Total global agricultural land footprint associated with UK food
supply 1986 - 2011
Henri de Ruiter1,2, Jennie I. Macdiarmid3, Robin B. Matthews1, Thomas Kastner4,5 Lee R, Lynd6,7 and
1Information and Computing Sciences Group, The James Hutton Institute, Craigiebuckler, Aberdeen
AB15 8QH, UK
2Institute of Biological and Environmental Sciences, University of Aberdeen, 23 St Machar Drive,
Aberdeen AB24 3UU, UK
3Public Health Nutrition Research Group, Rowett Institute, University of Aberdeen, Foresterhill,
Aberdeen AB25 2ZD, UK
4Institute of Social Ecology Vienna, Alpen-Adria Universität Klagenfurt, Wien, Graz,
Schottenfeldgasse 29, Vienna 1070, Austria
5 Senckenberg Biodiversity and Climate Research Centre (BiK-F), Senckenberganlage 25, 60325
Frankfurt am Main, Germany.
6 Thayer School of Engineering, Darmouth College, 14 Engineering Drive, Hanover, NH 03755, USA
7 BioEnergy Science Center, Oak Ridge, TN 37831, USA
H.R., J.M., R.B.M. and P.S. initiated and designed the study. H.R. carried out most of the analysis,
and wrote the draft and final version of the manuscript. T.K. performed the analysis on croplands for
feed and food and commented on draft versions of the paper. L.L. contributed the conceptual idea
behind Figure 5 and commented on draft versions of the paper. J.M., R.B.M. and P.S. supervised the
study and commented on draft versions of the manuscript. All authors gave final approval for
This work was supported by a University of Aberdeen Environment and Food Security Theme/the
James Hutton Institute PhD studentship, and contributes to the Scottish Food Security Alliance-Crops
and the Belmont Forum supported DEVIL project (NERC fund UK contribution: NE/M021327/1).
J.M. and R.B.M. acknowledge funding from the Scottish Government’s Rural and Environment
Science Analytical Services Strategic Research Programme. T.K. acknowledges funding from the
European Research Council Grant ERC-263522 (LUISE).
Over the coming decades the global demand for food, and especially for animal products is projected
to increase. At the same time, competition for agricultural land is projected to intensify due to a wide
range of drivers, including a growing world population, changes in food consumption patterns and
bioenergy production. It is therefore vital to understand the relationship between global agricultural
land use and the consumption of food. Here we use the United Kingdom as an example to show the
agricultural land footprint of a highly developed country over the period 1986 – 2011. Our analysis
shows that the total land footprint of the UK has decreased over time from 25,939 kha in 1987 (3-year
mean) to 23,723 kha in 2010 (3-year mean), due to a lower grassland footprint resulting from lower
ruminant meat supply. Cropland use has increased slightly from around 8,400 kha in 1987 to about
8,800 kha in 2010, but has decreased slightly on a per-capita basis as the UK’s population increased
over time. Our analysis shows that 85% of the UK’s total land footprint is associated with meat and
dairy production, but only 48% of total protein and 32% of total calories derive from livestock
products. Our results suggest that, if countries reduce their ruminant product consumption, land could
be freed up for other uses, including bio-energy production, forest regrowth, and biodiversity
Land use, food supply, croplands, grasslands, sustainable diets
Global agricultural land is a finite resource, and competition for land is projected to intensify in the
coming decades due to a range of drivers, including a growing world population, changes in global
food consumption patterns and increasing demand for bioenergy (Haberl 2015). It is projected that
food production has to double in the coming decades to keep up with increasing demand (Tilman et
al. 2011). In theory, an increase in food production can be achieved by expanding current agricultural
areas. Global assessments on the availability of suitable land for agricultural expansion, such as the
FAO report “World Agriculture Towards 2030/2050”, indicate that about 1,400 Mha are still available
for further expansion (Alexandratos and Bruinsma 2012). However, taking into account
comprehensive social and environmental constraints leads to much lower estimates of available land
(Lambin et al. 2013). Moreover, most of the historical expansion of cropland area has been at the
expense of forests (Gibbs et al. 2010), which play a critical role in safeguarding global environmental
sustainability. Forests deliver multiple ecosystem services, store large amounts of carbon and are vital
for preserving global biodiversity (Machovina et al. 2015), and thus further deforestation is an
extremely undesirable option. Hence, meeting increasing demand for food by expanding current
agricultural areas, while taking into account social and environmental constraints, will be very
challenging. On the other hand, it is also possible to increase food production without expanding
current agricultural areas, but in order to achieve this, agricultural yields need to increase considerably
or diets need to change to a much lower consumption of land-intensive livestock products (Erb et al.
2016, Tilman and Clark 2014). In a world without deforestation, human diets are the strongest
determinant of projected total available land area by 2050, implying that shifts towards diets with
lower consumption of animal products represent the best option to increase production while limiting
land use (Erb et al. 2016). However, the feasibility and desirability of such a global strategy towards
such a large reduction of animal product consumption is the subject of much debate.
It is widely accepted that current animal product consumption patterns, especially in Western
countries, are unsustainable and reductions in meat consumption are needed to decrease the pressure
on natural resources (Herrero et al. 2016). Therefore, studies on the environmental consequences of
agriculture have focused particularly on the environmental impact of livestock systems (Garnett
2009). Livestock systems occupy about 30% of the world’s ice-free surface and contribute about 15-
20% to all global greenhouse gas emissions (Steinfeld et al. 2006). It has been shown that diets rich in
meat and dairy products have generally higher associated greenhouse gas emissions and a higher
water and land use (Nijdam et al. 2012). Moreover, it is projected that changes in global dietary
patterns will soon overtake population growth as the main driver behind the increased pressure on
global agricultural land (Kastner et al. 2012, Alexander et al. 2015). As countries become wealthier,
populations tend to consume more meat and livestock products. This global increase in consumption
of livestock products has major implications for food security and the environment, because the
production of livestock products is less efficient compared to producing the same amount of calories
or protein from vegetable sources (Nijdam et al. 2012). More than one third of all crop calories
produced are currently fed to animals, with only 12% of those feed calories coming back as human
food (Cassidy et al. 2013).
