan J Loor, University of Illinois, UNITED STATES
There is extensive literature addressing acceptable practices of colostrum distribution to new-born calves; however, no economic analyses are available concerning the profitability of this practice. Moreover, the health standards associated with colostrum management have been defined through the observation of reference farms without explicit reference to economic assessments. The objective of the current study was to evaluate the profitability of farm colostrum management and to define the optimal economic situations for given livestock systems and farm situations. The herd-level net value was calculated using the value of calf products, the cost of passive transfer failure and the cost of prevention. This value was determined for various beef and dairy scenarios and the various time periods spent managing colostrum. The maximal net values defined the optimal economic situations and enabled the determination of the optimal times for colostrum management and respective health standards (i.e., the prevalence of disorders at optimum). The results showed that the optimal time farmers should spend on colostrum management is approximately 15 min per calf. Furthermore, farmers should err on the side of spending too much time (> 15 min) on colostrum management rather than not enough, unless the cost of labour is high. This is all the more true that potential long term consequences of passive transfer failure on milk yields were not accounted for here due to scarcity of data, leading to consider this time threshold (15 min) as a minimal recommendation. This potential underestimation may arise from the greater nutrient content and bioactive compounds identified in colostrum although the passive immune transfer is here defined through immunoglobulins only. The present results show that for small farms that cannot hire colostrum managers, this work can be performed by the farmer after subcontracting other tasks. Moreover, the method proposed hereÐthe definition of health standards through economic optimisationÐis a promising approach to analysing health conventions in the cattle industry.
Competing interests: The authors have declared
that no competing interests exist.
Abbreviations: CI, confidence interval; FPT, failure
of passive transfer; PFPT, prevalence of failure of
passive transfer; RR, relative risk.
The failure of neonatal calves to absorb adequate colostral immunoglobulins within the first
hours of life results in the failure of passive transfer (FPT). Depending on how FPT and
livestock systems are defined, approximately 20 to 40% of new-born calves experience FPT [1,2].
FPT leads to an increased risk of mortality as well as decreased health and longevity. A recent
meta-analysis reported the adjusted risks (and 95% confidence intervals [CIs]) for mortality,
bovine respiratory disease, diarrhoea and overall morbidity associated with FPT as 2.12 (1.43±
3.13), 1.75 (1.50±2.03), 1.51 (1.05±2.17) and 1.91 (1.63±2.24), respectively .
Ensuring that calves drink enough colostrum within a few hours of birth is a powerful
way to reduce FPT and its associated disorders. This advice is evidence-based [4±6] and is
found consistently in books, papers and technical communications regarding the
management of young livestock, calf morbidity and new-born nursing. Several practical guidelines
to prevent FPT have been proposed for use on farms [7,8]. The management practices that
are risk factors for FPT are also well known [1,9,10]. The minimum quantity of
immunoglobulins that a calf needs to absorb within a few hours after birth to prevent FPT is
approximately 150 g . Other bioactive compounds of colostrum, such as chemokines and
immune cells, are recognised as playing a role in neonatal calf immunity, butÐfor practical
reasonsÐthey are not used in the field to evaluate FPT . The average marginal cost of
FPT was recently assessed as €60±121 for dairy calves and €80±140 for beef calves . There
is a dramatic lack of economic consideration regarding the definition of the resources
allocated to managing new-born calves, especially with regard to the stakes of colostrum
ingestion. The aforementioned extensive literature concerning the risk of FPT does not include
The causes of calf mortality and morbidity are multifactorial; however, colostrum
distribution remains a key factor that influences these parameters. The evaluation of cattle health
status on farms includes mortality and morbidity rates, and acceptable thresholds are often
proposed in the grey literature. These technical indicators are used as objectives to guide farm
advisers and farmers in farm management. In this paper, we propose that these objective
thresholds be considered as health standards. The health standards proposed here do not refer
to private or legal regulations; rather, they are defined as an accepted or approved target
against which others are judged or measured. These health standards are then considered as
the ªgold standardº or as ªreferencesº for comparisons between farms. The concept underlying
the health standards proposed here approximates the definition of the quality convention
observed in institutional economics i.e., the formal or informal habits potentially linked to
given routines that enable efficient daily activities through the better coordination of actors
. Health conventions represent a set of health standards in a given situation. In accordance
with their definition, they differ greatly among livestock systems, including breeds, feeding
systems, soils, climates and supply chains.
