A MECHANISTIC NUTRITION MODEL
TO EVALUATE BEEF COW EFFICIENCY
Luis Orlindo Tedeschi *,1, Danny G. Fox1, Michael J. Baker1, and Keith L. Long2
1Department of Animal Science, Cornell University, Ithaca, NY 14850
2Bell Ranch, Solano, NM 87746
*Corresponding author: Email: lot1@cornell.edu, Phone: 607-255-7712, Fax: 607-255-9829
The beef cattle seedstock industry in the USA
is searching for ways to select for improved beef cow efficiency. Most
selection indexes for efficiency have a goal of using less resource
while obtaining the same outcome in a sustainable environment. However,
the inputs required to determine individual beef cow feed efficiency
are not readily available in practical conditions. A mathematical model
was developed to use inputs readily available in each production situation
to estimate the ratio of cow ME required to calf-weaning weight (WW)
for computing an energy efficiency index (EEI). This model ranks EEI
estimates and compares individual cow EEI with the range of expected
EEI using Monte Carlo (MC) methods to identify the upper and lower cutoff
values. It uses the National Research Council recommendations as implemented
in the Cornell Net Carbohydrate and Protein System for energy requirements
for maintenance, lactation, and pregnancy. Data containing varying levels
of milk and forage intake of individual calves during the first 200
d after birth was used to develop a submodel to estimate calf forage
and peak milk intake (PKM) based on calf BW and forage composition.
A database collected from the Bell Ranch, NM (N = 182) was used to evaluate
the ranking from most to least efficient cows. The simulation indicated
that as peak milk (PKM) increases, WW increases almost linearly, the
difference in the calf WW between small and large cows tended to increase,
and EEI estimates improve exponentially. As PKM increased, the EEI difference
between small- and large-size cows decreased. The model-predicted least
efficient cows were in agreement with culling decisions made prior to
evaluating the EEI ranking. The MC simulation based on the distribution,
mean, and variability of cow BW, PKM, and forage quality indicated that
cows having EEI lower than 30.6 or higher than 38 Mcal/kg are within
the 10% more and less efficient cows, respectively. Our analysis suggested
this model could assist beef producers in identifying the most and least
efficient cows for their resource, and can be used to simulate different
production scenarios to identify the best match of cow type to alternative
management systems.
INTRODUCTION
The
beef cattle seedstock industry in the USA is searching for ways to select
for improved beef cow efficiency to improve their competitiveness and
profitability. Increases in beef production have occurred due to enhancements
in reproduction indexes (e.g. calving frequency, age at first calving,
calving interval), nutrition concepts (e.g. strategic supplementation,
type of forage as well as quantity and quality), genetic selection (e.g.
bull selection, crossbreeding), and (or) ranch management (e.g. matching
breeding and calving seasons with availability of forage). Nonetheless, beef production still is a relatively inefficient process
from the standpoint of energy expenditure. Research has indicated that
beef cows are responsible for 60 to 70% of the total of energy expenditure
(Johnson, 1984) in beef production (Ferrell and Jenkins, 1985). Ideally, efficient beef cows use less resource
to obtain the same outcome in a sustainable environment. There are several
indexes used to identify efficient beef cows. Most are based on retaining
beef cows that routinely produce a weaned calf with fewer inputs, with
a high ratio of pounds of calf weaned per number of females exposed
to a bull. Additionally, beef cow maturation rate has also been shown
to be correlated with production efficiency and may be used to select
for efficient cows (Parker et al., 1972, Tedeschi et al., 2000a,
b).
Jenkins
and Ferrell (2002) concluded that to evaluate biological efficiency,
productivity must be expressed relative to some measure of input, and
feed energy required per unit of output is logical. We have developed
a beef cow model that uses this approach to identify differences in
efficiency among beef cows. The principal objective of this paper is
to present our model that estimates the Mcal of metabolizable energy
(ME) required by a cow per kg of weaned calf (energy efficiency index
– EEI). A second objective is to demonstrate how the model compares
the EEI computed for each beef cow to the range of expected EEI using
Monte Carlo simulation to identify the upper and lower cutoff EEI. A
third objective is to discuss the potential for identification and selection
of mitochondrial DNA (mtDNA) mutants in beef cows that have higher energy
efficiency.
