Access to this full-text is provided by Springer Nature.
Content available from The International Journal of Life Cycle Assessment
This content is subject to copyright. Terms and conditions apply.
CARBON FOOTPRINTING
Improving the accounting of field emissions in the carbon
footprint of agricultural products: a comparison of default IPCC
methods with readily available medium-effort modeling
approaches
Christiane Peter
1
&Angela Fiore
2
&Ulrike Hagemann
3
&Claas Nendel
1
&
Cristos Xiloyannis
2
Received: 29 May 2015 /Accepted: 4 February 2016 /Published online: 20 February 2016
#The Author(s) 2016. This article is published with open access at Springerlink.com
Abstract
Purpose The estimations of greenhouse gas (GHG) field
emissions from fertilization and soil carbon changes are chal-
lenges associated with calculating the carbon footprint (CFP)
of agricultural products. At the regional level, the IPCC
Guidelines for National Greenhouse Gas Inventories (2006a)
Tier 1 approach, based on default emission factors, insuffi-
ciently accounts for emission variability resulting from
pedo-climatic conditions or management practices.
However, Tier 2 and 3 approaches are usually considered
too complex to be practicable. In this paper, we discuss dif-
ferent readily available medium-effort methods to improve the
accuracy of GHG emission estimates.
Methods We present four case studies—two wheat crops in
Germany and two peach orchards in Italy—to test the perfor-
mance of Tier 1, 2, and 3 methodologies and compare the
estimated results with available field measurements. The
methodologies selected at Tier 2 and Tier 3 level are charac-
terized by simple implementation and data collection, for
which only a medium level of effort for stakeholders is re-
quired. The Tier 2 method consists of calculating direct and
indirect N
2
O, emissions from fertilization with a multivariate
empirical model which accounts for pedo-climatic and crop
management conditions. The Tier 3 method entails simulation
of soil carbon stock change using the Rothamsted carbon
model.
Results and discussion Relevant differences were found
among the tested methodologies: in all case studies,
the Tier 1 approach exceeded the Tier 2 estimations
for fertilizer-induced emissions (up to +50 %) and the
measurements. Using this higher Tier approach reduced
the estimated CFP calculation of annual crops by 4 and
21 % and that of the perennial crop by 7 %. Removals
related to positive soil carbon change calculated using
the Tier 1 approach also exceeded the Tier 3 calcula-
tions for the studied annual crops (up to +90 %) but
considerably underrated the Tier 3 estimations and mea-
surements for perennial crops (−75 %). In this case, the
impact of the selected Tier method on the final CFP
results was even more relevant: an increase of 194
and 88 % for the studied annual crops and a decrease
of 67 % for the perennial crop case study.
Conclusions The use of higher Tiers for the estimation of
land-based emissions is strongly recommended to improve
the accuracy of the CFP results. The suggested medium-
effort methods tested in this study represent a good compro-
mise between complexity reduction and accuracy improve-
ment and can be considered reliable for the assessment of
GHG mitigation potentials.
Keywords Carbon footprint .Crop management .Fertilizer
field emissions .GHG accounting .Regional variability .Soil
carbon change .Tier methodologies
Responsible editor: Ivan Muñoz
*Christiane Peter
christiane.peter@zalf.de
1
Leibniz Centre for Agricultural Landscape Research (ZALF) e.V.,
Institute of Landscape Systems Analysis, Eberswalder Straße 84,
15374 Müncheberg, Germany
2
Department of European and Mediterranean Cultures: Architecture,
Environment and Cultural Heritage, Università degli Studi della
Basilicata, Via San Rocco 3, 75100 Matera, Italy
3
Leibniz Centre for Agricultural Landscape Research (ZALF) e.V.,
Institute of Landscape Biogeochemistry, Eberswalder Straße 84,
15374 Müncheberg, Germany
Int J Life Cycle Assess (2016) 21:791–805
DOI 10.1007/s11367-016-1056-2
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1 Introduction
Agriculture accounts for one third of global greenhouse gas
(GHG) emissions (Lal and Kimble 1997). If demand for food
and biomasses continues to increase, annual GHG emissions
from agriculture may increase proportionally, along with the
vulnerability of agro-ecosystems to climate change (Xiong
and Khalil 2009). However, agriculture also has a significant
potential to reduce GHG emissions, as soils are the second
largest carbon (C) sink after oceans (Lal and Kimble 1997).
In general, there are three options for climate change mitiga-
tion in agriculture. First, GHG emissions can be reduced by
improving the management of C and nitrogen (N) flows in
agro-ecosystems. Second, increasing the level of temporary
C storage through improved agricultural management prac-
tices can increase soil C sequestration. Third, the replacement
of fossil fuels with renewable fuels such as residues from
agricultural crop production also is possible.
During the last decade, the interest of companies and
policy-makers in carbon footprint (CFP) as a supporting tool
to assess the impact of food and biomass production on global
warming processes and as a tool to design impact reduction
plans has grown steadily.
Based on the completeness principle stated in ISO 14067
(2013), the most recent international reference standard about
CFP, all GHG emissions and removals that provide a signifi-
cant contribution to the CFP of the analyzed product system
should be included in the study.
Although field emissions from fertilization and crop resi-
due management (CO
2
and N
2
O) can contribute considerably
to the GHG balance of food and bioenergy products, they are
often disregarded in CFP studies (Bessou et al. 2013). The
same applies to CO
2
fluxes occurring due to changes in soil
C stocks subsequent to crop management change (Brentrup
et al. 2000; Petersen et al. 2013), which, following the ISO
14067 (2013), should be included in CFP if not already cal-
culated as part of land use change.
For the accounting of field emissions at country level, the
IPCC guidelines for National GHG Inventories (IPCC 2006a)
provide, in the fourth volume dedicated to Agriculture,
Forestry and Other Land Use sector (AFOLU), three calcula-
tion pathways (Tiers) characterized by different degrees of
complexity: Tier 1 includes low-accuracy methodologies,
which can be applied by using the default emission factors
provided by the IPCC; Tier 2 methodologies require the use
of national emission factors reflecting local pedo-climatic
characteristics; finally, Tier 3 methodologies are based on
model simulations or in situ measurements. Tiers 2 and 3 are
referred to as the higher Tiers in the following text.
At present, the most common practice is using Tier 1 meth-
odologies for field emission calculation in life cycle assess-
ment (LCA) and carbon footprinting of food and energy crops.
However, Tier 1 methodologies are intended for use at large
spatial scales, and they can generate substantial errors in pre-
dictions at finer spatial scales. In fact, at regional and sub-
regional levels, Tier 1 methods are not always sufficiently
accurate to account for the spatial variability of GHG emis-
sions due to different soil, climate, and management practices.
Conversely, higher Tiers (Tiers 2 and 3) are usually considered
too complex and time-consuming to be practicable in the de-
velopment of LCA studies. Therefore, the IPCC guidelines
recommend using a higher Tier for the key emission catego-
ries and provide decision trees to support identification of the
most suitable Tier.
There is an urgent need for the application and validation of
appropriate higher tier methodologies at farm, project, or plan-
tation scales, to address local issues with bioenergy and food
sustainability and identify local mitigation potentials (Smith
et al. 2007; Smith et al. 2012).
The aim of the present paper is to provide to CFP practi-
tioners higher Tier methods Breadily available^and Beasy to
implement^(with a medium effort) to assess field emissions
from fertilization and from soil carbon change consequent to
crop management change for the inclusion into CFP
assessment studies of agricultural products. We selected
higher Tier methods which match these requirements,
choosing the Tier 2 method based on the Bouwman et al.
(2002a,b) approach for estimating field emissions from fer-
tilization and the Tier 3 method for soil organic carbon (SOC)
change assessment based on simulations with the Rothamsted
C (RothC) model (Coleman et al. 1997); both methods are
described, respectively, in Sects. 2.2 and 2.3. These methods
have been applied to four case studies and compared with Tier
1 results and with measurements, in order to test their perfor-
mance. The measurements were not intended to be used as
Tier 3 approach, because they are not always readily available
for LCA practitioners, but could be used to confirm the valid-
ity of the model for the examined agro-ecosystem conditions.
A further goal of the paper is to assess and compare the
influence of the variability of regional inventory data on CFP
results, depending on the adoption of Tier 1, Tier 2, or Tier 3
assessment methods. The examined case studies, two winter
wheat crops in Germany and two peach orchards in Italy, are
described in Sect. 2.1. The case studies were deliberately se-
lected to represent different soil characteristics, climate condi-
tions, and crop types (annual and perennial) in order to test
how the different methodologies handle this variability.