The definition of these health standards is obvious, at least for ruminants. Such standards
are often defined by experts who examine the best farms within the same type of livestock
system. These farms, sometimes called ªreference farmsº, often represent farms with satisfactory
technical results; however, the farm's financial situation is not considered as the key driver
when defining health standards. The way in which health standards are defined indicates that
some level of disease is acceptable and that a total lack of disease is not only unrealistic on
most farms, it is not an objective. This method has a certain value in this context. However, it
is of interest to evaluate the health standards associated with the economic optimum proposed
in the present study. Our objective was to determine the economic reasoning regarding
newborn calf health standards, including the resources allocated to colostrum management.
Materials and methods Economic model
Our economic reasoning was based on the change in the net value of farm production for a
one-year production cycle. The outcome variable of the economic model (NetValue) was the
net value for a herd (h) with a given prevalence (p) of FPT (PFPT) and a given amount of time
(t) spent on colostrum management, as indicated in Eq 1:
Products,h,p = 0 = the sum of the calf products for the herd (h) when PFPT = 0
CostPreventionFPT,ht = the cost of prevention linked to FPT for the herd (h) and time (t) spent
on each calf
CostFPT,hp = the cost of FPT for the herd (h) and PFPT (p)
MarginalCostFPT = the cost for a calf with FPT compared to no FPT
Products,h,p = 0 was calculated as indicated in Eq (3) for dairy (ProductsDairy,h,p = 0) or beef
(ProductsBeef,h,p = 0). This value stays fixed at PFPT = 0 for a given herd because MarginalCostFPT
includes losses not sold. The present model considers various values of MarginalCostFPT, but
FPT remained defined as binary (Yes/No), including all the expected consequences of FPT if
ProductsBeef ;h;p0 SellingPrice
SellingPrice and SellingWeight = the selling price and weight of beef calves
ConcPrice ConcQuantity = the price and quantity of the concentrate eaten for each calf
from birth to selling
MarketValue = the average value of a dairy calf on the market (half male and half female)
CostPreventionFPT,ht was defined based on the minimum legal French salary income, taxes
included (64%), as indicated in Eq (5):
NetSalaryHour = the minimal legal French wage per hour
Flex = the coefficient of flexibility accounting for time lost because the work performed is
discontinuous (i.e., depends on the timing of the calvings)
The production function linking time (t) and PFPT (p) is explained in Eq (6):
PFPT,h,t = 0 = PFTP for no time spent by the farmer on colostrum management (t = 0).
Efft = the ability of the farmer to reduce PFPT on average for a given time (t)
The term PFPT,h,t = 0 does not denote a lack of FPT but that PFPT is observed when the calf is
left alone without the intervention of the farmer. PFPT,h,t = 0 is a characteristic of the livestock
system because PFPT without farmer intervention can differ between systems. Importantly, the
time spent by farmers enables a reduced number of calves with FPT compared to no time
devoted, and the changes in related costs are obtained through Eq 2, considering the
MarginalCostFPT for each additional or fewer calf with FPT. The model does not consider changes in
mortality or morbidity only, nor does it consider changes in only one item of mobility; instead,
it considers an average improvement of all these consequences of FPT in the case of better
colostrum management through more time spent.
The mortality rate and the prevalence of morbidity (both denoted PMo) were calculated as a
function of t for herd h as the sum of the prevalence of the disorder in the population with FPT
(left, first bracket) and the prevalence of the disorder in the population without FPT (right,
second bracket) using Eqs (7) and (8):
PMo_FPT = the within-herd prevalence of mortality or morbidity in calves with FPT
PMo_NoFPT = the within-herd prevalence of mortality or morbidity in calves without FPT
RRMo = the relative risk of mortality or morbidity in calves with FPT compared with calves
The model was conducted using the open source software Scilab (www.scilab.org) with
10,000 iterations, and 95% prediction intervals (PIs) were calculated.
The input parameters are summarised in Table 1. All calculations were performed based on an
average herd, with 100 calves born. MarginalCostFPT has been assessed in a previous study ,
and the 2 values retained for dairy and beef (for both the Baseline and Alternative scenarios)
were considered to represent the most plausible situations farmers have to face.
