Policy Modeling for Greenhouses Gas Emissions
on Dairy Cattle Sector:
the Importance of the Milk Production Improvement
Atzori Alberto Stanislao', David Ford’, Tedeschi Luis Orlindo*, Cannas Antonello!
Dipartimento di Scienze Zootecniche, Universita di Sassari, Sassari, Italy!
Zachry Department of Civil Engineering, Texas A&M University, College Station, TX, USA?
Department of Animal Science, Texas A&M University, College Station, TX, usa?
1. ABSTRACT
More than 14.5 Billion of $ were spent in US by Federal R&D authority for 2011 for Climate
Change. Current focus of animal scientists is to identify the most viable solutions to improve farm
profitability and milk production while minimizing the environmental impact of livestock. The
main objective of this paper was to use the System Dynamics methodology to: 1) model the
importance of the milk production improvement to reduce greenhouses emissions in the dairy cattle
sector, and 2) focus a viable policy to minimize the dairy environmental impact.
The model consisted of 6 sub-models, as follows: (A) human population and milk market; (B)
cow population; (C) milk production; (D) energy and feed requirements by the animals; (E)
economics; and (F) potential environmental impacts and government policy; the last sub-model was
aimed to encourage and strengthen the reduction of GHG emissions in dairy cattle farms by
improving the milk production per cow. Considering the simulated scenarios, the carbon footprint
of the milk (expressed as kg of CO? equivalents emitted per kg of milk), decreased proportionally to
the milk production per cow; while the total amount of emissions was highest in the scenarios with
highest milk production rate.
Policies oriented to improve the milk production per cow per year stimulated the milk
consumption rate and reduced the carbon footprint of milk. Simulated scenarios showed that to
reach environmental goals, governmental policies should be oriented to stimulate dairy farm
productivity and taking advantages from minimizing resource utilization and farm inefficiencies
2. BACKGROUND
Food production and environmental impact are going to represent the most important issues
related to animal production in the near future (Tedeschi et al., 2011a). In this context, dairy cows
acts related the global warming potential of the entire world moreover for the contribution to the
greenhouses emissions in the milk production processes. Furthermore, the increase of world
population to over 8 billion people in 2030 will increase the demand for food around the world
(FAO, 2006). At the moment, the dairy cow sector provides about 83 % and 13 % of total world
milk and meat production, respectively (FAO, 2009); in quantitative terms and it is expected to
increase its energy demand to partially contribute to the food supply chain by increasing meat, milk,
and processed products; in particular, the demand of animal derivate food is expected to growth
very fast, especially in the developing countries (FAO, 2006).
Nonetheless, the livestock sector also contributed to about 18% of total anthropological
greenhouses gas in terms of CO> equivalent emissions (COzeq) per year (FAO, 2006). In 2007, the
world dairy cow sector emitted 22% of total livestock COzeq of which 15% was allocated to milk
production and the remaining 7% was assigned to meat production (Gerber et al., 2010). The
pollutant effect of dairy cattle, in terms of carbon footprint, was recently estimated in 2.4 + 26%
(varying from 1.3 to 7.5) kg of COzeq per kg of fat and protein corrected milk (FPCM; Gerber et al.,
2010). The lowest values of emissions were estimated for intensive livestock production systems of
the industrialized regions located in North America and Europe; values lower than | kg of COzeq
per kg of FPCM were also estimated for high productive systems where an average cow produces
more than 8,500 kg of milk per year (Rotz et al., 2010).