MODEL
DEVELOPMENT
Several
models have been developed to simulate cow/calf production systems (Boyd, 1977, Notter et al., 1979b, c, a, Long,
1972, Miller et al., 1980, Naazie et al., 1997, Fox et al., 1988). Fox et al. (1988) developed a nutritional model to evaluate
the match of the energy requirements of a cow/calf herd with forage
available each month to enhance profitability of the herd. Their model
computes a balance between energy requirements for maintenance, pregnancy,
lactation, and tissue mobilization and energy available from the forage;
thus, allowing one to match availability of forage with periods of higher
energy demand by the cow and calf. Reynoso-Campos et al. (2004) published an application of the Cornell Net
Carbohydrate and Protein System model for dual-purpose cattle that computes
daily energy balances between the herd requirements and forage available.
This model is based on those developed at Cornell University for beef
cows (Fox et al., 1988) and for dual-purpose cows (Reynoso-Campos et al., 2004), with modifications as described in this paper.
Figure
1 summarizes the structure of our model developed to estimate daily
energy requirements of the beef cow and the interactions between lactation
and weaning weight (WW) of the calf. The objectives of this model include:
(1) computing the energy requirements of individual beef cows each day
of the year and simulating the growth of the nursing calf given the
information available, (2) computing energy balances for the herd each
day of the year to evaluate the balance between herd numbers and requirements
with the forage available, and (3) identifying differences in efficiency
among individual beef cows in a herd.
Maintenance
Requirement
Energy
required for maintenance is based on body weight adjusted for conceptus
weight, environment (climate effects), physical activities, and physiological
stage (dry vs lactation) as recommended by the NRC (2000). Smooth curve adjustments using the cubic
spline technique are used during transition phases since this is a time-dependent
model. DiConstanzo et al. (1990) found that among non-pregnant non-lactating
Angus cows of similar fat masses, those with larger protein masses had
higher energy requirements for maintenance because the ME required to
maintain 1 kg of protein was 9.3 times higher than fat (192.9 ± 24.8
vs 20.7 ± 21.5 kcal, respectively). We account for this effect in our
growth model (Tedeschi et al., 2004) and plan on including a component for body
composition effects on maintenance requirement into our beef cow model.
Pregnancy
and Lactation Requirement
Energy
for pregnancy is based on the NRC (2000) recommendations that uses days pregnant to
derive energy concentration in the conceptus. The model assumes a fixed
calving interval of 365 d. The model computes milk production by changing
the peak milk until the WW predicted by the model matches the observed
WW. The energy requirement for lactation is computed based on NRC (2000) and Fox et al. (2004). Milk composition is used to compute net energy
of the milk, which drives the energy requirement for lactation. A fixed
value of 5.29 Mcal of ME/kg of milk (DM basis) is assumed in computing
intake of ME by the calf. The peak milk is used to plot the lactation
curve (George, 1984), which predicts the daily amount of milk available
for the nursing calf.
Forage
and Milk Intake of the Calf
The
data of Abdelsamei (1989) was used to develop equations for estimating
forage intake of the calf. In his experiment, the daily ad libitum intake of chopped alfalfa of
40 Holstein calves fed 5 levels of milk (peak milk at 59.5 DIM: 2.72,
5.44, 8.16, 10.88, and 13.6 kg) was measured for 200 d. We used this
data to derive five multiple regression equations to estimate forage
intake for the pre-peak milk phase as shown in Table 1.
The
intake of forage (kg/d) for the post-peak milk phase (Equation 1) is
computed using the surface response regression (R2 = 98.6%,
N = 394, RMSE = 281.24) in which CowMilk
is the amount of milk (DM basis, assumed to be milk production ´ 0.12), kg/d; CalfBW is the weight of the calf, kg; and PeakMilk is the milk production at the peak (as-fed basis), kg/d.
Cow
milk and peak milk have a negative correlation (-0.594 and -0.326, respectively)
whereas calf BW has a positive correlation (0.594) with forage intake.
Figure 2 shows a three-dimensional plot of the forage intake based on
calf BW and day post-peak milk for a peak milk of 3 kg/d. The amount
of cow milk was estimated based on the equation proposed by George (1984). The higher the peak milk, the lower is the
forage DMI at same calf BW and d post-peak milk (Figure 2), indicating
that calves increase their forage intake as milk availability is reduced.