In brief, the evaluation of MarginalCostFPT was based on the rationale that a calf with FPT is
more likely to face mortality, morbidity and decreased performances compared to a calf
without FPT. Because FPT is subclinical by definition, its cost was defined as the differences in the
economic consequences between calves with and without FPT. The meta-analysis allowed to
precisely determine the odd ratio linking FPT and its outcomes (mortality, diarrhoea,
respiratory diseases, omphalitis, septicaemia, average daily gain change). Many scenarios were
proposed in the stochastic economic evaluation, including several prices for extra input and lost
outputs, and confidence and prediction intervals of the costs were provided . The 4 values
MarginalCostFPT retained here represent the most frequent situations observed in the field
under European breeding systems. MarginalCostFPT has been evaluated using parameters from
S1 and S2 Tables: the differences between the 2 scenarios included (i) the diseases induced by
FPT that are considered or not considered (no omphalitis and septicaemia for baseline), (ii)
the differences in input parameters (relative risks and unit costs; S2 Table), (iii) the way some
unit costs were calculated (mortality for beef, daily breeding cost for dairy; S2 Table) and (iv)
the weight-adjusted therapy costs for diseases (S1 Table). Due to the scarcity of evidence, the
long term consequences of FPT on the milk production of dairy cows were not included in the
present evaluation, nor were other types of performance of in milk cows.
CostPreventionFPT,ht was defined as reported in Eq (5), with Net SalaryHour and Flex as defined
in Table 1 for scenarios S1 to S3. This value was also fixed for S4 to S7, as indicated in Fig 1.
Efft was defined as indicated in Fig 2 based on the data available in the literature. Efft first
increased slowly with t because the quantity of immunoglobulins was not expected to be
sufficient for full efficacy. The slope then increased as follows: the maximal marginal efficacy of the
time spent to decrease the default FPT was expected because of the maximal marginal effect of
the immunoglobulins ingested. Pinocytosis is expected to be high given the large quantity of
IgG absorbed and present in the blood. Then, the effect plateaus, such that more time spent
increases colostrum ingestion; however, the expected improvement in FPT is reduced. The
data in the literature related to the concept of early colostrum ingestion, large quantities of
ingested colostrum, and adequate hygiene regarding colostrum management suggest the
format of the production curve. The location of the curve's inflections on the x-axis and the value
at which the curve peaks were difficult to address because few data were available. The
threshold of maximal efficacy was fixed at approximately 15±20 min for colostrum distribution, as
Fig 1. The cumulative cost of labour (time spent on colostrum management) for a given herd under several scenarios.
suggested by the average time for a calf to ingest enough colostrum to reach adequate immune
transfer. Calves ingest 3 litres of colostrum in 15 min between 1 and 4 hours after birth when
the colostrum is delivered via a feeding bottle . This duration is the maximal time a farmer
should spend on colostrum distribution to a calf. Another study reported that 20 min were
needed for calves to ingest 2.5 litres of colostrum 1 hour after birth via a feeding bottle;
however, one third of calves refused to suckle this quantity within one hour . This finding
suggests that the value of Efft be capped at 0.9. Similarly, the time needed to intake 2.5 litres of
colostrum varies from a few minutes to one hour . This variation suggests the inclusion of
a wide range for the function of Efft, as indicated by the dotted lines (Fig 2).
The results clearly showed that the profiles of the net value curves (Figs 3 to 5) have a shape
close to Efft (Fig 2), demonstrating that the cost of labour (scenarios S1 to S7) has a small effect
on the net value. The net value dramatically increased, reached its maximal value for
t = optimal time (approximately 14±16 min) and then decreased slowly. The slight decrease
after the optimal time confirmed the small effect of the labour cost on the net value and
demonstrated that the profitability of the time spent on colostrum management is not reduced
when more than the optimal time is spent. The negative slope of the net value curve after the
economic optimum was more pronounced (i) for S2 and S3 than it was for S1 (Fig 3); (ii) for
Fig 2. The expected efficacy (Efft) of the time spent on colostrum management to reduce FPT. The dotted lines represent the ranges of the function
low PFPT,h,t = 0 than it was for high PFPT,h,t = 0 (Figs 4 and 5) and (iii) for the scenario baseline
than it was for scenario alternatives (S1, S2 and S3 Figs). The net return between the situations
without any time spent on colostrum management and the optimal time varied depending on
whether beef versus dairy was considered, the cost of labour and PFPT,h,t = 0.
The optimal time spent on colostrum distribution and the respective PFPT, PMortality,
PRespiratory and PDiarrhoea values are reported in Tables 2 and 3. As expected, PMortality, PRespiratory
and PDiarrhoea were lower for the optimal time and for optimal PFPT than they were for t = 0 for
any given PFPT,h,t = 0. No differences were observed in any of the indicators between S1 and S4
to S7, and only S1 was reported. The optimal time spent on colostrum management was 14±18
min, after considering all of the modelled situations (Tables 2 and 3).
The results clearly show that break-even points were reached for different combinations of
(i) baseline/alternative scenarios, (ii) labour costs (S1, S2 or S3) and (iii) the prevalence of FPT
for no time spent (PFPT,h,t = 0 = 30, 50 or 80%). The break-even points represent combinations
of parameters that are associated with no marginal profitability of allocating resources (time)
for colostrum management. The break-even points represent the situations beyond which
extra time was associated with a decreased prevalence of disorders and increased optimal PFPT.
