This is the author’s pre-print version of the article

published in: Agricultural Systems 108 42-49,

http://dx.doi.org/10.1016/j.agsy.2012.01.004, available

online at:

http://www.sciencedirect.com/science/article/pii/S0308521

X12000121

Comparing energy balances, greenhouse gas

balances and biodiversity impacts of contrasting

farming systems with alternative land uses

H.L. Tuomistoa*, I.D. Hodgeb, P. Riordana & D.W. Macdonalda

aWildlife Conservation Research Unit, University of Oxford, The

Recanati-Kaplan Centre, Tubney House, Abingdon Road, Tubney,

Oxon OX13 5QL, UK,

bDepartment of Land Economy, University of Cambridge, Cambridge

CB3 9EP, UK

*Corresponding author: htuomist@gmail.com, Tel: +39 0332

786731, Fax: +39 0332 785601

Abstract

1

Life cycle assessment (LCA) is commonly used for

comparing environmental impacts of contrasting farming

systems. However, the interpretation of agricultural LCA

studies may be flawed when the alternative land use

options are not properly taken into account. This study

compared energy and greenhouse gas (GHG) balances and

biodiversity impacts of different farming systems by

using LCA accompanied by an assessment of alternative

land uses. Farm area and food product output were set

equal across all of the farm models, and any land

remaining available after the food crop production

requirement had been met was assumed to be used for other

purposes. Three different management options for that

land area were compared: Miscanthus energy crop production,

managed forest and natural forest. The results illustrate

the significance of taking into account the alternative

land use options and suggest that integrated farming

systems have potential to improve the energy and GHG

balances and biodiversity compared to both organic and

conventional systems. Sensitivity analysis shows that the

models are most sensitive for crop and biogas yields and

2

for the nitrous oxide emission factors. This paper

provides an approach that can be further developed for

identifying land management systems that optimize food

production and environmental benefits.

Keywords: greenhouse gas emissions, organic farming,

conventional farming, integrated farming, anaerobic

digestion, Miscanthus

1 Introduction

Many studies have compared the environmental impacts of

organic and conventional farming (Feber et al., 2007;

Mondelaers et al., 2009; Williams et al., 2010). They

show wide variation in the environmental impacts within

both organic and conventional systems. Arguably, the

greatest weakness of organic farming is its low yields,

primarily resulting from lower levels of inputs, and

higher abundances of pests and weeds (Köpke et al.,

2008). Thus, organic farming requires more land for

producing the same volume of output than conventional

farming. Therefore, it is important to identify the

3

specific practices that can provide environmental

benefits and develop integrated farming systems that

utilise those practices while maintaining relatively high

levels of output per unit area.

Life cycle assessment (LCA) is commonly used for

assessing environmental impacts of agricultural

production (Nemecek et al., 2011; Stone et al., 2012;

Thomassen et al., 2008; Williams et al., 2010). LCA uses

a “cradle-to-grave” approach in accounting simultaneously

for several environmental aspects of a product or service

(ISO 14040, 2006). The impacts are allocated with respect

to a unit of product termed the functional unit (FU).

Generally agricultural LCAs use system boundaries from

input production (e.g. fertilizers, pesticides and fuels)

up to the farm gate and the FU is a unit of the

agricultural product studied leaving the farm gate. Due

to the complexity and high land use impacts of

agricultural systems, agricultural LCAs face some

specific challenges compared to industrial LCAs.

4

In agricultural LCA studies, there is scope for

misinterpretations if the alternative land use options

are not taken into account. Thus, some studies suggest

that extensive farming systems are more environmentally

sound than intensive systems (Cederberg, 1998; Hole et

al., 2005). However, land is a limited resource for which

there are always alternative potential uses. By

definition extensive farming systems require more land to

produce a given amount of product than do intensive

systems. Extensive systems may have lower energy need per

product unit due to low input use, but if the alternative

land use options are taken into account, it may be found

that the overall energy balance of the intensive system

is more favourable (Berlin and Uhlin, 2004).

If only a fraction of the land used in an intensive

system is needed to produce the same product output, the

land saved can be utilized for other purposes, e.g.

bioenergy production. Therefore, the intensive system

might produce more energy than is needed for the

production process and that excess energy could be used,

5

for instance, to replace oil in heating, electricity

production or transportation fuels. After taking account

of the alternative land use options, the overall energy

efficiency of the intensive farming system becomes more

favorable. The opportunity cost principle in the context

of agricultural land use has been introduced by Berlin

and Uhlin (2004), who compared a system producing organic

milk with a system producing conventional milk and willow

for energy production.

The alternative land use options are also relevant when

biodiversity conservation strategies are assessed. Many

studies have shown that organic farms have higher level

of biodiversity compared to conventional farms (Bengtsson

et al., 2005; Hole et al., 2005). However, some attempts

have been made to answer the question concerning whether

extensive farming can provide higher benefits than

intensive farming with land sparing (Fischer et al.,

2008; Green et al., 2005; Phalan et al., 2011). Green et

al. (2005) created a model to assess the trade-offs

between wildlife friendly farming systems having lower

6

levels of crop productivity and land sparing that

minimizes the demand of farmland by increasing yields.

Due to lack of data, they were not able to assess which

of the systems is better for biodiversity in the

developed countries. However, they suggested that in

developing countries high-yield farming might lead to

higher levels of biodiversity. Hodgson et al. (2010) used

butterflies as an indicator of biodiversity and found

that conventional cereal farming with nature reserves

provides higher biodiversity benefits than organic

farming when the organic crop yield is below 87% of

conventional yield.

The aim of this study is to compare greenhouse gas (GHG)

balances, energy balances, and biodiversity impacts of

organic, conventional and integrated farming systems

taking account of the alternative land use options. The

impacts are compared both at the farming system and the

farming practice level.

7

As the aim of the study is to compare different farming

systems and practices, modelling and secondary data were

chosen to be the most suitable tools for the analysis.

Modelling enables the assessment of different types of

farming practice combinations and does not limit the

study only to existing systems. The use of secondary data

based on average values also excludes biases that may

occur when site-specific data are used.

The integrated farming systems were designed so as to

combine the best practices for reducing environmental

impact, including a versatile crop rotation, use of

organic fertilisers, use of over winter cover crops, use

of pesticides only when needed in order to avoid crop

failures, integration of biogas production and recycling

of nutrients. The models do not represent average

integrated farming systems, but rather are designed to

enable the comparison of the impacts of farming systems

consisting of combinations of different farming

practices. Therefore, this study does not provide

information about the sustainability of the existing

8

integrated farming systems, but aims at examining the

impacts of different farming practices and systems on

energy use, GHG emissions and biodiversity when

alternative land use options are taken into account.