2 Materials and methods
2.1 Field experiments
Four field trials were selected based on crop type (annual
crops and perennial crops), soil, and climate conditions.
Characteristics of all sites are presented in Table 2.Whilewe
792 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
assumed that no land use change occurred in our case studies,
there were confirmed changes of the crop management
practices.
The two winterwheat crops were cultivated at two different
sites in Germany (sites 1 and 2) and for 2 years (2011–2012
and 2012–2013). They were sown after plowing at the end of
September and harvested the following summer (end of June).
Both cropping systems were rain fed. Site-specific rates of N
fertilization were calculated based on field-sampled soil min-
eral N content in springtime and crop-specific target values
from the official recommendation system, which reflects ex-
pected crop N uptake during the season, and split into 50 %
mineral (calcium ammonium nitrate applied in cultivation pe-
riod 2011–2012: 101 kg N ha
−1
at site 1 and 73 kg ha
−1
at site
2 and in period 2012–2013: 105 kg N ha
−1
at site 1 and
75 kg N ha
−1
at site 2) and 50 % organic N fertilizer (digestate
applied in cultivation period 2011–2012: 135 kg N ha
−1
at site
1 and 137 kg ha
−1
at site 2 and in period 2012–2013:
98 kg N ha
−1
at site 1 and 108 kg N ha
−1
at site 2) (Table 2).
At sites 1 and 2, straw management was changed in the first
year of the assessment period (2011). Before the change, all
straw was taken off the field to be used for energy production
or animal feeding. After the change, it was left on the field and
incorporated into the soil. Data on C inputs from straw and
applied digestate were direct field data, and C inputs from
roots (C
r
) and root exudates (C
e
) were calculated based on
yield (Y) data, harvest index (HI), and shoot to root ratio
(S:R), using the following equations (Farina et al. 2013):
Cr¼Y
HI⋅S=R
⋅0:45 ð1Þ
Ce¼0:09⋅Y
HI
⋅0:45 ð2Þ
At site 1, the mean yearly yield of the two considered crop
cycles (2011–2012, 2012–2013) was 6.8 t (dry matter grain)
and 3.9 t (dry matter straw), while at site 2, it was 8.0 t (dry
matter grain) and 5.4 t (dry matter straw). Tier 1 and Tier 2
estimates were compared with measured GHG emissions,
originating from a joint research project investigating GHG
emissions from energy crops fertilized with fermentation res-
idues. In situ measurements of N
2
OandNH
3
emissions were
conducted from sowing of winter wheat until the sowing of
the subsequent crop for the crop years 2011–2012 and 2012–
2013.
Periodic N
2
O measurements were conducted one to two
times per month out at three permanently installed soil collars
(0.75×0.75 m) at each site, with higher resolution following
fertilization events. Emissions of N
2
O were measured by tak-
ing four consecutive 100-ml gas samples from static non-
flow-through non-steady-stateopaquechambers(closure
time 60 min, vol. 0.296 m
3
; Livingston and Hutchinson
1995) and subsequently analyzed using a gas chromatograph.
N
2
O fluxes were calculated based on the ideal gas law using
the R package Bflux 0.2-2^(Jurasinski et al. 2012), using
linear regression analysis with stepwise backward elimination
of outliers. The calculated flux rates were then averaged for
the respective measurement day and linearly interpolated to
determine total N
2
O exchange.
Ammonia volatilization was measured for 2 to 5 days im-
mediately following fertilization using the open dynamic
chamber Dräger-tube method of Pacholski et al. (2006).
Four stainless steel chambers were placed on pre-installed
stainless steel rings (104 cm
2
). Chamber air was pumped
through the system with a constant air flow (1 L min
−1
), and
the actual NH
3
concentration in the chamber air was directly
determined in vol. ppm. Cumulative NH
3
losses were calcu-
lated by linear interpolation between measurements and in the
end were summarized.
The perennial crop field trials were conducted at two peach
orchards (sites 3a, 3b), located in Metapontino, the southern area
of the Basilicata region in Italy, which is devoted to fruit produc-
tion. They are characterized by the same rootstock and the same
training system of canopy, as well as similar pedo-climatic con-
ditions, orchard layout, and management regime. The orchard at
site 3a was planted in 2006 and the orchard at site 3b in 1996 and
removed after 15 years in 2011. In both orchards, a management
change occurred in the eighth year after plantation (2013 for site
3a, 2004 for site 3b) from conventional to sustainable manage-
ment. The conventional management regime consisted of soil
tillage (site 3b), chemical weed control (site 3a), and the burning
of pruning material. The sustainable management regime intro-
duced some innovative elements, including no tillage (site 3b),
mechanical weed control (grass cover mowed twice per year),
the chipping of pruning residues to be left in the field, and the
provision of 10 t of compost per hectare per year.
The life cycle of a fruit orchard can be divided into three
main stages: the young stage characterized by low yield and
grow of permanent structures, the mature stage characterized
by stable high yield, and the senescence stage characterized by
the decrease of yield. The farmers can decide in which stage
the orchard will be removed and replanted, and usually, it
happens at the end of the mature stage (Cerutti et al. 2010).
For peach trees, the young stage lasts 2 years and the mature
stage about 13 years. The amounts of C added to the soil
during the mature stage of the orchards (i.e., as crop residues
and organic fertilizers) were derived from direct field mea-
surements: figures regarding the dry matter of senescent
leaves, pruning residues, thinned fruit, and grass cover were
retrieved from field sampling performed at site 3b from 2004
to 2010 (from 8th to 14th year after establishment) and at site
3a in 2013–2014 (8th and 9th year after establishment), as-
suming a mean C content of 0.45 t C per ton of dry matter, and
a grass cover below-ground contribution of 20 % (Celano
et al. 2003). The C input during the young stage of the orchard
(senescent leaves and pruning material) was retrieved from an
Int J Life Cycle Assess (2016) 21:791–805 793
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
experiment performed in a peach orchard located in the same
area, with same rootstock, same training method, comparable
management regime, and adapting data to the different tree
density per hectare (Sofo et al. 2005). The C input from root
turnover was calculated as 30 % of the trees’above-ground
biomass turnover (senescent leaves, pruning material, and
fruit yield) for the first 3 years and as 25 % for all other years
(Buwalda 1993).
Data collected from site 3b were used to compare SOC
change estimate after crop management change, performed
with Tier 1 and Tier 3 methods with measurements of soil C
content, performed at the beginning of the experimental peri-
od (2004) and after 7 years of management shifted to sustain-
able practices (2010). Each time, 30 soil samples at 30-cm
depth were taken from a 1-ha field at different distances from
the tree row line. SOC was determined by the potassium-
dichromate oxidation procedure (Heanes 1984). Total per-
hectare C stocks in the topsoil (0–30 cm) were calculated as
the weighted average of SOC measured in the 30 soil samples.
For site 3a, the complete CFP was calculated using primary
data from the field logbook about the amount of productive
inputs used to perform all agricultural operations during the
orchard establishment, as well as the young and mature phase
of the orchard life cycle; the SOC change of site 3a was also
estimated using Tier 1 and Tier 3 methods.
2.2 Scope of the CFP assessment
GHG emissions of sites 1, 2, and 3a were calculated and
included in the CFP accounting according to ISO standards
14040 (2006), ISO 14044 (2006), and ISO 14067 (2013). Our
selected functional unit was not the unit of product, but the
surface unit of the cropland (1 hectare), since the focus of the
study was not on the environmental efficiency of the produc-
tion, but on the methodology comparison (Cerutti et al. 2015).
In fact, the results of Tier 1 methods are usually expressed as
emissions per unit of hectare, as their first application is
intended to analyze emissions from national crop production.