For the baseline scenario (beef, Table 2 or dairy, Table 3), the break-even points were reached
for S2, S3, and PFPT,h,t = 0 = 30%, with a lower optimal time and higher optimal PFPT than those
for S1. The break-even point was only reached for S3 for PFPT,h,t = 0 = 50% and was never
reached for PFPT,h,t = 0 = 80%. For the scenario alternative (with, among other characteristics, a
higher marginal cost of FPT, Table 1), the break-even point was reached for higher values of
Fig 3. The net value for the time spent on colostrum management expected for the baseline scenarios, the cost of labour S1 and S3 and the
prevalence of FPT without an intervention (PFPT,h,t = 0) of 50%. The dotted lines represent the results with the ranges of the function curve Efft (Fig
PFPT = 0 and higher costs of labour (Table 3). For example, the break-even point for PFPT = 0 =
50% and beef was not reached here, although it was achieved for S3 for the baseline scenarios.
The results were not sensitive to Products,h,p = 0 for beef, with slight changes in the net value
for changes in SellingPrice, SellingWeight, ConcPrice and ConcQuantity (Table 1).
Conversely, the results were sensitive to Products,h,p = 0 for dairy. The changes in the net value for
MarketValue = N(125,9) or N(330,15) were more important; however, the change in the
optimal time was not affected (S4 Fig).
Allocation of resources for colostrum management
The present study clearly demonstrated the profitability of colostrum management for dairy
and beef farms. Despite the diversity of the situations analysed, the results are homogeneous
and can be summarised as follows. First, the optimal time to be spent on colostrum
management is approximately 15 min per calf (14±18 min depending on field situations). Second,
farmers should err on the side of spending too much time (> 14±18 min) on colostrum
management rather than not enough, except when the cost of labour is high. The fact that potential
long term consequences of FPT (milk yields) were not accounted for due to scarcity of data
suggests to consider this time threshold as a minimal recommendation. The greater nutrient
content and bioactive compounds identified in colostrum may support other potential benefits
of colostrum ingestion that have not been clearly described up to now and have not been
Fig 4. The net value for the time spent on colostrum management expected for the baseline scenarios, the cost of labour S1 and S2 and the
prevalence of FPT without an intervention (PFPT,h,t = 0) of 30%. The dotted lines represent the results with the ranges of the function curve Efft (Fig
included in the present study. Although many studies, reviews and reports have highlighted
the usefulness of adequate immune transfer through the intake of large quantities of
high-quality colostrum shortly after birth, surprisingly, no previous economic assessments have been
The major limitation of the proposed model is the synchronisation between demand and
the offer of labour provided for colostrum management at the farm level. The proposed models
were based on the minimum salary in France, tax included, and we assumed that the farmer
would only pay for the worker's time spent on colostrum management. Because of limited
herd sizes and the unpredictable events of birth, this situation is not realistic. Scenarios S2 and
S3 addressed this point, with one-half and two-thirds of the time (salary cost) paid without any
work performed. In this situation, the optimal time spent on colostrum management was
slightly reduced; however, the primary conclusion did not change.
The present work focused on the profitability of colostrum management. The results should
not be confused with the total cost of FPT . For a given farm, the money saved based on the
total costs associated with FPT will always be higher than the net value calculated here because
the total costs do not include the expenditures needed to achieve better health situations.
Farmers and farm advisers often confuse these assessments. This confusion is particularly
pronounced for small-to-moderate size herds. We propose using the economic reasoning of
opportunity cost in this situation. Because employing a person for only 15 min per calf born is
impossible for these farms, it might be sufficient to rely on the farmer or a permanent
Fig 5. The net value for the time spent on colostrum management expected for the baseline scenarios, the cost of labour S1 and S3 and the
prevalence of FPT without an intervention (PFPT,h,t = 0) of 80%. The dotted lines represent the results with the ranges of the function curve Efft (Fig
employee to conduct colostrum management, which should be considered a strict priority.