Demand management for animal products is thus an important mitigation strategy in the context of
climate change (Bajzelj et al. 2014, Hedenus et al. 2014), and could also limit the amount of land and
water currently used to produce food. Furthermore, the restriction of animal products in the human
diet has received considerable attention since a lower consumption of animal products, particularly of
processed red meat, could be beneficial for human health (McMichael et al. 2007, Friel et al. 2009).
For instance, it has been suggested that a global shift toward a more plant-based diet, in line with
standard dietary guidelines, could reduce global mortality by 6-10% (Springmann et al. 2016).
Because of the potential for environmental and health benefits, it is vital to determine land use
associated with all food consumption in general, as well as with the consumption of livestock
products in particular.
While global analyses are important to highlight the potential of certain strategies, such as demand
management for animal products, studies at country-level are needed to inform national policies,
because local analyses may lead to different insights. For instance, we have recently shown that while
global trade is contributing to more efficient global land use (Kastner et al. 2014), UK trade patterns
are displacing cropland use to other countries (de Ruiter et al. 2016). This highlights the importance
of analysing environmental consequences at different scales.
In this study, we consider the UK as a case study to examine the total land footprint associated with
the total livestock product supply. The production of livestock depends on two broad land use
categories: croplands and grasslands: ruminants, such as cattle, dairy cows and sheep, use large areas
of grassland for grazing or use grasslands indirectly by consuming silage. Monogastrics such as pigs
and chickens, but also ruminants, depend on croplands for much of their feed crops, which primarily
consist of cereals and oil crops. Here we calculate the total livestock land footprint of UK supply by
combining cropland area required for feed crop production and grazing areas required to produce
ruminant products, such as milk and meat. We then compare the land footprint of livestock supply
with the total cropland footprint of crops directly consumed by humans to obtain the total land
footprint associated with UK food supply.
We consider three different types of land in this study: grasslands, croplands used to produce feed
crops, and croplands used to produce crops for human food (see Figure 1 for a summary of our
methodology). To calculate grassland area, we use an adapted version of the methodology developed
by Alexander et al. (2015). To analyse croplands for food and croplands for animal feed (hereafter,
simply referred to as “feed”), we use an updated version of the database described in de Ruiter et al.
(2016). Our analysis now extends to 2011 and hence covers the period 1986 – 2011. In addition, we
were able to split all crop supplies into use for food, feed and “other uses”, allowing us to determine
the cropland footprint for feed and food crops separately. The total calculated agricultural land
footprint is based on UK supply statistics; cropland and grassland ‘embodied’ in exports is not
2.1. Cropland for feed and food
Our analysis for croplands used for feed and food is based on the utilisation shares given by FAO’s
commodity balances (FAOSTAT 2012). For instance, commodity balances for wheat suggest that
about 44% of UK wheat was used for feed in 2011, while about 4% was used for “other uses”, and the
remainder was used for food or food processing. We assume that utilisation shares are the same for
production and imports. That is, if cereals are split 50/50 between food and feed use, we assume that
this is equally valid for domestically produced cereals as well as for imported cereals. However, one
important caveat is that feed use incorporated in imported animal products is added to the feed share
of imports. That is, if the UK imports 500 tons of chicken meat from, for example, the US, and the
production of these chickens from the US required 1000 tons of cereals, these 1000 tons are added to
the feed that is imported. Therefore, feed imports are generally higher than the FAO utilisation shares
for feed for the United Kingdom, since feed incorporated in imported animal products is added to the
feed imports. Animal feed use is based on Kastner et al. (2014).
2.2. Grassland appropriation
The analysis of grasslands areas is more complex, since multiple products can be delivered from the
same area of grassland simultaneously (e.g. ruminant meat and milk), and because data on grasslands
are in general less comprehensive and reliable than for croplands. Furthermore, quantifying grassland
productivity is not as straightforward as for cropland productivity. Definitions of grasslands vary
widely and grasslands are grazed at different intensities (Erb et al. 2007). To calculate the grassland
area associated with UK supply of animal products, we adapt a top-down methodology following
Alexander et al. (2015), in which all grassland areas in a country are assigned to livestock products
(see Figure 2 for a stylised example). We use this methodology to establish the grassland area
associated with animal products for all countries supplying animal products to the UK, including the
UK itself. Grassland areas in a country are allocated to the different animal products on the basis of
the relative difference in feed conversion rates (FCRs) and their total production. Feed conversion
ratios are a measure of how much feed mass (often expressed on a dry matter (DM) basis) is needed to
produce a kilogram of animal product. We use the relative difference between FCRs for typical UK
livestock systems obtained from Wilkinson (2011) for all supplying countries to the UK. Since, to our
knowledge, no reliable temporal information is available on FCRs, we have assumed the FCRs have
remained constant over time. The advantage of this method is that all grassland areas are allocated to
livestock production and total calculated grassland areas are, by definition, consistent with nationally
available grassland areas. Furthermore, using this methodology we are able to take into account trends
regarding intensification or extensification in the use of grasslands.
A limitation of this approach, however, is that no land use is assigned to non-food animal products,
such as leather or wool, and countries with large areas of marginal grasslands, or extensively grazed
areas, show very large grassland appropriation associated with livestock production. Therefore, our
method is likely to overestimate grassland use. However, most of the animal products consumed in
the UK are produced in countries with a relatively high level of intensification, and the role of
countries with vast areas of marginally grazed areas is limited for UK supply of livestock products. To
prevent the possibility that relatively unimportant countries for the UK supply with large marginal
grassland areas skew our findings, we have excluded countries that contribute less than 1.5% to total
UK supply for an individual animal product. This resulted in a total coverage of UK supply for each
animal product of over 95% in all years, except for 2001, which had a coverage of 94% for beef.