Therefore, the results of higher Tiers and the field measure-
ment were calculated per hectare size, in order to become
comparable with Tier 1 values. However, it is important to
highlight that for global issues such as food security and glob-
al warming impact of food production, GHG emission assess-
ment and mitigation potential per unit product are often more
useful than the absolute emissions per unit area (Bennetzen
et al. 2012). Since CFP calculations per unit of product can
take into account the variability related to yield differences,
they are particularly suitable for comparative CFP studies of
the same product cultivated in different locations or with dif-
ferent farming practices. System boundaries were fixed from
cradle to farm gate, starting with production of all productive
inputs, e.g., seeds, fertilizers, pesticides, agricultural machin-
ery, and fuels (indirect emissions), and ending with the harvest
of the crop, encompassing all emissions along the production
chain. Crop cultivation and processing of agricultural products
can lead to multiple outputs, e.g., straw and grain from cereal
harvesting or biogas and digestate from anaerobic biomass
digestion. In CFP calculations, there are different methods to
allocate the process emissions to different products (Benoist
et al. 2012). Manure and digestate are productive input for the
crop life cycle and residues (by-products) for the livestock and
bioenergylife cycle. They are re-used in the same form at field
(non-treated). As Rehl et al. (2012) stated, there are many
ways to allocate organic fertilizer but the most logical one is
the economic value. However, usually, manure and digestate
are not sold by the farmers; therefore, it is difficult to
determine prices because it does not exist. We followed the
approach by Rehl et al. (2012) and used the open-loop alloca-
tion procedure (ISO 14067 2013, Sect. 6.4.6.3 BAllocation
procedure for reuse and re cycling^)andtheeconomicindica-
tor with the market value and assumed that the by-products are
given away free of charge. Emissions from storage of organic
fertilizer and on field (application of fertilizer) were consid-
ered in the crop cultivation process where these emissions
occurring. The agri-footprint database (Blonk Agri-footprint
BV. 2014) also considers organic fertilizer (digestate and an-
imal manure) as residual products of biogas and animal pro-
duction system, so it does not include any emissions of the
biogas or animal production system, in order to avoid double
counting. Conversely, for compost, the production phase was
entirely included within the boundaries of the study, as it is not
reused in the same form (organic urban waste), but it is the
outcome of a recycling process, performed for agricultural
purpose only. For the peach case study 3a, the whole life cycle
of the orchard (site 3a) was included within the time bound-
aries, from orchard establishment until removal (15 years).
This approach is coherent with the most common practice of
LCA sectoral studies about fruit production from perennial
tree crops (Cerutti et al. 2010; Milà i Canals and Clemente
Polo 2003). The soil preparation prior to orchard establish-
ment, the establishment and removal phase, comprising the
production, and the disposal phases of materials constituting
the support structure (concrete and aluminum poles, steel
wire, and concrete blocks) were included in the boundaries,
extrapolating data about machinery operations for the removal
phase and end-of-life of trees’permanent structure from
farmers’experience.
For the winter wheat case studies, direct data from two crop
cycles were considered (2011–2012 and 2012–2013).
Inventory data on the amount of productive inputs (fertil-
izers, pesticides, fuels, machinery) were mostly retrieved di-
rectly from our on-farm experiments. GHG emissions related
to the production of these productive inputs were based on the
Ecoinvent database v.2.2 and v.3.0 (Ecoinvent 2013). The
calculation of CO
2
,CH
4
,andN
2
O emissions from diesel com-
bustion was based on IPCC Tier 1 (2006b).
794 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
The considered impact category was the global warming
potential (GWP)—100 years with the characterization factors
from the CML 2001 method corresponding with IPCC 2007
(Ecoinvent 2013).
CFP calculations were divided into three main parts: (i)
field emissions from fertilization, (ii) soil C stock change sub-
sequent to crop management change, and (iii) all other emis-
sions from agricultural operations. The consequences of meth-
odological choices were analyzed by comparing the CFP re-
sults based on Tier 2 and Tier 3 approach with the reference
CFP results based on IPCC Tier 1 default accounting methods.
2.3 Methodologies for the assessment
of fertilization-induced field emissions
The simple IPCC Tier 1 method (IPCC 2006a) for calculating
direct emission of nitrous oxide (N
2
O) from managed soils
simply takes into account 1 % (uncertainty range 0.3–3%)
of the anthropogenic N inputs (mineral fertilizer, organic
amendments, and crop residues) at the field level. Indirect
N
2
O emissions take place through two pathways. The first
pathway is volatilization of N as NH
3
and NO
x
and their
deposition onto soil and water, accounted by IPCC Tier 1
method as 10 % (0.3–3.0 % uncertainty) of kilogram N ap-
plied from mineral fertilizer and 20 % (0.5–5.0 % uncertainty)
from organic amendments expressed as kilogram NH
3
-N +
NO
x
-N. Only 1 % of these emissions from atmospheric depo-
sition of N volatilized from managed soils are accounted as
indirect N
2
O-N emissions. The second pathway of indirect
N
2
O emissions is leaching and runoff of N from fertilizer
application and crop residues, accounted by IPCC Tier 1
method only for regions where leaching and runoff occur as
30 % (10–80 % uncertainty) expressed as kilogram N. Only
0.75 % of these emissions leaching and runoff are accounted
as indirect N
2
O-N emissions. This Tier 1 approach completely
disregards any impact of crop type, fertilizer type, manage-
ment system, and local climate conditions on the GHG
emissions except for the calculation approach for leaching
and runoffs which takes the regional risk for leaching into
account. However, considering all or some of these
agricultural characteristics in the calculation of N
2
O, NO,
and NH
3
emissions would more accurately reflect the
heterogeneity of the environmental and management
conditions occurring in agriculture. This would better allow
the identification of local GHG emission hotspots and to
evaluate options of reduction.
We chose a Tier 2 level modeling approach used by
Bouwman et al. (2002b) to determine direct and indirect
N
2
O emissions and the approach of Bouwman et al. (2002a)
for NH
3
volatilization. A Tier 3 level modeling approach for
estimating field emissions from fertilization was not tested,
since to our knowledge, it does not exist at present any ap-
proach that matches our requirements to be readily available
and easily implementable by the user. Both tested Tier 2 ap-
proaches have been validated on a large global dataset from
measured agricultural field emissions encompassing 846 N
2
O
measurements from 126 studies, 99 NO measurements from
29 studies, and 1667 NH
3
measurements from 148 studies
(Bouwman et al. 2002a; Bouwman et al. 2002b). These
methods should therefore demonstrate better performance at
the local scale and under different agricultural management
systems than the Tier 1 methods, thus reducing the uncertainty
of the estimates with respect to the global emission factors
used in Tier 1 assessments (IPCC 2006a). However,
implementing the Tier 2 approach after Bouwman et al.
(2002a,b) requires more detailed data, as shown in Table 2.
The multivariate empirical model of Bouwman et al. (2002b)
classifies the parameters influencing N
2
O and NO emissions
into specific categories for each factor. For N
2
O, the signifi-
cant parameters are fertilizer type and application rate, crop
type, soil texture, SOC, soil drainage, soil pH, and climate
type, but only data regarding fertilizer type and application
rate, SOC, and soil drainage are needed to calculate the NO
emissions. The climate condition has less influence on the NO
emissions, since these emissions appear to be more concen-
trated during the crop-growing season than N
2
O emissions.
During the growing season, climate condition varies less be-
tween climate types than during other seasons (Bouwman
et al. 2001). The model for ammonia NH
3
emissions
(Bouwman et al. 2002a) is similar to the Bouwman et al.
(2002b) approach, but the significant parameters are fertilizer
type, fertilizer application rate and method, crop type, soil
texture, soil cation exchange capacity (CEC), soil pH, and
climate type. The amount of indirect emissions can be con-
verted to N
2
O-N by multiplying the NO-N and NH
3
-N emis-
sions with the default value 0.01 (based on IPCC 2006a). For
reporting purposes, the total N
2
O-N emissions can be convert-
ed to N
2
O by multiplying the kilogram of N
2
O-N by 44/28
(ratio of molecular weight of N and N
2
O). For the NH
3
emis-
sions induced by organic fertilizers (i.e., slurry and manure,
digestate, poultry manure), we used the more detailed model
approach by KTBL (2009) for organic fertilization to calculate
NH
3
-N emissions, with NH
3
volatilization depending on fer-
tilizer type, fertilizer application rate and method, daily tem-
perature, and a binary variable indicating whether the fertilizer
wasincorporatedwithin1h(Table1). CO
2
emissions from the
application of urea and liming were calculated based on Tier 1
IPCC (2006a)factors.
