J. Dairy Sci.

100:1–18
https://doi.org/10.3168/jds.2016-12325
© American Dairy Science Association®, 2017.

Got pasture? How grazing intensity affects the environmental

impacts of dairy systems
H. A. Aguirre-Villegas,*1 T. H. Passos-Fonseca,† D. J. Reinemann,* and R. Larson*
*Department of Biological Systems Engineering, and
†Agroecology Program, University of Wisconsin, Madison 53706

ABSTRACT 0.28 kg of CO2 equivalents/kg of FPCM and reduced

NEI to −1.26 MJ/kg of FPCM, indicating a net energy
Dairy products are major components of the human producing system and highlighting the potential of
diet but are also important contributors to global en- AD to improve the sustainability of confined systems.
vironmental impacts. This study evaluated greenhouse For scenarios that combined confinement and grazing,
gas (GHG) emissions, net energy intensity (NEI), and GHG emissions ranged from 0.84 to 0.92 kg of CO2
land use of confined dairy systems with increasing levels equivalents, NEI ranged from 1.42 to 1.59 MJ, and land
of pasture in the diet. A Wisconsin farm was modeled use ranged from 1.19 to 1.26 m2/kg of FPCM. All en-
to represent practices adopted by dairy operations in a vironmental impacts were minimized in scenarios that
humid continental climate typical in the Great Lakes supplemented enough feed to increase milk yield but
region and other climates that have large differences maintained dry matter intake from pasture at a level
in seasonal temperatures. Five grazing scenarios (all high enough to reduce material and energy use.
of which contained some portion of confinement) were Key words: grazing, greenhouse gas emission, energy
modeled based on different concentrations of dry mat- intensity, life cycle assessment
ter intake from pasture and feed supplementation from
corn grain, corn silage, and soybean meal. Scenarios
INTRODUCTION
that incorporate grazing consisted of 5 mo of pasture
feeding from May to September and 7 mo of confined Dairy products are a major component of the human
feeding from October to April. Environmental impacts diet, but the production of this important source of
were compared within the 5 scenarios that incorporate nutrients contributes 4% of global anthropogenic green-
grazing and across 2 entirely confined scenarios with house gas (GHG) emissions (FAO, 2010). With milk
and without on-farm electricity production through demand projected to double between 2000 and 2050,
anaerobic digestion (AD). To conduct a fair compari- the need to implement GHG mitigation strategies will
son, all scenarios were evaluated based on the same increase. Currently, 80% of the GHG emissions from
total amount of milk produced per day where resource dairy production comes from on-farm sources: enteric
inputs were adjusted according to the characteristics of methane (CH4) is the largest single contributor, fol-
each scenario. A cradle-to-farm gate life cycle assess- lowed by CH4 from manure management and nitrous
ment evaluated the environmental burdens that were oxide (N2O) from soils (IDF, 2009). Studies quantify-
partitioned by allocation between milk and meat and ing these emissions have adopted life cycle assessment
by system expansion when biogas-based electricity was (LCA) techniques, making this the preferred meth-
produced. Overall, results for all scenarios were com- odology for estimating the carbon footprint of dairy
parable. Enteric methane was the greatest contributor systems (O’Brien et al., 2014).
to GHG emissions, and the production of crops was Continuous grazing has been adopted by dairy farm-
the most energy-intense process. For the confined ers worldwide given the low costs of this feeding strat-
scenario without AD, GHG emissions were 0.87 kg of egy. However, confined systems usually achieve higher
CO2 equivalents, NEI was 1.59 MJ, and land use was milk yields, which might explain the predominance of
1.59 m2/kg of fat- and protein-corrected milk (FPCM). this feeding strategy in some countries. For example,
Anaerobic digestion significantly reduced emissions to in the United States, the largest cow milk–producing
country in the world (USDA, 2016), 59% of dairy op-
erations use confinement, 7% use grazing, and 26% use
a combination of both strategies (USDA, 2014).
Received November 18, 2016.
Accepted March 30, 2017. When cows are grazing, CH4 emissions that occur
1
Corresponding author: [email protected] from long-term storage during confinement (Aguirre-

1
2 AGUIRRE-VILLEGAS ET AL.

Villegas et al., 2014) are reduced. On the other hand, anaerobic digestion (AD) in the same agricultural
N2O emission factors for manure deposited in grass- context.
lands have been cited to double those of manure that
is applied to land after storage (IPCC, 2006a; Lesschen MATERIALS AND METHODS
et al., 2011). Studies have compared the carbon foot-
print of grazing and confined dairy systems, but results Goal and Scope
are not consistent. Some studies that found reduced
GHG emissions of grazing systems evaluated grazing The goals of this cradle to farm gate LCA were (1)
and confinement in different landscapes, making direct to develop life cycle inventory data for dairy systems
comparisons of these practices difficult (Flysjö et al., that use yearly confinement and systems that use a
2011; O’Brien et al., 2014). Studies that compared these combination of confinement and grazing at different
systems for the same location found that pasture-based levels; (2) to quantify GHG emissions, NEI, and land
systems resulted in higher GHG emissions (Capper et use of specific practices used in both types of systems;
al., 2008, 2009; de Léis et al., 2015). One study by and (3) to compare these environmental impacts within
Belflower et al. (2012) estimated greater GHG emis- systems.
sions from confined systems but also estimated greater The functional unit was defined as 1 kg of fat- and
milk production, suggesting that pasture systems could protein-corrected milk (FPCM) to 3.3% protein and
benefit from diets that increase milk production. 4.0% fat and was calculated with Equation A1 accord-
One of the main challenges of grazing systems is the ing to the recommendations of the International Dairy
lower milk production that they achieve compared with Federation (IDF, 2010) and the Food and Agriculture
confined systems (Powell et al., 2013; Stott and Gour- Organization (FAO, 2010). The time frame of the anal-
ley, 2016). Milk production from grazing cows is usually ysis was 1 yr, and the boundaries included pasture and
lower than that from grain-fed cows because pasture crop production; milk production, dairy herd structure,
has a lower energy protein content when comparing and animal nutrition; manure handling and biogas
the same feed quantities (NRC, 2001). Given that cows generation; and manufacturing of nutrients, chemicals,
can eat only a certain amount of pasture to get their machinery, and energy inputs, which included embed-
daily energy intake, a common practice for increasing ded GHG emissions and energy use (Figure 1).
milk production is to supplement rations with other
feeds such as corn or silage (Belflower et al., 2012). Life Cycle Impact Assessment
These supplements can increase direct GHG emissions
because additional steps and resources are needed for The impact categories evaluated in this study were
their production. For example, Doole (2014) evaluated direct land use, measured in meters squared; NEI, mea-
intensive and extensive grazing systems in New Zealand sured in megajoules; and GHG emissions, measured in
and found that strategies that reduce input use are cost kilograms of CO2 equivalents (CO2-eq). The NEI was
effective and reduce GHG emissions. defined as the difference between the energy inputs and
Northern regions with colder climates, such as Wis- energy outputs. When system expansion was applied,
consin, have to adopt some degree of animal confine- the avoided grid electricity displaced by biogas-based
ment and supplemental feeding because cows cannot electricity based on its energy content was discounted
graze year round. Confined systems require more re- (Equation A2). A negative NEI indicates an energy-
sources such as fertilizers, energy, and infrastructure producing system, and a positive NEI indicates an en-
(e.g., manure storage structures and harvesting equip- ergy-losing system. The GHG emissions were measured
ment) than grazing systems, suggesting that additional over a 100-yr time horizon, and characterization factors
environmental impacts are associated with the use of were 264 kg of CO2-eq for N2O and 28 kg of CO2-eq for
these resources. The LCA is a useful tool for evaluat- biotic CH4 based on the latest Intergovernmental Panel
ing how the level of feed supplementation affects the on Climate Change report (Myhre et al., 2013). It was
environmental impacts of milk production in combined assumed that biotic CO2 emissions were recaptured by
grazing and confinement systems and for identifying the plants that make up the dairy feed ration; thus,
farming practices that mitigate the environmental im- biotic CO2 emissions were not included in the GHG
pacts of dairy systems. emission count.
This study evaluated how different levels of feed
supplementation affect milk production, life cycle GHG Life Cycle Inventory
emissions, net energy intensity (NEI), and land use of
grazing systems. Moreover, grazing systems were com- Data on material and energy inputs needed for crop
pared with confinement systems both with and without production, manure management, and milk production

Journal of Dairy Science Vol. 100 No. 8, 2017

 ENVIRONMENTAL IMPACTS OF GRAZING DAIRY 3

Figure 1. Cradle to farm gate boundaries of the scenarios producing milk and meat (solid boundaries); the scenarios producing milk and
meat with soybean meal (SBM) supplementation (dotted boundaries); and the scenario producing milk, meat, and electricity from manure
through anaerobic digestion (AD; dashed boundaries). Allocation and system expansion decisions are shown with gray shaded boxes and italics.
A total of 7 scenarios representative of Wisconsin were modeled based on information collected through surveys sent to Wisconsin dairy farms:
2 confinement scenarios (C1 and C2AD) with no access to pasture year round and 5 scenarios combining confinement and grazing at varying
percentages of annual DMI from pasture (G1, G2, G3, G4, and G5, where G1 had the lowest DMI from pasture and G5 had the highest). Color
version available online.