2 Materials and methods

2.1 Goal, scope, functional unit and system boundaries

LCA with an assessment of alternative land use options

was used for comparing energy balances, GHG balances and

biodiversity loss of model farming systems under organic,

conventional and integrated management. The functional

unit (FU) was food crop output of 460 t potatoes (t =

tonne = 1000 kg), 88 t winter wheat, 60 t field beans and

66 t spring barley produced on a 100 ha (ha = hectare =

10,000 m2) farm. These crop outputs were determined by the

yield from a 20 ha of organic field available for each

crop under a standard organic rotation in lowland farming

in England. Higher yielding systems required less land

for producing the FU and the land area not needed for

production of the food crops for the FU and green manure

crops to sustain fertility was available for alternative

uses. Three alternative land uses for ‘the rest-of-the-9

land’ were included: cultivation of Miscanthus energy

grass, managed forest and natural forest. In some of the

systems biogas was produced from green manure, cover

crops and straw. The energy produced from Miscanthus, wood

and biogas was assumed to replace fossil fuels, and

therefore treated as negative energy input and GHG

emissions in the balance calculations.

The system boundaries included the production of farming

inputs (e.g. fuels, fertilizers and pesticides),

machinery, buildings and biogas production facility;

field operations and crop cooling and drying. Soil

nitrous oxide emissions were included in the study. The

soil carbon emissions and sequestration were not taken

into account, because net sequestration or emission only

occurs when the soil management type has been changed

until a new equilibrium level is reached. Energy inputs,

greenhouse gas emissions and biodiversity impacts were

calculated using Microsoft Excel spreadsheets.

10

2.2 Farming system models

The organic crop rotation was designed according to the

recommendations for an arable organic farm that does not

use external nitrogen inputs (Lampkin et al., 2008). The

model organic crop rotation was thus designed to be self-

sufficient in nitrogen, consisting of: 1. grass-clover

(GC); 2. potatoes (Solanium tuberosum); 3. winter wheat

(Triticum aestivum) + undersown overwinter cover crop (CC);

4. spring beans (Vicia faba) + CC; and 5. spring barley

(Hordeum vulgare) + undersown GC.

The model farming systems compared were:

1. Organic farm without biogas production (O). The GC, CC and

crop residues (CR) were incorporated into the soil.

Ploughing was used.

2. Organic farm with biogas production (OB). Otherwise similar

than O, but the GC, CC and CR (straw of wheat and

bean crops) were harvested for biogas production and

the digestate was spread to potatoes, winter wheat

and spring barley. Ploughing was used.

11

3. Conventional farm (C). Produced potatoes, winter wheat,

spring beans and spring barley using mineral

fertilizers and non-organic pesticides. The crop

rotation did not include GC or CC, and biogas was

not produced. Ploughing was used. Crop rotation

consisted of potatoes, winter wheat, spring beans

and spring barley.

4. Integrated farm (IF). The crop rotation and biogas

production were similar to the OB system, but non-

organic pesticides were applied. Ploughing was used.

5. Integrated Special (IFS). As IF but instead of GC municipal

biowaste was used as a fertilizer. Non-organic

pesticides and no-tillage were used. Crop rotation

consisted of potatoes, winter wheat, spring beans

and spring barley.

2.3 Nutrient supply

The O, OB and IF systems were designed to be self-

sufficient in nitrogen, whereas C and IFS systems used

external inputs. C system is the only one that uses

synthetic nitrogen fertilizers. IFS imports anaerobically

12

treated food waste from human communities in order to

close the nutrient cycle between fields and consumption.

The nutrient inputs in the systems are presented in Table

1. In order to ensure the nitrogen supply in the O, IF

and IFS systems, it was assumed that cover crops included

nitrogen fixing species. Additional phosphorus (P) and

potassium (K) were applied in all of the systems. P and K

inputs in IFS systems were assumed to be half of the

inputs in the C system as the other half were assumed to

be retrieved from the organic materials imported. In the

organic systems P and K fertilizers were supplied from

sources allowed under organic farming rules.

The fertilizer requirements for Miscanthus are low, because

the leaf mulch recycles nutrients and deep roots can

access minerals from deeper soil. The recommended amounts

of nutrients, 5 kg P/ha/yr and 30 kg K/ha/yr were assumed

to be used (Nix, 2009). These levels may not be

sufficient in the long term, but the recommended levels

were used in this study, reflecting common current

application levels used in practice.

13

14

Table 1. Nutrient inputs (kg ha-1 yr-1). Only new inputs into the system are listed here (not recycled nutrients).

O OB C IF IFSN INPUTSPotato seed 9.0 9.0 9.0 9.0 9.0Winter wheat seed 2.6 2.6 2.6 2.6 2.6Spring bean seed 15 15 15 15 15Spring barley seed 3.0 3.0 3.0 3.0 3.0Ley seed 1.1 1.1 1.1Cover crop seed 0.3 0.3 0.3 0.3Atmospheric deposition 35 35 35 35 35External N fertilizer1

Potatoes 201 139Winter wheat 219 158Spring beans 15 15Spring barley 123 161Nitrogen fixationSpring beans 200 200 200 200 200Ley 240 350 350Cover crops 42 42 42

P-K FERTILIZERS O OB/IF C

IFS Potatoes 9-129 14-156 17-195

8-80Winter wheat 10-41 13-47

18-36 9-18Spring beans 14-29 14-29

12-32 6-16Spring wheat 10-13 13-49

19-60 10-301 Synthetic nitrogen fertilizers were used in C system andimported anaerobically treated organic material was used in IFS system.

2.4 Yield data

15

The average organic and conventional yields in the UK

were chosen as reference values and the yields of the

other systems were adjusted based on those values. The

crop yields for the O and C models were calculated as an

average of the yields published in Organic Farm Incomes

in England and Wales 2001/02 – 2007/08 (Moakes and

Lampkin, 2003-2009) (Table 2).

In the OB model it was assumed that the yields increased

due to the enhanced nutrient management, by factors based

on experimental results from similar systems studied in

Germany (Stinner et al., 2008). It was assumed that in

IF systems the yields were increased from OB yields due

to the use of conventional pesticides, by factors

adjusted according to published field experiments

comparing yields with and without use of pesticides

(Cooper, 2008; Deike et al., 2008; Delin et al., 2008).

It was assumed that in IFS model, the yields were equal

to those in C models because of the higher nutrient

inputs and use of pesticides.

16

Table 2. Organic (O), conventional (C) and Integrated Special (IFS) crop yields (t wet weight ha-1, variation in the brackets) and multiplication factors used for calculating Organic Biogas (OB) and Integrated Farming (IF) yields (factor multiplied by the organic yield).