2.4 Methodologies for the assessment of emissions
from soil C stock change
The default international practice about GHG accounting in
the AFOLU sector (Tier 1) assumes that the soil carbon con-
tent is in equilibrium (steady state) when the last crop man-
agement change or land use change occurred more than
Int J Life Cycle Assess (2016) 21:791–805 795
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
20 years before the assessed time frame. In this case, it is
assumed that C outputs as CO
2
emissions from organic matter
decomposition equal the C inputs from organic material added
to soil. As soon as the landuse or management regime (tillage,
soil cover, carbon input level, irrigation) changes, C input and
outputs become imbalanced and either C emissions or C se-
questration will occur. It may take several decades before the
system returns to steady state at a new equilibrium, and the
default time set by Tier 1 method is 20 years. To include the
soil C change into CFP assessment, different methods are
available. We tested and compared methodological ap-
proaches from Tier 1 and Tier 3, since Tier 2 national emission
factors for SOC change are not always Bready available^for
LCA practitioners and other existing models for both C and N
cycle in soil do not match our requirement for Bmedium-effort
modeling approaches.^
The simple Tier 1 method, explained in Chapter 5 of
Volume 4 (cropland remaining cropland) of the IPCC guide-
lines (IPCC 2006a), requires the application of Eqs. 3and 4.
The soil organic C reference (SOC
ref
) under native vegetation
must be assigned based on six available soil types (high ac-
tivity clay, low activity clay, sandy, spodic, volcanic, wetland)
and nine climate regions (boreal, cold temperate dry, cold
temperate moist, warm temperate dry, warm temperate moist,
tropical dry, tropical moist, tropical wet, tropical montane).
Three relative stock change factors further describe the site:
(i) F
LU
is related to land use (long-term cultivated, paddy rice,
perennial/tree crop, set aside), (ii)F
MG
characterizes the tillage
regime (full, reduced, no tillage), and (iii) F
I
describes the
carbon input level (low, medium, high without manure, high
with manure). These factors come with individual error ranges
(between ±5 and ±50 %) and have to be defined for conditions
both before and after the change in management or land use
occurred.
Using Eq. 3, the soil organic carbon content before
(SOC
initial
) and after (SOC
final
) the change can be calculated
as follows:
SOCt¼SOCre f FLU FMG FIð3Þ
The difference between the final (new equilibrium,
SOC
final
) and the initial C stock (old equilibrium, SOC
initial
)
indicates the soil C stock change in the topsoil (0–30 cm) over
a period of 20 years, expressed as tons of C per hectare. This
amount can be converted to atmospheric CO
2
stored in or
emitted from the soil by multiplying the tons of C by 44/12
(ratio of molecular weight of CO
2
and C):
ΔSOC t C year‐1
¼SOCfinal‐SOCinitial
ðÞ=Tð4Þ
where
T¼default time period f or transition between equilibrium
SOC values;20 years
The selected Tier 3 methodology consists of a simulation of
the soil C turnover using the RothC model (26.3); Coleman
et al. 1997). To run the simulation, the model requires inputs
regarding the soil characteristics (i.e., clay content, considered
depth horizon, initial SOC), climate data (i.e., monthly aver-
age temperature, cumulative evapotranspiration, and rainfall),
and monthly soil C input (in tons of C per hectare), expressed
as net primary production (NPP). The output of the RothC
simulation of soil organic matter decomposition process is
the dynamic of C fluxes between soil carbon pools (resistant
and decomposable plant material, microbial biomass, and hu-
mified organic material) and the inferred CO
2
emissions in
atmosphere. The RothC simulations were run for different
time perspectives (T=20, 50, 100 years) and compared with
the default Tier 1 result (T= 20 years ). Equation 4was used as
well to include soil C change in the CFP case studies using a
Tier 3 method.
Before running the simulation, the initial value of the four
carbon pools (decomposable, resistant, biological, humus)
constituting the SOC in the RothC model was simulated using
the following procedure (initialization procedure):
The C stock of each C pool was set at 0 t C ha
−1
. The
simulation was then run to equilibrium using the C input esti-
mated for the management regime before management
change. The equilibrium values of the SOC pools were used
Table 1 Ammonia-volatilization rate (based on KTBL 2009)
Fertilizer type NH
3
volatilization in % of the applied NH
4
-N Application method-related reduction of NH
3
emission in %
5°C10°C15°C25°CDraghose
(uncovered)
Drag hose
(covered)
Drag shoe Manure
chisel plow
Incorporated
within 1 h
Cattle slurry 30 40 50 90 8 30 30 90 80
Digestate (thickened)
Pigslurry 1020257030 506090 80
Digestate (liquid)
Slurry 20
Mature manure 90
Poultry manure 90
796 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
as initial values for the simulation if the equilibrium obtained
for total organic C (TOC) stocks matched the measured initial
TOC; otherwise, the model was run in inverse mode to gen-
erate the required C input (CI
req
), using the equation suggested
by Mondini et al. (2012):
CIreq ¼CIi⋅Cmeas −IOM
Csim −IOM
ð5Þ
where CI
i
is the monthly C input used in the first equilibrium
run, C
meas
is the measured soil C stock to be matched, C
sim
is
the simulated soil C stock after the first equilibrium run, and
IOM is inert organic matter, the small soil organic C compart-
ment resistant to decomposition (with an equivalent radiocar-
bon age of more than 50,000 years), which, in absence of
radiocarbon data, can be roughly estimated from total SOC
using the equation provided by Falloon et al. (1998):
IOM ¼0:049⋅TOC1:139 ð6Þ
With the required C input (CI
req
), the model must be run again
to equilibrium to obtain the initial value of the SOC pools.
The RothC model can only simulate heterotrophic respiration
resulting from microbial decomposition of soil organic matter.
However, the autotrophic respiration of plants is implicitly in-
cluded as well, because the amounts of C inputs entered in the
model are expressed as net primary production (NPP) of bio-
mass, resulting from the balance between CO
2
absorbed through
photosynthesis and CO
2
released through dark respiration.
The crop management changes examined in our four
case studies concerned mainly the different amount of or-
ganic material returned or added to soil: In Table 3,the
carbon input stock change factors selected to evaluate
SOC change with Tier 1 method and the C input derived
from direct field data used to implement the RothC simu-
lation are summarized.
RothC can be considered as a reliable simulation tool of
carbon turnover in soil for arable land in cool or temperate
moist climates based on multiple validation campaigns
(Coleman et al. 1997; Falloon and Smith 2002;Ludwig
et al. 2007; Zimmermann et al. 2007), but there is a lack of
model application to perennial crops (Bessou et al. 2013).
Therefore, the results of RothC simulation were compared
with available measurements just at site 3b and not at sites 1
and 2, due also to the unavailability of medium-long-term
monitoring of SOC dynamic at these sites.
Tab le 2summarizes all data required for the implementa-
tion of the four tested methodologies.
3 Results and discussion
As shown in Fig. 1, in all case studies, new SOC equilibrium
after crop management change estimated using RothC model
(Tier 3) was reached in more than 20 years, the default time
period assumed in the Tier 1 approach. For site 1, the SOC
change at equilibrium is much lower if calculated using Tier 3
approach than using Tier 1. The Tier 3 value is outside the
range of uncertainty of the Tier 1 value (±51 %), resulting
from the propagation of errors declared for each stock change
factor in the IPCC (2006a) guidelines. For site 2, the equilib-
rium SOC change calculated using Tier 3 is also lower than
Tier 1 but falls within the forecasted Tier 1 error (±51 %).
The Tier 1 SOC change estimate is the same for the two
winter wheat case studies, as the same crop, soil (high activity
clay (HAC)), climate zone (cold temperate moist), and man-
agement practices were investigated, and thus, no difference
between the two sites is recognizable using the Tier 1 ap-
proach. In contrast, the curves resulting from the RothC sim-
ulation are very different, because site 1 is characterized by
lower soil clay content and a moister climate, which leads to a
slower rate of SOC rise due to the faster decomposition rate of
soil C pools. Moreover, the different amount of straw added to
soi1 as C input in the RothC simulations is higher at site 2 than
at site 1 in the considered time period 2011–2013, which leads
to different values of simulated SOC change. This difference
in C input is not appreciable using Tier 1 method because it is
only possible to choose between the qualitative C input levels
Blow,^Bmedium,^and Bhigh^(with or without manure), as
summarized in Table 3. Therefore, finer regional variation of
climate, yield, and soil texture cannot be represented using
Tier 1 methodology for SOC change estimates, in the case
of a crop management change.