It has been shown that forage intake per unit of BW prior to
weaning is consistently greater for calves receiving low quantities
of milk (Le Du et al., 1976a, b,
Broesder et al., 1990), and the consumption of milk reduces herbage
intake (Baker et al., 1976).
Body
Reserves
It
is well documented that BCS has an important role in beef production
and reproduction efficiency (Houghton et al., 1990, Mortimer et al., 1991). In our model, tissue mobilization and repletion
is used to compute energy available/required for body reserves based
on BCS changes, similar to that described by Reynoso-Campos (2004).
The coefficients of energy inter-conversion were derived by Moe et al. (1970) using multiple regression analysis of 126 and 224 lactating dairy cows with negative and positive energy balances, respectively. Several assumptions were made in their analysis.
When intake of energy is lower than energy required for milk production (NEl), it indicates a negative energy balance and energy reserves (NEr) are used for milk production. An efficiency of NEr to NEl of 82% is generally used (Moe, 1981, NRC, 2001, Fox et al., 1999). On the other hand, when intake of energy is greater than energy required for milk production, it indicates a positive energy balance in which energy intake above requirements was deposited as energy reserves. Commonly, in lactating cows an efficiency of ME to NEr of 75% and ME to NEl of 64.4% are assumed (Moe, 1981, NRC, 2001, Fox et al., 1999).
As reported by Moe et al. (1970), for lactating cows in negative energy balance an average efficiency of 66.1% and 84% were calculated for ME to NEl and NEr to NEl, respectively. However, an analysis of the variation indicated the true efficiency value is between 63.3 and 69.2% for ME to NEl and 81.7 to 86% for NEr to NEl, assuming a = 5% and a less rigid combination of coefficients. Similarly, for lactating cows in positive energy balance, Moe et al. (1970) reported an efficiency of 72.6% for ME to NEr and 63.5% for ME to NEl, with a confidence interval of 67.4 to 78.6% for ME to NEr and 61.2 to 65.8% for ME to NEl (a = 5%).
Our preliminary analysis indicated no statistical difference at P = 0.01 between the efficiency of NEr to ME and ME to NEl, suggesting the efficiency of using tissue energy for ME might be equivalent to the efficiency of using ME for milk production. Nonetheless, these coefficients are different at P = 0.05. The efficiency of use of mobilized tissue energy for milk production (84%) is statistically more efficient (P < 0.05) than the efficiency of use of ME for milk production (66.1%).
MODEL EVALUATION
Figure
3 shows the comparison of weaning weight and EEI of two cows (small,
450 kg, and large, 530 kg) with six peak milk levels. Birth weight was
assumed to be 6.5% of the mature weight of the cow. As peak milk increases,
WW increases almost linearly (Figure 3A) and the energy efficiency index
decreases exponentially (Figure 3B). Differences between the two cows’
sizes are due to differences in body weight at the same milk production.
Published data indicates that cow mature weight does not influence the
efficiency of energy use (Klosterman and Parker, 1976, Morris and Wilton,
1976, Ferrell and Jenkins, 1984a, b). Their studies indicate that as mature size
increases, milk production, weaning weight and finished weight increase
proportionally.
Figure
3B indicates that milk production is a determinant of calf WW and efficiency
of the cow. Figure 3A indicates the higher the milk production, the
greater the weaning weight, in agreement with
the studies of Abdelsamei (1989), Lewis et al. (1990), and Clutter and Nielsen (). As milk production increases, cow maintenance
requirement becomes increasingly diluted by the additional weaning weight
produced. However, it is well known that high-milk production cows have
higher energy requirements for maintenance because the internal organs
are larger and they have a faster metabolism compared to low milk production
cows (Ferrell and Jenkins, 1984a, b, 1985). This means that higher-milking cows require
more feed for maintenance and energy per pound of BW than lower-milking
cows (Montano-Bermudez and Nielsen, 1990). If feed available is adequate, this higher
maintenance requirement will be offset by an increased weaning weight
of the calf. Compensatory growth may also play a key role in the growth
of calves from cows that produce less milk. In agreement with the study
of Abdelsamei (1989), Lewis et al. (1990) found that post-weaning effects of increased
WW on ADG due to higher milk intake pre-weaning were small. They reported
that only calves from the low-milking group (5.6 kg/d) showed compensatory
growth. Miller et al. (1999) reported no effect of milk yield on biological
efficiency of Hereford, Charolais x Simmental x Maine-Anjou, and Tarentaise
x Pinzgauer x Gelbvieh x Angus calves from calving to harvest.