None of the countries included had land use values that were widely divergent from typical life-cycle
analysis (LCA) values (Nijdam et al. 2012) (see Tables 1 - 3), with the exception of grassland use for
mutton and lamb (referred to as “mutton”, hereafter) for Australia. The latter finding is explained by
large grassland areas of Australia and the low productivity levels.
A total of ten countries contribute more than 1.5% to UK’s beef supply in at least one of the years in
our study period. Table 1 shows that most of UK beef supply is domestically produced. The domestic
share in total UK beef supply peaked in 1993 (77%) and declined substantially until 2004 (55%), after
which the relative contribution of domestic production to total beef supply increased again. Ireland is
the most important foreign supplier of beef to the UK, with an increasing importance over recent
Most of the mutton is domestically produced, with a significant contribution from New Zealand (see
Table 2). Australia and Ireland have a much smaller contribution to total UK supply, but have a
contribution larger than 1.5% in one or more years during the study period. In total, these four
countries cover 98% or more of UK mutton supply in all years.
More than 75% of milk is domestically produced in all years, but the average share of domestic
production in total UK milk supply has decreased over time. Other contributions to UK milk supply
are from several European countries. Note that milk is obtained in primary product equivalents, which
implies that some this product may be supplied in other forms, such as cheese. Taken together, these
six countries cover more than 95% of total UK milk supply in all years during the studied period.
2.3. Allocation of feed crops to different animal products
FAO food balance sheets, which represent a comprehensive picture of the pattern of a country’s total
food supply and utilisation (FAOSTAT 2012), do not specify feed use for the different animal
products and therefore we developed a method to allocate the cropland footprint to the respective
animal products. The method depends on three variables:
1) The total supply of an animal product
2) The relative FCRs of the different animal products
3) The relative grass content of the feed
Relative FCRs are used based on the logic that if, for instance, the production of a kilogram of beef
requires twice as much feed than the production of a kilogram of pig meat, croplands for feed should
be allocated in the same ratio when total supply of both products is equal. However, allocating based
solely on FCRs does not take into account that the production of animal products require different
types of feed, most notably the difference in grass feed vs. concentrate/crop feed. To account for this,
we use data from Wilkinson (2011) to calculate the average amount of grass vs. crops in the different
livestock systems, expressed per kg of dry matter (DM) per kg of livestock product produced. We do
not take into account differences in crop feed composition between the different livestock systems.
For instance, it might be the case that poultry requires more cereals relative to pigs which would
require a different ratio to allocate cropland for cereals than their FCRs might suggest. However, this
would add another layer of complexity and we have assumed here that for all animals, crop feed
composition is equal. Table 4 shows the FCRs and feed content used in the current study. To our
knowledge, no reliable temporal information are available for FCRs and feed composition, therefore
we have assumed that FCRs and feed composition have remained unchanged over our study period.
In short, we assign cropland area for livestock products according to the following equations:
req × area
req × area
j , t
area j,t=reqj , t
j , t
where areaj,t is cropland area assigned to livestock product j for the year t; reqj crop demand for
livestock product j in year t; req the crop demand for all livestock products in year t; and areat the
total cropland area calculated for all livestock products in year t.
We calculate regj,t according to the equation:
j , t
where supplyj is the UK supply of livestock product j; FCRj the feed conversion ratio for livestock
product j; and Gcj the fractional grass content of the feed.
Based on this procedure, we estimate for 2009 that about 29% of crops assigned to feed utilisation
were being used for the production of pig meat, 24% for beef, 19% for milk, 16% for poultry, 7% for
mutton, and 5% for eggs. Cropland for feed is consequently allocated in the same ratios. Based on our
FCRs, composition of the feed, and total supply, we estimate that a total of 22,745 kton of feed was
required in 2010. Our analysis based on FBS feed use data estimates a total feed use of 22,636 kton of
feed. Both estimates are therefore very close, building confidence in the methods.
2.4. Calculation of nutritional content
We calculate calories and protein content associated with livestock and crop supply using FAO
nutritional factors for calories and protein (FAO 2001). In order to calculate how many calories are
produced on croplands and grasslands, we allocate calories based on the grass content of the feed. For
instance, based on a 75% grass content of cattle feed, we allocate 75% of beef calories to grasslands
and 25% to the croplands for feed.
Our results show that UK ruminant meat (i.e. mutton and beef) supply has decreased over the study
period, while the supply of other animal products has increased, particularly for pig meat and poultry
meat (see Table 5). On a per-capita basis, supply for all animal products, except pig meat and poultry
meat, decreased, with decreases in per-capita ruminant meat consumption of more than 20%. The
share of domestic production in the total UK supply has decreased for all animal products, except for
mutton where the domestic share increased slightly from 69% in 1987 to 72% in 2010 (all values
presented for an individual year are 3-year means around the respective year). In 2010, the domestic
share in the total UK supply was highest for eggs (88%), and lowest for pig meat (34%).
3.1. Cropland footprint for feed crops
Our analysis shows that about 38%, or 22,630 kton, of the total UK crop supply (i.e. domestic
production + imports – exports) in 2010 is used for animal feed (see Table 6). This percentage has
remained relatively constant over the study period. However, because total crop supply increased over
time, the absolute amount of crops used for animal feed has increased over this period. More than half
of the cereals, pulses and oil crops is used for animal feed, while the contributions of other crop
categories are negligible. Of all cereals, the share for animal feed is highest for barley, with around
87% of all barley going to animal feed in our analysis, while about 93% of all soya beans are used for
The cropland footprint associated with animal feed production increased between 2000 and 2008,
after which it declined (see Figure 3) such that the values in 1987 and 2010 are similar. Total cropland
footprint for feed was between 4,000 and 6,000 kha during the studied period. Expressed on a per-
capita basis, the cropland footprint for feed was about 826 m2 cap-1 in 2010. Cereals and oil crops are
the most important contributors to the cropland footprint for feed, while contributions from other crop
categories are negligible.
About 55% of the total cropland footprint for feed, or about 2,619 kha was located overseas in 1987
and this increased to 64%, or 3,293 kha in 2010.