For sites 3a and 3b, the Tier 3 SOC change at equilibrium is
much higher than the Tier 1 value and outside the forecasted
error range of Tier 1 (±172 %). In the Tier 3 simulation, the
succession of different peach orchard life cycles results in a
fluctuation of SOC within the orchard life cycle due to the
lower amount of crop residues during the establishment, the
young phase, the senescence phase, and the removal of the
orchard. The peak of SOC simulated at site 3a (Fig. 1)atthe
beginning of each orchard life cycle is due to the soil prepa-
ration with manure (60 t ha
−1
). Moreover, for both orchards,
the simulation reveals an overall increasing trend of SOC be-
yond the single orchard life cycle, if pursued with sustainable
management practices.
As can be noticed in Fig. 1, the comparison of SOC change
measurements with estimates at site 3b reveals that Tier 1
method underestimates the measured SOC change after 7 years
of sustainable management practices of 9.73 t C ha
−1
,while
RothC simulation better represents the SOC increasing trend,
with an overestimation of 6.6 t C ha
−1
. Tier 1 C input stock
change factors (Table 3) do not probably reflect the real
growth of soil C input after the incorporation of crop residues,
compost, and grass cover into the soil. RothC overestimation
could be due to the lack of consideration in the simulation of
soil condition variability across the orchard hectare: the
Int J Life Cycle Assess (2016) 21:791–805 797
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Tab le 2 Characteristics of experimental sites and comparison of data requirements by tested Tier methods
Site 1 Site 2 Site 3a–3b Data required to estimate
Field emissions from fertilization Soil carbon change
Country Germany (south) Germany (central) Italy (south) Tier 1 Tier 2 Tier 1 Tier 3
Geographical location 48° 59′N 51° 00′N 40° 14′N; 16° 42′E(site3a)
12° 39′E11°39′E 40° 23′N; 16° 42′E (site 3b)
Height above sea level (m) 431 247 16 (site 3a)
23 (site 3b)
Crop Crop type Wheat
j
Wheat
j
Peach
k
●●
Amount crop residues (t DM ha
−1
year
−1
) 3.9 5.4 6.8 Qualitatative ●
C content crop residues (% DM) 45 45 45 ●
Fertilization Fertilizer type CAN
a
, digestate
b
CAN
a
, digestate
b
ON
c
,AS
d
,NPK
e
,CN
f
, compost
g
●
Method of fertilizer application Broadcast, incorporated Broadcast, incorporated Broadcast ●
Fertilizer amount (kg ha
−1
)
h
2012: 375
a
, 2012: 270
a
, 2400
c
, 200
d
,●●
30,400
b
2013: 388
a
, 33,000
b
35,000
b
2013: 278
a
, 32,000
b
370
e
,370
f
, 80,000
g
N content of fertilizer (%) 2012: 27
a
,0.45
b
2013: 27
a
,0.3
b
2012: 27
a
,0.39
b
2013: 27
a
,0.34
b
12.5
c
,21
d
,20
e
,20
f
,0.62
g
C content of organic fertilizer (% FM) 2.5
b
2.5
b
40
c
,22.5
g
●
Soil Soil type (WRB classification) Stagnic Cambisol (HAC) Luvisol (HAC) Eutric Cambisol (HAC) ●●
Soil texture Loamy sand Silty clayey loam Sandy loam ●●
Soil bulk density (g/cm
3
) 1.7 1.5 1.5 (site 3a) 1.6 (site 3b) ●●
pH value 5.1 7.4 7.1 (site 3 a) 8.1 (site 3 b) ●
SOC before change 0.5 0.5 1.0 ●●
Soil drainage Good Good Good ●
Climate Climate region Temperate Temperate Temperate ●●
Precipitation (cumulative yearly in mm)
i
957 582 525 ●● Monthly value
Temperature (yearly in °C)
i
8.3 9.6 16.4 Specific days Monthly value
Evapotranspiration (cumulative
yearly in mm)
i
611 568 1262 ●● Monthly value
Management Tillage regime Converted to no tillage during
the experimental period
●●
a
CAN calcium ammonium nitrate
b
digestate fermentation of biomass from energy crops and animal slurry
c
ON organic nitrogen fertilizer
d
AS ammonium sulfate
e
NPK nitrogen (N) + phosphorus (P) + potassium (K) compound
f
CN calcium nitrate,
g
Compost with 40 % mineralization efficiency
h
Cumulative amount of fertilizers distributed within the time boundaries of the study
i
Average values over the experimental periods
j
Winter wheat (Triticum aestivum L.)
k
Peach (Prunus persica)
798 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
variable soil moisture with the distance from the tree line due
to drip irrigation and the concentrated distribution of compost
and roots along the tree line (Montanaro et al. 2012).
In Table 4, the yearly rates of SOC changes are reported for
differenttime horizons (20, 50, 100 years). Generally, it can be
stated that with a longer time horizon, the yearly rate of SOC
Fig. 1 Simulations of SOC
change subsequent to the shift to
sustainable crop management,
performed using IPCC (2006a)
Tier 1 approach and using RothC
26.3 model initialized with
measured SOC before
management change (the gray
line represents monthly
simulation and the black one
yearly average)
Int J Life Cycle Assess (2016) 21:791–805 799
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
change, expressed as CO
2
removed from atmosphere, de-
creases, since SOC change is always faster during the first
years after disturbance. This aspect has already been
highlighted in Petersen et al. (2013), where they suggested
using a 100-year time horizon when simulating SOC change
for CFP studies, based on a 100-year GWP calculation.
However, it is difficult to elaborate predictions in such a long
term, as many factors characterizing the agricultural sector
(e.g., land use, cropping systems, and management regimes)
are usually defined by highly volatile framework conditions
(e.g., consumer demand, economic trends, societal transfor-
mation, and public policy). For agricultural land use decision-
making, even 20-year continuous land use of the same kind is
not common and LUC should be considered at a more reason-
able time horizon. Furthermore, when changing the cultiva-
tion system each year, the effect of management change on the
SOC content is not stable and the uncertainty of the results is
very high. In order to more consistently compare Tier 1 and
Tier 3 methods, the yearly SOC change value derived from
20-year RothC simulation has been used and included into
CFP of the examined case studies (Fig. 2).
Figure 2shows the CFPs of sites 1, 2, and 3a and the
relative contribution of the three different GHG emission cat-
egories. The CO
2
removals from SOC change are reported
separately as prescribed by ISO 14067 (2013), because of
the temporary character of CO
2
storage in soil. For sites 1
and 2, the CFP represents the sum of all GHG emissions and
removals during a 1-year crop cycle, while for the site 3a CFP,
the entire 15-year peach orchard life cycle was taken into
account, with the last 8 years of management regime shifted
to sustainable practices. For CFP calculation of annual crops,
the whole crop rotation and the crop rotation-related effects
should be taken into consideration as stated by Brankatschk
and Finkbeiner (2015). Especially, crop residues can have a
great influence on the crop rotation effects, since crop residues
remaining on the field affect the subsequent crop through
influencing the physical, chemical, and biological soil proper-
ties and improving the soil fertility. The nutrients (N, P, K)
Table 3 Selected C input stock change factors selected to calculate SOC change with Tier 1 method and the C input derived from direct data used to
implement RothC simulation
Management practices Tier 1 C input stock change factor C input Tier 3 (t C ha
−1
a
−1
)
Site 1 Site 2 Site 3a Site 3b
Winter wheat
before change
Straw removed from
field + digestate addition
Medium
1.00
Representative for annual cropping with cereals
where all crop residues are returned to the
field. If residues are removed, then supplemental
organic matter (e.g., manure) is added. Also,
it requires mineral fertilization or N-fixing crop
in rotation.
2.22 2.24
Winter wheat
after change
Straw incorporated in
soil + manure addition
High with manure
1.44
(Error ±13 %)
Represents significantly higher C input over medium
C input cropping systems due to an additional
practice of regular addition of animal manure.
4.02 4.67
Peach orchard
before change
Pruning residues burned,
chemical weed control
(site 3a), and tillage
(site 3b)
Low
0.95
(Error ±13 %)
Low residue return occurs when there is due to
removal of residues (via collection or burning),
frequent bare fallowing, production of crops
yielding low residues (e.g., vegetables, tobacco,
cotton), no mineral fertilization, or N-fixing crops.