Additionally, the environmental impact of livestock is related to energy demand and vegetal
biomass use. The competition for these products could generate a raise of price of grains (Kim
2009). Therefore, the understanding of the interplay between aggregate demand for animal products,
production factors, demand, and their relationships to the environment is crucial to guide
governmental policy and industrial activity (Tedeschi et al., 2011b). However, because the cost to
?
maintain the herd is between the second or third greatest cost on many dairy farms after feed and
labor (De Vries 2005; 2006), milk production improvement has been associated with profitability,
with the reduction of resources utilization, and therefore with some environmental benefits (Chase
et al., 2003; Capper et al., 2007). The increased productivity of individual animals has decreased the
number of dairy cows in the United States from approximately 12.1 million dairy cows in 1970 to
9.1 million cows in 1999 (Cassel, 2001). A big limitation in the increase of productivity is the
ability of producers to invest in productivity that is profit related and affected by the commodities
and its price in the market demand (Meadows, 1970).
The Kyoto Protocol aimed to reach the reduction of greenhouses gas emissions; is already
noticeable a big effort of the governments, business firms and scientific community, in terms of
funds and capital resources to understand the keys variables of the global warming control. The
Federal R&D authority budget spent in US for 2011 for Climate Change was higher than 14.5
Billion of $ (Higgins, 2012). In this context, the current focus of animal scientists is to identify the
most viable solutions to improve farm profitability and milk production while minimizing the
environmental impact of the livestock. Without clear priorities between improvement of animal
productivity and environmental sustainability a dynamic approach to this issue is needed to
understand the possible effects of policy on GHG emission in dairy cattle farms. In particular,
system dynamic models could help to focus the eco-sustainability trade-off points, between
economic and ecological factors; from this point of view, the variable dynamics driving farmer’s
investment choices need to be studied and profundized. Simulation modeling developed in
conjunction with decision makers and stakeholders can generate credible and relevant assessments
of climate change impacts on farming systems (Rivington et al., 2007).
Therefore, the main objective of this paper was to use the system dynamics (SD)
methodology to model the milk production and consumption dynamics in the world studying the
importance of the milk production per cow improvement to reduce greenhouses emissions in the
dairy cattle sector and focus the way pro propose possible policies to minimize the environmental
impact of the milk production.
3. MATERIAL AND METHODS
3.1. Development of the model and causal diagram
The model was built using Vensim® version 5.9 DSS (Ventana System, Inc.). The complete
model was based on the dairy sector model reported by Atzori et al. (2011). It consisted of 6 sub-
models, as follows: (A) human population and milk market; (B) cow population; (C) milk
production; (D) energy and feed requirements by the animals; (E) economics; and (F) potential
environmental impacts and government policy.
The original model structure was modified substituting the milk market sub model with the
basic commodity model structures in Meadows (1970) and Sterman (2001). This structure was
integrated with a delay between milk production and milk availability at retail level, to take into
account the milk processing into different dairy products consumed at different aging time.
The effect of farm investments on milk production was modeled assuming a direct effect of profit
ratio (revenues/cost) on the variation of the herd size (changing the number of replacement or culled
cattle) or, alternatively on the destination of part of farm profit to improve breeding and
management of producing animals, as previously modeled by Pagel et al. (2002).
The dairy cow milk system was preliminary identified with two important balancing loops, as
follows: world milk demand (loop 1; MD; Figure 1) and world milk production (loops B2 and B3,;
MP; Figure 1). Loops B2 and B3 in the figure 1, indicate cow number (CN), cow productivity (CPr)
while the loop B4 and R6 both indicated the feed cost to produce milk and represented the energy
demand due to herd maintenance (FCmaint) and energy demand due to milk production (FC mit),
respectively.
The MD (kg/year) represents the desired state of the system; it mainly depends on the
number of people in the world (assumed as an exogenous variable to our model), and their annual
milk consumption per capita that was assumed almost constant in the world as reported in the
historical trend from 1970 to 2010 by Rouyer (2011). An increasing trend of milk consumption was
assumed to take into account the forecasted trend of milk consumption of developing countries in
the near future (FAO, 2006).MP can be varied both changing the CN (indirect way) or the CPr
(direct way).