In our analysis, we have assumed that only ruminant products are dependent on grasslands. Figure 4
shows that the total grassland footprint has decreased over the study period, mainly as a result of a
lower ruminant product supply. The grassland footprint is substantially larger than the cropland
footprint. In 2010, total grassland appropriation associated with UK supply was 14,890 kha, while
total cropland footprint for feed was 5,176 kha. In our analysis, most of the total grassland footprint is
allocated to beef production, or about 44% of the grassland footprint averaged over the study period.
Per-capita, the grassland footprint decreased from 3,095 m2 cap-1 to 2,376 m2 cap-1.
In 1987, an average kilogram of beef supplied to the UK required 64 m2 of grassland. In 2010, this
value decreased to about 52 m2. The average area per kilogram of mutton increased, while the area for
milk remained similar between 1987 and 2010 (~130 m2 kg-1 in 1987 to 141 m2 kg-1 in 2010, and ~2.5
m2 kg-1 in 1987 and 2010, respectively). The UK has, according to our analysis, seen an
extensification of livestock production on grasslands, as a lower quantity of ruminant products are
produced on the same amount of land. Other countries important for UK supply, however, have often
intensified production, most notably Brazil and Ireland, and hence a reduction in overall land
requirement per kilogram beef is seen.
The same is reflected in the domestic share of the total grassland footprint. Domestic share in the total
grassland footprint was about 59% in 1987 and this increased to 69% in 2010, despite the fact that the
UK is importing slightly more ruminant products from overseas. This suggests that countries abroad
have intensified their ruminant products, while UK production has seen an extensification.
3.3. Total land footprint of UK food supply
Combining grassland footprint, cropland footprint for feed, and cropland footprint for food shows that
the total land footprint of UK food supply has decreased slightly over recent decades, mainly as a
result of a lower grassland footprint (Figure 5). In 2010, the total land footprint of UK food supply
was 23,723 kha, while this footprint was 25,939 kha in 1987 (-9%). Total cropland footprint
increased: in 2010 the cropland footprint for both feed and food was 8,833 kha compared to 8,406 kha
in 1987. The total land footprint per capita decreased from 4,578 m2 cap-1 to 3,785 m2 cap-1, a drop of
3.4. Calorie and protein supply to the UK
While grasslands were responsible for 63% of the total land footprint in 2010, they only produced
14% of all calories for human consumption, and 22% of total protein (Figure 6). Croplands used for
feed represented 22% of the total land footprint and produced about 3.6 Mt of protein in feed, but only
22% of the protein ended up as animal product for human consumption (0.8 Mt of protein), which
equals 26% of total protein supply. Therefore, livestock products are responsible for 85% of the total
land footprint but produce only 48% of total protein for human consumption. Cereals are most
efficient for producing protein for human consumption; while cereals represent 6% of the total land
footprint, they produce about 33% of the protein for human consumption. The same is true for
calories: despite their large land footprint, grasslands contribute 14% to total calorie supply, and
croplands used for feed ultimately contribute 18% to total human calorie supply through the
production of livestock products. However, stimulants (coffee, cocoa beans and tea) also contribute
less to total protein and calorie supply than their share in the total land footprint would suggest (3% in
land footprint, but deliver 1% of total kcal and protein, respectively). Comparing different types of
land and different types of commodities on a calorie or protein basis may neglect some essential
differences in quality, which will be discussed in the next section.
Our analysis shows that the total land footprint of UK food supply has decreased over recent decades.
This differs from the findings of our previous analysis when we only considered the total cropland
footprint of UK food supply (de Ruiter et al. 2016), and did not include grasslands in our analysis.
Total per-capita ruminant meat supply has decreased as has milk supply, so that the total grassland
footprint associated with UK food supply has decreased over time, which accounts for the difference.
This highlights the importance of including all types of land use in analyses of land footprints.
Also, croplands and grasslands have seen divergent trends in terms of the domestic share in the total
land footprint. The overall trend of the domestic share in the total agricultural land footprint indicates
that throughout the studied period, the domestic share was relatively stable around 55%. However,
crops have increasingly been sourced from abroad, leading to a larger relative, and absolute, cropland
footprint abroad, both for crops for feed as well as for food. In 1987, the domestic share in the
cropland footprint for food and feed, was about 42% and 45%, respectively, and this share decreased
to 38% and 36%, respectively, in 2010. Grasslands, on the other hand, have seen the opposite, where
the relative domestic share in the total grassland footprint has increased over time (from 56% in 1987
to 65% in 2010), even though the absolute domestic grassland footprint has decreased slightly.
According to our analysis, the reason for this is that main exporters of ruminant products have
intensified their production while the UK has seen an extensification on grasslands. Since the 1980’s
there has been a consistent downward trend in cattle numbers, while grassland area has stayed
relatively stable. Reasons for a decrease in cattle numbers include changes in agricultural policies,
such as restrictions on milk production from milk quota, with the decline magnified by outbreaks of
BSE in the 1990’s and foot and mouth disease in the 2000’s (Zayed 2016).
Our analysis indicates that only 15% of UK’s land footprint in 2010 was associated with crops that are
directly grown for human food, while 22% of the total land footprint was used for growing feed crops,
with the remaining 63% due to grazing area. At the same time, this 85% of the land footprint used to
produce animal products only contributed about 32% to total calorie supply and 48% of total protein
supply. This illustrates the relative inefficiency of producing livestock products. However, croplands
and grasslands should not be treated equally, as many grasslands are not suitable for crop production
and livestock production on grassland does not always compete with food for human consumption
(Schader et al. 2015), and may be a good option, especially if grasslands are used for milk production
(Wilkinson 2011). Moreover, converting grasslands to cropland, if at all possible, may have negative
consequences for greenhouse gas emissions in the short term, because grasslands generally store more
carbon than croplands (Smith 2014). Nonetheless, grasslands that become available could be used to
regrow forests that could sequester carbon and benefit biodiversity, or be used to produce bio-energy
that does not compete directly with human food production (Lamb et al. 2016).