1.8 1.8
Peach orchard
after change
Pruning residues
incorporated in soil,
mechanical weed control,
no tillage, compost addition
High without manure
1.04
(Error ±13 %)
Represents significantly greater crop residue inputs
over medium C input cropping systems due to
additional practices, such as production of
high-residue-yielding crops, use of green
manures, cover crops, improved vegetated
fallows, irrigation, frequent use of perennial
grasses in annual crop rotations, but without
manure applied
5.8 6.8
800 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
remaining in crop residues on field can be used by the subse-
quent crop and can result in a reduced fertilizer dose for the
subsequent crop. This problem can be accounted in the CFP of
the subsequent crop in two ways: allocating the respective
environmental burdens to the subsequent crop or a credit can
be given for the current crop if a reduced fertilizer dose is
recommended for the subsequent crop. So far, there is no
agreement about a uniform approach on how to include the
effects of the crop rotation in the CFP calculation of an annual
crop (Brankatschk and Finkbeiner 2015). We tried to include
the crop rotation effect for the winter wheat crops by using the
real crop cultivation data provided by the researchers from the
experimental sites. The amount of farming operating material
for each crop applied at field was calculated in advance in
consideration of the local pedo-climatic conditions, the char-
acteristics of the previous crop (overall fertilization strategy
Table 4 Results of SOC change,
expressed as CO
2
removed from
atmosphere, calculated with
different time horizons and
different Tier approaches for the
experimental sites
Experimental site Assessment
approach
Time horizon (years) SOC change at
equilibrium (t C ha
−1
)
SOC change
(t CO
2
ha
−1
a
−1
)
1 Tier 3 RothC 1 −1.13
20 −0.33
50 −0.26
100 5.18 −0.19
Tier 1 20 28.8 (±14.7) −5.29
2 Tier 3 RothC 1 −7.80
20 −2.74
50 −1.87
100 38.15 −1.40
Tier 1 20 28.8 (±14.7) −5.29
3a Tier 3 RothC 20 2.99
50 2.63
100 55,2 2.03
Tier 1 20 7.4 (±12.7) −1.35
3b Tier 3 RothC 7 −9.91
20 −5.36
50 −2.94
100 43.87 −1.61
Tier 1 20 7.4 (±12.7) −1.35
Fig. 2 Comparison of CFP
calculated using the Tier 1 and
Tier 2/3 approach for the three
case studies (for sites 1 and 2,
mean values from two cultivation
years at each site, site 3 values are
totaled up over 15 years). SOC
change is calculated in a 20 years’
time horizon
Int J Life Cycle Assess (2016) 21:791–805 801
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
for the crop rotation), and the amount of nutrients available in
the soil (provided by the soil samples). Furthermore, the real
obtained yield and nutrient content at each site were used to
calculate the amount of SOC added to the field through crop
residues and digestate.
For site 1, the choice of either Tier 1 or Tier 3 for SOC
change estimation is a relevant decision, as SOC change cal-
culated with Tier 1 corresponds to 194 % of all other CFP
emissions (the agro-ecosystem is a C sink), while the Tier 3
estimate amounts to approximately 13 % of all other emis-
sions. For site 2, the Tier level selected for SOC change esti-
mation is less crucial, as in both cases, the agro-ecosystem is
considered to be a C sink, but the benefit calculated by Tier 3
is lower (145 against 221 % of all other CFP emissions offset
with Tier 1). In site 3a, the use of the Tier 3 method resulted in
a more realistic value of SOC change than Tier 1. During
8 years of sustainable management practices, 81 % of all other
CFP emissions can be offset through soil C storage, compared
to only 19 % estimated with Tier 1. Pursuing a sustainable
management regime, the Tier 3 simulation reveals that the
subsequent peach orchard life cycle would be a C sink, storing
152 % of the emissions from agricultural operations and field
emissions from fertilization in the soil.
For the annual crops, field emissions from fertilizer applica-
tion are an important factor, accounting for almost 50 % of the
emissions calculated using Tier 1 for sites 1 and 2. In contrast,
for the perennial peach crop, emissions from fertilizer applica-
tion contribute around 10 % (Fig. 2). In all case studies, the Tier
2 estimates of fertilizer-induced field emissions are lower than
the Tier 1 estimate (−9 % for wheat at site 1, −46 % for wheat at
site 2, and −65 % for peach at site 3) and within the uncertainty
range of ∼−70 to +325 % reported for the default global emis-
sion factor from Tier 1. The consequence of using a higher Tier
(Bouwman et al. 2002b) to calculate field GHG emissions from
fertilizer application is a CFP reduction of 4 % for wheat pro-
duction at site 1, 21 % for wheat production at site 2, and 7 %
for peach production at site 3a. As the yield from annual crops
is strictly dependent on nutrient availability and weather
conditions during a relatively short cultivation period, the
fertilizer management system is often more intensive for the
short annual crop cycle than for perennial crops. Consequently,
the fertilizer management system has a larger influence on
overall field emissions for annual crops than for perennial
crops, which feature considerably lower fertilizer input
throughout their entire life cycle.
As reported in the IPCC (2006a) guidelines, the global de-
fault values (Tier 1) are, in some cases, adequate to determine
fertilizer-induced field emissions—as confirmed by our wheat
case study (site 1). However, in most cases, these factors should
be specified based on environmental conditions (climate and
soil characteristics) as well as on crop management conditions
(fertilizer type, fertilizer application method, and rate) as in our
second wheat (site 2) and our peach (site 3) case study.
To evaluate these findings, wecompared the Tier 1 and Tier
2 estimates for N
2
OandNH
3
emissions from the two winter
wheat case studies with field-measured GHG emissions
(Fig. 3). For both sites and both gases, the Tier 1 and 2 calcu-
lations overestimated the measured fluxes. However, for NH
3
-
N emissions, Tier 2 estimates only deviate 1 and 10 % from
measured data of site 1 and site 2, respectively. In contrast,
Tier 1 calculated NH
3
-N emissions are 87 and 168 % higher
than measured emissions for site 1 and site 2, respectively.
Tier 2 estimates of fertilizer-induced N
2
O-N field emissions
are less accurate than estimates for NH
3
-N, but more accurate
than Tier 1 estimates. As Firestone and Davidson (1989)have
stated, the microbial processes of denitrification and nitrifica-
tion are the dominant sources of gaseous N emissions from
agricultural soil systems. Many factors associated with crop,
soil, water, climate, and fertilizer management can influence
soil turnover processes at all levels, e.g., organic matter de-
composition, denitrification, and nitrification. Consequently,
the heterogeneity of soil and weather conditions hamper a
sufficiently accurate representation of N
2
O, NO, and NH
3
emissions from a field using a model. As presumed, the Tier
2 estimates of fertilizer-induced N
2
O-N and NH
3
-N field
emissions were closer to the measurements than the Tier 1
estimations. Especially, the NH
3
-N field emission estimates
were very close to the measurements results. Most NH
3
-N
emissions on field arise from organic fertilizer application,
but the modeling approach from Bouwman et al. (2002a)does
not distinguish between different organic fertilizer types, but
the used KTBL (2009) calculation method for organic fertil-
izer takes the fertilizer type, the temperature during applica-
tion of the organic fertilizer, and the application method into
account. Therefore, through combining the two modeling ap-
proaches (for mineral fertilizer Bouwman et al. (2002a)and
for organic fertilizer KTBL (2009)), the accuracy of the
modeling results could be increased.
Comparing modeling data with measurements can be prob-
lematical since measured data and modeled data have also a
risk of uncertainty. As Bouwman et al. (2002c) pointed out in
their study, the amount of N
2
O and NO emissions is influ-
enced by the measurement technique, the length of measure-
ment period, and the frequency of measurements per day. For
N
2
O emissions are longer measurement periods (>300 days)
and intensive measurements (≥1 per day) better to detect the
fertilization effect on the N
2
O emissions. However, for NO
emission, no significant differences in measurement frequen-
cy classes could be found. The frequency of measurement and
continuity is a sensitive factor to detect the fertilization effect
on the N
2
O emissions. In our case, N
2
O measurements were
conducted one to two times per month with higher resolution
following fertilization events; therefore, they are not continu-
ous and include a risk of uncertainty. To reduce this uncertain-
ty, the N
2
O fluxes arising in the periods between the measure-
ments were calculated using linear regression analysis as
802 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
explained in the Sect. 2.1. Ammonia volatilization was measured
for 2 to 5 days immediately following fertilization. On both sites
(1 and 2), the experimental length exceeded the 300 days and,
correspondingly, our measurement period implying a low risk of
uncertainty. The considered time frame for data collection can
influence the N
2
O emission result. The emissions that occur
during mineralization of organic matter after harvest can be
charged to the harvested crop or to the subsequent crop. Both
modelingapproaches(Tiers1and2)arebasedonthesame
dataset of measurements, with a determined level of uncertainty
(Bouwman et al. 2002a,b). However, the Tier 1 modeling ap-
proach only considers the amount of N applied and is more
suitable for global- or national-scale calculation where the vari-
ability of environmental- and management-related factors is
not appreciable for the different regions (IPCC 2006a). Since
the Tier 2 modeling approach introduces more parameters to
account for the heterogeneity of local pedo-climatic and manage-
ment conditions, it is more suitable to represent field emissions at
farm or project scale.