Then, an external firm or service could perform other work on the farm (e.g., manure
spreading). The present assessment was based on the profitability of a measure, and it is only valid
when extra labour must be paid, either directly or indirectly. Thus, if the farmer has to spend
more time on colostrum management, then the cost of labour included in the present study
might merely represent the substitution of work because other types of work must be
subcontracted. Otherwise, asking the farmer to spend more time on colostrum management without
any decrease in other work indicates a zero cost for his or her labour, and the reasoning will be
based on the total cost of FPT. Thus, economic reasoning is biased by the lack of consideration
of the opportunity cost related to the time spent on colostrum management. In other words,
the profitability of advice is overestimated in accordance with the assumption that no
expenditures are spent on achieving a better output. For most modern farmers, it is not reasonable to
dismiss the opportunity costs. Spending an extra 15 min per new-born calf is a large
investment of time, even for moderate or small farms. Several popular examples (e.g., time spent on
oestrus detection, the detection of ill animals after arrival in the feedlot, general hygiene in the
barns) clarify that opportunity cost is not a theoretical construct; rather, the working time of
farmers is limited, and not accounting for opportunity costs leads farmers to switch from these
key but time-consuming activities to other activities. All this questioning around opportunity
costs suggests the need to prioritize an approach based on maximum net value over a formal
cost-benefit analysis, even if the latter approach may have been possible.
T = 0
T = 0
Beef, scenario Baseline
Importantly, the present study did not precisely define the tasks to be performed during the
15 min time slot. These tasks might vary among farmers, breeds, seasons and livestock systems.
They may include not only colostrum feeding but also colostrum management, including
colostrum quality checks, feeding colostrum from another dam, nursing with or feeding with a
bottle or tube, and colostrum storage. The current agreed-upon international
recommendation is to provide at least 150 g of IgG to new-born calves within the first hours of life .
Allowing calves fed by nipple bottle to ingest as much colostrum as they can within 12 hours of
age, with at least 2 L and 4 L of colostrum within 4 and 12 hours of age, respectively, appears to
be a good rule to control FPT . Importantly, leaving the calf with its dam for an extended
period is inferior to feeding colostrum directly to prevent FPT . Whatever is done during
this time slot, the present results are valid under the assumption that a 90% decrease in FPT is
achieved for 15 min spent per calf (Efft, Fig 2). The extensive available literature might be used
to help attain this goal . The belief that colostrum management does not take time away
from beef production because the calf is let with the dam is based on extensive farming systems
but cannot be extrapolated to all beef production. The stakes of colostrum intake in beef
productionÐin which all the revenue depends on the calfÐmake the time spent on this item not
as low as is often expected. The fact that the same optimal time is found for dairy and beef
calves does not mean that the time spent is used in similar ways. Because the labour costs did
not have a substantial influence on the results, and because the labour costs refer to France
(i.e., they are comparatively high), an easy extrapolation of the results from FranceÐto Europe
and elsewhereÐcan be easily performed, whatever the herd size. The fact that the results were
strongly influenced by the efficacy (Efft, Fig 2) suggests an easy extrapolation of the results to
farms with a different efficacy curve, provided that this curve is well-known (for instance, for
large farms). In a case where the 90% decrease in FPT (plateau of the curve, Fig 2) is obtained,
on average, before 15 min, then the optimal time to allocate for colostrum management may
decrease similarly (provided the labour costs are lower than the French labour costs). In a case
where the 90% decrease in FPT (Fig 2) is obtained, on average, after 15 min, then the optimal
time to allocate cannot be predicted because it will also depend on the labour costs.
The shape of the Efft curve (Fig 2) was based on the reasoning about IgG pinocytosis as
stated above and did not include the other bioactive compounds of colostrum. These
compounds are not likely to justify any different shape of the curve; in particular, they do not
generate more efficient action for the first few minutes of time spent. The usefulness of these
bioactive compounds is now agreed-upon, but their mode of action remains speculative, and
they have not been demonstrated to strongly reduce the usual FPT consequences in the case of
limited colostrum ingestion, i.e., spending only a few minutes, as in in Fig 2.
The results depend strongly on the value of PFPT,h,t = 0. The 3 situations proposed should be
considered as intrinsic characteristics of livestock systems or of subpopulations of animals,
depending on their risk of FPT. For example, an extensive beef livestock system with easy
births or a low prevalence of dystocia might correspond to PFPT,h,t = 0 = 30%. In contrast, blue
Belgian or Charolais calves after difficult births or late caesarean sections might have a high
risk of FPT without a specific intervention from the farmer (PFPT,h,t = 0 = 80%). However, the
present results are probably not of interest for very extensive beef livestock systems with calves
born outside of the surveillance of the farmer and with no intervention at all by the farmer
with regard to new-born calves.