The MP (kg/year) describes the system ways to meet the milk world demand. The MP can
be adjusted both varying the CN (indirect way) or the CPr (direct way). It respectively goes
throughout the number of dairy cows in the world and their annual milk productivity (kg/cow/year).
The gap between demand and offer of milk is represented by the difference between MD and MP. It
directly and negatively affects the milk price. Milk price changes can alter the desired state of the
system because of the negative relationship with the milk demand per capita and it can also
positively affect the farm profitability. The milk production rate strongly depends on the
profitability of the production process, which is positively related to the farm investments; they
could be used either to expand the number of cows in the herd or to increase their own milk
production level. As reported by Weller et al., (2011) the actual genetic gain is 100 kg/cow per year
in US; the same authors calculated a nominal economic value of 0.1$ per additional kg of milk
productivity improvement per cow on the annual basis; using a conservative assumption the cost of
milk production improvement per cow was considered equal to 0.3$ per kg/cow of annual gain; in
fact, as reported by Weller et al., (1994), a long term investment in breeding, could be justified if its
cost is lower then 3 times the nominal value of gain.
Greenhouses gas (GHG) emissions were modeled using IPCC (2006) assumptions. Methane
is produced by the enteric fermentation of organic biomass used for her feeding and by the waste
fermentation storage facilities. From storage facilities NxO and CO) were also emitted; other CO2
emissions comes from secondary sources (IPCC, 2006; Rotz et al., 2010). In the model, emissions
of carbon oxide (CO2), methane (CH4) and nitrous oxide (N2O), were considered and then
converted in CO equivalents (CO2eq) as reported by Gerber et al. (2010).
replacement rate
variation
herd size of cows +
and replacement
8)
cattle
I Cow annual
productivity
Ee | e
Energy needs for
cattle maintenance
/*; @ Milk produetion
Milk demand ’
Farm investment
Energy Mn for to increase farm
milk production om production
i Milk stock Milk consumption
per capita
| !
\ ;
\ ‘Milk price farm profit
government support to
reduce the carbon
\ ” Profit from milk: footprint
a “production es re
ad wn
Energy cost per
kg of milk
target GHG
emissions. —————p GHG emission’
+, surplus
GHGemissions
Figure 1. Causal loop diagram of the effect of the milk production improvement on the reduction of
greenhouses gas emission:
- world milk demand (desired state of the system, MD; loop B1);
- world milk production (MP; loops B2, B3, B4 and, R6; they indicate cow number (CN), cow
productivity (CPr), feed cost to maintain the herd (FCmain.) and feed cost to produce milk (FC mix)
respectively).
- GHG emissions and government policy in loop B5 (GHG).
Black arrows are involved in > 1 loops while Bold lines indicate a possible policy to
stimulate and strengthen the loop effect.
3.2. The dynamic hypothesis
Supposing a constant milk demand in the world, the increase of milk production per cow
should cause a reduction of the number of cows and replacement cattle, principally for economical
reasons; a reduction of about 2% per year in the US cow number was reported by Capper et al.
(2007), in conditions of constant milk production rate, from 1944 to 1980.
Assuming an increasing demand of milk in the world, the farm profit, it could be destined to
increase the amount of milk produced by a general farm can be used to increase the number of cows
in the herd or to improve their own annual milk production (Figure 1). Looking at the second way
of CPr (loop B3 in Figure 1), the investments in breeding programs are unlike other investments, in
that the gain are eternal and cumulative (Weller and Ron, 2011); it brings higher environmental and
economic benefits. In fact, looking at the Figure 1, more milk is produced by the farms, more milk
arrives to the milk market stock; it probably causes a reduction of the milk price, that in turn causes
the elimination of a proportional number of cows from the production process. It also generates a
proportional reduction of the energetic needs to produce a kg of milk, lowering the maintenance
cost of the herd. Energetic needs of milk production are positively related both with the GHG
emissions (moreover methane from enteric fermentation) and with the feeding cost. Consequently,
the improvement of the annual milk production per cow shall: reduce the carbon footprint of the
dairy sector and stimulate the milk consumption.