Gabrielle et al. (2006) also compared different modeling
approaches for N
2
O emission calculation from winter wheat
on regional scale. We come to the same conclusion as
Gabrielle et al. (2006) that in the case of fertilizer-induced
field emissions, a higher Tier approach with a focus on the
aforementioned regulating factors, especially regional envi-
ronmental conditions, could therefore be used to adequately
detect mitigation potentials. As Fig. 3shows, our suggested
Tier 2 approach can be a good solution to estimate fertilizer-
induced field NH
3
-N emissions, as it takes these regional en-
vironmental and management conditions into account.
However, the results for fertilizer-induced field N
2
O-N emis-
sions modeling with the different Tier approaches were not
convincing; therefore, we recommend to continue testing at
more sites and with more crops to confirm our hypothesis.
4Conclusions
The choice of the methodological approach (Tier level) can
considerably affect the CFP of agricultural products.
Therefore, sufficient transparency is required to inform rele-
vant parties about possible error and shortcomings introduced
by the selected method when applied to a case study. In this
paper, we identified appropriate, readily available, assessment
methods at the Tier 2 and Tier 3 level with medium efforts for
stakeholders and explored the consequences of these method-
ological choices on the CFPs of annual and perennial crops for
field GHG emissions from crop cultivation.
Only few site-specific data are needed to apply these higher
Tier approaches, which can be used to improve the accuracy
of the estimate of land-based GHG emissions from fertiliza-
tion and soil C change, thus supporting the assessment of the
agricultural mitigation potential and the development of GHG
reduction plans at farm level.
The results for fertilizer-induced field emission calculation
were consistent among the three case studies: using the higher
Tier (Tier 2) led to lower estimated field emissions from fer-
tilization at two sites and to almost equal emissions at one site
compared to results obtained with the Tier 1 approach and to a
more reliable estimate in agreement with field measurements.
Based on our results, we suggest the following recommenda-
tions: For annual crops, a higher Tier approach is particularly
important when estimating fertilizer-induced field emissions,
whereas for perennial crops, it has a minor impact on the CFP.
However, we cannot draw general conclusions on the efficacy
of default emission factors for annual and perennial crops
from this limited amount of data; therefore, further studies
are needed to confirm our findings.
Regarding soil C stock change, important differences were
found between results calculated with Tier 1 and Tier 3 meth-
odologies. Using the Tier 1 approach can lead to wrong esti-
mates due to equivocal interpretation of the carbon input stock
change factor (qualitative description of the amount of organic
material entered to soil) and to the lack ofspecification of local
pedo-climatic conditions. Tier 3 RothC simulations can con-
stitute a valid alternative, as local primary data about climate,
soil features, and carbon input can be entered in the model;
RothC simulation of SOC change after the modification of a
crop management routine showed more reliable results when
tested against available measurements than Tier 1 estimates. A
more frequent SOC monitoring campaign would be useful to
further test the model’s performance and calibrate it to or-
chards’features. The present study has underlined the rele-
vance of SOC change from crop management change on
Fig. 3 Comparison of fertilizer-
induced N
2
O-N and NH
3
-N
emissions on field calculated
using Tier 1 and 2 approach and
measured data, mean values from
two wheat cultivation periods at
each site
Int J Life Cycle Assess (2016) 21:791–805 803
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
CFP of perennial crops, which cannot be always adequately
represented using a Tier 1 approach. Concerning annual crops,
crop rotations were not included in the RothC simulation, as
SOC change after land use change (crop change) was not
included in the scope of the assessment. However, the influ-
ence of SOC change on CFP of 1-year crop cycle could be
strongly related to the long-term SOC dynamic, subsequent to
crop choice and to the management regime, which determine
the amount of organic residues returned to soil. Thus, for
annual crops, a simulation approach is also advisable to eval-
uate SOC change as the default Tier 1 does not allow to rep-
resent the change of different crops in the rotation. Further
investigation efforts are needed in this direction. Similarly to
what was done in this paper for default carbon input level
stock change factors, it would be interesting to assess how
the default land use stock change factors (long term cultivat-
ed, paddy rice, perennial crops, set aside) of Tier 1 method
influence the performance of SOC change estimate with re-
spect to higher Tier approaches.
The outcomes of the present paper suggest that it is neces-
sary to foster more awareness and consensus within LCA
practitioners and policy-makers about the importance of in-
cluding regional field emissions into CFP of agricultural prod-
ucts, as it can considerably affect the results of the analysis.
Moreover, it is recommendable to use modeling approaches
for field emissions’estimate, taking into account local pedo-
climatic and crop management conditions, because this can
significantly improve the reliability of GHG accounting for
agriculture at farm level.
The higher-tiered methodologies for the calculation of
field emissions from fertilization and SOC change require
little additional effort compared with default Tier 1
methods, and thus, their practical application is advisable.
However, the development of user-friendly, crop-specific
tools underpinning these modeling approaches could more
efficiently increase the usefulness of CFP for agricultural
sustainability assessment at farm and regional landscape
level.
Acknowledgments The authors want to express their gratitude to the
participants of the following research projects who provided the observed
data used for the CFP calculations: (i) BDevelopment and comparison of
optimized cropping systems for the agricultural production of energy
crops^(FKZ 22013008), funded by the German Federal Ministry of
Food and Agriculture through the Agency of Renewable Resources
(FNR); (ii) BInnovation for Quality and Sustainability of fruit and vege-
table production (IQuaSoPO),^funded by Measure 124 of the Rural
Development Program 2007–2013 of Basilicata Region (Italy); and (iii)
the joint research project BPotentials for the mitigation of climate-relevant
greenhouse gas emissions from energy crop cultivation for biogas
production^(FKZ 22021008), funded by the German Federal Ministry
of Food and Agriculture through the Agency of Renewable Resources
(FNR) e.V. We would especially like to thank Achim Seidel, Gawan
Heintze, and Madlen Pohl for their contribution of data on field NH
3
and N
2
O emissions and Anne-Katrin Prescher for reviewing the
manuscript.
Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made.
References
Bennetzen EH, Smith P, Soussana J-F, Porter JR (2012) Identity-based
estimation of greenhouse gas emissions from crop production: case
study from Denmark. Eur J Agron 41:66–72
Benoist A, Dron D, Zoughaib A (2012) Origins of the debate on the life-
cycle greenhouse gas emissions and energy consumption of first-
generation biofuels—a sensitivity analysis approach. Biomass
Bioenerg 40:133–142
Bessou C, Basset-Mens C, Tran T, Benoist A (2013) LCA applied to
perennial cropping systems: a review focused on the farm stage.
Int J Life Cycle Assess 18(2):340–361
Blonk Agri-footprint BV. (2014) Agri-footprint—part 2—description of
data—version 1.0. Blonk Consultants Gouda, the Netherlands.
Bouwman AF, Boumans LJM, Batjes NH (2001) Global estimates of
gaseous emission of NH
3
, NO and N
2
O from agricultural land.
International Fertilizer Industry Association and Food and
Agriculture Organization of the United Nations, Rome
Bouwman AF, Boumans LJM, Batjes NH (2002a) Estimation of global
NH
3
volatilization loss from synthetic fertilizers and animal manure
applied to arable lands and grasslands. Global Biogeochem Cy. doi:
10.1029/2000GB001389
Bouwman AF, Boumans LJM, Batjes NH (2002b) Modeling global an-
nual N
2
O and NO emissions from fertilized fields. Global
Biogeochem Cy 16:1080
Bouwman AF, Boumans LJM, Batjes NH (2002c) Emissions of N
2
Oand
NO from fertilizes fields: summary of available measurements data.