were based on the model developed by Aguirre-Villegas the lowest DMI from pasture and G5 had the highest;
et al. (2015). These data came from the literature and Table 1). Dairy diets for all scenarios met maintenance,
from the US Life Cycle Inventory (NREL, 2010) and growth, pregnancy, and milk production requirements
Ecoinvent 2.0 US (Ecoinvent Centre, 2007) databases (NRC, 2001). Scenario C1 considered that manure was
accessed through GaBi 4 Professional (Frischknecht et collected, stored for 6 mo, and applied to land in spring
al., 2005). The electricity generation mix was repre- and fall. Scenario C2AD considered the production of
sentative of Wisconsin: coal, 55%; natural gas, 19%; electricity from manure AD before land application.
nuclear, 21%; and renewables, 5% (Meier Engineering Scenarios that incorporated grazing also incorporated
Research, 2011). The GHG emissions were estimated varying contents of supplemental feedings to maximize
based on the integrated farm system model (Rotz et milk production and consisted of 5 mo of grazing (May
al., 2015) and emission factors recommended by the to September) and 7 mo of confinement (October to
Intergovernmental Panel on Climate Change. April). During grazing months, feed was supplemented
with corn grain, corn silage, and soybean meal. The lev-
Scenario Definition el of milk production has economic and environmental
impacts because in most dairy farms, milk is the most
A total of 7 scenarios representative of Wisconsin important or only product. As a result, increasing milk
were modeled based on information collected through production will most likely result in both economic and
surveys sent to Wisconsin dairy farms: 2 confinement environmental benefits for the dairy farm (O’Brien et
scenarios (C1 and C2AD) with no access to pasture al., 2015).
year round and 5 scenarios combining confinement During confined months, the dairy diet composition
and grazing at varying percentages of annual DMI remained unchanged and manure was stored and ap-
from pasture (G1, G2, G3, G4, and G5, where G1 had plied to crops. Scenario G1 maximized the number of

Journal of Dairy Science Vol. 100 No. 8, 2017

4 AGUIRRE-VILLEGAS ET AL.

animals per area for which DMI from pasture was mini- this feeding strategy, it was assumed that the compo-
mized. As a result, the supplemental feed input was sition and characteristics of the herd in all scenarios
high. This scenario was modeled to analyze whether were the same as opposed to a system that relies only
this marginal change in pasture DMI had any effects on on grazing, where cows tend to be smaller and of
environmental impacts. Scenarios G2 and G3 simulated different breeds. As a result, all scenarios had the
systems with moderate pasture utilization, but scenario same dry period of 62 d (which happened randomly
G2 had higher supplemental feed input. Scenarios G4 throughout the year), calving intervals of 14 mo (Ag-
and G5 focused on maximizing pasture intake, which Source, 2016), and annual mortality rates for weaned
constituted 60% of the total diet DMI during graz- and unweaned heifers of 1.6 and 4.2%, respectively
ing months and 40% of the weighted annual DMI. In (USDA, 2011), and animals reached the same BW
these scenarios, calving was concentrated in the spring, (milking cows, mature heifers, and dry cows = 650
meaning that lactating cows were not milked for 2 mo kg; growing heifers = 347 kg; calves = 52 kg; Powell
during winter. The differentiation factor between these et al., 2008). Cows achieved their mature weight at 21
2 scenarios was that scenario G4 assumed less feed mo of age, and all heifers needed to maintain the herd
supplementation than scenario G5. size were raised on farm as replacements. Nonlactat-
All scenarios targeted a daily milk production of ing animals were divided into heifers aged 1 to 11 mo,
10,000 kg at concentrations of 4.8% lactose, 3.1% pro- heifers aged 12 to 21 mo, and heifers aged >21 mo.
tein, and 3.7% fat (AgSource, 2016). Scenarios that The annual adult replacement rate (including mor-
incorporated grazing also incorporated different lev- tality and culling) was 40.3% (AgSource, 2016). All
els of supplement to maintain milk production levels scenarios maintained the same crop and forage yields
comparable with those of confinement systems. With and had high soil fertility.

Table 1. Seasonal cow diet composition and characteristics of scenarios that combine grazing and confinement and scenarios that use yearly
confinement

Combined grazing and confinement Confinement

Item G1 G2 G3 G4 G5 C1 C2AD1
Grazing season (May–September)
DMI from pasture (%) 9.0 49.0 54.0 56.0 62.0 0.0 0.0
Milk yield (kg/cow per day) 35.0 35.0 28.0 35.0 28.0 35.0 35.0
Diet for milking cows (kg/cow per day)
   Intensively managed pasture 2.0 12.0 12.0 14.0 14.0 — —
   Mature legume silage 5.0 3.0 3.0 2.0 2.0 5.0 5.0
  Corn silage 5.0 — — — — 8 8
  Corn grain 8.4 9.7 7.2 9.0 6.5 6.5 6.5
  Soybean meal 2.4 — — — — 2.5 2.5
  Total DMI 22.8 24.7 22.2 25.0 22.5 21.9 21.9
Confined season (October–April)
Milk yield (kg/cow per day) 35.0 35.0 35.0 35.0 35.0 35.0 35.0
Diet for milking cows (kg/cow per day)
   Intensively managed pasture 0.0 0.0 0.0 0.0 0.0 0.0 0.0
   Mature legume silage 5.0 5.0 5.0 5.0 5.0 5.0 5.0
  Corn silage 8.0 8.0 8.0 8.0 8.0 8.0 8.0
  Corn grain 6.5 6.5 6.5 6.5 6.5 6.5 6.5
  Soybean meal 2.5 2.5 2.5 2.5 2.5 2.5 2.5
  Total DMI 21.9 21.9 21.9 21.9 21.9 21.9 21.9
Year (12 mo)
No. of milking cows 286 286 315 286 315 286 286
No. of growing heifers 161 161 176 161 176 173 173
No. of mature heifers and dry cows 66 66 70 66 70 83 83
DMI from pasture (%) 4.0 22.0 23.0 40.0 43.0 0.0 0.0
Total milk production (kg/d per herd) 10,000 10,000 10,000 10,000 10,000 10,000 10,000
Weighted average annual DMI 22.3 23.1 22.0 23.2 22.1 21.9 21.9
   (kg/cow per day)
Diet for dry cows during dry period
   (kg/cow per day)
   Mature legume silage 4.0 4.0 4.0 4.0 4.0 4.0 4.0
  Corn silage 3.3 3.3 3.3 3.3 3.3 3.3 3.3
  Total DMI 7.3 7.3 7.3 7.3 7.3 7.3 7.3
1
Scenario C2AD includes anaerobic digestion (AD).

Journal of Dairy Science Vol. 100 No. 8, 2017

 ENVIRONMENTAL IMPACTS OF GRAZING DAIRY 5
Table 2. Material and energy inputs for pasture production

Value Primary energy

Input (kg/t of DM) Source (MJ/kg) Source
1
Seed 0.9 Rotz et al. (2010) 56.15 West and Marland (2002)
Pesticides 0.051 Rotz et al. (2010) 232 PE International (2010)
Nitrogen2 16.2 Laboski and Peters (2012) 41.2 NREL (2010)
Phosphorus2 3.3 Laboski and Peters (2012) 6.38 NREL (2010)
Potassium2 22.8 Laboski and Peters (2012) 6.13 PE International (2010)
1
Average for the United States.
2
Total nutrient requirements supplied by both manure and purchased fertilizers.

Feed Production. Material and energy inputs re- for manure application in 0.2-h cycles (Hadrich et al.,
quired for pasture production are presented in Table 2010), which demanded 0.4 L of diesel/ton of applied
2. Additional inputs and embedded energy required manure (ASABE, 2006). The total energy required for
for crop production during confined months included manure collection, transport to storage, and agitation
nutrients, seeds, herbicides, pesticides, lime, diesel, was estimated as 1.28 kWh/ton of manure (based on
gasoline, liquefied petroleum gas, electricity, natural survey data from 19 Wisconsin dairy farms; S. Sanford,
gas, and machinery (Tables A1 and A2 in the Ap- University of Wisconsin–Madison, personal communi-
pendix). Nitrogen (N), phosphorus (P), and potassium cation).
(K) were applied according to pasture and crops re- Scenario C2AD added an anaerobic digester that
quirements, prioritizing the use of dairy manure over processed manure into biogas and digestate after ma-
purchased fertilizers. For scenarios combining grazing nure collection. It is assumed that 0.268 m3 of biogas
and confinement, manure was collected and stored from was produced per 1 kg of volatile solids (VS; Berglund
October to April. It was assumed that all P and K and Börjesson, 2006) with a composition of 65% CH4
from manure applied to soils was available to plants but and 35% CO2 (Appels et al., 2011). Biogas was further
that only 50% of the applied N was available (Beegle processed into electricity in a reciprocating engine at
et al., 2008). Crop yields were 9,000 kg of DM/ha for 35% efficiency and 90% capacity factor (USEPA, 2008).
pasture (Laboski and Peters, 2012), 8,515 kg of DM/ha Approximately 28% of the produced biogas was used
for alfalfa silage, 13,332 kg of DM/ha for corn silage, to heat the digester to maintain a steady temperature
6,891 kg of DM/ha for corn grain, and 1,871 kg of DM/ of 38°C (Berglund and Börjesson, 2006). Digestate
ha for soybean based on the latest USDA census for was stored for 6 mo and applied to land using surface
Wisconsin (USDA-NASS, 2013). broadcast methods in spring and fall. Because of the
Manure Handling. Manure composition and char- reduction of TS in the manure during AD, it was as-
acteristics were estimated with the equations presented sumed that no crust was formed on top of the storage
in Appendix Table A3. Two manure management in this scenario.
strategies were considered: direct land application (all Estimation of GHG Emissions. At the dairy
scenarios except C2AD) and processing through AD farm level, GHG emissions included CH4 and N2O from
(scenario C2AD). Manure was managed the entire year enteric fermentation, manure management, and soils
in scenarios C1 and C2AD and for 7 mo in scenarios (Table 3) and CO2 emissions from the combustion of
G1 through G5. Manure was scraped daily and con- fossil fuels and the embedded energy in material and
tained 10% moisture chopped straw, which was used as energy inputs. Enteric CH4 emissions depend on the
bedding material. After collection, manure was stored chemical composition of the diet and were estimated
in uncovered pits for a period of 6 mo at TS content with the method developed by Moe and Tyrrell (1979).
higher than 10%, which promoted the formation of a Emissions from manure management were estimated
natural crust on the surface (Rotz et al., 2015). As a based on factors from the literature and equations from
result, it was assumed that manure was agitated just the integrated farm system model that capture the local
before emptying the storage. Manure was then applied conditions of Wisconsin. The CH4 and N2O were emit-
to land by surface broadcast during the months of ted from manure after excretion at the barn, during
April and October. This means that a load of 6 mo of storage, and after land application in confined months,
accumulated manure was applied during April but a whereas these emissions occurred only from the manure
load of only 1 mo of accumulated manure was applied excreted on pasture in grazing months. Emissions of
during October in scenarios that incorporated graz- NH3 were also estimated because they have an indirect
ing. A truck (179 kW, 240 HP, 35,900 m3) was used effect on N2O emissions. Scenario C2AD assumed that