4.1. Dietary change as a means to reduce land footprints
Our results indicate that the lower UK land footprint associated with food is due to a lower ruminant
product supply, which suggests that a lower meat consumption globally could lower land use. At the
global scale, there are some indications of an inverted U-shaped relation between income and meat
consumption, implying that with high income, people are starting to consume less meat, although the
evidence is mixed and highly heterogeneous amongst countries (Cole and McCoskey 2013, Vranken
et al. 2014, Bodirsky et al. 2015). Red meat consumption has been gradually declining in the
European Union and the USA over recent decades, while the supply of pig meat and poultry meat has
increased over the same period of time, similar to our UK data (FAOSTAT 2012, Daniel et al. 2011).
The relation between income and meat consumption, particularly red meat consumption, may have
positive implications for future land footprints, especially for high-income countries. At the same
time, it must be recognised that there is a general reluctance to reduce the consumption of meat in
developed countries because meat consumption is the normalised dietary habit (Macdiarmid et al.
2016). Strategies aimed at reducing the consumption of meat should consider social and cultural
norms and may be different for different types of consumers (de Boer et al. 2014). Our analysis
indicates that stimulants also have a lower contribution to protein and calorie supply than their share
in the land footprint would suggest, as is the case for sugar crops for protein supply. This may be a
result of the fact that these crops are not primarily consumed for protein and energy (for stimulants) or
for protein (for sugar crops). However, at the same time, these crops are not a necessary part of a
healthy diet, so strategies for a lower land footprint may include reducing consumption of stimulants,
though their total share in UK’s land footprint is only 3%. Moreover, the consumption of stimulants is
probably as ingrained in current food consumption patterns as meat, and might be equally hard to
The current study only considers calories and protein because only these nutrients were available in
the FAO data, while a healthy diet requires a diverse set of micronutrients. Achieving a healthy and
environmentally sustainable diet requires looking at the whole diet and not just single nutrients,
especially since the UK population is not deficient in protein or calories. Moreover, singling out
individual food items, such as animal products, to reduce the environmental impact of food, is an
oversimplification. Modelling individual diets to reduce greenhouse gas emissions while optimising
nutritional outcomes reveals that there are multiple options to achieve this, and that in some cases,
increasing meat intake can actually be the best strategy to ensure an acceptable nutritionally balanced
diet (Horgan et al. 2016). Therefore, it is important to consider whole diets and to extend our current
analysis using a wider range of nutrients.
4.2. Limitations of the current study
We rely mostly on FAOSTAT data for our food supply analysis, which has known reliability and
robustness issues, particularly for developing countries. However it is the best global-level dataset for
food supply and widely used for similar purposes as the current study. Moreover, our time series only
extends to the year 2011, while important changes may have happened since 2011. This is because the
release of particularly FAO utilisation shares data generally lags a few years. Furthermore, we
calculate cropland areas for feed and for food based on the same country-specific average yields for
crops, while in reality the yields of feed crops may differ from those of crops grown mainly for food.
Also, using the methodology by Kastner et al. (2014), we assume crops that are exported to the UK
have the same average yields as crops used for domestic consumption in the respective country. This
might lead to different results if, for instance, crops grown for exports are more intensively managed
and thus have higher yields. Moreover, a temporal trend may be present for these two issues (e.g.
yields of feed crops may have improved faster over time than for food crops) which we cannot detect
using national average yields. A more comprehensive dataset with different yields for feed and export
crops may overcome this issue, but will be difficult to establish in practice.
The calculation of grassland area is more complex than the relatively straight-forward approach for
cropland areas. Our current approach has a number of advantages: it assigns all grasslands to animal
products and calculated areas are therefore consistent with nationally available statistics. It is also able
to capture changes in intensification levels over time, and it is a relatively simple and transparent way
of assigning grassland areas to animal products. Results are generally in line with LCA studies (see
later section). However, it could be argued that this method does not distinguish meaningful changes
in levels of intensification. For instance, our method suggests that beef production in the UK required
11% more grassland per kilogram of beef in 2010 compared to 1987. It is unclear, however, whether
this reflects actual changes in the production of beef and associated level of intensity, or because
grassland areas was used for other purposes, e.g. set aside or used to produce other products.
4.3. Sensitivity analysis for FCR ratios
Our allocation of grassland and cropland areas to different livestock products is primarily based on the
FCR ratios obtained for typical UK livestock systems. We use these relative FCRs for all countries in
this study, which implies that all beef production is allocated a grassland area per kilogram that is ~17
times larger than milk production, which is based on the differences in respective FCR (18.5 vs 1.1).
Our assumption is that the relative difference between the FCRs has not changed over time, i.e. the
efficiency of the production of each animal product alters at the same rate. In reality, it is likely that
relative FCRs differ between countries, for instance when a country specialises in one type of product,
or that relative difference in FCRs has changed over time. Moreover, there is large uncertainty
involved with FCRs and the choice of FCRs will impact greatly on where grasslands and croplands
are allocated. For the original Alexander et al. (2015) study, this was not a problem, because the
authors were analysing relative changes in global grassland areas over time instead of absolute
grassland appropriation. For our purposes, this is an issue, because we are not assigning all global
grassland areas, only the areas that are linked to UK supply. We performed a sensitivity analysis to
show the effect of relative FCRs on grassland allocation, based on plausible FCRs obtained from the
literature. As Table 7 shows, the relative value of the FCRs has an influence, particularly for the
relative share of animal products in the total land use. Total grassland use is, in all cases, around
15,000 kha. Using different estimates for FCRs leads to a grassland for beef estimate ranging from
6,826 kha to 10,005 kha. Therefore, our sensitivity analysis shows that the relative FCR is important
when allocating grassland areas to different animal products. It also shows, however, that the total
grassland footprint associated with animal products does not change much (~ 6%) and is robust across
different assumptions regarding FCRs.