O1 OB2 IF2 C, IFS1

t/ha factor factort/ha

Grass-clover 46 (44 – 48) 1.00 1.00 -Potatoes 23 (14 – 29) 1.01 1.42 37 (34 – 40)Winter wheat 4.4 (3.1-5) 1.09 1.39

7.9 (7.2-8.8)Spring beans 3.0 (2-3.6) 1 1.15

4.0 (3.0 – 5.0)Spring barley 3.3 (1.9-4.0) 1.10 1.25

6.0 (5.4 – 6.8)1Moakes and Lampkin (2003-2009)2Conversion factors based on data explained in the text

2.5 Data for energy and GHG balance calculations

2.5.1 Crop production and balance calculations

The primary energy used for the field operations,

production of mineral fertilizers, pesticides and

machinery, and crop cooling and storage were based on the

data from Williams et al. (2006). The GHG emission factors

for machinery manufacturing and field diesel were based

on the data from Williams et al. (2006) (Table 3). The GHG

emissions were converted to carbon dioxide equivalents

(CO2-eq) using 100 year Global Warming Potential (GWP)

17

conversion factors (Table 3). The uncertainty ranges were

based on the coefficients of variation reported in the

literature: 40% for fuel use, 7% for manufacturing of

synthetic nitrogen fertilizers and 70% for soil N2O

emissions (Williams et al., 2006).

Table 3. Global warming potential (GWP) factorsSource

GWP emissions from machinery manufacturing 0.052 kg CO2-eq/MJ 1GWP emissions from field diesel 0.071 kg CO2-eq/MJ 1CO2-eq conversion factors for CO2, N2O and CH4 1, 298, 25

21Williams et al. (2006)2IPCC (2006)

The N2O emissions were calculated according to the IPCC

2006 guidelines (IPCC, 2006). The following sources of N2O

were taken into account: the use of inorganic

fertilizers, ploughing in crop residues, spreading biogas

digestate on land, indirect emissions from atmospheric

deposition of NOx and NH3 and indirect emissions from N

leaching.

In the balance calculations, the electricity generated by

biogas was assumed to replace the average UK electricity

18

mix. Production of this mix requires 3.2 MJ primary

energy per MJ electricity generated and emits 0.19 kg CO2-

eq/MJ (ELCD 2009). The heat produced from biogas, wood

and Miscanthus was assumed to replace light fuel oil. The

production of light fuel oil requires 1.2 MJ primary

energy per MJ and emits 0.09 kg CO2-eq/MJ (ELCD 2009).

2.5.2 Production of biogas

It was assumed that biogas was produced in a farm-scale

biogas plant using one-stage continuous digestion

technology operating at mesophilic temperatures. A wet

process was assumed and heat exchangers were used. Crop

biomass is commonly co-digested with livestock manure,

but digesters that use only crop biomass are also

possible (Bachmaier et al., 2010; Zielonka et al., 2010).

The construction and reconstruction of the biogas reactor

and storage facilities were estimated to account for 5%

of the electricity produced by the biogas plant and the

GHG emissions were estimated to account for 69 kg CO2-eq

per 1 GJ used for construction (Michel et al., 2010). The

energy input needed for heating the reactor was 240 MJ t-1

19

(uncertainty range 198-288 MJ t-1), and for pumping and

mixing 92 MJ t-1 of substrate (uncertainty range 73.6-

110.4 MJ t-1) (Börjesson and Berglund, 2006). It was

assumed that water was added into the reactor to reach a

10 % dry matter (DM) concentration of the substrate.

Therefore, 1 t of GC and CC with 23% DM concentration

produced 2.3 t of substrate for digestion and 1 t of

straw with 86% DM produced 8.6 t substrate for digestion.

The methane yields of GC and CC were assumed to be 10.6

GJ tDM-1 (8.5-12.7 GJ tDM-1) and for CR 7.1 GJ tDM-1 (5.7-

8.5 tDM-1) respectively (Berglund and Börjesson, 2006). It

was assumed that 30% of the energy was used for

electricity generation, 50% for heat and 20% was lost.

1.8% of the methane produced was lost in the atmosphere

(Michel et al., 2010).

The energy used for spreading and transportation of

biogas digestate was 25 MJ/t for liquid phase and 28 MJ/t

for solid phase (Berglund and Börjesson, 2006). It was

assumed that the digestate was transported 2 km by

tractor with an empty return. The transportation of

20

municipal biowaste by truck requires 1.6 MJ/t km with an

empty return transport. It was assumed that 35 t wet

weight municipal biowaste per hectare was applied in the

IFS model with an average transport distance of 20 km.

The transportation of municipal biowaste by truck

requires 1.6 MJ/t km with an empty return.

2.5.3 Managed forest, natural forest and Miscanthus

Two alternatives types of woodland management were

included: natural forest and a managed forest. In the

natural forest option it was assumed that the forest was

maintained in a natural state without any management.

When forest is planted on arable land it sequesters

carbon into the soil and vegetation during the first 100

years until it reaches an equilibrium state after which

there is no further net carbon sequestration (MacCarthy

et al., 2010). In the UK, broadleaved forest sequesters

2.8 t C/ha/yr into the vegetation, and approximately 0.62

t C/ha/yr into the soil when planted on arable land and

0.1 t C/ha/yr when planted on grassland (Dawson and

Smith, 2007). In this study, it was assumed that forest

was planted on grassland. Therefore, the annual carbon

21

mitigation of the natural forest was assumed at 10.63 t

CO2-eq/ha/yr (calculated by using C to CO2 conversion

factor 44/12) during the first 100 years after planting.

It was assumed that the managed forest option grows

conifer species. The average yields for conifer species

varies between 12 and 18 m3/ha/yr in the UK (Nix, 2009).

In this study it was assumed that 15 m3/ha/yr was

harvested. Based on data from Elsayed et al. (2003), it was

estimated that this volume of wood produced 0.57 t wood

chips (25% moisture content), 2.3 t composite board (10%

moisture content) and 0.39 t sawn timber. For the wood

chips, a heating value of 17.8 MJ/kg was used (Elsayed et

al., 2003) and the boiler was assumed to operate at 90%

efficiency. The life cycle energy use, 345 ± 36 MJ/t (75%

DM), and GHG emissions, 21 ± 2 kg CO2-eq/t (75% DM),

associated with harvesting the wood were based on the

data from Elsayed et al. (2003). It was assumed that the

wood chips replaced oil used for heating, and the

composite board and timber replaced steel. Production of

1 kg stainless steel requires 30.6 MJ primary energy and

22

emits 3.38 kg CO2-eq kg-1 (ELCD 2009). For the natural

forest option, it was assumed that the forest was planted

on grassland so that soil carbon stocks increase during

the first 100 years after planting by an average of 0.1 t

C/ha/yr (equals to 0.37 CO2-eq/ha/yr). In order to avoid

double counting, the carbon mitigation by the vegetation

above ground was not taken into account as the wood was

harvested and burned or used as materials which will

ultimately decompose.