Journal of Dairy Science Vol. 100 No. 8, 2017

6 AGUIRRE-VILLEGAS ET AL.

all CH4 from biogas was converted to CO2 through RESULTS AND DISCUSSION
combustion. For accounting purposes, the emissions
that occurred after the land application of manure and Seasonal System Results Before Allocation
digestate were assigned to the production of crops given
GHG Emissions. For all scenarios, enteric CH4 was
that manure was used as a nutrient source. Finally, fos-
the major contributor to GHG emissions, followed by
sil emissions from fuel combustion and resource use for
N2O from the production of crops for dairy feed, CH4
crop and pasture production were estimated based on
from manure handling, and fossil emissions from energy
emission factors from the literature (Appendix Table
consumption (Table 4). Enteric CH4 per kilogram of
A4).
FPCM was lower in grazing months for scenarios G1,
G2, and G5 but higher for grazing months in scenarios
Allocation of Coproducts G3 and G4 when compared with confined months. This
difference is attributable to both the composition of
At the input level, there was multifunctionality be-
the dairy diet that directly affected emissions and the
tween soybean meal and oil in scenarios C1, C2AD, and
higher milk production (as milk was the functional unit
G1. At the output level, there was multifunctionality
of this study) achieved through feed supplementation.
between milk and meat in all scenarios and between
Crop production was the second source of GHG emis-
milk, meat, and biogas-based electricity in scenario
sions and included CH4, N2O, and fossil emissions from
C2AD. To deal with multifunctionality, the Interna-
on-farm and off-farm processes. Emissions of CH4 after
tional Organization for Standardization recommends
excretion on pasture (G1–G5) were higher than after
first applying subdivision, then system expansion, and
manure land application (C1 and C2AD; Figure 2).
finally allocation when possible (ISO, 2006). Subdivi-
This is because the content of manure VS in grazing
sion cannot be applied in this system because the inputs
months was higher than in confinement months as VS
and outputs of several processes cannot be assigned en-
were not degraded in the barn or during storage. Emis-
tirely to a specific product. System expansion was ap-
sions of N2O from crops in confined months were lower
plied in scenario C2AD because biogas-based electricity
than in grazing months for 2 reasons. First, more N was
provides the same function as grid electricity (Azapagic
volatilized as NH3 during collection and storage in con-
and Clift, 1999). Under this method, it was assumed
fined months, reducing the available N to be potentially
that the production of biogas electricity displaced an
emitted as N2O after land application. Second, the N2O
equal amount of grid electricity on an energy content
emission factor for pasture doubled the emission fac-
basis (Curran, 2006; Finnveden et al., 2009; Suh et al.,
tor for confined systems in which manure was applied
2010). Negative environmental indicators showed that
to cropping fields (IPCC, 2006a; Lesschen et al., 2011;
the burdens of the system being replaced (grid electric-
O’Brien et al., 2014).
ity) were larger than the burdens of the system under
Manure management was the third source of GHG
study (milk, meat, and biogas electricity).
emissions in all scenarios except for C2AD, which
There are multiple potential substitutes for meat and
showed a significant reduction in GHG emissions in the
oil, which complicates applying system expansion at
forms of both CH4 and N2O (Figure 2). Emissions of
this point. As a result, allocation ratios were applied
CH4 were reduced as VS in the stored digestate and
between soybean meal and oil (mass) and between milk
were reduced after the digestion process. In addition,
and meat (nutritional content). For soybean meal, a
N2O emissions from storage were decreased in C2AD
ratio of 0.79 kg of meal/kg of soybean was estimated
because of the absence of a natural crust, which pro-
based on Wang et al. (2011). For milk and meat, the
moted a mix of aerobic and anaerobic conditions that
nutritional content from both protein and fat in bone-
are optimal for nitrification and denitrification. Graz-
less meat was compared with the nutritional content
ing months showed significantly higher emissions from
of milk produced during the same time period; 98.9%
manure storage as the production of 1 mo of manure
of the impacts were attributed to milk and 1.1% were
was stored for 6 mo before its land application. This
attributed to meat (Equation A3). Boneless meat pro-
time period coincided with summer, and CH4 emissions
duction for consumption was 195 kg per culled adult
from storage are highly dependent on temperature
dairy cow (Cederberg and Stadig, 2003) with 6.5% fat
(Rotz et al., 2015).
and 20.7% protein (Franco et al., 2009; Minchin et al.,
The GHG emissions from material and energy in-
2009). An allocation ratio based on the nutritional con-
puts were divided into on-farm and off-farm emissions.
tent of these products was chosen over other physical
It was assumed that milking and manure operations
ratios (e.g., mass) because both milk and meat are val-
in the barn used electricity, and because electricity is
ued for their fat and protein contents as food products
produced outside of the farm, on-farm emissions came
(Aguirre-Villegas et al., 2015).
Journal of Dairy Science Vol. 100 No. 8, 2017
 Table 3. Description of emission factors from biotic sources

Emission and source Equation Reference

CH4
Enteric fermentation CH4 = 0.814 + (0.122 × SR) + (0.415 × hemicellulose) + (0.633 × cellulose), Moe and Tyrrell (1979)
where CH4 (enteric emissions) is in megacalories and soluble residue (SR), hemicellulose, and cellulose are in
kilograms.
Manure application confined CH4 = [(0.17 × FVFA) + 0.026] × Acrop × 0.032, Chianese et al. (2009)
where CH4 = methane from application (kg of CH4/d), FVFA = VFA in manure (mmol/kg of manure), and Acrop =
area of application (Ha).
Manure on barn floor A Chianese et al. (2009)
CH4 = max (0, 0.13 × T) × barn ,
1, 000
where CH4 = daily emissions from barn (kg of CH4/d), T = temperature, and Abarn = barn area exposed to manure
(m2).
Manure storage   E     E   Chianese et al. (2009)
ln(A)−    ln(A)− 
 24 × VSd × b1  RT    24 × VSnd × b2   RT  
CH4 =   × e   + ×e   ,
 1, 000    1, 000  
   
where CH4 = methane emissions from storage (g of CH4/d), VSd = degradable volatile solids (g), VSnd =
nondegradable volatile solids (g), b1 and b2 = rate-correcting factors (b1 = 1, b2 = 0.01; dimensionless), A =
Arrhenius parameter (g of CH4/kg of volatile solids per hour), ln(A) = 43.33, E = apparent activation energy
(112,700 J/mol), R = gas constant (8.314 J/K per mole), and T = temperature (K).
Manure on pasture CH4 = VS × 0.24 × 0.67 × 0.01, IPCC (2006b)
where CH4 = methane emissions from manure on pasture (kg of CH4/d) and VS = volatile solids in manure (kg/d).
N2O direct
Manure and fertilizer 10 g of N2O-N/kg of N IPCC (2006a)
  application confined
Manure on barn floor 2.1 × 10−3 g of N2O-N/animal per day Wheeler et al. (2008)
Manure storage 13 g of N2O-N/m3 of manure Amon et al. (2006)
Manure on pasture 20 g of N2O-N/kg of N IPCC (2006a)
N2O indirect
Ammonia volatilization 10 g of N2O-N/kg of NH3-N IPCC (2006b)
ENVIRONMENTAL IMPACTS OF GRAZING DAIRY

Leaching1 7.5 g of N2O-N/kg of N IPCC (2006b)

NH3
Fertilizer application 100 g of NH3-N/kg of N IPCC (2006a)
Manure application   days  17  Jokela et al. (2004)
TAN × (20 + 5 × TS) ×   ×  
NH3 =   days + 0.3   14 
,
100
where NH3 = emissions after application (kg of NH3), TAN = total ammonia N in manure (kg of NH3-N), TS =
total solids in manure (%), and days = days to incorporate manure (if not incorporated, days >7).
Manure on barn floor 0.05 kg of NH3-N/kg of N intake Harper et al. (2009)
Manure storage 0.03 kg of NH3-N/kg of N intake Harper et al. (2009)
1
Leaching fraction = 0.3 kg of N (IPCC, 2006b).

Journal of Dairy Science Vol. 100 No. 8, 2017

7
8 AGUIRRE-VILLEGAS ET AL.