Methods and calibration
A weakness regarding the economics of bovine health arises from the difficulties of quantifying
the effects if using alternative and substitute practices. A lack of epidemiologic data limits our
understanding of the economics of animal health. This situation is all the more true for open
livestock systems with long cycles of production and limited sizes, such as bovine production
systems in major European countries. Among the methodological possibilities of addressing
this issue, a cost effectiveness analysis might enable the comparison of measures without any
reference to their financial cost. This approach might be useful for addressing the problem of
the cost of labour on small or moderate sized farms, where unofficial familial labour might be
available. The production function that defines work efficacy remains difficult, and this
complication might hinder the modelling of a cost-effectiveness analysis, including the comparison
of 2 competitive time-consuming practices. The present study addressed this issue by defining
the break-even points associated with several combinations of situations. This allowed for the
definition of the limit of profitability of the time spent at a given labour cost or the cost of
labour under which the measure is profitable.
Because FPT appears at birth and generally has short-term consequences, the static
approach proposed is likely adequate. A dynamic model might be useful but would lead to
significant difficulties in calibration. Because cattle production systems are complex, with long
production cycles and low levels of information regarding the production function and inputs
at the farm level (especially the difficulties in assessing diet costs), a partial economic reasoning
is proposed here. Reasoning through opportunity costÐas proposed aboveÐsolves the issue
of a partial analysis, enabling substitutions among the inputs of different farm units and
authorising a more holistic farm-level reasoning.
Health standards defined through economic optimisation
Economic optimisation should be considered as a reference method for defining health
standards. Despite the limitations mentioned, the present study proposed various indicators as
health standards (Table 2) for different situations, including livestock systems (PFPT,h,t = 0),
labour costs (S1, S2 and S3) and product prices (Table 3). These results can be used in the field
to assess and improve the situation of a given herd in a particular context.
The proposed approach offers advantages over preferences based on observations and
expert opinion. This way of reasoning is similar to how biological or medical standards, such
as blood parameters, are traditionally defined, with the observation of a healthy population
and ranges defined with the exclusion of the 5% extreme values. On the one hand, in the
absence of exhaustive datasets, this method depends heavily on the farms selected for
observation and is not sensitive to the contextual market situation. On the other hand, the proposed
approach might help farmers decide how to adapt their management practices to market
changes because health standards account for input and output prices. For a given livestock
system (PFTP,h,t = 0 = 30% for instance), Optimal_PFPT changed depending on the labour costs
(Tables 2 and 3) and the prices of products (S4 Fig). The use of observational methods (and
not economic optimisation reasoning) to define health standards might not allow for reactions
to agricultural price volatility, which is an increasing tendency in a liberalised economy.
Pmortality and Pmorbidity were not highly sensitive to changes in labour cost or product price.
Pmortality and Pmorbidity values corresponding to the optimal economic situation but differing
between different situations of t = 0 (Tables 2 and 3) could be retained as health standards in
the corresponding livestock systems.
The production of such health conventions, as a shared and legitimised understanding of
the activity, is part of the responsibility of a professional group [19,20]. Many health topics still
lack agreed-upon health conventions; thus, not all actors pursue the same sought-after health
standards. The production of economic indicators corresponds to the elaboration of an
evaluation device and might contribute to a legitimisation process. This production might facilitate
coordination among farmers, their advisers and downstream agents in the supply chain. From
this perspective, the proposed approach might be useful to define health conventions and health
standards for various health topics. One example of interest is the convention concerning the
use of antimicrobials in farms. The use of antimicrobials significantly differs among farms, and
the correlation between the use of antibiotics and the level of disease is low . The acceptable
levels of use of antimicrobials must be determined from societal, political, and micro-economic
(farm-level) points of view. Studies of trade-offs between health standards (mortality and
morbidity) and standards of antimicrobial use have to be performed. The results of an optimisation
of antimicrobial use may produce different health standards from those proposed here.
Decreases in FPT are expected to be associated with a decrease in antimicrobial use. Because
the present results suggest an optimal situation at low FPT, such a situation is expected to
already match with low antimicrobial use. In cases of diseases among pre-weaned calves,
antibiotic therapies were demonstrated to limit mortality and morbidity compared to non-antibiotic
therapies . Non-antibiotic prevention, such as colostrum distribution, remains a powerful
way to reduce antibiotic use without any deterioration in operational performances.
Somatic cell counts are another area of interest in which a health convention should be
defined with economic consideration. The proposed approach might help to assess which
thresholds represent the farm-level economic optimum when a decrease in yield exists for
cows with high somatic cell counts  and at which thresholds extra culling would be based
on somatic cell counts. The same reasoning, applied to dairy manufacturers, accounts for the
decrease in yield processing for high SCC milk and might enable the comparison of optimal
somatic cell count thresholds for farmers or manufacturers .