Global Biogeochem Cy 16:1058
Brankatschk G, Finkbeiner M (2015) Modeling crop rotation in agricul-
tural LCAs—challenges and potential solutions. Agr Syst 138:66–
76
Brentrup F, Küsters J, Lammel J, Kuhlmann H (2000) Methods to esti-
mate on-field nitrogen emissionsfrom crop production as an input to
LCA studies in the agricultural sector. Int J Life Cycle Assess 5:
349–357
Buwalda JG (1993) The carbon costs of root systems of perennial fruit
crops. Environ Exp Bot 33:131–140
Celano G, Palese AM, Xiloyannis C, (2003) Gestione del suolo. In:
Fiorino P (Ed.), Olea-Trattato di Olivicoltura. Il Sole 24 ORE
Edagricole S.r.L.. Calderini, Bologna, Italy, p 349–363
Cerutti AK, Bagliani M, Beccaro GL, Bounous G (2010) Application of
ecological footprint analysis on nectarine production: methodologi-
cal issues and results from a case study in Italy. J Clean Prod 18:
771–776
Cerutti AK, Beccaro GL, Bosco S, De Luca AI, Falcone G, Fiore A,
Iofrida N, Lo Giudice A, Strano A (2015). Life cycle assessment
in fruit sector. In: Notarnicola B et al. (eds) Life cycle assessment in
the agri-food sector, Springer International Publishing Switzerland.
DOI 10.1007/978-3-319-11940-3_6.
Coleman K, Jenkinson DS, Crocker GJ, Grace PR, Klir J, Korschens M,
Poulton PR, Richter DD (1997) Simulating trends in soil organic
carbon in long-term experiments using RothC-26.3. Geoderma
81(1–2):29–44
Ecoinvent (Weidema BP, Bauer Ch, Hischier R, Mutel Ch, Nemecek T,
Reinhard J,Vadenbo CO, Wernet G) (2013) The ecoinvent database:
804 Int J Life Cycle Assess (2016) 21:791–805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
overview and methodology, data quality guideline for the ecoinvent
database version 3, www.ecoinvent.org.
Falloon P, Smith P (2002) Simulating SOC changes in long-term
experiments with RothC and CENTURY: model evaluation
for a regional scale application. Soil Use Manage 18(2):
101–111
Falloon P, Smith P, Coleman K, Marshall S (1998) Estimating the size of
the inert organic matter pool for use in the Rothamsted carbon mod-
el. Soil Biol Biochem 30:1207–1211
Farina R, Coleman K, Whitmore AP (2013) Modification of the RothC
model for simulations ofsoil organic C dynamics in dryland regions.
Geoderma 200:18–30
Firestone MK, Davidson EA (1989) Microbiological basis for NO and
N2O production and consumption in soils. In: Andreae M, Schimel
ODS (eds) Exchange of trace gases between terrestrial ecosystems
and the atmosphere. Wiley, New York, pp 7–21
Gabrielle B, Laville P, Duval O, Nicoullaud B, Germon JC, Hénault C
(2006) Process-based modeling of nitrous oxide emissions from
wheat-cropped soils at the subregional scale. Global Biogeochem
Cy 20. doi:10.1029/2006GB002686.
Heanes DL (1984) Determination of total organic-C in soils by an im-
proved chromic acid digestion and spectrophotometric procedure.
Commun Soil Sci Plant Anal 15:1191–1213
IPCC (Intergovernmental Panel on Climate Change) (2006a) In:
Egglestone HS, Buendia L, Miwa K, Ngara T, Tanabe K (eds)
Guidelines for National Greenhouse Gas Inventories. Volume 4:
agriculture, forestry and other land use. Prepared by the National
Greenhouse Gas Inventories Program. IGES, Japan.
IPCC (Intergovernmental Panel on Climate Change) (2006b) In:
Egglestone HS, Buendia L, Miwa K, Ngara T, Tanabe K (eds)
Guidelines for National Greenhouse Gas Inventories. Volume 2:
energy. Prepared by the National Greenhouse Gas Inventories
Program. IGES, Japan.
IPCC (Intergovernmental Panel on Climate Change) (2007) In: Solomon
S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M,
Miller HL (eds.) Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate
Change, 2007. Cambridge University Press, Cambridge, and New
York .
ISO 14040 (International Organization for Standardization) (2006)
Environmental management—life cycle assessment: principles and
framework. ISO 14040, Geneva.
ISO 14044 (International Organization for Standardization) (2006)
Environmental management—life cycle assessment: requirement
and guidelines. ISO 14044, Geneva.
ISO 14067 (International Organization for Standardization) (2013)
Greenhouse gases—carbon footprint of products: requirements and
guidelines for quantification and communication. ISO 14067,
Geneva.
Jurasinski G, Koebsch F, Hagemann U (2012) Flux rate calculation from
dynamic closed chamber measurements, R package version 0.2–1,
available at: http://CRAN.R-project.org/15 package=flux/.
KTBL (Kuratorium für Technik und Bauwesen in der Landwirtschaft)
(2009) Faustzahlen Biogas. Vol 2. KTBL, Darmstadt.
Lal R, Kimble JM (1997) Conservation tillage for carbon sequestration.
Nutr Cycl Agroecosys 49:243–253
Livingston GP, Hutchinson GL (1995) Enclosure-based measurement of
trace gas exchange: applications and sources of error. In: Matson PA,
Harris RC (eds) Methods in ecology. Biogenic Trace Gases,
Measuring emissions from soil and water. Blackwell Science
Malden, pp 14–51
Ludwig B, Schulz E, Rethemeyer J, Merbach I, Flessa H (2007)
Predictive modelling of C dynamics in the long-term fertilization
experiment at Bad Lauchstadt with the Rothamsted Carbon
Model. Eur J Soil Sci 58(5):1155–1163
Milà i Canals L, Clemente Polo G (2003) Life cycle assessment of fruit
production. In: Mattsson B, Sonesson U (eds) Environmentally-
friendly food processing. Woodhead Publishing Limited and CRC
Press LLC, Cambridge and Boca Raton, pp 29–53
Mondini C, Coleman K, Whitmore AP (2012) Spatially explicit model-
ling of changes in soil organic C in agricultural soils in Italy, 2001–
2100: potential for compost amendment. Agr Ecosyst Environ 153:
24–32
Montanaro G, Dichio B, Bati CB, Xiloyannis C (2012) Soil management
affects carbon dynamics and yield in a Mediterranean peach orchard.
Agri Ecosyst Environ 161:46–54
PacholskiA, Cai G, Nieder R, Richter J, Fan X, Zhu Z, Roelcke M (2006)
Calibration of a simple method for determining ammonia volatiliza-
tion in the field—comparative measurements in Henan Province,
China. Nutr Cycl Agroecosyst 74:259–273
Petersen BM, Knudsen MT, Hermansen JE, Halberg N (2013) An ap-
proach to include soil carbon changes in life cycle assessments. J
Clean Prod 52:217–224
Rehl T, Lansche J, Müller J (2012) Life cycle assessment of energy
generation from biogas—attributional vs. consequential approach.
Renew Sust Energy Rev 16:3766–3775
Smith P, Martino D, Cai Z, Gwary D, Janzen H, Kumar P, McCarl B,
Ogle S, O’Mara F, Rice C, Scholes B, Sirotenko O (2007) Chapter 8.
Agriculture. In: Metz B, Davidson OR, Bosch PR, Dave R, Meyer
LA (eds) Climate change 2007: mitigation of climate change.
Contribution of Working Group III to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change.
Cambridge University Press, Cambridge and New York
Smith P, Davies CA, Ogle S, Zanchi G, Bellarby J, Bird N, Boddey RM,
McNamara NP, Powlson D, Cowie A, van Noordwijk M, Davis SC,
Richter DB, Kryzanowski L, van Wijk MT, Stuart J, Kirton A, Eggar
D, Newton-Cross G, Adhya TK, Braimoh AK (2012) Towards an
integrated global framework to assess the impacts of land use and
management change on soil carbon: current capability and future
vision. Glob Change Biol 18(7):2089–2101
Sofo A, Nuzzo V, Palese AM, Xiloyannis C, Celano G, Zukowskyj P,
Dichio B (2005) Net CO
2
storage in Mediterranean olive and peach
orchards. Sci Hortic-Amsterdam 107:17–24
Xiong ZQ, Khalil MAK (2009) Greenhouse gases from crop fields.
Environ Sci Eng 113–132
Zimmermann M, Leifeld J, Schmidt MWI, Smith P, Fuhrer J (2007)
Measured soil organic matter fractions can be related to pools in
the RothC model. Eur J Soil Science 58(3):658–667
Int J Life Cycle Assess (2016) 21:791–805 805
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Content uploaded by Angela Fiore
Author content
All content in this area was uploaded by Angela Fiore on Feb 23, 2016
Content may be subject to copyright.