Table 4. Greenhouse gas (GHG) emissions, net energy intensity (NEI), and land use during grazing (May–September) and confined (October–
April) months for scenarios that combine grazing and confinement and during the entire year for confined scenarios

Grazing months Confined months

Process1 G1 G2 G3 G4 G5 G1–G5 C1 C2AD2

GHG emissions (kg of CO2-eq/kg of FPCM)
Enteric emissions 0.46 0.45 0.54 0.54 0.45 0.48 0.48 0.48
Manure management3 0.10 0.11 0.13 0.13 0.11 0.06 0.11 0.03
Crops production 0.18 0.21 0.26 0.28 0.22 0.14 0.14 0.14
On-farm energy use 0.07 0.05 0.06 0.05 0.04 0.08 0.08 0.08
Off-farm emissions from resource use 0.06 0.05 0.06 0.06 0.05 0.06 0.06 0.06
Total 0.88 0.87 1.04 1.05 0.88 0.83 0.87 0.79
NEI (MJ/kg of FPCM)
Milking and barn operations4 0.46 0.45 0.54 0.54 0.45 0.48 0.48 0.48
Manure management3,4 0.10 0.11 0.13 0.13 0.11 0.06 0.11 0.03
Feed crops on farm 0.18 0.21 0.26 0.28 0.22 0.14 0.14 0.14
Feed crops off farm 0.07 0.05 0.06 0.05 0.04 0.08 0.08 0.08
Energy from biogas5 0.06 0.05 0.06 0.06 0.05 0.06 0.06 0.06
Total 0.88 0.87 1.04 1.05 0.88 0.83 0.87 0.79
Land use (m2/kg of FPCM) 1.25 1.15 1.31 1.31 1.15 1.25 1.25 1.25
1
CO2-eq = CO2 equivalents; FPCM = fat- and protein-corrected milk.
2
Scenario C2AD includes anaerobic digestion (AD).
3
Manure management includes only collection and storage of manure because land application is assigned to the production of crops on farm.
4
Energy for milking and barn operations and manure handling (collection and agitation) is in the form of electricity.
5
The energy used to heat the digester is discounted from the produced biogas.

entirely from crop production, which included lime pasture and high feed supplementation in G1. Scenarios
application and fossil liquid fuel combustion (Figure G3 and G4 were also high in terms of off-farm NEI
2). As expected, on-farm emissions were higher in even though they included less corn in the diet because
confined months because of the use of machinery for these scenarios had the lowest milk production rates.
crop production. On-farm emissions from G1 during Scenario G5 had the lowest NEI because it had high
grazing months were higher compared with the rest pasture content and enough feed supplementation to
of the grazing scenarios because of the farm’s higher maintain high-producing cows.
level of corn and alfalfa silage supplementation. Finally, Direct Land Use. Direct land use depends on cow
off-farm emissions reflect the embedded emissions from productivity and cropping yields. Scenarios G2 and G5
material (e.g., seeds, synthetic fertilizers) and energy required less land for the same milk production during
inputs needed for milk, manure, and cropping opera- grazing months than confined scenarios C1 and C2AD
tions. Most off-farm emissions were also from crop pro- (Table 4). The high direct land use of confined sce-
duction, and in some scenarios they matched on-farm narios was attributable to the low yields of soybeans
emissions, highlighting the importance of embedded compared with other crops and the high soybean meal
emissions in farm activities. content in confined diets. In general, crops that pro-
NEI. The production of feed crops off farm and on duce higher protein than soybean meal per unit of area
farm was the major contributor to NEI (Table 4), high- could decrease direct land use. One common alternative
lighting the significant amount of energy that goes into source of dietary protein is dried distillers grains with
manufacturing inputs (e.g., fertilizers, fuels). Off-farm solubles (DDGS). Studies have shown an increase in
NEI was higher in some scenarios that incorporated direct land use of dairy operations despite DDGS being
grazing when synthetic fertilizers were purchased be- a by-product of ethanol (where only a portion of the
cause manure was not available year round in these land area is allocated to DDGS production) due to the
scenarios. The NEI from the production of crops on lower yields and protein content of DDGS compared
farm was significantly lower during grazing when there with soybean meal (Aguirre-Villegas et al., 2015). Fur-
was less intervention from machinery. When compar- ther research using a common methodology is needed
ing grazing and confinement months, scenario C2AD to complete a detailed evaluation on integrating DDGS
was the less energy intense because of the production in combined grazing and confined dairy systems. A sec-
of biogas-based electricity. However, scenarios C1 and ond factor affecting land use results was pasture yield
G1 were the most energy intense because AD was not (9 t/ha), which was on the higher end of the range for
considered in C1 and because there was low DMI from Wisconsin (0.5–5 tons/acre; Laboski and Peters, 2012).

Journal of Dairy Science Vol. 100 No. 8, 2017

 ENVIRONMENTAL IMPACTS OF GRAZING DAIRY 9

If a yield of 2 tons/acre was used for pasture, direct est GHG emitters because they were influenced by
land use would be increased by 4% (G1) to 62% (G5) in their lower milk yields. Scenarios G1 and G2 produced
scenarios that combine confinement and grazing. similar GHG emissions even though scenario G1 used
the least amount of pasture. When comparing scenarios
Allocated Yearly Results After Weighting Grazing with moderate DMI from pasture (22–23%), scenario
and Confined Months G2 resulted in lower GHG emissions than scenario G3
because its level of feed supplementation was higher,
After weighting the 5 grazing months and 7 confined which increased milk yield. This same trend applied
months and after applying allocation and system ex- for scenarios G4 and G5, which had high DMI from
pansion, results for GHG emissions and NEI followed pasture (40–43%); scenario G5 with higher feed supple-
the same trend as the system results (Figure 3). The mentation had lower GHG emissions.
GHG emissions were comparable for all scenarios, but The NEI in scenarios combining grazing and confine-
trends suggested that scenarios that combine grazing ment (except for G1) was lower than those in C1 but
and confinement with high milk production (G1, G2, higher than those in C2AD, which included AD (Figure
and G5) were marginally lower GHG emitters than 3b). The addition of a digester in scenario C2AD re-
yearly confined scenarios with the exception of scenario sulted in a positive net energy production, which is in-
C2AD. The addition of a digester decreased emissions dicated by the negative sign. Negative NEI values were
significantly as biogas-based electricity displaced grid- reported by several studies exploring energy production
based electricity, which is produced mostly by coal in using the system expansion methodology (Jury et al.,
Wisconsin. As a result, C2AD was the lowest GHG 2010; Michel et al., 2010; Lim et al., 2011). In addition
emission scenario. Scenarios G3 and G4 were the high- to biogas-based electricity, energy credits resulted from

Figure 2. System life cycle greenhouse gas (GHG) emissions before allocation for scenarios that combine grazing and confinement and sce-
narios with yearly confinement during the 5 grazing months (May–September). A total of 7 scenarios representative of Wisconsin were modeled
based on information collected through surveys sent to Wisconsin dairy farms: 2 confinement scenarios (C1 and C2AD) with no access to pasture
year round and 5 scenarios combining confinement and grazing at varying percentages of annual DMI from pasture (G1, G2, G3, G4, and G5,
where G1 had the lowest DMI from pasture and G5 had the highest). CO2-eq = CO2 equivalents; AD = anaerobic digestion; FPCM = fat- and
protein-corrected milk.

Journal of Dairy Science Vol. 100 No. 8, 2017

10 AGUIRRE-VILLEGAS ET AL.

Figure 3. Life cycle (a) greenhouse gas (GHG) emissions and (b) net energy intensity (NEI) after allocation, system expansion, and weighting
grazing and confined months in 1 yr. A total of 7 scenarios representative of Wisconsin were modeled based on information collected through
surveys sent to Wisconsin dairy farms: 2 confinement scenarios (C1 and C2AD) with no access to pasture year round and 5 scenarios combining
confinement and grazing at varying percentages of annual DMI from pasture (G1, G2, G3, G4, and G5, where G1 had the lowest DMI from
pasture and G5 had the highest). CO2-eq = CO2 equivalents; AD = anaerobic digestion; FPCM = fat- and protein-corrected milk.