It was assumed that Miscanthus was harvested for 15 year

after planting. The Miscanthus yield in the UK varies

between 12-15 oven dried tons (odt, 15% moisture

content)/ha, but during the first year it is not

harvested and in the second year the harvestable yield is

about 50% of the mature yield (Nix, 2009). Therefore the

average yield over the whole growing period varies

between 10.8-13.5 odt/ha, equivalent to 9.2-11.5

tDM/ha/yr. An average value of 10.4 tDM/ha was used in

the base calculations. The energy yield of Miscanthus was

calculated by using the lower heating value of 17.6 MJ

23

kgDM-1 (Hillier et al., 2009) and a 90% efficient boiler.

The energy input required for production of Miscanthus was

based on data from Gaunt and Lehmann (2008): crop

establishment including input production 4935 MJ/ha,

harvesting 3853 MJ/ha and transportation and processing

5078 MJ/ha. When the energy inputs in crop establishment

were divided over the whole growing period, the total

energy input amounted to 9260 MJ/ha/yr.

2.6 System variations

The impacts of some specific practices on GWP and energy

use were separately quantified. In this study, these

practices were assumed to be applied only on the food

crop and ley area if applicable. The impacts of the

practices were calculated both with and without taking

into account the alternative land use options. The

assessment excluding the alternative land use options

included food crop production, ley production and

production of biogas. When the alternative land use

options were taken into account it was assumed that the

area that was not needed for crop production was used for

24

Miscanthus production. The practices included in the

analysis are explained in Table 4.

Table 4. Specific practices whose impact on energy and

GHG emissions were separately analyzed.

Practice Explanation of calculation

Use of pesticides Difference between OB and

IF systems

Replacing synthetic nitrogen Difference between C

and

fertilizers with nitrogen fixing ley OB systems

Replacing synthetic nitrogen fertilizers Difference between the

energy use and

with imported food waste digestate GHG emissions related to

the fertilization in C and

IFS systems

Yield increase as a result of plant breeding Yield improvements in

C model based on Defra’s

25

estimates1 by 2025 and 2050

respectively: wheat and

barley 40% and 60%;

potatoes 10% and 20%; beans

30% and 60%. Input use and

other parameters remained

unchanged.

Use of nitrification inhibitors and It was assumed that

nitrification

slow-release fertilizers inhibitors and slow-release

fertilizers reduced N2O

emissions by 38% and 35%,

respectively2. Other

parameters remained

unchanged.

Reduced tillage and no-tillage Reduced tillage and no-

tillage practices used for

winter wheat, spring beans

26

and spring barley in the C

model.

Introducing biogas production into Difference between O

and OB

system systems.

1 Defra (2005)

2 Akiyama et al. (2010)

2.7 Biodiversity impacts

The method for the biodiversity impact assessment was

adapted from De Schryver et al. (2010). This method

assesses the ecosystem damage by using the potentially

disappeared fraction (PDF) of species as an indicator.

The indicator describes the change in vascular plant

species richness within the occupied area as compared

with the baseline. The baseline was assumed to be natural

forest, because that is the land type that would arise

without human distortion in the area concerned. The

method allows consideration of both local and regional

damage, but in this study only the local damage was taken

27

into account. The local damage score (DS), which is the

relative change in species richness for the occupied

area, can be calculated as follows:

DSi=CFi∗Ai∗ti (1)

where CF is the characterization factor of land use type i

(Table 5); and Ai the area occupied by land use type i; and

ti the time of occupation by land use type i (Table 6). The

CF is the PDF of species. De Schryver et al. (2010) used

data from the British Countryside Survey 2000 (Defra,

2000) for calculating the CFs for various land use types.

Three different approaches were used: the

individualistic, the egalitarian and the hierarchist

perspectives. In this study, the CFs for individualistic

perspective were used (Table 5), because De Schryver et al.

(2010) only provided the CFs for arable land uses in this

approach. Table 6 shows the area used and time of

occupation for each crop. In the C system, where cover

crops were not used, the score for the previous crop was

used as long as the next crop was cultivated.

28

29

Table 5. Characterization factor (CF in the Equation 1) of land use (De Schryver et al., 2010)

Median 95% confidence levelConventional arable land 0.79 0.73-0.83Integrated arable land 0.44 0.31-0.54Organic arable land 0.36 0.15-0.51Intensive fertile grassland 0.65 0.56-0.72Less intensive fertile grassland 0.36 0.14-0.52Organic fertile grassland -0.01 -0.18-0.15Intensive woodland 0.55 0.44-0.65Baseline: natural forest 0.00

Table 6. Area requirements (ha) and cropping time for each crop (mo=number of months)

O OB C IF IFSmo ha mo ha mo ha mo ha mo ha

Crass-clover 17 20 17 20 0 0 17 20 00

Potatoes 7 20 7 20 7 12 7 14 7 12Winter wheat 12 20 1 18 21 11 12 14 12

11Spring beans 7 20 7 20 12 15 7 17 7

15Spring barley 7 20 7 18 12 11 7 16 7

11Cover crops 5 40 5 38 0 0 5 32 5

37

The grass-clover leys in the organic and IF farming

systems were regarded as intensive grassland, because of

their repeated cutting. Cover crops were regarded as

organic fertile grassland and Miscanthus energy grass as

intensive fertile grassland. Some evidence suggests that 30

Miscanthus has higher biodiversity than cereal fields

(Bellamy et al., 2009).

2.8 Sensitivity and uncertainty analyses

In the sensitivity analysis the impacts of different

factors on the results were assessed by changing the base

values in the primary data within the uncertainty ranges

reported earlier in this paper and in Appendix A, Tables

A.1 and A.2. Monte Carlo analysis was used for an

uncertainty analysis. The models were simulated over

50,000 replications with randomly generated input values.

Microsoft Excel 2007 software was used for random number

generation from a uniform distribution within the

estimated uncertainty ranges of the input values. SPSS

14.0 software was used for the statistical analyses.

Wilcoxon Signed Rank tests were used for calculating the

significance of the differences between the systems.

3 Results3.1 Land use

In the O model the entire land area (100 ha) was needed

for production of the food crops and green manure,

31

whereas, in the OB model, 3.7 ha was available for the

energy crop due to the higher food crop yields (Table 7).

The C and IFS models required only 50% of the land area

for production of food crops.

Table 7. Land use for different purposes (ha)Food crops Green manure Rest-of-the-land1

O 80.0 20.0 0.0OB 76.3 20.0 3.7C 49.6 0.0 50.4IF 61.9 20.0 21.0IFS 49.6 0.0 50.41Miscanthus, managed forest or natural forest

32

3.2 Energy and GHG balances

Figure 1 shows the impacts of whole farming systems under

different rest-of-the-land use options. The O and

C_Natural forest systems were net energy users, whereas

all other systems were net energy producers, and

mitigated more GHG emissions than were emitted. The

addition of biogas production into the organic system was

enough to convert the system from a net energy user to a

net energy producer with net GHG mitigation. IFS_Miscanthus

had the highest net energy production and GHG mitigation.