A policy program based on total GHG emissions of dairy cattle sector was included in this model in
addition to the previous version of the model reported by Atzori et al. (2011; Figure 1); a balancing
loop of government support to the milk production improvement was simulated in order to
encourage and strengthen the reduction of GHG emissions in dairy cattle farms. An environmental
policy could also stimulate the effect of the reinforcing loop (R6; Figure 2): in fact if more milk
production improvement causes a reduction of the cost of the milk, it will generate more profit and
more probability of farm investments to improve the annual milk production per cow.
3.3. Model parameterization and calibration
Values from the literature were used as initial values of the biological auxiliary
variables and constants. The initial values of principal stocks were setup using data reported by the
Food and Agricultural Organization (FAO; _ http://faostat.fao.org/default.aspx). Model
parameterization and calibration were also performed using data reported by FAO. The FAO
provides historical trends and forecasts of human population census and livestock primary
production in terms of producing animals, inventories of products, and animal productivity. Model
calibration was executed on the lookup functions until the model estimations matches the reported
and forecasted FAO trends, from 1999 to 2030, for cow number, cow productivity, and total milk
production (Figure 2). The base scenario was simulated considering that 50% of the total annual
farm profit was destined to the investments for milk production improvement of the same farm; the
investment was then equally divided between increase of farm herd size and improvement of cow
annual milk production (Figure 2).
A) B)
300! 2800 SEAQ data A model estimation
280 oat | 2700 .
260 anaes a
= 140 guy inaaanaan 2600 ccs
Ss » > 2500 pane
=
23 200 2400 ase
§ 180 2300 agesaeta
EA 160 2200 “
é& 140
120 ™FAO data 4 model estimation 2100
100 + 2000 +
1990 2000 2010 2020 2030 1990 2000 2010 2020 2030
Year Year
Cc)
800
Wow . 7 7
700 wa Figure 2. Model estimation and FAO
aan’
“ une! observed and forecasted data for cow number
Oa . .
ou (A)), milk production per cow (B)) and total
milk production (C)) in the entire world
FAO data model estimation
300 + between 1999 and 2030.
1990 2000 2010 2020 2030
Year
3.4. Model application: simulated case studies
The simulation time horizon was 31 years, between 1999 and 2030; unit for time was years with a
time step of 0.0625; the Range-Kutta4 method for integration was used.
Five scenarios with positive profitability were simulated to study the sector dynamics in
relationship with GHG emissions;
- 3 scenarios were simulated varying the profit partitioning between herd size (CN) and cow annual
productivity improvement:
a) CNO.5: base scenario with 50% of positive profit destined to increase the herd size and 50%
to improve cow productivity;
b) CNO0.3: scenario with 30% of positive profit destined to increase the herd size and 70% to
improve cow productivity;
c) CNO.7: scenario with 70% of positive profit destined to increase the herd size and 30% to
improve cow productivity.
2 other scenarios simulated the GHG policy from the year 2014:
d) CNO0.5_policy : base scenario + government support to farmers associations in order to
reduce GHG emissions; it consisted in 0.0045 $/kg of milk oriented to improve milk
production per cow.
e) CNO0.3_policy: government support to farmers associations in order to reduce GHG
emissions; it consisted in 0.0045 $/kg of milk applied to a system already oriented to
improve milk production per cow.
4. RESULTS AND DISCUSSIONS
The base scenario described the observed situation of the past years and the forecasted trends
reported by FAO. It resulted in a milk production improvement of 15 kg/cow per year (from 2200
to 2400 kg per year from the year 1999 to the 2014; figure 3) it represents the average trend of the
milk production improvement for the cows of the world herd size.