the avoided burdens of producing grid electricity by DMI from pasture (to reduce resource use) as seen in
system expansion. The significant amount of credits scenarios G2 and G5 (Figure 4), the differences in re-
from replacing grid electricity highlights the substan- sults were small. This suggests that integrating grazing
tial energy resources that are needed to produce grid into confined systems in climates such as Wisconsin,
electricity based on fossil fuels. Within scenarios that which prevent the incorporation of yearlong grazing,
combine grazing and confinement, NEI was the low- will likely not improve environmental impact indica-
est in scenario G5, where feed supplementation (that tors enough to warrant a change in practices. Moreover,
increased milk yield) and DMI content from pasture management practices are usually driven by profitabil-
(that reduced machinery use) were balanced. On the ity to the farmer, which was not evaluated in this study.
contrary, additional feed supplementation (G1) or addi- O’Brien et al. (2015) found that both farm profitability
tional DMI from pasture with limited supplementation and GHG emissions in Irish dairy farms would improve
(G3 and G4) increased NEI. by increasing milk production per hectare from grazed
Weighted and allocated results for direct land use grass instead of feed supplementation. More research is
were comparable for all scenarios and were 1.23 m2/kg needed to evaluate how the practices evaluated in this
of FPCM for scenarios G1, C1, and C2AD; 1.26 m2/ study would affect the economic performance of dairy
kg of FPCM for scenarios G3 and G4; and 1.19 m2/ farms, as more profits would likely drive more farms to
kg of FPCM for scenarios G2 and G5. As with system shift to integrated grazing and confinement.
results, scenarios with lower milk yield required more O’Brien et al. (2014) compared high-performance
land to achieve equivalent milk production. grazed and confined systems and reported 0.84 kg of
CO2-eq/kg of ECM for grazing systems in Ireland and
Broad Factors Affecting Environmental Impacts 0.90 kg of CO2-eq/kg of ECM for confined systems in
the United States, suggesting that increasing the per-
Like Belflower et al. (2012), who reported very simi- formance of a grazing system may increase the differ-
lar emissions of 0.88 and 0.87 kg of CO2-eq/kg of ECM ence in these systems. However, the authors evaluated
for pasture and confined systems, respectively, in the different locations and practices suggesting that the
United States, this study found small differences in differences may not actually be attributable to graz-
GHG emissions per unit of milk produced for most sce- ing. For example, O’Brien et al. (2014) assumed land
narios but a slight increase for some grazing scenarios. application of manure only once per year. Our study as-
Although GHG emissions, NEI, and land use were sumed that 1 mo of manure was stored for 6 mo during
minimized by combining the right amount of feed sup- grazing months as land application of manure occurred
plementation (to increase milk production) and enough 2 times per year. As this storage happened during the
Journal of Dairy Science Vol. 100 No. 8, 2017
 ENVIRONMENTAL IMPACTS OF GRAZING DAIRY 11

summer, CH4 emissions from storage were significantly

increased. By considering only 1 manure application
per year, GHG emissions would be reduced by 5 to 6%
in scenarios combining grazing and confinement. These
results suggest that manure management significantly
affects emission outcomes because storing manure for
1 mo in hot temperatures emits more CH4 than does
storing manure for 6 mo in cold temperatures.
The differentiation of on-farm and off-farm impacts
highlights the significant energy resources embedded
in producing material and energy inputs for crop pro-
duction and feed supplementation. This is shown by
scenarios C1 (no DMI from pasture) and G1 (4% of
DMI from pasture) as these scenarios resulted in the
highest NEI despite their high milk yield. Doole (2014)
highlighted the potential reduction of GHG emissions
by reducing input use, but NEI was not analyzed. As
modeled in scenario C2AD, AD is a promising tech-
nology for reducing both GHG emissions and NEI in
confined dairy systems that collect manure. However,
AD is an expensive technology, and not all farmers
can afford the high maintenance costs. Faulhaber et
al. (2012) reported that AD is economically unfeasible
for farms with fewer than 400 cows and that even for
farms with 1,000 cows it is challenging to compete
with low natural gas prices. These results suggest that
small farms might benefit from incorporating grazing,
but larger farms that can afford AD might be able to
reduce their environmental impacts even further.
An important factor affecting GHG emissions was
the N2O emission factor of manure directly deposited
in pasture, as it doubled the N2O emission factor of
manure applied to crops. Previous studies did not use
(or do not state) different emission factors for N2O,
with the exception of O’Brien et al. (2014) and Capper
et al. (2009). This difference in factors might be related
to the higher ammoniacal N available in just-excreted
manure as N2O emissions from manure have been
cited to be directly related to ammoniacal N content
(Chadwick et al., 2011). In yearly confined systems,
more than 50% of N can be volatilized as NH3 in the
barn, storage, and after land application, significantly
reducing the potential of N2O emissions (IPCC, 2006b;
Hristov et al., 2011). If the same N2O emission fac-
tor was considered for grazing and confined months,
total GHG emissions during grazing months would be
Figure 4. Relating (a) greenhouse gas (GHG) emissions, (b) net
energy intensity (NEI), and (c) land use with DMI from pasture for all reduced by 4 to 5%.
scenarios with the exception of the scenario with anaerobic digestion The allocation method has been shown to significant-
(scenario C2AD). A total of 7 scenarios representative of Wisconsin ly affect results of dairy LCA studies (Cederberg and
were modeled based on information collected through surveys sent to
Wisconsin dairy farms: 2 confinement scenarios (C1 and C2AD) with Stadig, 2003; Thomassen et al., 2008a; Aguirre-Villegas
no access to pasture year round and 5 scenarios combining confine- et al., 2012). When applying an allocation ratio be-
ment and grazing at varying percentages of annual DMI from pasture tween milk and meat, the literature supports the idea
(G1, G2, G3, G4, and G5, where G1 had the lowest DMI from pasture
and G5 had the highest). CO2-eq = CO2 equivalents; FPCM = fat- that milk receives the majority of the environmental
and protein-corrected milk. Color version available online. burdens and benefits (Cederberg and Stadig, 2003;
Journal of Dairy Science Vol. 100 No. 8, 2017
12 AGUIRRE-VILLEGAS ET AL.