Cows
selected for improved efficiency in a certain environment may not express
their potential efficiency in another environment (Ferrell and Jenkins, 1985). When forage availability is not limiting,
cattle with higher milk and growth potential can utilize the extra feed
to wean heavier calves, therefore increasing weight sold for the forage
available. However, when forage is limited, those with lower milk and
growth potential can wean more calves for the same forage because there
is a higher proportion of the energy intake above maintenance available
for maintaining reproductive efficiency.
We
conclude the cow mature size should be determined by the optimum weight
for the calves at the target carcass composition, and the milk production
level should be based on the forage available.
A PRACTICAL APPLICATION OF THE MODEL
Using
Monte Carlo simulation (Winston, 1993), one can simulate the expected outcomes of
the EEI given the distribution, mean, and variability of the parameters
used by the model. Figure 4 shows a simulation with a model at a selected
cow BW, peak milk, and forage quality. For that scenario, the model
indicated that cows that had EEI lower than 29.67 or higher than 39.27
Mcal ME/kg WW are within the 10% more and less efficient groups of cows,
respectively. Therefore, one could use this model to assist in identifying
efficient cows: cows having less than 29.67 Mcal ME/kg WW to increase
the efficiency of the herd, or conversely, one could cull those cows
having EEI higher than 39.27 Mcal ME/kg WW.
A
second application of the model is the simulation of availability and
quality of forage throughout the year relative to forage requirements
for different cattle type. Figure 5A shows a potential range in forage
ME for each month for a particular ranch and Figure 5B shows the energy
balance (requirement minus supply) across this range in forage ME for
each month. This simulation was performed using a cow mature size of
530 kg with a peak milk of 8.35 kg/d. The simulation indicated that
improved forage quality, change in cattle type or numbers, or supplementation
is needed during the months of July through December.
A
third application is to determine the highest level of milk production
for the target cow size that can be supported by the forage available.
IDENTIFYING DIFFERENCES IN BEEF COW EFFICIENCY
IN THE FUTURE
The
European Association on Animal Production published a report (Ostergaard et al., 1990) that provides definitions of efficiency for
primary and secondary traits for dairy cow efficiency, which also applies
to beef cow efficiency. They summarized as follows: “The improvement in biological efficiency is important, and research has
to be focused on the underlying processes such as rumen function, utilization
of digested and metabolized energy, and the partitioning of feed energy
between milk and body tissue. Knowledge about genetic variation between
animals for these different biological processes is very limited, and
should be studied in relation to the composition of feed ration, the
feeding strategy and the physiological state of the animal”.
However, the identification of differences among individual animals
in these biological processes is difficult, particularly given the information
typically available on farms and ranches. Australian scientists have
used the residual feed intake (RFI) analysis of post-weaning growth
of individually fed progeny to identify efficient bulls (Archer and Bergh, 2000, Archer et al., 1999). The main problems with this technique are
the need for measurement of individual intake and the tendency to select
for leaner animals as an undesirable consequence. We are currently evaluating
our mathematical model (Cornell Value Discovery System – CVDS;
Tedeschi et al., 2004) for ranking individual animals fed in groups
on the basis of feed required for the observed growth and body weight.
The CVDS estimates required intake given each animal’s performance
and adjusts gain and intake for body composition (degree of maturity).
This model is currently available for growing/finishing animals and
may be downloaded at http://www.cncps.cornell.edu. The beef
cow model described in this paper is being incorporated into the CVDS.
We are evaluating its potential for providing information that can be
used to rank individual cows on the basis of their EEI. Our goal is
to determine if the output of these biological models can be used to
develop EPD’s for feed efficiency for use in genetic selection
programs.