Both energy and GHG balances were significantly different

(P < 0.001) between all of the systems (Appendix B, Table

B.1). The main sources of uncertainty for the

uncertainties of the energy balances were crop yields,

whereas for GHG balances crop yields and nitrous oxide

emissions were the main contributors to uncertainty

(Table 8).

When only the inputs in crop production were compared, it

was found that IFS system had 14% lower and C system 29%

higher energy inputs compared to O system (Table 9). The

33

GHG emissions from crop production were 32% lower from

the IFS system and 26% lower from the C system compared

to the O system. The OB and IF systems had 10% and 19%

higher energy inputs and 21% and 9% higher GHG emissions

compared to the O system, respectively. Use of machinery

and crop cooling and drying, especially potato storage,

made the highest contribution to the total energy input

in crop production (Table 9). In the C system the

production of fertilizers accounted for 30% of the total

energy input in crop production. The highest source of

GHG emissions was soil N2O emissions in all of the

systems.

O

OB_Managed forest

C_Miscanthus

C_Natural forest

IF_Managed forest

IFS_Miscanthus

IFS_Natural forest

-15 -10 -5 0 5

CropBiogasRest of the land

Energy inputs and outputs (TJ/FU)

A

34

OOB_Miscanthus

OB_Managed forestOB_Natural forest

C_MiscanthusC_Managed forestC_Natural forest

IF_MiscanthusIF_Managed forestIF_Natural forest

IFS_MiscanthusIFS_Managed forestIFS_Natural forest

-14 -12 -10 -8 -6 -4 -2 0 2 4Energy balance (TJ/FU)

B

O

OB_Managed forest

C_Miscanthus

C_Natural forest

IF_Managed forest

IFS_Miscanthus

IFS_Natural forest

-1000 -500 0 500

Crop

Biogas

Rest of the land

Greenhouse gas emissions (t CO2-eq/FU)

C

O

OB_Managed forest

C_Miscanthus

C_Natural forest

IF_Managed forest

IFS_Miscanthus

IFS_Natural forest

-1000 -800 -600 -400 -200 0 200Greenhouse gas balance (t CO2-eq/FU)

D

Figure 1. Comparison of energy inputs (positive value) and outputs (negative value) (A), median energy balances (inputs-outputs) (B), greenhouse gas emissions (positive value) and mitigation (negative value) (C) and median greenhouse gas balances (emissions – mitigation) (D) of the studied farming systems with alternative uses for ‘the rest-of-the-land’ category per functional unit (FU, farm). The error bars represent the 25 and 75 percentiles based on the Monte Carlo analysis.

35

36

Table 8. Change in whole farm net energy production (input-output) and greenhouse gas balance (emissions – mitigation) afterincreasing or decreasing the input parameters of the model systems (Miscanthus option used as an alternative land use for each system).Change O OB C IF IFS

TJ t CO2-eq TJ t CO2-eq TJ t CO2-eq TJ t CO2-eq TJ t CO2-eqCrop yields +23% -3.49 -259 -3.05 -158 -2.29 -113 -2.5 -130 -1.98 -102 Field diesel use +40% 0.31 22 0.34 25 0.29 21 0.36 26 0.26

18Biogas yield +20% -0.77 -55 -0.72 -52 -0.28 -20Miscanthus yield +12% -0.10 -4.4 -1.34 -60 -0.48 -22 -1.34 -60Soil N2O +70% 101 118 62 97 56

Table 9. Contribution of different production stages to the total energy use and GHG emissions of crop production (as % of the total energy use or GHG emissions of crop production in each system).

O OB C IF IFSEnergy GHG Energy GHG Energy GHG Energy GHG

Energy GHGMachinery use1 56 18 54 16 30 13 43 16 31 13Crop cooling and drying 35 12 34 10 31 14 37 13 46 21Fertilizer manufacture 9 2 12 3 30 25 8 3 8 3

37

Pesticide manufacture 1 1 1 1 9 5 12 7 15 7Nitrous oxide emissions 67 70 43 61 561 Include fuels use and machinery manufacture

38

The results of the system variations showed that when the

alternative land use options were not taken into account

the replacement of synthetic nitrogen fertilizers with

clover ley improved both net energy production and net

GHG mitigation (Table 10). However, when the alternative

land use options were taken into account the system using

synthetic nitrogen fertilizers mitigated more GHGs than

the one with clover ley. Replacement of synthetic

nitrogen fertilizers with food waste digestate improved

the net energy production and GHG mitigation both with

and without taking into account the alternative land use

options. Also yield improvement scenarios showed benefits

in both of the approaches. Integration of biogas

production into the O system had the highest benefits in

terms of net energy production and GHG mitigation

compared to the other options. Changes in tillage methods

and reduction of N2O emissions by using nitrification

inhibitors or slow release fertilizers had the same

impacts both with and without taking into account the

alternative land use options, because those methods did

39

not have an impact on the area of land required for crop

production.

40

Table 10. The effect of different farming practices on energy balance (inputs – outputs) and greenhouse gas balance (GHG, emissions-mitigation) of the system with and without taking into account the alternative land use (‘the rest-of-the-land’).

Without the rest-of-the-land With the rest-of-the-landEnergy GHG Energy GHGTJ/Farm t CO2-eq/Farm TJ/Farm t CO2-eq/Farm

Direct drilling instead of ploughing -0.11 -7.09 -0.11 -7.09Reduced tillage instead of ploughing -0.02 -1.32 -0.02 -1.32Use of pesticides (IF compared to OB) 0.66 19.69 -2.10 -220.02Yield improvements by 2025 -0.22 -15.26 -2.31 -196.14Yield improvements by 2050 -0.34 -22.54 -3.39 -287.61Clover ley instead of synthetic N -7.23 -316.42 -1.06 218.60Food waste instead of synthetic N -0.31 -22.01 -3.19 -154.46N2O emissions -38% -18.77 -18.77Biogas production (OB compared to O) -7.57 -355.96 -8.27 -416.72

41

3.3 Biodiversity impacts

IF_Natural forest and IFS_Natural forest were the only

two systems that had a lower biodiversity loss index than

the organic systems (Figure 2); the IFS_Natural forest

had a 51% lower biodiversity loss index compared to the O

system. C_Miscanthus had the highest biodiversity loss

score, being a factor of 2.4 times higher than the score

for the O system and a factor of 4.9 times higher than

that for the IFS_Natural forest system. The C_Natural

forest system had a score 1.29 times higher than the O

system and 2.6 times higher than the IFS_Natural forest

system. The uncertainty analysis showed that the

differences between all of the systems were significant

(P < 0.001, Appendix B, Table B.2).