Casey and Holden, 2005; Thomassen et al., 2008b; use of material and energy resources for crop produc-
Rotz et al., 2010; Mc Geough et al., 2012; Thoma et tion and the emissions from manure storage. Anaerobic
al., 2013). However, the magnitude of the allocation digestion significantly reduced GHG emissions both
ratio can change significantly based on the criteria used from decreased VS during manure storage and from the
for its calculation. This study applied an allocation avoided emissions from displaced grid-based electricity.
ratio based on nutritional content because the purpose Moreover, incorporating AD in the confined scenario
of dairy products is to provide a complete source of changed the farm from a net energy consumer to a net
nutrients and proteins for human consumption. This energy producer. This suggests that large farms might
reasoning has also been applied by other LCA studies achieve significant environmental benefits from install-
(FAO, 2010) and has been used to develop the FPCM ing a digester if they can afford the costs, whereas small
functional unit, which has become universal for milk farms could benefit from grazing, which does not repre-
LCA studies. A mass allocation ratio can be calcu- sent as large an investment.
lated based on the following assumptions: 40.3% adult
culling and death replacement (USDA, 2011), 49% of ACKNOWLEDGMENTS
calves being male (AgSource, 2016), and a production
of 195 kg of bone-free meat per adult dairy cow (Ced- This study is based on work supported by the Grazing
erberg and Stadig, 2003). Almost 10 calves per year are Land Conservation Initiative of the Wisconsin Depart-
maintained at an average weight of 57 kg. With these ment of Natural Resources (Madison) through its 2010
assumptions, 97% of the impacts would be attributed Grazing Land Conservation Initiative Research Grant
to milk and 3% would be attributed to meat. This is (award no. 10309) and by the National Institute of
in line with O’Brien et al. (2014), who allocated GHG Food and Agriculture, USDA (Washington, DC; award
emissions based on mass, with 98% of emissions going no. 2013-68002-20525). Any opinions, findings, conclu-
to milk and 2% going to meat. sions, or recommendations expressed in this publication
Finally, this study did not consider carbon sequestra- are those of the authors and do not necessarily reflect
tion of grassland soils, which has been an important the view of the USDA.
factor in the evaluation of GHG emissions in previous
studies. Belflower et al. (2012) and O’Brien et al. (2014) REFERENCES
found that GHG emissions in grazing systems can be
reduced by 10 and 9%, respectively, when accounting AgSource. 2016. DHI Benchmarks Breed Averages. AgSource Coop-
erative Services, Verona, WI.
for the sequestration of carbon. Aguirre-Villegas, H. A., R. A. Larson, and D. J. Reinemann. 2014.
From waste-to-worth: Energy, emissions, and nutrient implications
of manure processing pathways. Biofuels Bioprod. Biorefin. 8:770–
CONCLUSIONS 793. https://doi.org/10.1002/bbb.1496.
Aguirre-Villegas, H. A., F. X. Milani, S. Kraatz, and D. J. Reine-
In this study, GHG emissions, NEI, and land use of mann. 2012. Life cycle impact assessment and allocation meth-
2 confined dairy systems with and without AD and 5 ods development for cheese and whey processing. Trans. ASABE
55:613–627.
dairy systems that combine confinement and grazing Aguirre-Villegas, H. A., T. H. Passos-Fonseca, D. J. Reinemann, L. E.
were quantified and compared based on different pas- Armentano, M. A. Wattiaux, V. E. Cabrera, J. M. Norman, and
ture intensities. For all scenarios, CH4 from enteric fer- R. A. Larson. 2015. Green cheese: Partial life cycle assessment of
greenhouse gas emissions and energy intensity of integrated dairy
mentation and N2O from crop production contributed production and bioenergy systems. J. Dairy Sci. 98:1571–1592.
the most to GHG emissions. In addition, the production https://doi.org/10.3168/jds.2014-8850.
of crops was the most energy intense process. Overall, Amon, B., V. Kryvoruchko, T. Amon, and S. Zechmeister-Boltenstern.
2006. Methane, nitrous oxide and ammonia emissions during stor-
results for all scenarios were comparable, showing that age and after application of dairy cattle slurry and influence of
the evaluated feeding strategies can include pasture slurry treatment. Agric. Ecosyst. Environ. 112:153–162. https://
without drastically affecting milk yield and environ- doi.org/10.1016/j.agee.2005.08.030.
Appels, L., J. Lauwers, J. Degrève, L. Helsen, B. Lievens, K. Willems,
mental impacts. The comparable results also suggest J. Van Impe, and R. Dewil. 2011. Anaerobic digestion in global
that any environmental benefits from grazing are small bio-energy production: Potential and research challenges. Renew.
in regions that cannot support continuous grazing. De- Sustain. Energy Rev. 15:4295–4301. https://doi.org/10.1016/j.
rser.2011.07.121.
spite the similarity in results, environmental benefits ASABE (American Society of Agricultural and Biological Engineers).
across impact categories of GHG emissions, NEI, and 2005. Manure Production and Characteristics. ASAE D384.2
land use were achieved in scenarios that supplemented MAR2005. ASABE, St. Joseph, MI.
ASABE (American Society of Agricultural and Biological Engineers).
enough feed to increase milk yield but maintained a 2006. Agricultural Machinery Management. Standards EP496.3.
high enough DMI from pasture to reduce material and ASABE, St. Joseph, MI.
energy use. These trends suggest that yearly confined Azapagic, A., and R. Clift. 1999. Allocation of environmental burdens
in multiple-function systems. J. Clean. Prod. 7:101–119.
scenarios could benefit from strategies that reduce the
Journal of Dairy Science Vol. 100 No. 8, 2017
 ENVIRONMENTAL IMPACTS OF GRAZING DAIRY 13
Beegle, D., K. Kelling, and M. Schmitt. 2008. Nitrogen from animal loin from the meat of Holstein-Fresian cull cows. Meat Sci. 83:484–
manures. Pages 823–882 in Nitrogen in Agricultural Systems. J. 491. https://doi.org/10.1016/j.meatsci.2009.06.030.
Schepers and W. Raun, ed. Agron. Monographs, Crop Science So- Hadrich, J. C., T. M. Harrigan, and C. A. Wolf. 2010. Economic com-
ciety of America and Soil Science Society of America, Madison, parison of liquid manure transport and land application. Appl.
WI. Eng. Agric. 26:743–758.
Belflower, J. B., J. K. Bernard, D. K. Gattie, D. W. Hancock, L. Harper, L. A., T. K. Flesch, J. M. Powell, W. K. Coblentz, W. E.
M. Risse, and C. A. Rotz. 2012. A case study of the potential Jokela, and N. P. Martin. 2009. Ammonia emissions from dairy
environmental impacts of different dairy production systems production in Wisconsin. J. Dairy Sci. 92:2326–2337. https://doi.
in Georgia. Agric. Syst. 108:84–93. https://doi.org/10.1016/j. org/10.3168/jds.2008-1753.
agsy.2012.01.005. Hristov, A. N., M. Hanigan, A. Cole, R. Todd, T. A. McAllister, P. M.
Berglund, M., and P. Börjesson. 2006. Assessment of energy perfor- Ndegwa, and C. A. Rotz. 2011. Review: Ammonia emissions from
mance in the life-cycle of biogas production. Biomass Bioenergy dairy farms and beef feedlots. Can. J. Anim. Sci. 91:1–35. https://
30:254–266. https://doi.org/10.1016/j.biombioe.2005.11.011. doi.org/10.4141/cjas10034.
Capper, J. L., R. A. Cady, and D. E. Bauman. 2009. The environmen- IDF (International Dairy Federation). 2009. Environmental/ecological
tal impact of dairy production: 1944 compared with 2007. J. Anim. impact of the dairy sector: Literature review on dairy products
Sci. 87:2160–2167. https://doi.org/10.2527/jas.2009-1781. for an inventory of key issues. Bulletin 436/2009. IDF, Brussels,
Capper, J. L., E. Castañeda-Gutiérrez, R. A. Cady, and D. E. Bau- Belgium.
man. 2008. The environmental impact of recombinant bovine so- IDF (International Dairy Federation). 2010. A common carbon foot-
matotropin (rbST) use in dairy production. Proc. Natl. Acad. Sci. print approach for dairy. The IDF guide to standard lifecycle as-
USA 105:9668–9673. https://doi.org/10.1073/pnas.0802446105. sessment methodology for the dairy sector. Bulletin 445/2010.
Casey, J. W., and N. M. Holden. 2005. Analysis of greenhouse gas IDF, Brussels, Belgium.
emissions from the average Irish milk production system. Agric. IPCC (Intergovernmental Panel on Climate Change). 2006a. N2O
Syst. 86:97–114. https://doi.org/10.1016/j.agsy.2004.09.006. emissions from managed soils, and CO2 emissions from lime and
Cederberg, C., and M. Stadig. 2003. System expansion and allocation urea application. Pages 11.1–11.54 in IPCC Guidelines for Nation-
in life cycle assessment of milk and beef production. Int. J. Life al Greenhouse Gas Inventories. Vol. 4: Agriculture, Forestry and
Cycle Assess. 8:350–356. Other Land Use. H. Eggleston, L. Buendia, K. Miwa, T. Ngara,
Chadwick, D., S. Sommer, R. Thorman, D. Fangueiro, L. Cardenas, B. and K. Tanabe, ed. Institute for Global Environmental Strategies
Amon, and T. Misselbrook. 2011. Manure management: Implica- (IGES), Hayama, Japan.
tions for greenhouse gas emissions. Anim. Feed Sci. Technol. 166– IPCC (Intergovernmental Panel on Climate Change). 2006b. Emis-
167:514–531. https://doi.org/10.1016/j.anifeedsci.2011.04.036. sions from livestock and manure management. Pages 10.1–10.87 in
Chianese, D. S., C. A. Rotz, and T. L. Richard. 2009. Simulation of IPCC Guidelines for National Greenhouse Gas Inventories. Vol. 4:
methane emissions from dairy farms to assess greenhouse gas re- Agriculture, Forestry and Other Land Use. H. Eggleston, L. Buen-
duction strategies. Trans. ASABE 52:1313–1323. dia, K. Miwa, T. Ngara, and K. Tanabe, ed. Institute for Global
Curran, M. A. 2006. Life cycle assessment: Principles and practice. US Environmental Strategies (IGES), Hayama, Japan.
Environmental Protection Agency, Washington, DC. ISO (International Organization for Standardization). 2006. Envi-
de Léis, C. M., E. Cherubini, C. F. Ruviaro, V. P. da Silva, V. do Nas- ronmental management—Life cycle assessment—Principles and
cimento Lampert, A. Spies, and S. R. Soares. 2015. Carbon foot- framework. Standard number 14040:2006. ISO, Geneva, Switzer-
print of milk production in Brazil: A comparative case study. Int. land.
J. Life Cycle Assess. 20:46–60. https://doi.org/10.1007/s11367- Jokela, B., F. Magdoff, S. Barlett, S. Bosworth, and D. Ross. 2004. Nu-
014-0813-3. trient recommendations for field crops in Vermont. Accessed Aug.
Doole, G. J. 2014. Least-cost greenhouse gas mitigation on New Zea- 20, 2016. http://pss.uvm.edu/vtcrops/?Page=nutrientmanure.
land dairy farms. Nutr. Cycl. Agroecosyst. 98:235–251. https:// html.
doi.org/10.1007/s10705-014-9608-y. Jury, C., E. Benetto, D. Koster, B. Schmitt, and J. Welfring. 2010.
Ecoinvent Centre. 2007. Ecoinvent Reports No. 1–25. Accessed Aug. Life cycle assessment of biogas production by monofermentation of
20, 2016. www.ecoinvent.org. energy crops and injection into the natural gas grid. Biomass Bio-
ECW (Energy Center of Wisconsin). 2009. Efficiency and customer- energy 34:54–66. https://doi.org/10.1016/j.biombioe.2009.09.011.
sited renewable energy: Achievable potential in Wisconsin 2012– Kim, S., and B. E. Dale. 2004. Cumulative energy and global warming
2018. ECW Rep. No. 244-1. Accessed Aug. 20, 2016. https://psc. impact from the production of biomass for biobased products. J.
wi.gov/reports/documents/WIPotentialFinal.pdf. Ind. Ecol. 7:147–162.
FAO (Food and Agriculture Organization of the United Nations). Kraatz, S., D. J. Reinemann, and W. Berg. 2009. Energy inputs for
2010. Greenhouse gas emissions from the dairy sector—A life corn production in Wisconsin and Germany. Appl. Eng. Agric.
cycle assessment. Accessed Aug. 20, 2016. http://www.fao.org/ 25:653–662.
agriculture/lead/themes0/climate/emissions/en/. Laboski, C. A. M., and J. B. Peters. 2012. Nutrient application guide-
Frischknecht, R., N. Jungbluth, H. J. Althaus, G. Doka, R. Dones, T. lines for field, vegetable, and fruit crops in WI (A2890). Accessed
Heck, S. Hellweg, R. Hischier, T. Nemecek, G. Rebitzer, and M. Aug. 6, 2016. https://learningstore.uwex.edu/Nutrient-Applica-
Spielmann. 2005. The ecoinvent database: Overview and method- tion-Guidelines-for-Field-Vegetable-and-Fruit-Crops-in-Wiscon-
ological framework. Int. J. Life Cycle Assess. 10:3–9. sin-P185.aspx.
Faulhaber, C. R., D. R. Raman, and R. Burns. 2012. An engineering- Lesschen, J. P., G. Velthof, W. de Vries, and J. Kros. 2011.
economic model for analyzing dairy plug-flow anaerobic digesters: Differentiation of nitrous oxide emission factors for agricultural
Cost structures and policy implications. Trans. ASABE 55:201– soils. Environ. Pollut. 159:3215–3222. https://doi.org/10.1016/j.
209. envpol.2011.04.001.
Finnveden, G., M. Z. Hauschild, T. Ekvall, J. Guinée, R. Heijungs, Lim, S., K. T. Lee, and P. Pinang. 2011. Parallel production of bio-
S. Hellweg, A. Koehler, D. Pennington, and S. Sangwon. 2009. diesel and bioethanol in palm-oil-based biorefineries: Life cycle as-
Recent developments in life cycle assessment. J. Environ. Manage. sessment on the energy and greenhouse gases. Biofuels Bioprod.
91:1–21. https://doi.org/10.1016/j.jenvman.2009.06.018. Biorefin. 5:132–150. https://doi.org/10.1002/bbb.
Flysjö, A., M. Henriksson, C. Cederberg, S. Ledgard, and J. Englund. Mc Geough, E. J., S. M. Little, H. H. Janzen, T. A. McAllister, S.
2011. The impact of various parameters on the carbon footprint M. McGinn, and K. A. Beauchemin. 2012. Life-cycle assessment
of milk production in New Zealand and Sweden. Agric. Syst. of greenhouse gas emissions from dairy production in Eastern
104:459–469. https://doi.org/10.1016/j.agsy.2011.03.003. Canada: A case study. J. Dairy Sci. 95:5164–5175. https://doi.
Franco, D., E. Bispo, L. González, J. A. Vazquez, and T. Moreno. org/10.3168/jds.2011-5229.
2009. Effect of finishing and ageing time on quality attributes of