4.4. Comparison with other studies
Nijdam et al. (2012) compared the land use of a wide range of different LCA studies, and included
estimates of grassland use per kilogram of livestock product. They estimated, depending on the
livestock system, a grassland use for beef ranging from 2 to 420 m2 kg-1, 18 to 30 m2 kg-1 for mutton,
and circa 1 m2 kg-1 for milk. Our analysis indicates comparable grassland values for the 10 countries
that have been included using the 1.5% cut-off, as our values range from 7 m2 kg-1 (Belgium) to 396
m2 kg-1 (Brazil) for beef; and less comparable values for mutton & lamb, as our values range from 52
(Ireland) to 2,136 (Australia) m2 kg-1. However, about 95% of UK supply falls within the range of 72
(UK) to 126 (NZ). Our values for milk are quite similar, ranging from 0.2 m2/kg (Denmark) to 3.4 m2/
kg (UK). Audsley et al. (2009) estimate, based on a LCA study, typical UK grassland factors for beef
to be 12.5 m2 kg-1, about 93 m2 kg-1 for mutton, and about 0.1 m2 kg-1 for milk, compared to our
estimates for the UK of 44-57 m2 kg-1 for beef, 75-98 m2 kg-1 for mutton, and 2.6 – 3.4 m2 kg-1 for
milk. This shows that our grassland values for beef and milk are higher, while our values for mutton
are comparable. Some of these changes can be attributed to the fact that our method assigns all
grasslands in the UK, including grasslands that may not be used for livestock production to livestock
products in a top-down way, whereas Audsley et al. (2009) arrive at their values using a bottom-up
LCA approach. Nevertheless, conclusions about the total land footprint will differ substantially using
these two different approaches.
Lesschen et al. (2011) estimated EU-27 land use values for crop feed and forage to be 37.3 m2 kg-1 for
beef (vs. our estimate of 65.7 m2 kg-1), 2.4 m2 kg-1 for milk (vs. 3.9 m2 kg-1), 11.7 m2 kg-1 for pork (vs.
8.4 m2 kg-1), 9.2 m2 kg-1 for poultry (vs. 4.6 m2 kg-1 ), and 9.0 m2 kg-1 for eggs (vs. 5.1 m2 kg-1). Our
land use values for milk and beef are generally higher because we assign all grassland areas to
livestock production and because the UK relies relatively heavily on pasture area.
Lastly, Audsley et al. (2010) estimated, using a detailed LCA model, the following land uses
associated with UK food supply for the year 2005: 4,561 kha for human crops (vs. our estimate for
2005 of 3,657 kha), 3,265 kha for feed crops (vs. 5,433 kha) and 13,172 kha for grasslands (vs. 16,577
kha), roughly in line with our estimates. Differences could be explained by the fact that Audsley et al.
(2010) used a bottom-up approach to estimate feed use and grassland use.
In conclusion, our study shows that the total land footprint associated with UK food supply has
decreased over time, particularly because of a decrease in the grassland footprint. The decrease in the
grassland footprint is caused by a decline in ruminant meat consumption. Our findings highlight the
importance of including all types of land in the total land footprint and might suggest that, if there is
an inverted U-shaped relation between income and ruminant meat consumption, more land could
become available for other uses as incomes rise.
Alexander, P., Rounsevell, M.D.A., Dislich, C., Dodson, J.R., Engström, K. and Moran, D. (2015)
Drivers for global agricultural land use change: The nexus of diet, population, yield and bioenergy.
Global Environ. Change 35, pp. 138-147 http://dx.doi.org/10.1016/j.gloenvcha.2015.08.011.
Alexandratos, N. and Bruinsma, J., 2012. World Agriculture: Towards 2015/2030. The 2012
Revision. London: Earthscan Publications.
Audsley, E., Angus, A., Chatterton, J., Graves, A., Morris, J., Murphy-Bokern, D., Pearn, K., Sandars,
D. and Williams, A., 2010. Food, land and greenhouse gases. The effect of changes in UK food
consumption on land requirements and greenhouse gas emissions. The Committee on Climate
Audsley, E., Brander, M., Chatterton, J., Murphy-Bokern, D., Webster, C. and Williams, A., 2009.
How low can we go? An assessment of greenhouse gas emissions from the UK food system and the
scope to reduce them by 2050. WWF UK.
Bajzelj, B., Richards, K.S., Allwood, J.M., Smith, P., Dennis, J.S., Curmi, E. and Gilligan, C.A.
(2014) Importance of food-demand management for climate mitigation. Nature Clim. Change 4, pp.
Bodirsky, B.L., Rolinski, S., Biewald, A., Weindl, I., Popp, A. and Lotze-Campen, H. (2015) Global
Food Demand Scenarios for the 21st Century. PLoS ONE 10, pp. e0139201.
Cassidy, E.S., West, P.C., Gerber, J.S. and Foley, J.A. (2013) Redefining agricultural yields: from
tonnes to people nourished per hectare. Environ. Res. Lett. 8, pp. 034015
Cole, J.R. and McCoskey, S. (2013) Does global meat consumption follow an environmental Kuznets
curve? Sustainability: Science, Practice, & Policy 9.
Daniel, C.R., Cross, A.J., Koebnick, C. and Sinha, R. (2011) Trends in meat consumption in the USA.
Public Health Nutr. 14, pp. 575-583 http://dx.doi.org/10.1017%2FS1368980010002077.
de Boer, J., Schösler, H. and Aiking, H. (2014) “Meatless days” or “less but better”? Exploring
strategies to adapt Western meat consumption to health and sustainability challenges. Appetite 76, pp.
de Ruiter, H., Macdiarmid, J.I., Matthews, R.B., Kastner, T. and Smith, P. (2016) Global cropland and
greenhouse gas impacts of UK food supply are increasingly located overseas. J R Soc Interface 13,
Erb, K., Gaube, V., Krausmann, F., Plutzar, C., Bondeau, A. and Haberl, H. (2007) A comprehensive
global 5 min resolution land-use data set for the year 2000 consistent with national census data. J
Land use Sci. 2, pp. 191-224 http://dx.doi.org/10.1080/17474230701622981.