The other scenarios reported different increases in milk production per cow, within the range of
2880 kg for the highest production (CN03_GHG policy) and 2486 kg for the lowest (CN07),
measured at the year 2030. Considering the comparison between scenarios, the policy had an higher
effect on the CNOS5 scenario with respect to the CN03 scenario (Figure 3).
Relatively to the number of cows (Figure 4) in respect to the milk production per cow (Figure 3)
the scenarios were inversely ranked, highlighting the balancing effect of the system structure on the
milk production rate. In fact, the total milk production is in equilibrium with human milk demand;
consequently, higher production per cow is counterbalanced by lower number of producing animals
(Figure 4) as already observed in US (Capper et al., 2007). However the market response to the
milk production rate is reflected by the milk availability in the stocks, which causes variations in
milk price and, in turn affects the supply chain. The scenario (CN07) with lowest milk production
per cow and higher number of cows, also showed the lowest milk production rate between
simulated scenarios (Figure 5). CN07 could be considered as the worst scenario for the future of the
dairy sector because of its low ability to produce food for humans.
In economical terms, the market structure of this model was developed in order to replicate
oscillating patterns of the milk price in the market and at the farm gate, as reported in different
countries for fresh milk (IDF, 2011). Was not objective of this model to estimate the milk price
variability and volatility. In fact, as reported by Nicholson and Fiddaman (2003), milk price
volatility is often caused by a large number of factors than the usual supply-chain explanation, like
the behavioral response of industry participants to price and inventory signals, long-term
1n
contracting, speculative hoarding, coupled long-term cycles of processing capacity and herd size. A
more complex model, as approached by Pagel et al., (2002) accounting for many different variables,
such as seasonality of supply, multiple derivate products, minimum price, market premiums,
government offers to support prices, international trade policies, etc, must be used to estimate the
milk price variability. Moreover, observing that butter, skim powder of milk, cheese and cream are
the standard indicators of milk price, the use of the only synthetic price of raw milk for the entire
world could be considered an oversimplification of the same commodity market.
As expected, all the scenarios showed oscillating behavior for milk inventory, milk
consumption, milk expected consumption rate, inventory coverage, milk price at consumers, total
profit. In the base scenario the milk price at consumers varied from 1.1 $/kg to 0.8 $/kg showing
peaks every 4 years approximately; the oscillation period was very similar to that observed in US
milk price between 1990 and 2010 (Gould, 2011). A little variation between scenarios was also
observed. Considering the environmental performances of the system in the studied scenarios the
total GHG emissions increased proportionally to the milk production rate and the milk annual
production per cow (Figure 6). In fact, the carbon footprint of the milk, expressed in kg of emitted
CO; per kg of milk, decreased proportionally to the milk production per cow (Figure 6) while the
total amount of emissions were highest in the scenarios with highest milk production rate (Figure 5).
The scenarios were governmental policies were applied to the system showed that: the
public effort on GHG reduction could have a good exit if associated with farm goals of milk
production improvement; similar results could be achieved in the farms already managed in order to
reach efficiency and productivity; in fact, farm profitability resulted negatively associated with the
carbon footprint and the total GHG emissions of the dairy sector. A total reduction of total GHG
emissions, was not observed in these ranges of annual milk production improvement per cow; may
be a stronger increase of milk yield per cow per year could be helpful to reach these results, as
observed by Capper et al., (2007). Thus, the applied policy was not able to change the pattern of
behavior of the studied variables, but it was not included in the policy objectives; in fact, the policy
11
was capable to stimulate farmers to produce milk with high level of technical efficiency (i.e
increase milk production per livestock unit) and showed as forcing the system trough the B3 and R6
loops it is possible to obtain environmental benefits without significantly change the market
conditions.
The model limits model are related to the milk market description and complexity; this model
simplified the market structure in order to focus the relationship between the environmental aspect
and the farm efficiency. Other models, as the dairy model proposed by Pagel et al., (2002),
considered more adapted to study the economic aspect of the milk market, could be integrated with
this environmental sub-model in order to better understand the relationships between environmental
and economical aspects of the dairy sector.