Journal of Dairy Science Vol. 100 No. 8, 2017

14 AGUIRRE-VILLEGAS ET AL.

Meier Engineering Research. 2011. MyPower TM electricity-sector sessment. J. Dairy Sci. 93:1266–1282. https://doi.org/10.3168/
model. Wisconsin Energy Institute, Madison, WI. jds.2009-2162.
Michel, J., A. Weiske, and K. Möller. 2010. The effect of biogas diges- Stott, K. J., and C. J. P. Gourley. 2016. Intensification, nitrogen use
tion on the environmental impact and energy balances in organic and recovery in grazing-based dairy systems. Agric. Syst. 144:101–
cropping systems using the life-cycle assessment methodology. 112. https://doi.org/10.1016/j.agsy.2016.01.003.
Renew. Agric. Food Syst. 25:204–218. https://doi.org/10.1017/ Suh, S., B. Weidema, J. H. Schmidt, and R. Heijungs. 2010. General-
S1742170510000062. ized make and use framework for allocation in life cycle assess-
Minchin, W., F. Buckley, D. A. Kenny, F. J. Monahan, L. Shalloo, ment. J. Ind. Ecol. 14:335–353. https://doi.org/10.1111/j.1530-
and M. O’Donovan. 2009. Effect of grass silage and concentrate 9290.2010.00235.x.
based finishing strategies on cull dairy cow performance, carcass Thoma, G., J. Popp, D. Shonnard, D. Nutter, M. Matlock, R. Ulrich,
and meat quality characteristics. Meat Sci. 81:93–101. https://doi. W. Kellogg, D. S. Kim, Z. Neiderman, N. Kemper, F. Adom, and
org/10.1016/j.meatsci.2008.07.001. C. East. 2013. Regional analysis of greenhouse gas emissions from
Moe, P., and H. Tyrrell. 1979. Methane production in dairy cows. J. USA dairy farms: A cradle to farm-gate assessment of the Ameri-
Dairy Sci. 62:1583–1586. can dairy industry circa 2008. Int. Dairy J. 31(Suppl. 1):S29-S40.
Myhre, G., D. Shindell, F. M. Bréon, W. Collins, J. Fuglestvedt, J. https://doi.org/10.1016/j.idairyj.2012.09.010
Huang, D. Koch, J. Lamarque, D. Lee, B. Mendoza, T. Nakajima, Thomassen, M. A., R. Dalgaard, R. Heijungs, and I. Boer. 2008a.
A. Robock, G. Stephens, T. Takemura, and H. Zhan. 2013. 2013: Attributional and consequential LCA of milk production. Int. J.
Anthropogenic and natural radiative forcing. Pages 207–282 in Life Cycle Assess. 13:339–349. https://doi.org/10.1007/s11367-
Climate Change 2013: The Physical Science Basis. Contribution 008-0007-y.
of Working Group I to the Fifth Assessment Report of the Inter- Thomassen, M. A., K. J. van Calker, M. C. J. Smits, G. L. Iepema,
governmental Panel on Climate Change. T. Stocker, D. Qin, G. and I. de Boer. 2008b. Life cycle assessment of conventional and
Plattner, M. Tignor, S. Allen, J. Boschung, A. Nauels, Y. Xia, V. organic milk production in the Netherlands. Agric. Syst. 96:95–
Bex, and P. Midgle, ed. Cambridge University Press, Cambridge, 107. https://doi.org/10.1016/j.agsy.2007.06.001.
United Kingdom. USDA (US Department of Agriculture). 2011. Dairy heifer raiser,
Nennich, T. D., J. H. Harrison, L. M. VanWieringen, D. Meyer, A. 2011. Accessed Feb. 8, 2017. https://www.aphis.usda.gov/animal_
J. Heinrichs, W. P. Weiss, N. R. St-Pierre, R. L. Kincaid, D. L. health/nahms/dairy/downloads/dairyheifer11/HeiferRaiser.pdf.
Davidson, and E. Block. 2005. Prediction of manure and nutrient USDA (US Department of Agriculture). 2014. Dairy 2014: Dairy cattle
excretion from dairy cattle. J. Dairy Sci. 88:3721–3733. https:// management practices in the United States, 2014. Accessed Feb. 8,
doi.org/10.3168/jds.S0022-0302(05)73058-7. 2017. https://www.aphis.usda.gov/animal_health/nahms/dairy/
NRC (National Research Council). 2001. Nutrient Requirements of downloads/dairy14/Dairy14_dr_PartI.pdf.
Dairy Cattle. 7th rev. ed. Natl. Acad. Press, Washington, DC. USDA (US Department of Agriculture). 2016. Dairy: World markets
NREL (National Renewable Energy Laboratory). 2010. U.S. Life- and trade. Accessed Aug. 20, 2016. http://www.fas.usda.gov/
Cycle Inventory. National Renewable Energy Laboratory, Denver, data/dairy-world-markets-and-trade.
CO. Accessed Aug. 6, 2016. https://www.lcacommons.gov/nrel/ USDA-NASS (US Department of Agriculture National Agricultur-
search. al Statistics Service). 2013. Quick stats. Accessed Feb. 8, 2017.
O’Brien, D., J. L. Capper, P. C. Garnsworthy, C. Grainger, and L. https://quickstats.nass.usda.gov/.
Shalloo. 2014. A case study of the carbon footprint of milk from USDA-NRCS (US Department of Agriculture National Resources Con-
high-performing confinement and grass-based dairy farms. J. servation Service). 2009. Agricultural waste management field hand-
Dairy Sci. 97:1835–1851. https://doi.org/10.3168/jds.2013-7174. book. Accessed Aug. 20, 2016. http://www.nrcs.usda.gov/wps/
O’Brien, D., T. Hennessy, B. Moran, and L. Shalloo. 2015. Relating portal/nrcs/detailfull/national/water/?&cid=stelprdb1045935.
the carbon footprint of milk from Irish dairy farms to economic USEPA (US Environmental Protection Agency). 2008. Technology
performance. J. Dairy Sci. 98:7394–7407. https://doi.org/10.3168/ characterization: Reciprocating engines. Accessed Aug. 20,
jds.2014-9222. 2016. http://www.epa.gov/chp/documents/catalog_chptech_
Patzek, T. 2004. Thermodynamics of the corn-ethanol biofuel cycle. reciprocating_engines.pdf.
Crit. Rev. Plant Sci. 23:519–567. Wang, M., H. Huo, and S. Arora. 2011. Methods of dealing with co-
PE International. 2010. GaBi 4 Professional. PE International, Stutt- products of biofuels in life-cycle analysis and consequent results
gart, Germany. within the U.S. context. Energy Policy 39:5726–5736. https://doi.
Pimentel, D., and T. W. Patzek. 2005. Ethanol production using corn, org/10.1016/j.enpol.2010.03.052.
switchgrass, and wood; Biodiesel production using soybean and Wattiaux, M. A., and K. L. Karg. 2004. Protein level for alfalfa and
sunflower. Nat. Resour. Res. 14:65–76. https://doi.org/10.1007/ corn silage-based diets: II. Nitrogen balance and manure charac-
s11053-005-4679-8. teristics. J. Dairy Sci. 87:3492–3502. https://doi.org/10.3168/jds.
Powell, J. M., G. A. Broderick, and T. H. Misselbrook. 2008. Seasonal S0022-0302(04)73484-0.
diet affects ammonia emissions from tie-stall dairy barns. J. Dairy West, T., and G. Marland. 2002. A synthesis of carbon sequestration,
Sci. 91:857–869. https://doi.org/10.3168/jds.2007-0588. carbon emissions, and net carbon flux in agriculture: Comparing
Powell, J. M., M. Macleod, T. V. Vellinga, C. Opio, A. Falcucci, G. tillage practices in the United States. Agric. Ecosyst. Environ.
Tempio, H. Steinfeld, and P. Gerber. 2013. Feed–milk–manure ni- 91:217–232.
trogen relationships in global dairy production systems. Livest. Wheeler, E. F., M. A. Adviento-Borbe, P. A. Topper, N. E. Brown,
Sci. 152:261–272. https://doi.org/10.1016/j.livsci.2013.01.001. and G. Varga. 2008. Ammonia and greenhouse gas emissions
Pradhan, A., D. Shrestha, A. Mcaloon, W. Yee, M. Haas, J. A. Duff- from dairy freestall barn manure. Pages 1–15 in Proc. AS-
ield, and H. Shapouri. 2011. Energy life-cycle assessment of soy- ABE Ann. Int. Meet., Providence, RI. ASABE, St. Joseph, MI.
bean biodiesel revisited. Trans. ASABE 53:1031–1039. 10.13031/2013.24649.
Rotz, C. A., M. S. Corson, D. S. Chianese, S. D. Hafner, R. Jarvis, and Wilkerson, V. A., D. R. Mertens, and D. P. Casper. 1997. Prediction
C. U. Coiner. 2015. The integrated farm system model (IFSM): of excretion of manure and nitrogen by Holstein dairy cattle. J.
Reference manual, version 4.2. US Department of Agriculture Ag- Dairy Sci. 80:3193–3204.
ricultural Research Service, University Park, PA.
Rotz, C. A., F. Montes, and D. S. Chianese. 2010. The carbon foot-
print of dairy production systems through partial life cycle as-