Erb, K., Lauk, C., Kastner, T., Mayer, A., Theurl, M.C. and Haberl, H. (2016) Exploring the
biophysical option space for feeding the world without deforestation. Nat Commun 7,
FAO., 2001. FOOD BALANCE SHEETS - A Handbook. Rome: Food and Argicultural Organization
of the United Nationas.
FAOSTAT., 2012. FAO Statistical Database - FAOSTAT. Available from: http://faostat.fao.org/.
Friel, S., Dangour, A.D., Garnett, T., Lock, K., Chalabi, Z., Roberts, I., Butler, A., Butler, C.D.,
Waage, J., McMichael, A.J. and Haines, A. (2009) Public health benefits of strategies to reduce
greenhouse-gas emissions: food and agriculture. Lancet 374, pp. 2016-2025 http://dx.doi.org/10.1016/
Garnett, T. (2009) Livestock-related greenhouse gas emissions: impacts and options for policy
makers. Environ.Sci.& Policy 12, pp. 491-503 http://dx.doi.org/10.1016/j.envsci.2009.01.006.
Gibbs, H.K., Ruesch, A.S., Achard, F., Clayton, M.K., Holmgren, P., Ramankutty, N. and Foley, J.A.
(2010) Tropical forests were the primary sources of new agricultural land in the 1980s and 1990s.
Proc. Natl. Acad. Sci. USA 107, pp. 16732-16737 http://dx.doi.org/10.1073/pnas.0910275107.
Haberl, H. (2015) Competition for land: A sociometabolic perspective. Ecol.Econ. 119, pp. 424-431
Hedenus, F., Wirsenius, S. and Johansson, D.A. (2014) The importance of reduced meat and dairy
consumption for meeting stringent climate change targets. Clim. Change 124, pp. 79-91
Herrero, M., Henderson, B., Havlik, P., Thornton, P.K., Conant, R.T., Smith, P., Wirsenius, S.,
Hristov, A.N., Gerber, P., Gill, M., Butterbach-Bahl, K., Valin, H., Garnett, T. and Stehfest, E. (2016)
Greenhouse gas mitigation potentials in the livestock sector. Nature Clim.Change 6, pp. 452-461
Horgan, G.W., Perrin, A., Whybrow, S. and Macdiarmid, J.I. (2016) Achieving dietary
recommendations and reducing greenhouse gas emissions: modelling diets to minimise the change
from current intakes. Int J Behav Nutr Phys Act. 13, pp. 1-11 http://dx.doi.org/10.1186/s12966-016-
Kastner, T., Erb, K.H. and Haberl, H. (2014) Rapid growth in agricultural trade: effects on global area
efficiency and the role of management. Environ. Res. Lett. 9, pp. 034015
Kastner, T., Rivas, M.J.I., Koch, W. and Nonhebel, S. (2012) Global changes in diets and the
consequences for land requirements for food. Proc. Natl. Acad. Sci. USA
Lamb, A., Green, R., Bateman, I., Broadmeadow, M., Bruce, T., Burney, J., Carey, P., Chadwick, D.,
Crane, E., Field, R., Goulding, K., Griffiths, H., Hastings, A., Kasoar, T., Kindred, D., Phalan, B.,
Pickett, J., Smith, P., Wall, E., zu Ermgassen, Erasmus K. H. J. and Balmford, A. (2016) The potential
for land sparing to offset greenhouse gas emissions from agriculture. Nature Clim.Change 6, pp. 488-
Lambin, E.F., Gibbs, H.K., Ferreira, L., Grau, R., Mayaux, P., Meyfroidt, P., Morton, D.C., Rudel,
T.K., Gasparri, I. and Munger, J. (2013) Estimating the world's potentially available cropland using a
bottom-up approach. Global Environ.Change 23, pp. 892-901
Lesschen, J.P., van den Berg, M., Westhoek, H.J., Witzke, H.P. and Oenema, O. (2011) Greenhouse
gas emission profiles of European livestock sectors. Anim.Feed Sci.Technol. 166–167, pp. 16-28
Macdiarmid, J.I., Douglas, F. and Campbell, J. (2016) Eating like there's no tomorrow: Public
awareness of the environmental impact of food and reluctance to eat less meat as part of a sustainable
diet. Appetite 96, pp. 487-493 http://dx.doi.org/10.1016/j.appet.2015.10.011.
Machovina, B., Feeley, K.J. and Ripple, W.J. (2015) Biodiversity conservation: The key is reducing
meat consumption. Sci.Total Environ. 536, pp. 419-431
McMichael, A.J., Powles, J.W., Butler, C.D. and Uauy, R. (2007) Food, livestock production, energy,
climate change, and health. The Lancet 370, pp. 1253-1263 http://dx.doi.org/10.1016/S0140-
Nijdam, D., Rood, T. and Westhoek, H. (2012) The price of protein: Review of land use and carbon
footprints from life cycle assessments of animal food products and their substitutes. Food Policy 37,
pp. 760-770 http://dx.doi.org/10.1016/j.foodpol.2012.08.002.
Schader, C., Muller, A., Scialabba, N.E., Hecht, J., Isensee, A., Erb, K., Smith, P., Makkar, H.P.S.,
Klocke, P., Leiber, F., Schwegler, P., Stolze, M. and Niggli, U. (2015) Impacts of feeding less food-
competing feedstuffs to livestock on global food system sustainability. J R Soc Interface 12,
Smith, P. (2014) Do grasslands act as a perpetual sink for carbon? Global Change Biol. 20, pp. 2708-
Springmann, M., Godfray, H.C.J., Rayner, M. and Scarborough, P. (2016) Analysis and valuation of
the health and climate change cobenefits of dietary change. Proc. Natl. Acad. Sci. USA 113, pp. 4146-
Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M. and de Haan, C., 2006. Livestock’s
long shadow: environmental issues and options. Rome: Food and Agriculture Organisation.
Tilman, D., Balzer, C., Hill, J. and Befort, B.L. (2011) Global food demand and the sustainable
intensification of agriculture. Proc. Natl. Acad. Sci. USA 108, pp. 20260-20264
Tilman, D. and Clark, M. (2014) Global diets link environmental sustainability and human health.