5. CONCLUSIONS
In environmental terms, the milk production process is a source of emissions of CO) and its
environmental impact is higher as milk production rate increase in the world. From this point of
view, the milk production improvement, could help the dairy sector to fit the main objectives of the
sustainability pursued by the Life Cycle Assessment approaches (IPCC, 2006; FAO, 2006): to
produce enough food for the growing human population and reducing, at the same time, the
environmental cost per unit of products. Simulated scenarios showed that to reach environmental
goals, governmental policies should be oriented to stimulate dairy farm productivity and taking
advantages from minimizing resource utilization and farm inefficiencies.
19
6. FIGURES
3000
2900
2800
2700
2600
2500
2400
2300
2200
2100
2000
Kg of milk per year
<
SS
ANNUAL MILK PRODUCTION PER COW
—Base 05
—CN03
—CNO7
—CN03_GHG POLICY
—=CNO05_GHG POLICY
KVP MPH Vd Vad AMO OM rAaWeN Vea alel\adaaS
SPA AGEL HO RNAS AO LOLS OSES?
PaPPapPPapPrpay Pt
Figure 3. Model estimates of cow annual milk production for the simulated scenarios.
300
290
280
270
260
250
240
230
Millions of heads
MILKING COWS IN THE WORLD
—CNO03
—~CN07
—=CN03_GHG POLICY
—=CN05_GHG POLICY
DD. MPM PHA SND AMD AOM SAA DaNel} Orf\nbnQa®
PDMS NWP INIA N ALN IER PIL ENS
PVPREVY PPV PVRPVY yaaa
aPaPpapapap
Figure 4. Model estimates of cow number for the simulated scenarios.
12
800
Billions of kg of milk per year
wn
ral
i)
RATE OF MILK PRODUCTION
—Base 05
—CNO03
—CNO07
—CN03_GHG POLICY
—=CNO05_GHG POLICY
\r) OOS YA AD
BYWAY HPIHHP\H|H| HH
y
Figure 5. Model estimates of the world milk production rate for the simulated scenarios.
1700
1600
1500
1400
1300
1200
Billions of kg of CO,,,, per year
1100
1000
o, 6
AS SONA SY ONEQAHHHHY SASS KS
Figure 6. Mo
TOTAL GHG EMISSIONS FROM DAIRY COWS
—Base 05
—CNO03
—CNO07
—CN03_GHG POLICY
—=CN05_GHG POLICY
SS
del estimates of the total GHG emissions for the simulated scenarios.
1A
N
in
| CARBON FOOTPRINT OF DAIRY SECTOR
N
aS
N
is)
N
i)
—Base 05
—CNO03
—CN07 =
—CN03_GHG POLICY “~~
—CN05_GHG POLICY
yoy
aa
Billions of kg CO,,,/kg of milk per year
S
oe
S
wey
Figure 7. Model estimates of the carbon footprint for the simulated scenarios.
7. REFERENCES
1. Atzori A.S, Tedeschi L.O., Cannas, A. (2011). The dynamics in the dairy cattle sector:
policies on cow milk production can reduce greenhouses gas emissions and land use.
Proceeding of the 29th International Conference of the System Dynamics Society.
Washington, DC, USA.
2. Capper J., Cady R., Bauman D., 2009. The environmental impact of dairy production: 1944
compared with 2007. Journal of Animal Science 87, 2160-2167.
3. Cassel B.G., 2001. Optimal genetic improvement for the high producing dairy cows. Journal
of Animal Science 84 (suppl.), E144-E150.
4. Chase L.E., 2003. Nitrogen utilization in dairy cows what are the limits of efficiency? In
Proceedings of Cornell Nutrition Conference for Feed Manufacturers, pp. 233-245. Syracuse,
NY: New York State College of Agriculture and Life Science, Cornell University.