42

O

OB_Managed forest

C_Miscanthus

C_Natural forest

IF_Managed forest

IFS_Miscanthus

IFS_Natural forest

-10 0 10 20 30 40 50 60 70 80

Green manure

Rest of the land

Crop

Biodiversity loss index

Figure 2. Results of the whole farm biodiversity loss index. The error bars

represent the 25 and 75 percentiles based on the Monte Carlo analysis.

4 Discussion

The choice of the land use for the rest of the land

category had a substantial impact on the results. The

potential options for this land area are endless and more

options could be assessed. Furthermore, the assumption

made about the use of the biomass in the managed forest

and Miscanthus options also has a notable impact on the

43

results. Here it was assumed that the biomass was simply

used for heat production. However, the demand for heat

may not always be high enough for it to be fully utilized

on farm whenever it can be generated and other options

for the use of the biomass would have to be considered,

which could reduce its utilizable energy output. For

example, if biomass was converted to electricity, the net

energy output of the biomass would be reduced, as more

energy is required for the conversion than if the biomass

was simply burned.

Miscanthus or wood biomass can also be used for biochar

production. In that case the energy production would be

reduced, but long-term storage of carbon may be possible

in the form of biochar, and therefore, more GHG could be

mitigated (Gaunt and Lehmann, 2008). Furthermore, as a

result of changing the land management by converting

arable land into Miscanthus production could achieve a net

accumulation of soil carbon until a new soil carbon

equilibrium level is reached (Styles and Jones, 2008).

However, it has been shown that when Miscanthus is planted

44

on grassland the change in soil carbon levels is low

(Hillier et al., 2009)

The GHG balances of the selected land use types were very

similar, but energy balances and biodiversity impacts had

a wide variation. For forests, the carbon sequestration

depends on whether the forest is already in existence or

whether it is planted on arable or grassland. Forests

sequester carbon only during the first 100 years after

planting and after that carbon mitigation is only

achieved when wood products substitute other products

(MacCarthy et al., 2010). However, it is likely that over

100 years more alternatives for fossil fuels will be

developed, and therefore, the substitution mechanism may

no longer be effective. At that point other aspects of

land use, such carbon storage and biodiversity impacts,

will become more relevant.

In a future scenario where fossil fuels are not used and

the supply of alternative energy sources is abundant, it

may be argued that there would be no reason to avoid

45

using synthetic nitrogen fertilizers. However, aspects

other than energy and GHG balances would still support

the recycling of nutrients. First of all, if nitrogen is

not recycled it is very likely that it will end up in

waterways causing eutrophication. Secondly, the recycling

of biomass may become critical given the diminishing

levels of phosphorus reserves (Franz, 2008).

The results of biodiversity impacts analysis in this

study contradicts the findings of Hodgson et al. (2010) who

concluded that conventional cereal farming with nature

reserves provides higher biodiversity benefits than

organic farming when the organic crop yield is below 87%

of conventional yield. This may be explained by the

different biodiversity indicators used. Hodgson et al.

(2010) focused on butterflies whereas vascular plants

were used as the indicator in this study. Another

explanation could be that natural forest that was taken

as a baseline here, does not necessarily provide the

highest biodiversity benefits. Some other type of nature

reserves or natural meadows may have a higher level of

46

biodiversity than natural forests. However, the aim is

not to maximize biodiversity, but to have an appropriate

level and type of biodiversity. Forests support different

species than meadows.

The main weakness of the biodiversity impact assessment

used in this paper is that the species richness of

vascular plants does not always correlate closely with

other taxonomic groups (Weibull et al., 2003), and the

method does not make a distinction between scarce/normal

and desired/non-desired species (Schmidt, 2008). There is

thus a need to develop better biodiversity indicators

that can be used in LCA and similar studies. It would for

instance be possible to develop the current methodology

to include weighting factors for different species, so

that the indicator would take into account the scarcity

and desirability of different species.

5 Conclusions

The results clearly illustrate the importance of taking

into account the alternative land use options when LCA is

47

used for comparing impacts of different farming systems.

Even though the conventional systems had the highest

energy inputs and GHG emissions per food product output,

the whole farm energy and GHG balances were far more

favourable for the conventional systems compared to the

organic systems once the availability of extra land was

taken into account. The results also suggest that

integrated farming systems that use the best practices

for producing high yields while using environmentally

beneficial farming practices can lead to more favourable

whole farm energy and GHG balances and the lowest

negative impacts on biodiversity compared to organic and

conventional systems.

This paper provides an approach that can be further

developed for identifying land management systems that

optimize food production and environmental impacts. More

research is needed for studying the wider environmental

impacts of the different farming practices, for

developing and testing technologies that improve the

48

sustainability of farming and for determining their

comparative financial viability.

Acknowledgements

We thank Holly Hill Charitable Trust for funding the

project and Tom Curtis (LandShare) for commenting on the

paper.

References

Akiyama, H., Yan, X.Y., Yagi, K., 2010. Evaluation of effectiveness of enhanced-efficiency fertilizers as mitigation options for N2O and NO emissions from agricultural soils: meta-analysis. Global Change Biology 16, 1837-1846.

Bachmaier, J., Effenberger, M., Gronauer, A., 2010. Greenhouse gas balance and resource demand of biogas plants in agriculture. Engineering in Life Sciences 10, 560-569.

Bellamy, P.E., Croxton, P.J., Heard, M.S., Hinsley, S.A.,Hulmes, L., Hulmes, S., Nuttall, P., Pywell, R.F., Rothery, P., 2009. The impact of growing miscanthus for biomass on farmland bird populations. Biomass and Bioenergy 33, 191-199.

Bengtsson, J., Ahnstrom, J., Weibull, A.C., 2005. The effects of organic agriculture on biodiversity and abundance: a meta-analysis. Journal of Applied Ecology 42, 261-269.

49

Berglund, M., Börjesson, P., 2006. Assessment of energy performance in the life-cycle of biogas production. Biomass and Bioenergy 30, 254-266.

Berlin, D., Uhlin, H.-E., 2004. Opportunity cost principles for life cycle assessment: toward strategic decision making in agriculture. Progress in Industrial Ecology 1, 187-202.

Börjesson, P., Berglund, M., 2006. Environmental systems analysis of biogas systems - Part I: Fuel-cycle emissions. Biomass and Bioenergy 30, 469-485.

Cederberg, C., 1998. Life cycle assessment of milk production - a comparison of conventional and organic farming, in: Biotechnology, T.S.I.f.F.a. (Ed.), SIK Report No. 643, Gothenburg (Sweden).