The present study showed that the optimal time to be spent on colostrum management is
approximately 15 min per calf. Furthermore, farmers should err on the side of spending too
much time (> 15 min) on colostrum management rather than not enough, unless the cost of
labour is high. For small farms with no possibility of part-time employment, the farmer can
perform colostrum management if his or her other tasks are easier to subcontract. The method
proposedÐthe definition of health standards through economic optimisationÐis a promising
approach to analysing heath conventions in the cattle industry. Further work is needed to
analyse how vaccination and FPT may interact economically.
S2 Table. Input parameters used to calculate the marginal cost of FPT (MarginalCostFPT).
S1 Fig. The net value for expected time spent on colostrum management for the alternative
scenarios, labour costs S1 and S3 and the prevalence of FPT without an intervention
(PFPTt = 0) of 50%.
S2 Fig. The net value for the expected time spent on colostrum management for the
alternative scenarios, the labour costs S1 and S3 and the prevalence of FPT without an
intervention (PFPT,h,t = 0) of 30%.
S3 Fig. The net value for the expected time spent on colostrum management for the
alternative scenarios, the labour cost S1 and S3 and the prevalence of FPT without an
intervention (PFPT,h,t = 0) of 80%.
S4 Fig. The net value for the expected time spent on colostrum management for the
alternative scenarios, the labour costs S1 and S3, the prevalence of FPT without an intervention
(PFPT,h,t = 0) of 50% and a market value for dairy of N(330,15) instead of N(125,9).
Investigation: Pauline Trillat, CeÂlia Cahuzac.
Methodology: Didier Raboisson, Marie DervilleÂ.
Project administration: Didier Raboisson, Elise MaigneÂ.
Supervision: Elise MaigneÂ.
Validation: Didier Raboisson, Marie DervilleÂ.
Visualization: Pauline Trillat, CeÂlia Cahuzac.
Writing ± original draft: Didier Raboisson.
Writing ± review & editing: Pauline Trillat, Marie DervilleÂ, CeÂlia Cahuzac, Elise MaigneÂ.
1. Beam AL , Lombard JE , Kopral CA , Garber LP , Winter AL , Hicks JA , et al. Prevalence of failure of passive transfer of immunity in newborn heifer calves and associated management practices on US dairy operations . J Dairy Sci . 2009 ; 92 : 3973±3980. https://doi.org/10.3168/jds. 2009-2225 PMID: 19620681
2. National Animal Health Monotoring Sytem (NAHMS). National Dairy Heifer Evaluation Project . Dairy herd managment practices focusing on preweaned heifers . Fort Collins: USDA-APHIS Veterinary Services ; 1993 .
3. Raboisson D , Trillat P , Cahuzac C. Failure of passive immune transfer in calves: A meta-analysis on the consequences and assessment of the economic impact . PLoS One . 2016 ; 11 : e0150452. https:// doi.org/10.1371/journal. pone.0150452 PMID: 26986832
4. Chigerwe M , Coons DM , Hagey JV . Comparison of colostrum feeding by nipple bottle versus oroesophageal tubing in Holstein dairy bull calves . J Am Vet Med Assoc . 2012 ; 241 : 104±109. https://doi.org/ 10.2460/javma.241.1. 104 PMID: 22720994
5. Chigerwe M , Tyler JW , Middleton JR , Spain JN , Dill JS , Steevens BJ . Comparison of four methods to assess colostral IgG concentration in dairy cows . J Am Vet Med Assoc . 2008 ; 233 : 761±766. https://doi. org/10.2460/javma.233.5. 761 PMID: 18764714
6. Chigerwe M , Tyler JW , Summers MK , Middleton JR , Schultz LG , Nagy DW . Evaluation of factors affecting serum IgG concentrations in bottle-fed calves . J Am Vet Med Assoc . 2009 ; 234 : 785±789. https:// doi.org/10.2460/javma.234.6. 785 PMID: 19284346
7. Pritchett LC , Gay CC , Besser TE , Hancock DD . Management and production factors influencing immunoglobulin G1 concentration in colostrum from Holstein cows . J Dairy Sci . 1991 ; 74 : 2336±2341. https:// doi.org/10.3168/jds. S0022-0302(91)78406-3 PMID: 1894821
8. Godden S. Colostrum management for dairy calves . Vet Clin North Am Food Anim Pract . 2008 ; 24 : 19± 39. https://doi.org/10.1016/j.cvfa. 2007 . 10.005 PMID: 18299030
9. Quigley JD 3rd, Drewry JJ . Nutrient and immunity transfer from cow to calf pre- and postcalving . J Dairy Sci . 1998 ; 81 : 2779±2790. https://doi.org/10.3168/jds. S0022-0302(98)75836-9 PMID: 9812284
10. Trotz-Williams LA , Leslie KE , Peregrine AS . Passive immunity in Ontario dairy calves and investigation of its association with calf management practices . J Dairy Sci . 2008 ; 91 : 3840±3849. https://doi.org/10. 3168/jds. 2007-0898 PMID: 18832206
11. Chigerwe M , Tyler JW , Schultz LG , Middleton JR , Steevens BJ , Spain JN . Effect of colostrum administration by use of oroesophageal intubation on serum IgG concentrations in Holstein bull calves . Am J Vet Res . 2008 ; 69 : 1158±1163. https://doi.org/10.2460/ajvr.69.9. 1158 PMID: 18764687
12. Blum JW , Baumrucker CR . Colostral and milk insulin-like growth factors and related substances: Mammary gland and neonatal (intestinal and systemic) targets . Domest Anim Endocrinol . 2002 ; 23 : 101± 110. PMID: 12142230
13. Raboisson D. Approche institutionnelle de la santeÂ animale: place des compeÂtences, des territoires et des collectifs dans la gestion de la santeÂ des bovins en France . Toulouse: UniversiteÂ de Toulouse, INPT ; 2011 .