We
are also working on a genomic-modeling project that involves the mapping
and identification of mitochondrial DNA mutants (mtDNA) that are more
energetically efficient. The presence of maternal genetic effects has
long been hypothesized to have an effect on traits of economic importance
in beef cattle. However, little support has been found in the common
statistical analysis of genetic breeders (Gibson et al., 1997). Mitochondria are a likely source of some
of this “unexplained” variation since they contain their own
DNA and are only maternally inherited. It is well known that mtDNA variation
may cause bias in the estimation of variance components (Boettcher et al., 1996b). Therefore, a positive mitochondrial effect
is desirable for dams of cows, but not for dams of sires, since they
are not passed on to male progeny.
Mitochondrial DNA has been extensively used in phylogeny to identify
cattle lineages using DNA displacement loop sequence variation (Loftus et al., 1994, Bradley et al., 1996). Additionally, mtDNA has also been used to
characterize substitutions that could be responsible for several economically
important traits, including meat quality (Mannen et al., 2003), milk production and animal health (Boettcher et al., 1996a, Schutz et al., 1994,
Boettcher et al., 1996c). The basic hypothesis is that a lineage of cattle
that is more energetically efficient might exist due to certain arrangements
in the mtDNA that permit the mitochondria to be more efficient. This
energetic efficiency of the mitochondria is reflected in the bioenergetics
of the whole animal and is responsible for some variation found among
progeny of the same sire but different dams. External effects that might
regulate mitochondria efficiency have also been reported, such as acetyl-L
carnitine (Iossa et al., 2002) and fatty acids (Jezek et al., 1998, Clarke et al., 2000, Schrijver
and Privett, 1984).
In
broilers, low feed efficiency is related to defects in electron leak
in muscle mitochondria (Bottje et al., 2002). In plants, ATP Synthase is a key enzyme in
providing energy since it uses a transmembrane electrochemical proton
gradient to drive synthesis of ATP. The enzyme complexes function as
miniature rotary engines, ensuring energy coupling with very high efficiency
(Bunney et al., 2001). In rats, a low mitochondrial proton leak
rate may partially explain the abnormally lower heat production and
bioenergetics efficiencies of the obese Zucker rat (21% lower than leaner
animals) as reported by Ramsey et al. (1996).
Mitochondrial
proton leak may be responsible for at least 20% of the resting oxygen
consumption in mammals (Ramsey et al., 2001). It is also documented that uncoupling protein
1 homologue, UCP3, is responsible for a decrease in efficiency of energy
metabolism because of the dissipation of energy as heat due to an uncoupling
of adenosine triphosphate (ADP) production from mitochondrial respiration
process (Schrauwen, 2002). Therefore; mutants that have a lower mitochondrial
proton leak or have lower concentration of UCP3 will be more energetically
efficient.
Mathematical
models can be used to assist in the identification of efficient cows
and simulation of different production scenarios to identify optimum
management systems for beef cows to maximize profits on a given land
base. In identifying the most efficient beef cow type, the cow mature
weight should be determined by the optimum weight for the calves at
the target carcass composition, and the milk production level should
be based on the forage available. More
work is needed to account for protein availability and quality in the
current model. Once this is attained, this model can also be applied
to select best strategy for forage management and supplementation to
minimize costs and environment impacts of N.
ABDELSAMEI, A. H. (1989) Forage intake, growth and final composition of
suckling calves fed five levels of milk preweaning. Animal Science. Ithaca, NY, Cornel University.
ARCHER, J. A. & BERGH, L. (2000) Duration
of performance tests for growth rate, feed intake and feed efficiency
in four biological types of beef cattle. Livestock
Production Science, 65,
47-55.
ARCHER, J. A., RICHARDSON, E. C., HERD, R. M.
& ARTHUR, P. F. (1999) Potential for selection to improve efficiency
of feed use in beef cattle: a review. Australian
Journal of Agricultural Research, 50,
147-161.
BAKER, R. D., LE DU, Y. L. P. & BARKER,
J. M. (1976) Milk-fed calves. 1. The effect of milk intake upon the
herbage intake and performance and grazing calves. Journal
of Agricultural Science, 87,
187-196.
BOETTCHER, P. J., FREEMAN, A. E., JOHNSTON, S. D., SMITH, R. K., BEITZ, D. C. & MCDANIEL, B. T. (1996a) Relationships between polymorphism for mitochondrial deoxyribonucleic acid and yield traits of Holstein cows. Journal of Dairy Science, 79,