5. De Vries A., 2005. Economic importance of genetic improvements in milk production,
reproduction, and productive life. EDIS Document AN158.
6. De Vries A., 2006. Ranking dairy cows for future profitability and culling decisions. In:
Proceedings third Annual Florida & Georgia Dairy Road Show, pp. 91-109. Okeechobee,
Mayo, Chipley, FL, and Tifton, GA.
7. FAO, 2006. Livestock's Long Shadow: Environmental Issues and Options. Food and
Agricultural Organization, UN, Rome, Italy.
8. FAO, 2012. http://faostat.fao.org. Data download on March 2012.
9. Gerber P., Vellinga T., Opio C., Henderson B., Steinfeld H., 2010. Greenhouse gas emissions
from the dairy sector. A life cycle assessment. Food and Agricultural Organization, UN,
Rome, Italy.
10
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Gould B., 2011. Understanding dairy markets: all milk prices. Available at
http://future.aae.wisc.edu/data/annual_values/by_area/10. Data download, March 2011.
Higgins P.A.T., 2012. Climate Change in the FY 2011 Budget. In: AAAS Report XXXV,
Research and development FY 2011. Available at
http:/Avww.aaas.org/spp/rd/rdreport201 1/Higgins, 2012
IDF or International Dairy Federation, 2010. Bulletin of the International dairy federation: the
world dairy situation. 446/2010.
IPCC or Intergovernmental Panel On Climate Change, 2006. Guidelines for National
Greenhouse Inventories. Vol. 4: Agriculture, Forestry and other Land Use. //www.ipce-
nggip.iges.or.jp/public/2006gl/vol4.html. Accessed February 18, 2011.
Kim G.R., 2009. Analysis of Global Food Market and Food-Energy Price Links - Based on
System Dynamics Approach. Proceeding of the 27th International Conference of the System
Meadows D.L., 1970. Dynamics of Commodity Production Cycles. Wright-Allen Press,
Cambridge, MA.
Nicholson C.F., Fiddaman T., 2003. Dairy Policy and Price Volatility. Paper pre- sented at the
10th Annual Workshop for Dairy Economists and Policy Analysts, Mem- phis, TN, April 23-
24, 2003.
Pagel E., Nicholson C., Newton P., 2002. Structural Changes in U.S. Dairy Farming and the
Role of Government Policy; proceedings of the System Dynamics Society Conference,
Palermo, Italy, July 29-31, 2002.
Rivington M., Matthews K.B., Bellocchi G., Buchan K., Stéckle C.O., Donatelli M., 2007.
An integrated assessment approach to conduct analyses of climate change impacts on whole-
farm systems. Environmental Modelling and Software 22, 202-210.
Rotz C.A., Montes F., Chianese D.S., 2010. The carbon footprint of dairy production systems
through partial life cycle assessment. Journal of Dairy Science 93, 1266-1282.
Rouyer B., 2011. World Dairy Situation. Proceedings of the FIL-IDF Canada. March 22-23,
2011.
Sterman J.D., 2001. Business Dynamics: Systems Thinking and Modeling for a Complex
World. Boston: Irwin McGraw-Hill.
Tedeschi L.O., Nicholson C.F., Rich E., 2011a. Using System Dynamics modelling approach
to develop management tools for animal production with emphasis on small ruminants. Small
Ruminant Research 98,102-110.
Tedeschi L.O., Callaway T.R., Muir, J.P., Anderson R., 2011b. Potential environmental
benefits of feed additives and other strategies for ruminant production. Brazilian Journal of
Animal Science 40, 291-309.
Weller J.L., Ron M., 2011. Invited review: Quantitative trait nucleotide determination in the
era of genomic selection. J. Dairy Sci. 94:1082—1090.
Weller J.L, 1994. Economic Aspects of Animal Breeding. Chapman & Hall. London. 244 pp.
14é