Journal of Dairy Science Vol. 100 No. 8, 2017

 ENVIRONMENTAL IMPACTS OF GRAZING DAIRY 15

APPENDIX off-farm inputs for crops and biogas-based electricity

production (MJ); EO = energy generated in the form of
Equations biogas (MJ); EA = energy avoided due to displacement
of grid electricity by biogas-based electricity (MJ); and
The equation used to estimate fat- and protein-
FPCM = fat- and protein-corrected milk (kg). The
corrected milk (FPCM), used as the functional unit in
following equation was used to estimate an allocation
this study, is as follows:
ratio (based on meat and milk fat and protein) to as-
FPCM = sign the environmental impacts of the study between
milk and meat:
(0.029 × MF) × (0.0547 × MP) × (0.0395 × ML)
M× ,
0.7436 AMM = meat (kg)
[A1]
(0.0920 × meat fat) + (0.0547 × meat protein)
× cows ,
where FPCM = fat- and protein-corrected milk (kg), M (0.0920 × FPCM fat) + (0.0547 × FPCM protein)
= milk production (kg), MF = milk fat (%), MP = milk [A3]
protein (%), and ML = milk lactose (%; from NRC,
2001). The following equation was used to estimate net where AMM = adjusted meat mass based on its nutri-
energy intensity (NEI), which indicates the total net tional content compared with the nutritional content of
energy used in the farm per kilogram of FPCM: the produced milk (kg of boneless meat per culled cow),
cows = number of culled cows for human consumption,
E I − EO − E A meat fat = fat content in milk, meat protein = protein
NEI = , [A2]
FPCM content in meat, FPCM fat = fat content in FPCM,
and FPCM protein = protein content in FPCM.
where NEI = net energy intensity (MJ/kg of FPCM); EI
= energy inputs to the system, including on-farm and

Table A1. Inputs for supplemental feed and feed used (per ton of DM) in confined months

Input Unit Alfalfa silage Corn silage Corn grain Soybeans

1 1 2
Seed kg 0.9 1.7 1.78 15.73
Herbicides kg 0.424 NA5 0.324 0.354
Pesticides kg 0.11 0.31 0.282 NA
Lime kg 1626 29.76 7.706 74.96
Diesel L 171 191 7.622 19.74
Gasoline L NA NA 1.692 1.14
Liquefied petroleum gas L NA NA 2.542 0.264
Electricity kWh NA NA 4.522 0.324
Natural gas m3 NA NA 1.712 NA
Machinery kg 31 5.51 1.182 8.387
Nitrogen8 kg 0 7.159 16.19 0
Phosphorus8 kg 2.8410 2.2510 3.5110 6.6910
Potassium8 kg 24.910 9.8410 5.0910 22.310
1
Average for the United States (Rotz et al., 2010).
2
Average for Wisconsin (Kraatz et al., 2009).
3
Average for Wisconsin (Pradhan et al., 2011).
4
Average for Wisconsin (Kim and Dale, 2004).
5
NA = not applicable.
6
Calculated based on lime requirements to supply calcium removal in harvested material and neutralize acidi-
fication rate and on amounts of manure and nitrogen fertilizer applied.
7
Average for the United States (Pimentel and Patzek, 2005).
8
Total nutrient requirements supplied by both manure and purchased fertilizers.
9
Data sourced from USDA-NRCS (2009).
10
Data sourced from Laboski and Peters (2012).

Journal of Dairy Science Vol. 100 No. 8, 2017

16 AGUIRRE-VILLEGAS ET AL.

Table A2. Primary energy of inputs required for milking and crops production

Input Primary energy Reference

Housing and milking
Electricity 0.062 MJ/kg of milk ECW (2009)
Crop production
Herbicides 267 MJ/kg West and Marland (2002)
Agricultural lime 1.47 MJ/kg Kim and Dale (2004)
Diesel 46.1 MJ/L PE International (2010)
Gasoline 41.2 MJ/L PE International (2010)
Liquefied petroleum gas 28.8 MJ/L PE International (2010)
Oil 63.85 MJ/L PE International (2010)
Electricity 11.0 MJ/kWh NREL (2010)
Natural gas 42.3 MJ/m3 PE International (2010)
Machinery manufacture 45.9 MJ/kg Patzek (2004)

Journal of Dairy Science Vol. 100 No. 8, 2017

 Table A3. Description of the calculation method used to determine manure characteristics for each scenario

Manure characteristic Calculation method Reference

Manure production (kg/animal per day)
Lactating cows ME = (DMI × 2.63) + 9.4, [A4] Nennich et al. (2005)
where manure excretion (ME) and DMI are in kilograms.
Growing heifers ME = (DMI × 4.158) − BW × 0.0246, [A5] Nennich et al. (2005)
where manure excretion (ME), DMI, and BW are in kilograms.
Mature heifers and dry cows   CPin   NDFin   Wilkerson et al. (1997)
ME = (0.00711 × BW) + 32.4 ×  + 25.9 ×  + 8.05 , [A6]
     DMI  
 DMI   
where manure excretion (ME), BW, CP intake in diet (CPin), DMI, and NDF in diet (NDFin) are in
kilograms.
TS (% of excreted manure)
Lactating cows TS excreted = (DMI × 0.356) + 0.8, [A7] Nennich et al. (2005)
where TS excreted and DMI are in kilograms.
Growing heifers TS = 17% of excreted manure (from Table 5a) ASABE (2005)
Mature heifers and dry cows TS excreted = (DMI × 0.178) + 2.733, [A8] ASABE (2005)
where TS excreted and DMI are in kilograms.
Nitrogen (kg/animal per day)
Lactating cows Nmanure = N intake − N in milk Wattiaux and Karg
milk protein (2004)
Nmilk = + MUN + (milk yield × 0.203) , [A9]
6.38
where Nmilk and milk protein are in grams. MUN (g) = 0.0148 × N intake (g) + 2.16.
Growing heifers (DMI × dietary CP × 78.39) + 51.4 Nennich et al. (2005)
Nmanure = , [A10]
1, 000
where Nmanure and DMI are in kilograms and dietary CP is in grams per gram of DM.
Mature heifers and dry cows Same as N intake
Phosphorus (kg/animal per day)
Lactating cows Pmanure = P intake − P in milk NRC (2001)
0.9 × milk production
Pmilk = , [A11]
1, 000
where Pmilk and milk production are in kilograms.
ENVIRONMENTAL IMPACTS OF GRAZING DAIRY

Growing heifers DMI × dietary P × 622.03 Nennich et al. (2005)

Pmanure = , [A12]
1, 000
where Pmanure and DMI are in kilograms and dietary P is in grams per gram of DM.
Mature heifers and dry cows Same as P intake
Potassium (kg/animal per day)
Lactating cows Kmanure = K intake − K in milk NRC (2001)
1.5 × milk production
Kmilk = , [A13]
1, 000
where Kmilk and milk production are in kilograms.
Growing heifers (0.038 × BW) + (6.1 × DMI) NRC (2001)
Kmanure = , [A14]
1, 000
where Kmanure, BW, and DMI are in kilograms.
Mature heifers and dry cows Same as K intake

Journal of Dairy Science Vol. 100 No. 8, 2017

17
18 AGUIRRE-VILLEGAS ET AL.

Table A4. Description of emission factors from fossil sources

Source Emission factor1 Reference

Lime application 1.4 kg of CO2-eq/kg of CaCO3 PE International (2010)
Methane combustion 0.002 kg of CO2-eq/L CH4 + 2O2 → CO2 + 2H2O
Gasoline combustion 2.17 kg of CO2-eq/L NREL (2010)
Liquefied petroleum gas combustion 1.76 kg of CO2-eq/L NREL (2010)
Natural gas combustion 0.002 kg of CO2-eq/L NREL (2010)
Petro-diesel combustion 2.73 kg of CO2-eq/L NREL (2010)
N fertilizer production 1,968 g of CO2-eq/kg NREL (2010)
P fertilizer production 394 g of CO2-eq/kg NREL (2010)
K fertilizer production 533 g of CO2-eq/kg Kim and Dale (2004)
Herbicide production 17,242 g of CO2-eq/kg West and Marland (2002)
Pesticide production 18,534 g of CO2-eq/kg Kim and Dale (2004)
Lime production 42.1 g of CO2-eq/kg Kim and Dale (2004)
Diesel production 11.7 g of CO2-eq/MJ NREL (2010)
Gasoline production 11.6 g of CO2-eq/MJ NREL (2010)
Liquefied petroleum gas production 8.12 g of CO2-eq/MJ NREL (2010)
Electricity production 207 g of CO2-eq/MJ NREL (2010)
Natural gas production 7.01 g of CO2-eq/MJ NREL (2010)
Machinery manufacture 3,540 g of CO2-eq/kg Rotz et al. (2010)
1
CO2-eq = CO2 equivalent.

Journal of Dairy Science Vol. 100 No. 8, 2017