14. Faber SN , Faber NE , McCauley TC , Ax RL . Case study: Effects of colostrum ingestion on lactational performance . The Professional Animal Scientist . 2005 ; 21 : 420± 425 .
15. Besser TE , Gay CC , Pritchett L. Comparison of three methods of feeding colostrum to dairy calves . J Am Vet Med Assoc . 1991 ; 198 : 419±422. PMID: 1901305
16. Gay C. Colostrum research says feed 4 quarts for healthier calves . Hoard's Dairyman . 1994 ; 139 : 256 .
17. Hedegaard CJ , Heegaard PM . Passive immunisation, an old idea revisited: Basic principles and application to modern animal production systems . Vet Immunol Immunopathol . 2016 ; 174 : 50±63. https:// doi.org/10.1016/j.vetimm. 2016 . 04.007 PMID: 27185263
18. McAloon CG , Whyte P , O'Grady L , Lorenz I , Green MG , Hogan I , et al. Relationship between selected perinatal paratuberculosis management interventions and passive transfer of immunity in dairy calves . Vet Rec . 2016 ; 179 : 47. https://doi.org/10.1136/vr.103547 PMID: 27259758
19. Allaire G . Conventions professionnelles et reÂgimes de responsabiliteÂ . In: Eymard-Duvernay F, editor. L' economie des conventions: meÂthodes et reÂsultats . Paris: La deÂcouverte; 2006 . pp. 279 ± 293 .
20. Favereau O , Lazega E. Conventions and structures in economic organization: Markets, networks and hierarchy . Cheltenham: Edward Elgar ; 2002 .
21. Lhermie G , Raboisson D , Krebs S , Dupraz P. Facteurs deÂterminants et leviers de reÂduction de l'usage des antibiotiques en productions animales . Econ Rural . 2015 ; 348 : 3± 22 .
22. Berge AC , Lindeque P , Moore DA , Sischo WM . A clinical trial evaluating prophylactic and therapeutic antibiotic use on health and performance of preweaned calves . J Dairy Sci . 2005 ; 88 : 2166±2177. https://doi.org/10.3168/jds. S0022-0302(05)72892-7 PMID: 15905446
23. Hortet P , Seegers H. Calculated milk production losses associated with elevated somatic cell counts in dairy cows: Review and critical discussion . Vet Res . 1998 ; 29 : 497±510. PMID: 9851006
24. Smith L , Hogan JD . Milk qualityÐa worldwide perspective. 1998 [cited . In: National Mastitis Council 1998 Annual Meeting Proceedings [Internet] . http://www.nmconline.org/articles/keynote98.htm.
25. URSSAF. URSSAF. 2016 [ cited 2016] . In: URSSAF [Internet]. https://www.urssaf.fr/portail/home/tauxet-baremes/smic.html. Access on April 2015 .
26. French Livestock Institut [Insitut de l'Elevage]. ReÂsultats 2011 des exploitations bovins lait . In: ReÂsultats nationaux. Collection reÂsultats annuels. ReÂseaux d'eÂlevage pour le conseil et la prospective . 2013 . pp. 52 .
27. French Livestock Institut [Insitut de l'Elevage]. ReÂsultats 2011 des exploitations bovins viande . Paris: French Livestock Institut [Insitut de l'Elevage]; 2013 .