Cooper, J., 2008. Yield differences between organic and conventional agriculture: causes and solutions. The University of New Castle. Presentaion at The 16th IFOAM Organic World Congress: Cultivate the Future 16.6.-20.6.2008.

Dawson, J.J.C., Smith, P., 2007. Carbon losses from soil and its consequences for land-use management. Science of the Total Environment 382, 165-190.

De Schryver, A., Goedkoop, M., Leuven, R., Huijbregts, M., 2010. Uncertainties in the application of the speciesarea relationship for characterisation factors of land occupation in life cycle assessment. The International Journal of Life Cycle Assessment 15, 682-691.

Defra, 2000. Countryside Survey 2000, in: Department for Environment, Affairs, F.a.R. (Eds.).

Defra, 2005. Yields of UK crops and livestock: physiological and technological constraints, and expectations of progress to 2010, Final Report on Defra Project IS0210, p. 21.

Deike, S., Pallutt, B., Christen, O., 2008. Investigations on the energy efficiency of organic and

50

integrated farming with specific emphasis on pesticide use intensity. European Journal of Agronomy 28, 461-470.

Delin, S., Nyberg, A., Lindén, B., Ferm, M., Torstensson,G., Lerenius, C., Gruvaeus, I., 2008. Impact of crop protection on nitrogen utilisation and losses in winter wheat production. European Journal of Agronomy 28, 361-370.

Elsayed, M.A., Matthews, R., Mortimer, N.D., 2003. Carbonand energy balances for a range of biofuels options, Project Number B/B6/00784/REP URN 03/836. Recources Research Unit, Sheffield Hallam University.

Feber, R.E., Johnson, P.J., Firbank, L.G., Hopkins, A., Macdonald, D.W., 2007. A comparison of butterfly populations on organically and conventionally managed farmland. Journal of Zoology 273, 30-39.

Fischer, J., Brosi, B., Daily, G.C., Ehrlich, P.R., Goldman, R., Goldstein, J., Lindenmayer, D.B., Manning, A.D., Mooney, H.A., Pejchar, L., Ranganathan, J., Tallis,H., 2008. Should agricultural policies encourage land sparing or wildlife-friendly farming? Frontiers in Ecology and the Environment 6, 382-387.

Franz, M., 2008. Phosphate fertilizer from sewage sludge ash (SSA). Waste Management 28, 1809-1818.

Gaunt, J.L., Lehmann, J., 2008. Energy Balance and Emissions Associated with Biochar Sequestration and Pyrolysis Bioenergy Production. Environ. Sci. Technol. 42, 4152-4158.

Green, R.E., Cornell, S.J., Scharlemann, J.P.W., Balmford, A., 2005. Farming and the fate of wild nature. Science 307, 550-555.

Hillier, J., Whittaker, C., Dailey, G., Aylott, M., Casella, E., Richter, G.M., Riche, A., Murphy, R., Taylor, G., Smith, P., 2009. Greenhouse gas emissions from four bioenergy crops in England and Wales: Integrating spatial estimates of yield and soil carbon

51

balance in life cycle analyses. Global Change Biology Bioenergy 1, 267-281.

Hodgson, J.A., Kunin, W.E., Thomas, C.D., Benton, T.G., Gabriel, D., 2010. Comparing organic farming and land sparing: optimizing yield and butterfly populations at a landscape scale. Ecology Letters 13, 1358-1367.

Hole, D.G., Perkins, A.J., Wilson, J.D., Alexander, I.H.,Grice, F., Evans, A.D., 2005. Does organic farming benefit biodiversity? Biological Conservation 122, 113-130.

IPCC, 2006. Guidelines for National Greenhouse Gas Inventories. Volume 4 Agriculture, Forestry and Other Land Use.

ISO 14040, 2006. Environmental management - Life cycle assessment -Principles and framework, Geneva. International Organisation for Standardisation (ISO)

Köpke, U., Cooper, J., Lindhard Petersen, H., van der Burgt, G.J.H.M., Tamm, L., 2008. QLIF Workshop 3: Performance of Organic and Low Input Crop Production Systems, 16th IFOAM Organic World Congress, Modena, Italy, June 16-20, 2008.

Lampkin, N., Measures, M., Padel, S., 2008. 2009 Organic Farm Management Handbook. University of Wales, Aberystwyth.

MacCarthy, J., Thomas, J., Choudrie, S., Passant, N., Thistletwaite, G., Murrels, T., Watterson, J., Cardenas, L., Thomson, A., 2010. UK Greenhouse Gas Inventory, 1990 to 2008. Annual Report for Submission under the FrameworkConvention on Climate Change. AEA Tecnology plc. , Didcot, p. 330.

Michel, J., Weiske, A., Möller, K., 2010. The effect of biogas digestion on the environmental impact and energy balances in organic cropping systems using the life-cycleassessment methodology. Renewable Agriculture and Food Systems 25, 204-218.

52

Moakes, S., Lampkin, N., 2003-2009. Organic Farm Incomes in England and Wales 2001/02-2007/08. Aberystwyth University, Aberystwyth.

Mondelaers, K., Aertsens, J., Van Huylenbroeck, G., 2009.A meta-analysis of the differences in environmental impacts between organic and conventional farming. BritishFood Journal 111, 1098-1119.

Nemecek, T., Dubois, D., Huguenin-Elie, O., Gaillard, G.,2011. Life cycle assessment of Swiss farming systems: I. Integrated and organic farming. Agricultural Systems 104,217-232.

Nix, J., 2009. Farm Management Pocketbook 2010, 38 ed. The Andersons Centre.

Phalan, B., Onial, M., Balmford, A., Green, R.E., 2011. Reconciling Food Production and Biodiversity Conservation: Land Sharing and Land Sparing Compared, pp.1289-1291.

Schmidt, J.H., 2008. Development of LCIA characterisationfactors for land use impacts on biodiversity. Journal of Cleaner Production 16, 1929-1942.

Stinner, W., Moller, K., Leithold, G., 2008. Effects of biogas digestion of clover/grass-leys, cover crops and crop residues on nitrogen cycle and crop yield in organicstockless farming systems. European Journal of Agronomy 29, 125-134.

Stone, J.J., Dollarhide, C.R., Benning, J.L., Gregg Carlson, C., Clay, D.E., 2012. The life cycle impacts of feed for modern grow-finish Northern Great Plains US swine production. Agricultural Systems 106, 1-10.

Styles, D., Jones, M.B., 2008. Miscanthus and willow heatproduction - An effective land-use strategy for greenhouse gas emission avoidance in Ireland? Energy Policy 36, 97-107.

Thomassen, M.A., van Calker, K.J., Smits, M.C.J., Iepema,G.L., de Boer, I.J.M., 2008. Life cycle assessment of

53

conventional and organic milk production in the Netherlands. Agricultural Systems 96, 95-107.