Nature 515, pp. 518-522 http://dx.doi.org/10.1038/nature13959.
Vranken, L., Avermaete, T., Petalios, D. and Mathijs, E. (2014) Curbing global meat consumption:
Emerging evidence of a second nutrition transition. Environ.Sci.& Policy 39, pp. 95-106
Wilkinson, J.M. (2011) Re-defining efficiency of feed use by livestock. Animal 5, pp. 1014-1022
Zayed, Y., 2016. Agriculture: historical statistics. Briefing Paper.
Figure 1. Summary of data sources and methodology used in the current study.
Figure 2. Stylised example of top-down method used to assign grassland areas to livestock products,
based on Alexander et al. (2015). Three inputs are required: total grassland area in a country, its total
livestock production, and FCRs for the different livestock products. First, total grassland area in a
country is obtained. Hereafter, total production of a ruminant livestock product is multiplied by its
respective feed conversion ratio (FCR) to obtain the total grass requirements. Total grassland area is
then allocated to the different livestock products based on the relative difference in total grass
Figure 3. Cropland footprint associated with UK feed supply for the period 1986 - 2011.
Figure 4. Grassland footprint associated with UK food supply for the period 1986 - 2011.
Figure 5. Total land footprint associated with total UK food supply for the period 1986 - 2011.
Figure 6. Total protein supply (left panel) and total calorie supply (right panel) as cumulative
percentage produced on the different types of land (as cumulative percentage of total land footprint).
For instance, grasslands make up 63% of the total land footprint, but produce only 22% and 14% of
protein and calories, respectively. “Livestock croplands” represents the amount of calories and protein
delivered in livestock products from feed crops, while “livestock grasslands” represent the calories
and protein delivered on grasslands. All other crop categories are used for human food. Note that
protein and calories delivered are not necessarily consumed. For instance, we calculate total protein
from oil crops based on protein content of oil crops, but some oil crops may only be used for
vegetable oil production.
Table 1. Countries contributing > 1.5% to UK beef supply during at least one year of the period 1986
- 2011. To give an indication of the trend in intensification levels, 3-year means of the average land
use value per kg at the end (2010) and beginning (1987) of the study were compared.
Table 2. Countries contributing > 1.5% to UK mutton supply during at least one year of the period
1986 - 2011. To give an indication of the trend in intensification levels, 3-year means of the average
land use value per kg at the end (2010) and beginning (1987) of the study were compared.
Table 3. Countries contributing > 1.5% to UK milk supply during at least one year of the period 1986
- 2011. To give an indication of the trend in intensification levels, 3-year means of the average land
use value per kg at the end (2010) and beginning (1987) of the study were compared.
Table 4. Feed conversion ratios and feed composition used in this study.
Table 5. UK animal product supply: total supply; supply with only countries contributing > 1.5%
included; domestic supply. Values are 3-year means around the respective year.
Table 6. UK crop supply and percentage used for feed in 2010 (3-year mean).
Table 7. Sensitivity analysis for different FCRs. Values are averaged over the period 1986-2012. Total
may be different due to rounding.
FROM 1987 TO
ON TO TOTAL
ARGENTINA 265 332 5% 1.4%
7 9 N/A 0.8%
BRAZIL 171 396 -54% 5.3%
FRANCE 29 33 -3% 0.9%
GERMANY 13 17 9% 1.6%
IRELAND 30 54 -30% 14.5%
NETHERLANDS 7 10 -19% 1.6%
UNITED KINGDOM 44 57 10% 68%
UNITED STATES OF
138 162 -9% 1.4%
URUGUAY 176 316 -39% 1%
% CHANGE FROM
1987 TO 2010 (3-
TO TOTAL UK
AUSTRALIA 1,631 2,638 -33% 2.6%
IRELAND 52 92 -30% 0.9%
72 126 -36% 22.3%
75 98 10% 72.9%
% CHANGE FROM
1987 TO 2010 (3-
TO TOTAL UK
DENMARK 0.2 0.5 12% 1.6%
FRANCE 1.7 2.0 -3% 3.0%
GERMANY 0.8 1.0 9% 2.8%
IRELAND 1.8 3.2 -30% 4.8%
NETHERLANDS 0.4 0.6 -19% 1.3%
2.6 3.4 10% 82.7%
ANIMAL PRODUCT FCR % GRASS IN FEED
PIG 3.6 0%
POULTRY 2 0%
MILK 1.1 75%
BEEF 18.5* 75%
EGGS 2.2 0%
MUTTON 31.7* 90%
* Weighted average based on relative importance of different
production systems for cattle and sheep (e.g. upland vs. lowland
> 1.5% (KTON)
SUPPLY 1987 –
1987 2010 1987 2010 1987 2010
BEEF 1,348 1,173 1,297 1,123 943 813 - 13%
MUTTON 385 331 384 325 256 239 - 14%
MILK 14,880 16,106 14,710 15,416 13,312 12,150 + 8%
PIG 1,521 1,823 n/a n/a 955 618 + 20%
POULTRY 1,042 1,844 n/a n/a 689 638 + 77%
EGGS 710 723 n/a n/a 954 1323 + 2%
CROP CATEGORY TOTAL SUPPLY
CEREALS 25,099 15,335 61
ROOTS AND TUBERS 7,389 355 5
SUGAR CROPS 2,029 93 5
PULSES 688 494 72
NUTS 47 0 0
OIL CROPS 11,404 6,216 55
VEGETABLES 7,343 135 2
FRUITS 5,792 1 0
SPICES 58 0 0
STIMULANTS 448 0 0
TOTAL 60,297 22,630 38
PRODUCT BEEF MUTTON MILK TOTAL
TOTAL SUPPLY (KTON) 1105 371 14932
ALEXANDER 2015 25 12 0.6
TOTAL KHA 10,005 2,258 2,678 14,941
WILKINSON 2011 18.5 31.7 1.1
TOTAL KHA 6,826 4,888 3,942 15,655
BOUWMAN 2005 24 27 1.1
TOTAL KHA 7,824 3,865 3,669 15,358