Weibull, A.-C., Östman, Ö., Granqvist, Å., 2003. Species richness in agroecosystems: the effect of landscape, habitat and farm management. Biodiversity and Conservation 12, 1335-1355.

Williams, A.G., Audsley, E., Sandars, D.L., 2006. Determining the environmental burdens and resource use inthe production of agricultural and horticultural commodities, Main Report. Defra Research Project IS0205, Bedford.

Williams, A.G., Audsley, E., Sandars, D.L., 2010. Environmental burdens of producing bread wheat, oilseed rape and potatoes in England and Wales using simulation and system modelling. International Journal of Life CycleAssessment 15, 855-868.

Zielonka, S., Lemmer, A., Oechsner, H., Jungbluth, T., 2010. Energy balance of a two-phase anaerobic digestion process for energy crops. Engineering in Life Sciences 10, 515-519.

54

Appendix A

Table A.1. Minimum and maximum yield, energy use and nitrous oxide (N2O) input values used in the Monte Carlo analysis.

O OB C IF IFSYields (t/ha) min max min max min max min max min maxGrass-clover 9.0 12.0 9.0 12.0 9.0 12.0

Potatoes14.0 29.0 14.1 29.3 34.0 40.0 19.9 41.2 34.0 40.0

Winter wheat 3.1 5.0 3.4 5.5 7.2 8.8 4.3 7.0 7.2 8.8Spring beans 2.0 3.6 2.0 3.6 3.0 5.0 2.3 4.1 3.0 5.0Spring barley 1.9 4.0 2.1 4.4 5.4 6.8 2.4 5.0 5.4 6.8Miscanthus 12 15 12 15 12 15 12 15Energy use (MJ/ha)

Grass-clover1073 2503 3584 8362 3584 8362

Potatoes17371 40532 17848 41645 40778 86507 29144 68003 32253 75258

Winter wheat4149 9681 4002 9338 14473 23509 5144 12004 4200 9801

Spring beans4240 9894 4240 9894 4753 11090 4327 10097 3227 7529

Spring barley3726 8694 4203 9807 12884 24456 4952 11554 4461 10409

N2O (kg CO2-eq/ha)Grass-clover 430 2436 0 0 0 0Potatoes 49 280 382 2162 442 2506 402 2278 332 1884Winter wheat 179 1012 371 2101 423 2400 371 2101 287 1629

55

Spring beans 303 1715 0 0 129 729 0 0 27 155Spring barley 172 974 209 1182 243 1379 208 1181 293 1660

56

Table A.2. Minimum and maximum values of the input values used in the Monte Carlo analysis (odt=oven dried tons).

min maxCover crops (N2O emissions kg CO2-eq/ha) 9783 22826

N fertilizer manufacture for conventionalcrops (MJ/ha)Potatoes 7306 8406Winter wheat 8674 9980Spring beans 0 0Spring barley 4739 5453

Miscanthus, biogas, forestsMiscanthus yield (odt/ha) 12 15Miscanthus energy yield (MJ/odt) 10863 16295Heat input in biogas reactor (MJ/t raw material) 192 288Electricity input in biogas reactor (MJ/traw material) 73.6 110.4Methane yield from GC and CC (MJ/tDM) 8480 12720Methane yield from straw (MJ/tDM) 5680 8520Managed forest net energy yield (MJ/ha/yr) 60836

120763

Managed forest net GHG emission mitigation (kg CO2-eq/ha/a) 6764 13428Natural forest C mitigation (kg CO2-eq/ha/a) 7122 14138

Biodiversity indicator valuesConventional arable land 0.73 0.83Integrated arable land 0.31 0.54Organic arable land 0.15 0.51Intensive fertile grassland 0.56 0.72Less intensive fertile grassland 0.14 0.52Organic fertile grassland -0.18 0.15Intensive woodland 0.44 0.65Baseline: natural forest 0 0

57

Appendix B

Table B.1. Results of Wilcoxon Signed rank tests for energy and GHG balances. Energy OBMF - OBMI OBNF - OBMI IFMF - OBMI IFSMF - OBMI OBNF - OBMF IFMF - OBMFZ -174.9(a) -193.7(a) -193.7(b) -28.3(a) -104.5(a) -193.7(b)P 0.000 0.000 0.000 0.000 0.000 0.000Energy IFSMF - OBMF OBNF - IFMF OBNF - IFSMF IFSMF - IFMF CNF - O IFMI - CMI

Z -67.4(b) -193.650(a) -108.225(a) -193.649(a) -173.604(a) -99.233(b)P 0.000 0.000 0.000 0.000 0.000 0.000

GHG OBMF - OBMI OBNF - OBMI OBNF - OBMF IFMI - CMI IFSMF - CMI IFSNF - CMIZ -193.6(a) -193.5(a) -31.2(b) -169.2(a) -95.0(a) -59.3(a)P 0.000 0.000 0.000 0.000 0.000 0.000

GHG IFSMF - IFMIIFSNF -IFMI

IFSNF -IFSMF CNF - CMF IFNF - IFMF IFSNF - CNF

Z -19.703(a) -57.245(a) -31.198(a) -31.198(a) -31.198(a) -193.650(a)P 0.000 0.000 0.000 0.000 0.000 0.000

First letters describes the farming system studied: O = Organic, OB = Organic biogas, C = Conventional, IF = Integrated, IFS = Integrated Special. The last two letters describe the rest-of-the-land use option: MI = Miscanthus, MF = Managed forest and NF = Natural forest, a = based on negative ranks, b =based on positive ranks

58

59

Table B.2. Results of Wilcoxon Signed rank tests for biodiversity impacts

OBMI - OMIOBMF -OMI

OBNF -OMI

CNF -OMI

IFNF -OMI

Z-46.343(a)

-44.218(a)

-58.541(b)

-25.078(a)

-3.901(b)

P 0.000 0.000 0.000 0.000 0.000

OBMF - OBMIOBNF -OBMI

CNF -OBMI

IFNF -OBMI

CMF -IFSMI

Z-40.507(a)

-51.973(a)

-27.434(a)

-30.219(a)

-6.752(b)

P 0.000 0.000 0.000 0.000 0.000

OBMF - OBNFOBMF -CNF

OBMF -IFNF

CNF -OBNF

IFNF -OBNF

IFNF -CNF

Z-51.937(a)

-23.310(a)

-25.902(a)

-21.043(a)

-16.692(a)

-6.787(b)

P 0.000 0.000 0.000 0.000 0.000 0.000First letters describes the farming system studied: O = Organic, OB = Organicbiogas, C = Conventional, IF = Integrated, IFS = Integrated Special. The lasttwo letters describe the rest-of-the-land use option: MI = Miscanthus, MF =Managed forest and NF = Natural forest, a = based on negative ranks, b=based on positive ranks

60

61